diff --git a/.github/ISSUE_TEMPLATE/01_bugs.md b/.github/ISSUE_TEMPLATE/01_bugs.md index 255a5241e..f0d0ba912 100644 --- a/.github/ISSUE_TEMPLATE/01_bugs.md +++ b/.github/ISSUE_TEMPLATE/01_bugs.md @@ -10,7 +10,7 @@ about: Use this template if you came across a bug or unexpected behaviour differ ## Your Environment - + * Operating System: * Python Version Used: * spaCy Version Used: diff --git a/.github/azure-steps.yml b/.github/azure-steps.yml index c7722391f..cc0247b3a 100644 --- a/.github/azure-steps.yml +++ b/.github/azure-steps.yml @@ -10,6 +10,7 @@ steps: inputs: versionSpec: ${{ parameters.python_version }} architecture: ${{ parameters.architecture }} + allowUnstable: true - bash: | echo "##vso[task.setvariable variable=python_version]${{ parameters.python_version }}" @@ -27,7 +28,7 @@ steps: - script: python -m mypy spacy displayName: 'Run mypy' - condition: ne(variables['python_version'], '3.10') + condition: ne(variables['python_version'], '3.6') - task: DeleteFiles@1 inputs: diff --git a/.github/workflows/autoblack.yml b/.github/workflows/autoblack.yml index 8d0282650..3ad4cf408 100644 --- a/.github/workflows/autoblack.yml +++ b/.github/workflows/autoblack.yml @@ -12,10 +12,10 @@ jobs: if: github.repository_owner == 'explosion' runs-on: ubuntu-latest steps: - - uses: actions/checkout@v2 + - uses: actions/checkout@v3 with: ref: ${{ github.head_ref }} - - uses: actions/setup-python@v2 + - uses: actions/setup-python@v3 - run: pip install black - name: Auto-format code if needed run: black spacy @@ -23,10 +23,11 @@ jobs: # code and makes GitHub think the action failed - name: Check for modified files id: git-check - run: echo ::set-output name=modified::$(if git diff-index --quiet HEAD --; then echo "false"; else echo "true"; fi) + run: echo modified=$(if git diff-index --quiet HEAD --; then echo "false"; else echo "true"; fi) >> $GITHUB_OUTPUT + - name: Create Pull Request if: steps.git-check.outputs.modified == 'true' - uses: peter-evans/create-pull-request@v3 + uses: peter-evans/create-pull-request@v4 with: title: Auto-format code with black labels: meta diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index b959262e3..df59697b1 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -6,7 +6,7 @@ repos: language_version: python3.7 additional_dependencies: ['click==8.0.4'] - repo: https://gitlab.com/pycqa/flake8 - rev: 3.9.2 + rev: 5.0.4 hooks: - id: flake8 args: diff --git a/README.md b/README.md index d9ef83e01..abfc3da67 100644 --- a/README.md +++ b/README.md @@ -8,7 +8,7 @@ be used in real products. spaCy comes with [pretrained pipelines](https://spacy.io/models) and -currently supports tokenization and training for **60+ languages**. It features +currently supports tokenization and training for **70+ languages**. It features state-of-the-art speed and **neural network models** for tagging, parsing, **named entity recognition**, **text classification** and more, multi-task learning with pretrained **transformers** like BERT, as well as a @@ -16,7 +16,7 @@ production-ready [**training system**](https://spacy.io/usage/training) and easy model packaging, deployment and workflow management. spaCy is commercial open-source software, released under the MIT license. -💫 **Version 3.4.0 out now!** +💫 **Version 3.4 out now!** [Check out the release notes here.](https://github.com/explosion/spaCy/releases) [![Azure Pipelines](https://img.shields.io/azure-devops/build/explosion-ai/public/8/master.svg?logo=azure-pipelines&style=flat-square&label=build)](https://dev.azure.com/explosion-ai/public/_build?definitionId=8) @@ -79,7 +79,7 @@ more people can benefit from it. ## Features -- Support for **60+ languages** +- Support for **70+ languages** - **Trained pipelines** for different languages and tasks - Multi-task learning with pretrained **transformers** like BERT - Support for pretrained **word vectors** and embeddings diff --git a/azure-pipelines.yml b/azure-pipelines.yml index f475b7fdd..eea07cb7a 100644 --- a/azure-pipelines.yml +++ b/azure-pipelines.yml @@ -31,7 +31,7 @@ jobs: inputs: versionSpec: "3.7" - script: | - pip install flake8==3.9.2 + pip install flake8==5.0.4 python -m flake8 spacy --count --select=E901,E999,F821,F822,F823,W605 --show-source --statistics displayName: "flake8" @@ -76,15 +76,24 @@ jobs: # Python39Mac: # imageName: "macos-latest" # python.version: "3.9" - Python310Linux: - imageName: "ubuntu-latest" - python.version: "3.10" + # Python310Linux: + # imageName: "ubuntu-latest" + # python.version: "3.10" Python310Windows: imageName: "windows-latest" python.version: "3.10" - Python310Mac: - imageName: "macos-latest" - python.version: "3.10" + # Python310Mac: + # imageName: "macos-latest" + # python.version: "3.10" + Python311Linux: + imageName: 'ubuntu-latest' + python.version: '3.11.0-rc.2' + Python311Windows: + imageName: 'windows-latest' + python.version: '3.11.0-rc.2' + Python311Mac: + imageName: 'macos-latest' + python.version: '3.11.0-rc.2' maxParallel: 4 pool: vmImage: $(imageName) diff --git a/requirements.txt b/requirements.txt index 3e8501b2f..9d6bbb2c4 100644 --- a/requirements.txt +++ b/requirements.txt @@ -15,7 +15,7 @@ pathy>=0.3.5 numpy>=1.15.0 requests>=2.13.0,<3.0.0 tqdm>=4.38.0,<5.0.0 -pydantic>=1.7.4,!=1.8,!=1.8.1,<1.10.0 +pydantic>=1.7.4,!=1.8,!=1.8.1,<1.11.0 jinja2 langcodes>=3.2.0,<4.0.0 # Official Python utilities @@ -28,11 +28,12 @@ cython>=0.25,<3.0 pytest>=5.2.0,!=7.1.0 pytest-timeout>=1.3.0,<2.0.0 mock>=2.0.0,<3.0.0 -flake8>=3.8.0,<3.10.0 +flake8>=3.8.0,<6.0.0 hypothesis>=3.27.0,<7.0.0 -mypy>=0.910,<0.970; platform_machine!='aarch64' +mypy>=0.980,<0.990; platform_machine != "aarch64" and python_version >= "3.7" types-dataclasses>=0.1.3; python_version < "3.7" types-mock>=0.1.1 +types-setuptools>=57.0.0 types-requests types-setuptools>=57.0.0 black>=22.0,<23.0 diff --git a/setup.cfg b/setup.cfg index 2dc5e7042..c2653feba 100644 --- a/setup.cfg +++ b/setup.cfg @@ -56,7 +56,7 @@ install_requires = tqdm>=4.38.0,<5.0.0 numpy>=1.15.0 requests>=2.13.0,<3.0.0 - pydantic>=1.7.4,!=1.8,!=1.8.1,<1.10.0 + pydantic>=1.7.4,!=1.8,!=1.8.1,<1.11.0 jinja2 # Official Python utilities setuptools diff --git a/setup.py b/setup.py index c4138aa93..243554c7a 100755 --- a/setup.py +++ b/setup.py @@ -30,7 +30,9 @@ MOD_NAMES = [ "spacy.lexeme", "spacy.vocab", "spacy.attrs", - "spacy.kb", + "spacy.kb.candidate", + "spacy.kb.kb", + "spacy.kb.kb_in_memory", "spacy.ml.parser_model", "spacy.morphology", "spacy.pipeline.dep_parser", diff --git a/spacy/__init__.py b/spacy/__init__.py index d60f46b96..c3568bc5c 100644 --- a/spacy/__init__.py +++ b/spacy/__init__.py @@ -31,9 +31,9 @@ def load( name: Union[str, Path], *, vocab: Union[Vocab, bool] = True, - disable: Union[str, Iterable[str]] = util.SimpleFrozenList(), - enable: Union[str, Iterable[str]] = util.SimpleFrozenList(), - exclude: Union[str, Iterable[str]] = util.SimpleFrozenList(), + disable: Union[str, Iterable[str]] = util._DEFAULT_EMPTY_PIPES, + enable: Union[str, Iterable[str]] = util._DEFAULT_EMPTY_PIPES, + exclude: Union[str, Iterable[str]] = util._DEFAULT_EMPTY_PIPES, config: Union[Dict[str, Any], Config] = util.SimpleFrozenDict(), ) -> Language: """Load a spaCy model from an installed package or a local path. diff --git a/spacy/about.py b/spacy/about.py index 843c15aba..ce86e6294 100644 --- a/spacy/about.py +++ b/spacy/about.py @@ -1,6 +1,6 @@ # fmt: off __title__ = "spacy" -__version__ = "3.4.1" +__version__ = "3.4.2" __download_url__ = "https://github.com/explosion/spacy-models/releases/download" __compatibility__ = "https://raw.githubusercontent.com/explosion/spacy-models/master/compatibility.json" __projects__ = "https://github.com/explosion/projects" diff --git a/spacy/cli/_util.py b/spacy/cli/_util.py index 3506bdde7..ddd6c10ce 100644 --- a/spacy/cli/_util.py +++ b/spacy/cli/_util.py @@ -596,3 +596,12 @@ def walk_directory(path: Path, suffix: Optional[str] = None) -> List[Path]: # It's good to sort these, in case the ordering messes up cache. locs.sort() return locs + + +def _format_number(number: Union[int, float], ndigits: int = 2) -> str: + """Formats a number (float or int) rounding to `ndigits`, without truncating trailing 0s, + as happens with `round(number, ndigits)`""" + if isinstance(number, float): + return f"{number:.{ndigits}f}" + else: + return str(number) diff --git a/spacy/cli/debug_data.py b/spacy/cli/debug_data.py index bd05471b1..963d5b926 100644 --- a/spacy/cli/debug_data.py +++ b/spacy/cli/debug_data.py @@ -9,7 +9,7 @@ import typer import math from ._util import app, Arg, Opt, show_validation_error, parse_config_overrides -from ._util import import_code, debug_cli +from ._util import import_code, debug_cli, _format_number from ..training import Example, remove_bilu_prefix from ..training.initialize import get_sourced_components from ..schemas import ConfigSchemaTraining @@ -989,7 +989,8 @@ def _get_kl_divergence(p: Counter, q: Counter) -> float: def _format_span_row(span_data: List[Dict], labels: List[str]) -> List[Any]: """Compile into one list for easier reporting""" d = { - label: [label] + list(round(d[label], 2) for d in span_data) for label in labels + label: [label] + list(_format_number(d[label]) for d in span_data) + for label in labels } return list(d.values()) @@ -1004,6 +1005,10 @@ def _get_span_characteristics( label: _gmean(l) for label, l in compiled_gold["spans_length"][spans_key].items() } + spans_per_type = { + label: len(spans) + for label, spans in compiled_gold["spans_per_type"][spans_key].items() + } min_lengths = [min(l) for l in compiled_gold["spans_length"][spans_key].values()] max_lengths = [max(l) for l in compiled_gold["spans_length"][spans_key].values()] @@ -1031,6 +1036,7 @@ def _get_span_characteristics( return { "sd": span_distinctiveness, "bd": sb_distinctiveness, + "spans_per_type": spans_per_type, "lengths": span_length, "min_length": min(min_lengths), "max_length": max(max_lengths), @@ -1045,12 +1051,15 @@ def _get_span_characteristics( def _print_span_characteristics(span_characteristics: Dict[str, Any]): """Print all span characteristics into a table""" - headers = ("Span Type", "Length", "SD", "BD") + headers = ("Span Type", "Length", "SD", "BD", "N") + # Wasabi has this at 30 by default, but we might have some long labels + max_col = max(30, max(len(label) for label in span_characteristics["labels"])) # Prepare table data with all span characteristics table_data = [ span_characteristics["lengths"], span_characteristics["sd"], span_characteristics["bd"], + span_characteristics["spans_per_type"], ] table = _format_span_row( span_data=table_data, labels=span_characteristics["labels"] @@ -1061,8 +1070,18 @@ def _print_span_characteristics(span_characteristics: Dict[str, Any]): span_characteristics["avg_sd"], span_characteristics["avg_bd"], ] - footer = ["Wgt. Average"] + [str(round(f, 2)) for f in footer_data] - msg.table(table, footer=footer, header=headers, divider=True) + + footer = ( + ["Wgt. Average"] + ["{:.2f}".format(round(f, 2)) for f in footer_data] + ["-"] + ) + msg.table( + table, + footer=footer, + header=headers, + divider=True, + aligns=["l"] + ["r"] * (len(footer_data) + 1), + max_col=max_col, + ) def _get_spans_length_freq_dist( diff --git a/spacy/cli/package.py b/spacy/cli/package.py index b8c8397b6..324c5d1bb 100644 --- a/spacy/cli/package.py +++ b/spacy/cli/package.py @@ -299,8 +299,8 @@ def get_meta( } nlp = util.load_model_from_path(Path(model_path)) meta.update(nlp.meta) - meta.update(existing_meta) meta["spacy_version"] = util.get_minor_version_range(about.__version__) + meta.update(existing_meta) meta["vectors"] = { "width": nlp.vocab.vectors_length, "vectors": len(nlp.vocab.vectors), diff --git a/spacy/cli/project/dvc.py b/spacy/cli/project/dvc.py index 83dc5efbf..a15353855 100644 --- a/spacy/cli/project/dvc.py +++ b/spacy/cli/project/dvc.py @@ -25,6 +25,7 @@ def project_update_dvc_cli( project_dir: Path = Arg(Path.cwd(), help="Location of project directory. Defaults to current working directory.", exists=True, file_okay=False), workflow: Optional[str] = Arg(None, help=f"Name of workflow defined in {PROJECT_FILE}. Defaults to first workflow if not set."), verbose: bool = Opt(False, "--verbose", "-V", help="Print more info"), + quiet: bool = Opt(False, "--quiet", "-q", help="Print less info"), force: bool = Opt(False, "--force", "-F", help="Force update DVC config"), # fmt: on ): @@ -36,7 +37,7 @@ def project_update_dvc_cli( DOCS: https://spacy.io/api/cli#project-dvc """ - project_update_dvc(project_dir, workflow, verbose=verbose, force=force) + project_update_dvc(project_dir, workflow, verbose=verbose, quiet=quiet, force=force) def project_update_dvc( @@ -44,6 +45,7 @@ def project_update_dvc( workflow: Optional[str] = None, *, verbose: bool = False, + quiet: bool = False, force: bool = False, ) -> None: """Update the auto-generated Data Version Control (DVC) config file. A DVC @@ -54,11 +56,12 @@ def project_update_dvc( workflow (Optional[str]): Optional name of workflow defined in project.yml. If not set, the first workflow will be used. verbose (bool): Print more info. + quiet (bool): Print less info. force (bool): Force update DVC config. """ config = load_project_config(project_dir) updated = update_dvc_config( - project_dir, config, workflow, verbose=verbose, force=force + project_dir, config, workflow, verbose=verbose, quiet=quiet, force=force ) help_msg = "To execute the workflow with DVC, run: dvc repro" if updated: @@ -72,7 +75,7 @@ def update_dvc_config( config: Dict[str, Any], workflow: Optional[str] = None, verbose: bool = False, - silent: bool = False, + quiet: bool = False, force: bool = False, ) -> bool: """Re-run the DVC commands in dry mode and update dvc.yaml file in the @@ -83,7 +86,7 @@ def update_dvc_config( path (Path): The path to the project directory. config (Dict[str, Any]): The loaded project.yml. verbose (bool): Whether to print additional info (via DVC). - silent (bool): Don't output anything (via DVC). + quiet (bool): Don't output anything (via DVC). force (bool): Force update, even if hashes match. RETURNS (bool): Whether the DVC config file was updated. """ @@ -105,6 +108,14 @@ def update_dvc_config( dvc_config_path.unlink() dvc_commands = [] config_commands = {cmd["name"]: cmd for cmd in config.get("commands", [])} + + # some flags that apply to every command + flags = [] + if verbose: + flags.append("--verbose") + if quiet: + flags.append("--quiet") + for name in workflows[workflow]: command = config_commands[name] deps = command.get("deps", []) @@ -118,14 +129,26 @@ def update_dvc_config( deps_cmd = [c for cl in [["-d", p] for p in deps] for c in cl] outputs_cmd = [c for cl in [["-o", p] for p in outputs] for c in cl] outputs_nc_cmd = [c for cl in [["-O", p] for p in outputs_no_cache] for c in cl] - dvc_cmd = ["run", "-n", name, "-w", str(path), "--no-exec"] + + dvc_cmd = ["run", *flags, "-n", name, "-w", str(path), "--no-exec"] if command.get("no_skip"): dvc_cmd.append("--always-changed") full_cmd = [*dvc_cmd, *deps_cmd, *outputs_cmd, *outputs_nc_cmd, *project_cmd] dvc_commands.append(join_command(full_cmd)) + + if not dvc_commands: + # If we don't check for this, then there will be an error when reading the + # config, since DVC wouldn't create it. + msg.fail( + "No usable commands for DVC found. This can happen if none of your " + "commands have dependencies or outputs.", + exits=1, + ) + with working_dir(path): - dvc_flags = {"--verbose": verbose, "--quiet": silent} - run_dvc_commands(dvc_commands, flags=dvc_flags) + for c in dvc_commands: + dvc_command = "dvc " + c + run_command(dvc_command) with dvc_config_path.open("r+", encoding="utf8") as f: content = f.read() f.seek(0, 0) @@ -133,26 +156,6 @@ def update_dvc_config( return True -def run_dvc_commands( - commands: Iterable[str] = SimpleFrozenList(), flags: Dict[str, bool] = {} -) -> None: - """Run a sequence of DVC commands in a subprocess, in order. - - commands (List[str]): The string commands without the leading "dvc". - flags (Dict[str, bool]): Conditional flags to be added to command. Makes it - easier to pass flags like --quiet that depend on a variable or - command-line setting while avoiding lots of nested conditionals. - """ - for c in commands: - command = split_command(c) - dvc_command = ["dvc", *command] - # Add the flags if they are set to True - for flag, is_active in flags.items(): - if is_active: - dvc_command.append(flag) - run_command(dvc_command) - - def check_workflows(workflows: List[str], workflow: Optional[str] = None) -> None: """Validate workflows provided in project.yml and check that a given workflow can be used to generate a DVC config. diff --git a/spacy/cli/project/run.py b/spacy/cli/project/run.py index d42d95465..ebab7471e 100644 --- a/spacy/cli/project/run.py +++ b/spacy/cli/project/run.py @@ -1,5 +1,8 @@ -from typing import Optional, List, Dict, Sequence, Any, Iterable +from typing import Optional, List, Dict, Sequence, Any, Iterable, Tuple +import os.path from pathlib import Path + +import pkg_resources from wasabi import msg from wasabi.util import locale_escape import sys @@ -71,6 +74,12 @@ def project_run( commands = {cmd["name"]: cmd for cmd in config.get("commands", [])} workflows = config.get("workflows", {}) validate_subcommand(list(commands.keys()), list(workflows.keys()), subcommand) + + req_path = project_dir / "requirements.txt" + if config.get("check_requirements", True) and os.path.exists(req_path): + with req_path.open() as requirements_file: + _check_requirements([req.replace("\n", "") for req in requirements_file]) + if subcommand in workflows: msg.info(f"Running workflow '{subcommand}'") for cmd in workflows[subcommand]: @@ -310,3 +319,32 @@ def get_fileinfo(project_dir: Path, paths: List[str]) -> List[Dict[str, Optional md5 = get_checksum(file_path) if file_path.exists() else None data.append({"path": path, "md5": md5}) return data + + +def _check_requirements(requirements: List[str]) -> Tuple[bool, bool]: + """Checks whether requirements are installed and free of version conflicts. + requirements (List[str]): List of requirements. + RETURNS (Tuple[bool, bool]): Whether (1) any packages couldn't be imported, (2) any packages with version conflicts + exist. + """ + + failed_pkgs_msgs: List[str] = [] + conflicting_pkgs_msgs: List[str] = [] + + for req in requirements: + try: + pkg_resources.require(req) + except pkg_resources.DistributionNotFound as dnf: + failed_pkgs_msgs.append(dnf.report()) + except pkg_resources.VersionConflict as vc: + conflicting_pkgs_msgs.append(vc.report()) + + if len(failed_pkgs_msgs) or len(conflicting_pkgs_msgs): + msg.warn( + title="Missing requirements or requirement conflicts detected. Make sure your Python environment is set up " + "correctly and you installed all requirements specified in your project's requirements.txt: " + ) + for pgk_msg in failed_pkgs_msgs + conflicting_pkgs_msgs: + msg.text(pgk_msg) + + return len(failed_pkgs_msgs) > 0, len(conflicting_pkgs_msgs) > 0 diff --git a/spacy/errors.py b/spacy/errors.py index f55b378e9..e0628819d 100644 --- a/spacy/errors.py +++ b/spacy/errors.py @@ -212,6 +212,8 @@ class Warnings(metaclass=ErrorsWithCodes): W121 = ("Attempting to trace non-existent method '{method}' in pipe '{pipe}'") W122 = ("Couldn't trace method '{method}' in pipe '{pipe}'. This can happen if the pipe class " "is a Cython extension type.") + W123 = ("Argument {arg} with value {arg_value} is used instead of {config_value} as specified in the config. Be " + "aware that this might affect other components in your pipeline.") class Errors(metaclass=ErrorsWithCodes): @@ -538,6 +540,8 @@ class Errors(metaclass=ErrorsWithCodes): E199 = ("Unable to merge 0-length span at `doc[{start}:{end}]`.") E200 = ("Can't set {attr} from Span.") E202 = ("Unsupported {name} mode '{mode}'. Supported modes: {modes}.") + E203 = ("If the {name} embedding layer is not updated " + "during training, make sure to include it in 'annotating components'") # New errors added in v3.x E853 = ("Unsupported component factory name '{name}'. The character '.' is " @@ -709,9 +713,9 @@ class Errors(metaclass=ErrorsWithCodes): "`nlp.enable_pipe` instead.") E927 = ("Can't write to frozen list. Maybe you're trying to modify a computed " "property or default function argument?") - E928 = ("A KnowledgeBase can only be serialized to/from from a directory, " + E928 = ("An InMemoryLookupKB can only be serialized to/from from a directory, " "but the provided argument {loc} points to a file.") - E929 = ("Couldn't read KnowledgeBase from {loc}. The path does not seem to exist.") + E929 = ("Couldn't read InMemoryLookupKB from {loc}. The path does not seem to exist.") E930 = ("Received invalid get_examples callback in `{method}`. " "Expected function that returns an iterable of Example objects but " "got: {obj}") @@ -937,10 +941,17 @@ class Errors(metaclass=ErrorsWithCodes): E1040 = ("Doc.from_json requires all tokens to have the same attributes. " "Some tokens do not contain annotation for: {partial_attrs}") E1041 = ("Expected a string, Doc, or bytes as input, but got: {type}") - E1042 = ("Function was called with `{arg1}`={arg1_values} and " - "`{arg2}`={arg2_values} but these arguments are conflicting.") + E1042 = ("`enable={enable}` and `disable={disable}` are inconsistent with each other.\nIf you only passed " + "one of `enable` or `disable`, the other argument is specified in your pipeline's configuration.\nIn that " + "case pass an empty list for the previously not specified argument to avoid this error.") E1043 = ("Expected None or a value in range [{range_start}, {range_end}] for entity linker threshold, but got " "{value}.") + E1044 = ("Expected `candidates_batch_size` to be >= 1, but got: {value}") + E1045 = ("Encountered {parent} subclass without `{parent}.{method}` " + "method in '{name}'. If you want to use this method, make " + "sure it's overwritten on the subclass.") + E1046 = ("{cls_name} is an abstract class and cannot be instantiated. If you are looking for spaCy's default " + "knowledge base, use `InMemoryLookupKB`.") # Deprecated model shortcuts, only used in errors and warnings diff --git a/spacy/kb/__init__.py b/spacy/kb/__init__.py new file mode 100644 index 000000000..1d70a9b34 --- /dev/null +++ b/spacy/kb/__init__.py @@ -0,0 +1,3 @@ +from .kb import KnowledgeBase +from .kb_in_memory import InMemoryLookupKB +from .candidate import Candidate, get_candidates, get_candidates_batch diff --git a/spacy/kb/candidate.pxd b/spacy/kb/candidate.pxd new file mode 100644 index 000000000..942ce9dd0 --- /dev/null +++ b/spacy/kb/candidate.pxd @@ -0,0 +1,12 @@ +from .kb cimport KnowledgeBase +from libcpp.vector cimport vector +from ..typedefs cimport hash_t + +# Object used by the Entity Linker that summarizes one entity-alias candidate combination. +cdef class Candidate: + cdef readonly KnowledgeBase kb + cdef hash_t entity_hash + cdef float entity_freq + cdef vector[float] entity_vector + cdef hash_t alias_hash + cdef float prior_prob diff --git a/spacy/kb/candidate.pyx b/spacy/kb/candidate.pyx new file mode 100644 index 000000000..c89efeb03 --- /dev/null +++ b/spacy/kb/candidate.pyx @@ -0,0 +1,74 @@ +# cython: infer_types=True, profile=True + +from typing import Iterable +from .kb cimport KnowledgeBase +from ..tokens import Span + +cdef class Candidate: + """A `Candidate` object refers to a textual mention (`alias`) that may or may not be resolved + to a specific `entity` from a Knowledge Base. This will be used as input for the entity linking + algorithm which will disambiguate the various candidates to the correct one. + Each candidate (alias, entity) pair is assigned a certain prior probability. + + DOCS: https://spacy.io/api/kb/#candidate-init + """ + + def __init__(self, KnowledgeBase kb, entity_hash, entity_freq, entity_vector, alias_hash, prior_prob): + self.kb = kb + self.entity_hash = entity_hash + self.entity_freq = entity_freq + self.entity_vector = entity_vector + self.alias_hash = alias_hash + self.prior_prob = prior_prob + + @property + def entity(self) -> int: + """RETURNS (uint64): hash of the entity's KB ID/name""" + return self.entity_hash + + @property + def entity_(self) -> str: + """RETURNS (str): ID/name of this entity in the KB""" + return self.kb.vocab.strings[self.entity_hash] + + @property + def alias(self) -> int: + """RETURNS (uint64): hash of the alias""" + return self.alias_hash + + @property + def alias_(self) -> str: + """RETURNS (str): ID of the original alias""" + return self.kb.vocab.strings[self.alias_hash] + + @property + def entity_freq(self) -> float: + return self.entity_freq + + @property + def entity_vector(self) -> Iterable[float]: + return self.entity_vector + + @property + def prior_prob(self) -> float: + return self.prior_prob + + +def get_candidates(kb: KnowledgeBase, mention: Span) -> Iterable[Candidate]: + """ + Return candidate entities for a given mention and fetching appropriate entries from the index. + kb (KnowledgeBase): Knowledge base to query. + mention (Span): Entity mention for which to identify candidates. + RETURNS (Iterable[Candidate]): Identified candidates. + """ + return kb.get_candidates(mention) + + +def get_candidates_batch(kb: KnowledgeBase, mentions: Iterable[Span]) -> Iterable[Iterable[Candidate]]: + """ + Return candidate entities for the given mentions and fetching appropriate entries from the index. + kb (KnowledgeBase): Knowledge base to query. + mention (Iterable[Span]): Entity mentions for which to identify candidates. + RETURNS (Iterable[Iterable[Candidate]]): Identified candidates. + """ + return kb.get_candidates_batch(mentions) diff --git a/spacy/kb/kb.pxd b/spacy/kb/kb.pxd new file mode 100644 index 000000000..1adeef8ae --- /dev/null +++ b/spacy/kb/kb.pxd @@ -0,0 +1,10 @@ +"""Knowledge-base for entity or concept linking.""" + +from cymem.cymem cimport Pool +from libc.stdint cimport int64_t +from ..vocab cimport Vocab + +cdef class KnowledgeBase: + cdef Pool mem + cdef readonly Vocab vocab + cdef readonly int64_t entity_vector_length diff --git a/spacy/kb/kb.pyx b/spacy/kb/kb.pyx new file mode 100644 index 000000000..ce4bc0138 --- /dev/null +++ b/spacy/kb/kb.pyx @@ -0,0 +1,108 @@ +# cython: infer_types=True, profile=True + +from pathlib import Path +from typing import Iterable, Tuple, Union +from cymem.cymem cimport Pool + +from .candidate import Candidate +from ..tokens import Span +from ..util import SimpleFrozenList +from ..errors import Errors + + +cdef class KnowledgeBase: + """A `KnowledgeBase` instance stores unique identifiers for entities and their textual aliases, + to support entity linking of named entities to real-world concepts. + This is an abstract class and requires its operations to be implemented. + + DOCS: https://spacy.io/api/kb + """ + + def __init__(self, vocab: Vocab, entity_vector_length: int): + """Create a KnowledgeBase.""" + # Make sure abstract KB is not instantiated. + if self.__class__ == KnowledgeBase: + raise TypeError( + Errors.E1046.format(cls_name=self.__class__.__name__) + ) + + self.vocab = vocab + self.entity_vector_length = entity_vector_length + self.mem = Pool() + + def get_candidates_batch(self, mentions: Iterable[Span]) -> Iterable[Iterable[Candidate]]: + """ + Return candidate entities for specified texts. Each candidate defines the entity, the original alias, + and the prior probability of that alias resolving to that entity. + If no candidate is found for a given text, an empty list is returned. + mentions (Iterable[Span]): Mentions for which to get candidates. + RETURNS (Iterable[Iterable[Candidate]]): Identified candidates. + """ + return [self.get_candidates(span) for span in mentions] + + def get_candidates(self, mention: Span) -> Iterable[Candidate]: + """ + Return candidate entities for specified text. Each candidate defines the entity, the original alias, + and the prior probability of that alias resolving to that entity. + If the no candidate is found for a given text, an empty list is returned. + mention (Span): Mention for which to get candidates. + RETURNS (Iterable[Candidate]): Identified candidates. + """ + raise NotImplementedError( + Errors.E1045.format(parent="KnowledgeBase", method="get_candidates", name=self.__name__) + ) + + def get_vectors(self, entities: Iterable[str]) -> Iterable[Iterable[float]]: + """ + Return vectors for entities. + entity (str): Entity name/ID. + RETURNS (Iterable[Iterable[float]]): Vectors for specified entities. + """ + return [self.get_vector(entity) for entity in entities] + + def get_vector(self, str entity) -> Iterable[float]: + """ + Return vector for entity. + entity (str): Entity name/ID. + RETURNS (Iterable[float]): Vector for specified entity. + """ + raise NotImplementedError( + Errors.E1045.format(parent="KnowledgeBase", method="get_vector", name=self.__name__) + ) + + def to_bytes(self, **kwargs) -> bytes: + """Serialize the current state to a binary string. + RETURNS (bytes): Current state as binary string. + """ + raise NotImplementedError( + Errors.E1045.format(parent="KnowledgeBase", method="to_bytes", name=self.__name__) + ) + + def from_bytes(self, bytes_data: bytes, *, exclude: Tuple[str] = tuple()): + """Load state from a binary string. + bytes_data (bytes): KB state. + exclude (Tuple[str]): Properties to exclude when restoring KB. + """ + raise NotImplementedError( + Errors.E1045.format(parent="KnowledgeBase", method="from_bytes", name=self.__name__) + ) + + def to_disk(self, path: Union[str, Path], exclude: Iterable[str] = SimpleFrozenList()) -> None: + """ + Write KnowledgeBase content to disk. + path (Union[str, Path]): Target file path. + exclude (Iterable[str]): List of components to exclude. + """ + raise NotImplementedError( + Errors.E1045.format(parent="KnowledgeBase", method="to_disk", name=self.__name__) + ) + + def from_disk(self, path: Union[str, Path], exclude: Iterable[str] = SimpleFrozenList()) -> None: + """ + Load KnowledgeBase content from disk. + path (Union[str, Path]): Target file path. + exclude (Iterable[str]): List of components to exclude. + """ + raise NotImplementedError( + Errors.E1045.format(parent="KnowledgeBase", method="from_disk", name=self.__name__) + ) diff --git a/spacy/kb.pxd b/spacy/kb/kb_in_memory.pxd similarity index 92% rename from spacy/kb.pxd rename to spacy/kb/kb_in_memory.pxd index a823dbe1e..825a6bde9 100644 --- a/spacy/kb.pxd +++ b/spacy/kb/kb_in_memory.pxd @@ -1,14 +1,12 @@ """Knowledge-base for entity or concept linking.""" -from cymem.cymem cimport Pool from preshed.maps cimport PreshMap from libcpp.vector cimport vector from libc.stdint cimport int32_t, int64_t from libc.stdio cimport FILE -from .vocab cimport Vocab -from .typedefs cimport hash_t -from .structs cimport KBEntryC, AliasC - +from ..typedefs cimport hash_t +from ..structs cimport KBEntryC, AliasC +from .kb cimport KnowledgeBase ctypedef vector[KBEntryC] entry_vec ctypedef vector[AliasC] alias_vec @@ -16,21 +14,7 @@ ctypedef vector[float] float_vec ctypedef vector[float_vec] float_matrix -# Object used by the Entity Linker that summarizes one entity-alias candidate combination. -cdef class Candidate: - cdef readonly KnowledgeBase kb - cdef hash_t entity_hash - cdef float entity_freq - cdef vector[float] entity_vector - cdef hash_t alias_hash - cdef float prior_prob - - -cdef class KnowledgeBase: - cdef Pool mem - cdef readonly Vocab vocab - cdef int64_t entity_vector_length - +cdef class InMemoryLookupKB(KnowledgeBase): # This maps 64bit keys (hash of unique entity string) # to 64bit values (position of the _KBEntryC struct in the _entries vector). # The PreshMap is pretty space efficient, as it uses open addressing. So diff --git a/spacy/kb.pyx b/spacy/kb/kb_in_memory.pyx similarity index 90% rename from spacy/kb.pyx rename to spacy/kb/kb_in_memory.pyx index ae1983a8d..485e52c2f 100644 --- a/spacy/kb.pyx +++ b/spacy/kb/kb_in_memory.pyx @@ -1,8 +1,7 @@ # cython: infer_types=True, profile=True -from typing import Iterator, Iterable, Callable, Dict, Any +from typing import Iterable, Callable, Dict, Any, Union import srsly -from cymem.cymem cimport Pool from preshed.maps cimport PreshMap from cpython.exc cimport PyErr_SetFromErrno from libc.stdio cimport fopen, fclose, fread, fwrite, feof, fseek @@ -12,85 +11,28 @@ from libcpp.vector cimport vector from pathlib import Path import warnings -from .typedefs cimport hash_t -from .errors import Errors, Warnings -from . import util -from .util import SimpleFrozenList, ensure_path - -cdef class Candidate: - """A `Candidate` object refers to a textual mention (`alias`) that may or may not be resolved - to a specific `entity` from a Knowledge Base. This will be used as input for the entity linking - algorithm which will disambiguate the various candidates to the correct one. - Each candidate (alias, entity) pair is assigned to a certain prior probability. - - DOCS: https://spacy.io/api/kb/#candidate_init - """ - - def __init__(self, KnowledgeBase kb, entity_hash, entity_freq, entity_vector, alias_hash, prior_prob): - self.kb = kb - self.entity_hash = entity_hash - self.entity_freq = entity_freq - self.entity_vector = entity_vector - self.alias_hash = alias_hash - self.prior_prob = prior_prob - - @property - def entity(self): - """RETURNS (uint64): hash of the entity's KB ID/name""" - return self.entity_hash - - @property - def entity_(self): - """RETURNS (str): ID/name of this entity in the KB""" - return self.kb.vocab.strings[self.entity_hash] - - @property - def alias(self): - """RETURNS (uint64): hash of the alias""" - return self.alias_hash - - @property - def alias_(self): - """RETURNS (str): ID of the original alias""" - return self.kb.vocab.strings[self.alias_hash] - - @property - def entity_freq(self): - return self.entity_freq - - @property - def entity_vector(self): - return self.entity_vector - - @property - def prior_prob(self): - return self.prior_prob +from ..tokens import Span +from ..typedefs cimport hash_t +from ..errors import Errors, Warnings +from .. import util +from ..util import SimpleFrozenList, ensure_path +from ..vocab cimport Vocab +from .kb cimport KnowledgeBase +from .candidate import Candidate as Candidate -def get_candidates(KnowledgeBase kb, span) -> Iterator[Candidate]: - """ - Return candidate entities for a given span by using the text of the span as the alias - and fetching appropriate entries from the index. - This particular function is optimized to work with the built-in KB functionality, - but any other custom candidate generation method can be used in combination with the KB as well. - """ - return kb.get_alias_candidates(span.text) - - -cdef class KnowledgeBase: - """A `KnowledgeBase` instance stores unique identifiers for entities and their textual aliases, +cdef class InMemoryLookupKB(KnowledgeBase): + """An `InMemoryLookupKB` instance stores unique identifiers for entities and their textual aliases, to support entity linking of named entities to real-world concepts. - DOCS: https://spacy.io/api/kb + DOCS: https://spacy.io/api/kb_in_memory """ def __init__(self, Vocab vocab, entity_vector_length): - """Create a KnowledgeBase.""" - self.mem = Pool() - self.entity_vector_length = entity_vector_length + """Create an InMemoryLookupKB.""" + super().__init__(vocab, entity_vector_length) self._entry_index = PreshMap() self._alias_index = PreshMap() - self.vocab = vocab self._create_empty_vectors(dummy_hash=self.vocab.strings[""]) def _initialize_entities(self, int64_t nr_entities): @@ -104,11 +46,6 @@ cdef class KnowledgeBase: self._alias_index = PreshMap(nr_aliases + 1) self._aliases_table = alias_vec(nr_aliases + 1) - @property - def entity_vector_length(self): - """RETURNS (uint64): length of the entity vectors""" - return self.entity_vector_length - def __len__(self): return self.get_size_entities() @@ -286,7 +223,10 @@ cdef class KnowledgeBase: alias_entry.probs = probs self._aliases_table[alias_index] = alias_entry - def get_alias_candidates(self, str alias) -> Iterator[Candidate]: + def get_candidates(self, mention: Span) -> Iterable[Candidate]: + return self.get_alias_candidates(mention.text) # type: ignore + + def get_alias_candidates(self, str alias) -> Iterable[Candidate]: """ Return candidate entities for an alias. Each candidate defines the entity, the original alias, and the prior probability of that alias resolving to that entity. diff --git a/spacy/lang/ca/lemmatizer.py b/spacy/lang/ca/lemmatizer.py index 2fd012912..0f15e6e65 100644 --- a/spacy/lang/ca/lemmatizer.py +++ b/spacy/lang/ca/lemmatizer.py @@ -72,10 +72,10 @@ class CatalanLemmatizer(Lemmatizer): oov_forms.append(form) if not forms: forms.extend(oov_forms) - if not forms and string in lookup_table.keys(): - forms.append(self.lookup_lemmatize(token)[0]) + + # use lookups, and fall back to the token itself if not forms: - forms.append(string) + forms.append(lookup_table.get(string, [string])[0]) forms = list(dict.fromkeys(forms)) self.cache[cache_key] = forms return forms diff --git a/spacy/lang/char_classes.py b/spacy/lang/char_classes.py index 1d204c46c..37c58c85f 100644 --- a/spacy/lang/char_classes.py +++ b/spacy/lang/char_classes.py @@ -280,7 +280,7 @@ _currency = ( _punct = ( r"… …… , : ; \! \? ¿ ؟ ¡ \( \) \[ \] \{ \} < > _ # \* & 。 ？ ！ ， 、 ； ： ～ · । ، ۔ ؛ ٪" ) -_quotes = r'\' " ” “ ` ‘ ´ ’ ‚ , „ » « 「 」 『 』 （ ） 〔 〕 【 】 《 》 〈 〉' +_quotes = r'\' " ” “ ` ‘ ´ ’ ‚ , „ » « 「 」 『 』 （ ） 〔 〕 【 】 《 》 〈 〉 〈 〉 ⟦ ⟧' _hyphens = "- – — -- --- —— ~" # Various symbols like dingbats, but also emoji diff --git a/spacy/lang/fr/lemmatizer.py b/spacy/lang/fr/lemmatizer.py index c6422cf96..a7cbe0bcf 100644 --- a/spacy/lang/fr/lemmatizer.py +++ b/spacy/lang/fr/lemmatizer.py @@ -53,11 +53,16 @@ class FrenchLemmatizer(Lemmatizer): rules = rules_table.get(univ_pos, []) string = string.lower() forms = [] + # first try lookup in table based on upos if string in index: forms.append(string) self.cache[cache_key] = forms return forms + + # then add anything in the exceptions table forms.extend(exceptions.get(string, [])) + + # if nothing found yet, use the rules oov_forms = [] if not forms: for old, new in rules: @@ -69,12 +74,14 @@ class FrenchLemmatizer(Lemmatizer): forms.append(form) else: oov_forms.append(form) + + # if still nothing, add the oov forms from rules if not forms: forms.extend(oov_forms) - if not forms and string in lookup_table.keys(): - forms.append(self.lookup_lemmatize(token)[0]) + + # use lookups, which fall back to the token itself if not forms: - forms.append(string) + forms.append(lookup_table.get(string, [string])[0]) forms = list(dict.fromkeys(forms)) self.cache[cache_key] = forms return forms diff --git a/spacy/lang/grc/__init__.py b/spacy/lang/grc/__init__.py index e83f0c5a5..019b3802e 100644 --- a/spacy/lang/grc/__init__.py +++ b/spacy/lang/grc/__init__.py @@ -1,11 +1,15 @@ from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS from .stop_words import STOP_WORDS from .lex_attrs import LEX_ATTRS +from .punctuation import TOKENIZER_PREFIXES, TOKENIZER_SUFFIXES, TOKENIZER_INFIXES from ...language import Language, BaseDefaults class AncientGreekDefaults(BaseDefaults): tokenizer_exceptions = TOKENIZER_EXCEPTIONS + prefixes = TOKENIZER_PREFIXES + suffixes = TOKENIZER_SUFFIXES + infixes = TOKENIZER_INFIXES lex_attr_getters = LEX_ATTRS stop_words = STOP_WORDS diff --git a/spacy/lang/grc/punctuation.py b/spacy/lang/grc/punctuation.py new file mode 100644 index 000000000..8f3589e9a --- /dev/null +++ b/spacy/lang/grc/punctuation.py @@ -0,0 +1,46 @@ +from ..char_classes import LIST_PUNCT, LIST_ELLIPSES, LIST_QUOTES, LIST_CURRENCY +from ..char_classes import LIST_ICONS, ALPHA_LOWER, ALPHA_UPPER, ALPHA, HYPHENS +from ..char_classes import CONCAT_QUOTES + +_prefixes = ( + [ + "†", + "⸏", + ] + + LIST_PUNCT + + LIST_ELLIPSES + + LIST_QUOTES + + LIST_CURRENCY + + LIST_ICONS +) + +_suffixes = ( + LIST_PUNCT + + LIST_ELLIPSES + + LIST_QUOTES + + LIST_ICONS + + [ + "†", + "⸎", + r"(?<=[\u1F00-\u1FFF\u0370-\u03FF])[\-\.⸏]", + ] +) + +_infixes = ( + LIST_ELLIPSES + + LIST_ICONS + + [ + r"(?<=[0-9])[+\-\*^](?=[0-9-])", + r"(?<=[{al}{q}])\.(?=[{au}{q}])".format( + al=ALPHA_LOWER, au=ALPHA_UPPER, q=CONCAT_QUOTES + ), + r"(?<=[{a}]),(?=[{a}])".format(a=ALPHA), + r"(?<=[{a}0-9])(?:{h})(?=[{a}])".format(a=ALPHA, h=HYPHENS), + r"(?<=[{a}0-9])[:<>=/](?=[{a}])".format(a=ALPHA), + r"(?<=[\u1F00-\u1FFF\u0370-\u03FF])—", + ] +) + +TOKENIZER_PREFIXES = _prefixes +TOKENIZER_SUFFIXES = _suffixes +TOKENIZER_INFIXES = _infixes diff --git a/spacy/lang/ru/__init__.py b/spacy/lang/ru/__init__.py index c118c26ff..7d17628c4 100644 --- a/spacy/lang/ru/__init__.py +++ b/spacy/lang/ru/__init__.py @@ -28,7 +28,7 @@ class Russian(Language): assigns=["token.lemma"], default_config={ "model": None, - "mode": "pymorphy2", + "mode": "pymorphy3", "overwrite": False, "scorer": {"@scorers": "spacy.lemmatizer_scorer.v1"}, }, diff --git a/spacy/lang/ru/lemmatizer.py b/spacy/lang/ru/lemmatizer.py index 85180b1e4..c37a3a91a 100644 --- a/spacy/lang/ru/lemmatizer.py +++ b/spacy/lang/ru/lemmatizer.py @@ -19,11 +19,11 @@ class RussianLemmatizer(Lemmatizer): model: Optional[Model], name: str = "lemmatizer", *, - mode: str = "pymorphy2", + mode: str = "pymorphy3", overwrite: bool = False, scorer: Optional[Callable] = lemmatizer_score, ) -> None: - if mode == "pymorphy2": + if mode in {"pymorphy2", "pymorphy2_lookup"}: try: from pymorphy2 import MorphAnalyzer except ImportError: @@ -33,6 +33,16 @@ class RussianLemmatizer(Lemmatizer): ) from None if getattr(self, "_morph", None) is None: self._morph = MorphAnalyzer() + elif mode == "pymorphy3": + try: + from pymorphy3 import MorphAnalyzer + except ImportError: + raise ImportError( + "The Russian lemmatizer mode 'pymorphy3' requires the " + "pymorphy3 library. Install it with: pip install pymorphy3" + ) from None + if getattr(self, "_morph", None) is None: + self._morph = MorphAnalyzer() super().__init__( vocab, model, name, mode=mode, overwrite=overwrite, scorer=scorer ) @@ -104,6 +114,9 @@ class RussianLemmatizer(Lemmatizer): return [analyses[0].normal_form] return [string] + def pymorphy3_lemmatize(self, token: Token) -> List[str]: + return self.pymorphy2_lemmatize(token) + def oc2ud(oc_tag: str) -> Tuple[str, Dict[str, str]]: gram_map = { diff --git a/spacy/lang/sl/__init__.py b/spacy/lang/sl/__init__.py index 9ddd676bf..0070e9fa1 100644 --- a/spacy/lang/sl/__init__.py +++ b/spacy/lang/sl/__init__.py @@ -1,9 +1,17 @@ +from .lex_attrs import LEX_ATTRS +from .punctuation import TOKENIZER_INFIXES, TOKENIZER_SUFFIXES, TOKENIZER_PREFIXES from .stop_words import STOP_WORDS +from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS from ...language import Language, BaseDefaults class SlovenianDefaults(BaseDefaults): stop_words = STOP_WORDS + tokenizer_exceptions = TOKENIZER_EXCEPTIONS + prefixes = TOKENIZER_PREFIXES + infixes = TOKENIZER_INFIXES + suffixes = TOKENIZER_SUFFIXES + lex_attr_getters = LEX_ATTRS class Slovenian(Language): diff --git a/spacy/lang/sl/lex_attrs.py b/spacy/lang/sl/lex_attrs.py new file mode 100644 index 000000000..958152e37 --- /dev/null +++ b/spacy/lang/sl/lex_attrs.py @@ -0,0 +1,145 @@ +from ...attrs import LIKE_NUM +from ...attrs import IS_CURRENCY +import unicodedata + + +_num_words = set( + """ + nula ničla nič ena dva tri štiri pet šest sedem osem + devet deset enajst dvanajst trinajst štirinajst petnajst + šestnajst sedemnajst osemnajst devetnajst dvajset trideset štirideset + petdeset šestdest sedemdeset osemdeset devedeset sto tisoč + milijon bilijon trilijon kvadrilijon nešteto + + en eden enega enemu ennem enim enih enima enimi ene eni eno + dveh dvema dvem dvoje trije treh trem tremi troje štirje štirih štirim štirimi + petih petim petimi šestih šestim šestimi sedmih sedmim sedmimi osmih osmim osmimi + devetih devetim devetimi desetih desetim desetimi enajstih enajstim enajstimi + dvanajstih dvanajstim dvanajstimi trinajstih trinajstim trinajstimi + šestnajstih šestnajstim šestnajstimi petnajstih petnajstim petnajstimi + sedemnajstih sedemnajstim sedemnajstimi osemnajstih osemnajstim osemnajstimi + devetnajstih devetnajstim devetnajstimi dvajsetih dvajsetim dvajsetimi + """.split() +) + +_ordinal_words = set( + """ + prvi drugi tretji četrti peti šesti sedmi osmi + deveti deseti enajsti dvanajsti trinajsti štirinajsti + petnajsti šestnajsti sedemnajsti osemnajsti devetnajsti + dvajseti trideseti štirideseti petdeseti šestdeseti sedemdeseti + osemdeseti devetdeseti stoti tisoči milijonti bilijonti + trilijonti kvadrilijonti nešteti + + prva druga tretja četrta peta šesta sedma osma + deveta deseta enajsta dvanajsta trinajsta štirnajsta + petnajsta šestnajsta sedemnajsta osemnajsta devetnajsta + dvajseta trideseta štirideseta petdeseta šestdeseta sedemdeseta + osemdeseta devetdeseta stota tisoča milijonta bilijonta + trilijonta kvadrilijonta nešteta + + prvo drugo tretje četrto peto šestro sedmo osmo + deveto deseto enajsto dvanajsto trinajsto štirnajsto + petnajsto šestnajsto sedemnajsto osemnajsto devetnajsto + dvajseto trideseto štirideseto petdeseto šestdeseto sedemdeseto + osemdeseto devetdeseto stoto tisočo milijonto bilijonto + trilijonto kvadrilijonto nešteto + + prvega drugega tretjega četrtega petega šestega sedmega osmega + devega desetega enajstega dvanajstega trinajstega štirnajstega + petnajstega šestnajstega sedemnajstega osemnajstega devetnajstega + dvajsetega tridesetega štiridesetega petdesetega šestdesetega sedemdesetega + osemdesetega devetdesetega stotega tisočega milijontega bilijontega + trilijontega kvadrilijontega neštetega + + prvemu drugemu tretjemu četrtemu petemu šestemu sedmemu osmemu devetemu desetemu + enajstemu dvanajstemu trinajstemu štirnajstemu petnajstemu šestnajstemu sedemnajstemu + osemnajstemu devetnajstemu dvajsetemu tridesetemu štiridesetemu petdesetemu šestdesetemu + sedemdesetemu osemdesetemu devetdesetemu stotemu tisočemu milijontemu bilijontemu + trilijontemu kvadrilijontemu neštetemu + + prvem drugem tretjem četrtem petem šestem sedmem osmem devetem desetem + enajstem dvanajstem trinajstem štirnajstem petnajstem šestnajstem sedemnajstem + osemnajstem devetnajstem dvajsetem tridesetem štiridesetem petdesetem šestdesetem + sedemdesetem osemdesetem devetdesetem stotem tisočem milijontem bilijontem + trilijontem kvadrilijontem neštetem + + prvim drugim tretjim četrtim petim šestim sedtim osmim devetim desetim + enajstim dvanajstim trinajstim štirnajstim petnajstim šestnajstim sedemnajstim + osemnajstim devetnajstim dvajsetim tridesetim štiridesetim petdesetim šestdesetim + sedemdesetim osemdesetim devetdesetim stotim tisočim milijontim bilijontim + trilijontim kvadrilijontim neštetim + + prvih drugih tretjih četrthih petih šestih sedmih osmih deveth desetih + enajstih dvanajstih trinajstih štirnajstih petnajstih šestnajstih sedemnajstih + osemnajstih devetnajstih dvajsetih tridesetih štiridesetih petdesetih šestdesetih + sedemdesetih osemdesetih devetdesetih stotih tisočih milijontih bilijontih + trilijontih kvadrilijontih nešteth + + prvima drugima tretjima četrtima petima šestima sedmima osmima devetima desetima + enajstima dvanajstima trinajstima štirnajstima petnajstima šestnajstima sedemnajstima + osemnajstima devetnajstima dvajsetima tridesetima štiridesetima petdesetima šestdesetima + sedemdesetima osemdesetima devetdesetima stotima tisočima milijontima bilijontima + trilijontima kvadrilijontima neštetima + + prve druge četrte pete šeste sedme osme devete desete + enajste dvanajste trinajste štirnajste petnajste šestnajste sedemnajste + osemnajste devetnajste dvajsete tridesete štiridesete petdesete šestdesete + sedemdesete osemdesete devetdesete stote tisoče milijonte bilijonte + trilijonte kvadrilijonte neštete + + prvimi drugimi tretjimi četrtimi petimi šestimi sedtimi osmimi devetimi desetimi + enajstimi dvanajstimi trinajstimi štirnajstimi petnajstimi šestnajstimi sedemnajstimi + osemnajstimi devetnajstimi dvajsetimi tridesetimi štiridesetimi petdesetimi šestdesetimi + sedemdesetimi osemdesetimi devetdesetimi stotimi tisočimi milijontimi bilijontimi + trilijontimi kvadrilijontimi neštetimi + """.split() +) + +_currency_words = set( + """ + evro evra evru evrom evrov evroma evrih evrom evre evri evr eur + cent centa centu cenom centov centoma centih centom cente centi + dolar dolarja dolarji dolarju dolarjem dolarjev dolarjema dolarjih dolarje usd + tolar tolarja tolarji tolarju tolarjem tolarjev tolarjema tolarjih tolarje tol + dinar dinarja dinarji dinarju dinarjem dinarjev dinarjema dinarjih dinarje din + funt funta funti funtu funtom funtov funtoma funtih funte gpb + forint forinta forinti forintu forintom forintov forintoma forintih forinte + zlot zlota zloti zlotu zlotom zlotov zlotoma zlotih zlote + rupij rupija rupiji rupiju rupijem rupijev rupijema rupijih rupije + jen jena jeni jenu jenom jenov jenoma jenih jene + kuna kuni kune kuno kun kunama kunah kunam kunami + marka marki marke markama markah markami + """.split() +) + + +def like_num(text): + if text.startswith(("+", "-", "±", "~")): + text = text[1:] + text = text.replace(",", "").replace(".", "") + if text.isdigit(): + return True + if text.count("/") == 1: + num, denom = text.split("/") + if num.isdigit() and denom.isdigit(): + return True + text_lower = text.lower() + if text_lower in _num_words: + return True + if text_lower in _ordinal_words: + return True + return False + + +def is_currency(text): + text_lower = text.lower() + if text in _currency_words: + return True + for char in text: + if unicodedata.category(char) != "Sc": + return False + return True + + +LEX_ATTRS = {LIKE_NUM: like_num, IS_CURRENCY: is_currency} diff --git a/spacy/lang/sl/punctuation.py b/spacy/lang/sl/punctuation.py new file mode 100644 index 000000000..b6ca1830e --- /dev/null +++ b/spacy/lang/sl/punctuation.py @@ -0,0 +1,84 @@ +from ..char_classes import ( + LIST_ELLIPSES, + LIST_ICONS, + HYPHENS, + LIST_PUNCT, + LIST_QUOTES, + CURRENCY, + UNITS, + PUNCT, + LIST_CURRENCY, + CONCAT_QUOTES, +) +from ..char_classes import CONCAT_QUOTES, ALPHA_LOWER, ALPHA_UPPER, ALPHA +from ..char_classes import merge_chars +from ..punctuation import TOKENIZER_PREFIXES as BASE_TOKENIZER_PREFIXES + + +INCLUDE_SPECIAL = ["\\+", "\\/", "\\•", "\\¯", "\\=", "\\×"] + HYPHENS.split("|") + +_prefixes = INCLUDE_SPECIAL + BASE_TOKENIZER_PREFIXES + +_suffixes = ( + INCLUDE_SPECIAL + + LIST_PUNCT + + LIST_ELLIPSES + + LIST_QUOTES + + LIST_ICONS + + [ + r"(?<=°[FfCcKk])\.", + r"(?<=[0-9])(?:{c})".format(c=CURRENCY), + r"(?<=[0-9])(?:{u})".format(u=UNITS), + r"(?<=[{al}{e}{p}(?:{q})])\.".format( + al=ALPHA_LOWER, e=r"%²\-\+", q=CONCAT_QUOTES, p=PUNCT + ), + r"(?<=[{au}][{au}])\.".format(au=ALPHA_UPPER), + # split initials like J.K. Rowling + r"(?<=[A-Z]\.)(?:[A-Z].)", + ] +) + +# a list of all suffixes following a hyphen that are shouldn't split (eg. BTC-jev) +# source: Obeliks tokenizer - https://github.com/clarinsi/obeliks/blob/master/obeliks/res/TokRulesPart1.txt +CONCAT_QUOTES = CONCAT_QUOTES.replace("'", "") +HYPHENS_PERMITTED = ( + "((a)|(evemu)|(evskega)|(i)|(jevega)|(jevska)|(jevskimi)|(jinemu)|(oma)|(ovim)|" + "(ovski)|(e)|(evi)|(evskem)|(ih)|(jevem)|(jevske)|(jevsko)|(jini)|(ov)|(ovima)|" + "(ovskih)|(em)|(evih)|(evskemu)|(ja)|(jevemu)|(jevskega)|(ji)|(jinih)|(ova)|" + "(ovimi)|(ovskim)|(ema)|(evim)|(evski)|(je)|(jevi)|(jevskem)|(jih)|(jinim)|" + "(ove)|(ovo)|(ovskima)|(ev)|(evima)|(evskih)|(jem)|(jevih)|(jevskemu)|(jin)|" + "(jinima)|(ovega)|(ovska)|(ovskimi)|(eva)|(evimi)|(evskim)|(jema)|(jevim)|" + "(jevski)|(jina)|(jinimi)|(ovem)|(ovske)|(ovsko)|(eve)|(evo)|(evskima)|(jev)|" + "(jevima)|(jevskih)|(jine)|(jino)|(ovemu)|(ovskega)|(u)|(evega)|(evska)|" + "(evskimi)|(jeva)|(jevimi)|(jevskim)|(jinega)|(ju)|(ovi)|(ovskem)|(evem)|" + "(evske)|(evsko)|(jeve)|(jevo)|(jevskima)|(jinem)|(om)|(ovih)|(ovskemu)|" + "(ovec)|(ovca)|(ovcu)|(ovcem)|(ovcev)|(ovcema)|(ovcih)|(ovci)|(ovce)|(ovcimi)|" + "(evec)|(evca)|(evcu)|(evcem)|(evcev)|(evcema)|(evcih)|(evci)|(evce)|(evcimi)|" + "(jevec)|(jevca)|(jevcu)|(jevcem)|(jevcev)|(jevcema)|(jevcih)|(jevci)|(jevce)|" + "(jevcimi)|(ovka)|(ovke)|(ovki)|(ovko)|(ovk)|(ovkama)|(ovkah)|(ovkam)|(ovkami)|" + "(evka)|(evke)|(evki)|(evko)|(evk)|(evkama)|(evkah)|(evkam)|(evkami)|(jevka)|" + "(jevke)|(jevki)|(jevko)|(jevk)|(jevkama)|(jevkah)|(jevkam)|(jevkami)|(timi)|" + "(im)|(ima)|(a)|(imi)|(e)|(o)|(ega)|(ti)|(em)|(tih)|(emu)|(tim)|(i)|(tima)|" + "(ih)|(ta)|(te)|(to)|(tega)|(tem)|(temu))" +) + +_infixes = ( + LIST_ELLIPSES + + LIST_ICONS + + [ + r"(?<=[0-9])[+\-\*^](?=[0-9-])", + r"(?<=[{al}{q}])\.(?=[{au}{q}])".format( + al=ALPHA_LOWER, au=ALPHA_UPPER, q=CONCAT_QUOTES + ), + r"(?<=[{a}]),(?=[{a}])".format(a=ALPHA), + r"(?<=[{a}0-9])(?:{h})(?!{hp}$)(?=[{a}])".format( + a=ALPHA, h=HYPHENS, hp=HYPHENS_PERMITTED + ), + r"(?<=[{a}0-9])[:<>=/](?=[{a}])".format(a=ALPHA), + ] +) + + +TOKENIZER_PREFIXES = _prefixes +TOKENIZER_SUFFIXES = _suffixes +TOKENIZER_INFIXES = _infixes diff --git a/spacy/lang/sl/stop_words.py b/spacy/lang/sl/stop_words.py index c9004ed5d..8491efcb5 100644 --- a/spacy/lang/sl/stop_words.py +++ b/spacy/lang/sl/stop_words.py @@ -1,326 +1,84 @@ # Source: https://github.com/stopwords-iso/stopwords-sl -# Removed various words that are not normally considered stop words, such as months. STOP_WORDS = set( """ -a -ali -b -bi -bil -bila -bile -bili -bilo -biti -blizu -bo -bodo -bolj -bom -bomo -boste -bova -boš -brez -c -cel -cela -celi -celo -d -da -daleč -dan -danes -do -dober -dobra -dobri -dobro -dokler -dol -dovolj -e -eden -en -ena -ene -eni -enkrat -eno -etc. +a ali + +b bi bil bila bile bili bilo biti blizu bo bodo bojo bolj bom bomo +boste bova boš brez + +c cel cela celi celo + +č če često četrta četrtek četrti četrto čez čigav + +d da daleč dan danes datum deset deseta deseti deseto devet +deveta deveti deveto do dober dobra dobri dobro dokler dol dolg +dolga dolgi dovolj drug druga drugi drugo dva dve + +e eden en ena ene eni enkrat eno etc. + f -g -g. -ga -ga. -gor -gospa -gospod -h -halo -i -idr. -ii -iii -in -iv -ix -iz -j -jaz -je -ji -jih -jim -jo -k -kadarkoli -kaj -kajti -kako -kakor -kamor -kamorkoli -kar -karkoli -katerikoli -kdaj -kdo -kdorkoli -ker -ki -kje -kjer -kjerkoli -ko -koderkoli -koga -komu -kot -l -le -lep -lepa -lepe -lepi -lepo -m -manj -me -med -medtem -mene -mi -midva -midve -mnogo -moj -moja -moje -mora -morajo -moram -moramo -morate -moraš -morem -mu -n -na -nad -naj -najina -najino -najmanj -naju -največ -nam -nas -nato -nazaj -naš -naša -naše -ne -nedavno -nek -neka -nekaj -nekatere -nekateri -nekatero -nekdo -neke -nekega -neki -nekje -neko -nekoga -nekoč -ni -nikamor -nikdar -nikjer -nikoli -nič -nje -njega -njegov -njegova -njegovo -njej -njemu -njen -njena -njeno -nji -njih -njihov -njihova -njihovo -njiju -njim -njo -njun -njuna -njuno -no -nocoj -npr. -o -ob -oba -obe -oboje -od -okoli -on -onadva -one -oni -onidve -oz. -p -pa -po -pod -pogosto -poleg -ponavadi -ponovno -potem -povsod -prbl. -precej -pred -prej -preko -pri -pribl. -približno -proti -r -redko -res -s -saj -sam -sama -same -sami -samo -se -sebe -sebi -sedaj -sem -seveda -si -sicer -skoraj -skozi -smo -so -spet -sta -ste -sva -t -ta -tak -taka -take -taki -tako -takoj -tam -te -tebe -tebi -tega -ti -tista -tiste -tisti -tisto -tj. -tja -to -toda -tu -tudi -tukaj -tvoj -tvoja -tvoje + +g g. ga ga. gor gospa gospod + +h halo + +i idr. ii iii in iv ix iz + +j jaz je ji jih jim jo jutri + +k kadarkoli kaj kajti kako kakor kamor kamorkoli kar karkoli +katerikoli kdaj kdo kdorkoli ker ki kje kjer kjerkoli +ko koder koderkoli koga komu kot kratek kratka kratke kratki + +l lahka lahke lahki lahko le lep lepa lepe lepi lepo leto + +m majhen majhna majhni malce malo manj me med medtem mene +mesec mi midva midve mnogo moj moja moje mora morajo moram +moramo morate moraš morem mu + +n na nad naj najina najino najmanj naju največ nam narobe +nas nato nazaj naš naša naše ne nedavno nedelja nek neka +nekaj nekatere nekateri nekatero nekdo neke nekega neki +nekje neko nekoga nekoč ni nikamor nikdar nikjer nikoli +nič nje njega njegov njegova njegovo njej njemu njen +njena njeno nji njih njihov njihova njihovo njiju njim +njo njun njuna njuno no nocoj npr. + +o ob oba obe oboje od odprt odprta odprti okoli on +onadva one oni onidve osem osma osmi osmo oz. + +p pa pet peta petek peti peto po pod pogosto poleg poln +polna polni polno ponavadi ponedeljek ponovno potem +povsod pozdravljen pozdravljeni prav prava prave pravi +pravo prazen prazna prazno prbl. precej pred prej preko +pri pribl. približno primer pripravljen pripravljena +pripravljeni proti prva prvi prvo + +r ravno redko res reč + +s saj sam sama same sami samo se sebe sebi sedaj sedem +sedma sedmi sedmo sem seveda si sicer skoraj skozi slab sm +so sobota spet sreda srednja srednji sta ste stran stvar sva + +š šest šesta šesti šesto štiri + +t ta tak taka take taki tako takoj tam te tebe tebi tega +težak težka težki težko ti tista tiste tisti tisto tj. +tja to toda torek tretja tretje tretji tri tu tudi tukaj +tvoj tvoja tvoje + u -v -vaju -vam -vas -vaš -vaša -vaše -ve -vedno -vendar -ves -več -vi -vidva -vii -viii -vsa -vsaj -vsak -vsaka -vsakdo -vsake -vsaki -vsakomur -vse -vsega -vsi -vso -včasih -x -z -za -zadaj -zadnji -zakaj -zdaj -zelo -zunaj -č -če -često -čez -čigav -š -ž -že + +v vaju vam vas vaš vaša vaše ve vedno velik velika veliki +veliko vendar ves več vi vidva vii viii visok visoka visoke +visoki vsa vsaj vsak vsaka vsakdo vsake vsaki vsakomur vse +vsega vsi vso včasih včeraj + +x + +z za zadaj zadnji zakaj zaprta zaprti zaprto zdaj zelo zunaj + +ž že """.split() ) diff --git a/spacy/lang/sl/tokenizer_exceptions.py b/spacy/lang/sl/tokenizer_exceptions.py new file mode 100644 index 000000000..3d4109228 --- /dev/null +++ b/spacy/lang/sl/tokenizer_exceptions.py @@ -0,0 +1,272 @@ +from typing import Dict, List +from ..tokenizer_exceptions import BASE_EXCEPTIONS +from ...symbols import ORTH, NORM +from ...util import update_exc + +_exc: Dict[str, List[Dict]] = {} + +_other_exc = { + "t.i.": [{ORTH: "t.", NORM: "tako"}, {ORTH: "i.", NORM: "imenovano"}], + "t.j.": [{ORTH: "t.", NORM: "to"}, {ORTH: "j.", NORM: "je"}], + "T.j.": [{ORTH: "T.", NORM: "to"}, {ORTH: "j.", NORM: "je"}], + "d.o.o.": [ + {ORTH: "d.", NORM: "družba"}, + {ORTH: "o.", NORM: "omejeno"}, + {ORTH: "o.", NORM: "odgovornostjo"}, + ], + "D.O.O.": [ + {ORTH: "D.", NORM: "družba"}, + {ORTH: "O.", NORM: "omejeno"}, + {ORTH: "O.", NORM: "odgovornostjo"}, + ], + "d.n.o.": [ + {ORTH: "d.", NORM: "družba"}, + {ORTH: "n.", NORM: "neomejeno"}, + {ORTH: "o.", NORM: "odgovornostjo"}, + ], + "D.N.O.": [ + {ORTH: "D.", NORM: "družba"}, + {ORTH: "N.", NORM: "neomejeno"}, + {ORTH: "O.", NORM: "odgovornostjo"}, + ], + "d.d.": [{ORTH: "d.", NORM: "delniška"}, {ORTH: "d.", NORM: "družba"}], + "D.D.": [{ORTH: "D.", NORM: "delniška"}, {ORTH: "D.", NORM: "družba"}], + "s.p.": [{ORTH: "s.", NORM: "samostojni"}, {ORTH: "p.", NORM: "podjetnik"}], + "S.P.": [{ORTH: "S.", NORM: "samostojni"}, {ORTH: "P.", NORM: "podjetnik"}], + "l.r.": [{ORTH: "l.", NORM: "lastno"}, {ORTH: "r.", NORM: "ročno"}], + "le-te": [{ORTH: "le"}, {ORTH: "-"}, {ORTH: "te"}], + "Le-te": [{ORTH: "Le"}, {ORTH: "-"}, {ORTH: "te"}], + "le-ti": [{ORTH: "le"}, {ORTH: "-"}, {ORTH: "ti"}], + "Le-ti": [{ORTH: "Le"}, {ORTH: "-"}, {ORTH: "ti"}], + "le-to": [{ORTH: "le"}, {ORTH: "-"}, {ORTH: "to"}], + "Le-to": [{ORTH: "Le"}, {ORTH: "-"}, {ORTH: "to"}], + "le-ta": [{ORTH: "le"}, {ORTH: "-"}, {ORTH: "ta"}], + "Le-ta": [{ORTH: "Le"}, {ORTH: "-"}, {ORTH: "ta"}], + "le-tega": [{ORTH: "le"}, {ORTH: "-"}, {ORTH: "tega"}], + "Le-tega": [{ORTH: "Le"}, {ORTH: "-"}, {ORTH: "tega"}], +} + +_exc.update(_other_exc) + + +for exc_data in [ + {ORTH: "adm.", NORM: "administracija"}, + {ORTH: "aer.", NORM: "aeronavtika"}, + {ORTH: "agr.", NORM: "agronomija"}, + {ORTH: "amer.", NORM: "ameriško"}, + {ORTH: "anat.", NORM: "anatomija"}, + {ORTH: "angl.", NORM: "angleški"}, + {ORTH: "ant.", NORM: "antonim"}, + {ORTH: "antr.", NORM: "antropologija"}, + {ORTH: "apr.", NORM: "april"}, + {ORTH: "arab.", NORM: "arabsko"}, + {ORTH: "arheol.", NORM: "arheologija"}, + {ORTH: "arhit.", NORM: "arhitektura"}, + {ORTH: "avg.", NORM: "avgust"}, + {ORTH: "avstr.", NORM: "avstrijsko"}, + {ORTH: "avt.", NORM: "avtomobilizem"}, + {ORTH: "bibl.", NORM: "biblijsko"}, + {ORTH: "biokem.", NORM: "biokemija"}, + {ORTH: "biol.", NORM: "biologija"}, + {ORTH: "bolg.", NORM: "bolgarski"}, + {ORTH: "bot.", NORM: "botanika"}, + {ORTH: "cit.", NORM: "citat"}, + {ORTH: "daj.", NORM: "dajalnik"}, + {ORTH: "del.", NORM: "deležnik"}, + {ORTH: "ed.", NORM: "ednina"}, + {ORTH: "etn.", NORM: "etnografija"}, + {ORTH: "farm.", NORM: "farmacija"}, + {ORTH: "filat.", NORM: "filatelija"}, + {ORTH: "filoz.", NORM: "filozofija"}, + {ORTH: "fin.", NORM: "finančništvo"}, + {ORTH: "fiz.", NORM: "fizika"}, + {ORTH: "fot.", NORM: "fotografija"}, + {ORTH: "fr.", NORM: "francoski"}, + {ORTH: "friz.", NORM: "frizerstvo"}, + {ORTH: "gastr.", NORM: "gastronomija"}, + {ORTH: "geogr.", NORM: "geografija"}, + {ORTH: "geol.", NORM: "geologija"}, + {ORTH: "geom.", NORM: "geometrija"}, + {ORTH: "germ.", NORM: "germanski"}, + {ORTH: "gl.", NORM: "glej"}, + {ORTH: "glag.", NORM: "glagolski"}, + {ORTH: "glasb.", NORM: "glasba"}, + {ORTH: "gled.", NORM: "gledališče"}, + {ORTH: "gost.", NORM: "gostinstvo"}, + {ORTH: "gozd.", NORM: "gozdarstvo"}, + {ORTH: "gr.", NORM: "grški"}, + {ORTH: "grad.", NORM: "gradbeništvo"}, + {ORTH: "hebr.", NORM: "hebrejsko"}, + {ORTH: "hrv.", NORM: "hrvaško"}, + {ORTH: "ide.", NORM: "indoevropsko"}, + {ORTH: "igr.", NORM: "igre"}, + {ORTH: "im.", NORM: "imenovalnik"}, + {ORTH: "iron.", NORM: "ironično"}, + {ORTH: "it.", NORM: "italijanski"}, + {ORTH: "itd.", NORM: "in tako dalje"}, + {ORTH: "itn.", NORM: "in tako naprej"}, + {ORTH: "ipd.", NORM: "in podobno"}, + {ORTH: "jap.", NORM: "japonsko"}, + {ORTH: "jul.", NORM: "julij"}, + {ORTH: "jun.", NORM: "junij"}, + {ORTH: "kit.", NORM: "kitajsko"}, + {ORTH: "knj.", NORM: "knjižno"}, + {ORTH: "knjiž.", NORM: "knjižno"}, + {ORTH: "kor.", NORM: "koreografija"}, + {ORTH: "lat.", NORM: "latinski"}, + {ORTH: "les.", NORM: "lesna stroka"}, + {ORTH: "lingv.", NORM: "lingvistika"}, + {ORTH: "lit.", NORM: "literarni"}, + {ORTH: "ljubk.", NORM: "ljubkovalno"}, + {ORTH: "lov.", NORM: "lovstvo"}, + {ORTH: "m.", NORM: "moški"}, + {ORTH: "mak.", NORM: "makedonski"}, + {ORTH: "mar.", NORM: "marec"}, + {ORTH: "mat.", NORM: "matematika"}, + {ORTH: "med.", NORM: "medicina"}, + {ORTH: "meh.", NORM: "mehiško"}, + {ORTH: "mest.", NORM: "mestnik"}, + {ORTH: "mdr.", NORM: "med drugim"}, + {ORTH: "min.", NORM: "mineralogija"}, + {ORTH: "mitol.", NORM: "mitologija"}, + {ORTH: "mn.", NORM: "množina"}, + {ORTH: "mont.", NORM: "montanistika"}, + {ORTH: "muz.", NORM: "muzikologija"}, + {ORTH: "nam.", NORM: "namenilnik"}, + {ORTH: "nar.", NORM: "narečno"}, + {ORTH: "nav.", NORM: "navadno"}, + {ORTH: "nedol.", NORM: "nedoločnik"}, + {ORTH: "nedov.", NORM: "nedovršni"}, + {ORTH: "neprav.", NORM: "nepravilno"}, + {ORTH: "nepreh.", NORM: "neprehodno"}, + {ORTH: "neskl.", NORM: "nesklonljiv(o)"}, + {ORTH: "nestrok.", NORM: "nestrokovno"}, + {ORTH: "num.", NORM: "numizmatika"}, + {ORTH: "npr.", NORM: "na primer"}, + {ORTH: "obrt.", NORM: "obrtništvo"}, + {ORTH: "okt.", NORM: "oktober"}, + {ORTH: "or.", NORM: "orodnik"}, + {ORTH: "os.", NORM: "oseba"}, + {ORTH: "otr.", NORM: "otroško"}, + {ORTH: "oz.", NORM: "oziroma"}, + {ORTH: "pal.", NORM: "paleontologija"}, + {ORTH: "papir.", NORM: "papirništvo"}, + {ORTH: "ped.", NORM: "pedagogika"}, + {ORTH: "pisar.", NORM: "pisarniško"}, + {ORTH: "pog.", NORM: "pogovorno"}, + {ORTH: "polit.", NORM: "politika"}, + {ORTH: "polj.", NORM: "poljsko"}, + {ORTH: "poljud.", NORM: "poljudno"}, + {ORTH: "preg.", NORM: "pregovor"}, + {ORTH: "preh.", NORM: "prehodno"}, + {ORTH: "pren.", NORM: "preneseno"}, + {ORTH: "prid.", NORM: "pridevnik"}, + {ORTH: "prim.", NORM: "primerjaj"}, + {ORTH: "prisl.", NORM: "prislov"}, + {ORTH: "psih.", NORM: "psihologija"}, + {ORTH: "psiht.", NORM: "psihiatrija"}, + {ORTH: "rad.", NORM: "radiotehnika"}, + {ORTH: "rač.", NORM: "računalništvo"}, + {ORTH: "rib.", NORM: "ribištvo"}, + {ORTH: "rod.", NORM: "rodilnik"}, + {ORTH: "rus.", NORM: "rusko"}, + {ORTH: "s.", NORM: "srednji"}, + {ORTH: "sam.", NORM: "samostalniški"}, + {ORTH: "sed.", NORM: "sedanjik"}, + {ORTH: "sep.", NORM: "september"}, + {ORTH: "slabš.", NORM: "slabšalno"}, + {ORTH: "slovan.", NORM: "slovansko"}, + {ORTH: "slovaš.", NORM: "slovaško"}, + {ORTH: "srb.", NORM: "srbsko"}, + {ORTH: "star.", NORM: "starinsko"}, + {ORTH: "stil.", NORM: "stilno"}, + {ORTH: "sv.", NORM: "svet(i)"}, + {ORTH: "teh.", NORM: "tehnika"}, + {ORTH: "tisk.", NORM: "tiskarstvo"}, + {ORTH: "tj.", NORM: "to je"}, + {ORTH: "tož.", NORM: "tožilnik"}, + {ORTH: "trg.", NORM: "trgovina"}, + {ORTH: "ukr.", NORM: "ukrajinski"}, + {ORTH: "um.", NORM: "umetnost"}, + {ORTH: "vel.", NORM: "velelnik"}, + {ORTH: "vet.", NORM: "veterina"}, + {ORTH: "vez.", NORM: "veznik"}, + {ORTH: "vn.", NORM: "visokonemško"}, + {ORTH: "voj.", NORM: "vojska"}, + {ORTH: "vrtn.", NORM: "vrtnarstvo"}, + {ORTH: "vulg.", NORM: "vulgarno"}, + {ORTH: "vznes.", NORM: "vzneseno"}, + {ORTH: "zal.", NORM: "založništvo"}, + {ORTH: "zastar.", NORM: "zastarelo"}, + {ORTH: "zgod.", NORM: "zgodovina"}, + {ORTH: "zool.", NORM: "zoologija"}, + {ORTH: "čeb.", NORM: "čebelarstvo"}, + {ORTH: "češ.", NORM: "češki"}, + {ORTH: "člov.", NORM: "človeškost"}, + {ORTH: "šah.", NORM: "šahovski"}, + {ORTH: "šalj.", NORM: "šaljivo"}, + {ORTH: "šp.", NORM: "španski"}, + {ORTH: "špan.", NORM: "špansko"}, + {ORTH: "šport.", NORM: "športni"}, + {ORTH: "štev.", NORM: "števnik"}, + {ORTH: "šved.", NORM: "švedsko"}, + {ORTH: "švic.", NORM: "švicarsko"}, + {ORTH: "ž.", NORM: "ženski"}, + {ORTH: "žarg.", NORM: "žargonsko"}, + {ORTH: "žel.", NORM: "železnica"}, + {ORTH: "živ.", NORM: "živost"}, +]: + _exc[exc_data[ORTH]] = [exc_data] + + +abbrv = """ +Co. Ch. DIPL. DR. Dr. Ev. Inc. Jr. Kr. Mag. M. MR. Mr. Mt. Murr. Npr. OZ. +Opr. Osn. Prim. Roj. ST. Sim. Sp. Sred. St. Sv. Škofl. Tel. UR. Zb. +a. aa. ab. abc. abit. abl. abs. abt. acc. accel. add. adj. adv. aet. afr. akad. al. alban. all. alleg. +alp. alt. alter. alžir. am. an. andr. ang. anh. anon. ans. antrop. apoc. app. approx. apt. ar. arc. arch. +arh. arr. as. asist. assist. assoc. asst. astr. attn. aug. avstral. az. b. bab. bal. bbl. bd. belg. bioinf. +biomed. bk. bl. bn. borg. bp. br. braz. brit. bros. broš. bt. bu. c. ca. cal. can. cand. cantab. cap. capt. +cat. cath. cc. cca. cd. cdr. cdre. cent. cerkv. cert. cf. cfr. ch. chap. chem. chr. chs. cic. circ. civ. cl. +cm. cmd. cnr. co. cod. col. coll. colo. com. comp. con. conc. cond. conn. cons. cont. coop. corr. cost. cp. +cpl. cr. crd. cres. cresc. ct. cu. d. dan. dat. davč. ddr. dec. ded. def. dem. dent. dept. dia. dip. dipl. +dir. disp. diss. div. do. doc. dok. dol. doo. dop. dott. dr. dram. druž. družb. drž. dt. duh. dur. dvr. dwt. e. +ea. ecc. eccl. eccles. econ. edn. egipt. egr. ekon. eksp. el. em. enc. eng. eo. ep. err. esp. esq. est. +et. etc. etnogr. etnol. ev. evfem. evr. ex. exc. excl. exp. expl. ext. exx. f. fa. facs. fak. faks. fas. +fasc. fco. fcp. feb. febr. fec. fed. fem. ff. fff. fid. fig. fil. film. fiziol. fiziot. flam. fm. fo. fol. folk. +frag. fran. franc. fsc. g. ga. gal. gdč. ge. gen. geod. geog. geotehnol. gg. gimn. glas. glav. gnr. go. gor. +gosp. gp. graf. gram. gren. grš. gs. h. hab. hf. hist. ho. hort. i. ia. ib. ibid. id. idr. idridr. ill. imen. +imp. impf. impr. in. inc. incl. ind. indus. inf. inform. ing. init. ins. int. inv. inšp. inštr. inž. is. islam. +ist. ital. iur. iz. izbr. izd. izg. izgr. izr. izv. j. jak. jam. jan. jav. je. jez. jr. jsl. jud. jug. +jugoslovan. jur. juž. jv. jz. k. kal. kan. kand. kat. kdo. kem. kip. kmet. kol. kom. komp. konf. kont. kost. kov. +kp. kpfw. kr. kraj. krat. kub. kult. kv. kval. l. la. lab. lb. ld. let. lib. lik. litt. lj. ljud. ll. loc. log. +loč. lt. ma. madž. mag. manag. manjš. masc. mass. mater. max. maxmax. mb. md. mech. medic. medij. medn. +mehč. mem. menedž. mes. mess. metal. meteor. meteorol. mex. mi. mikr. mil. minn. mio. misc. miss. mit. mk. +mkt. ml. mlad. mlle. mlr. mm. mme. množ. mo. moj. moš. možn. mr. mrd. mrs. ms. msc. msgr. mt. murr. mus. mut. +n. na. nad. nadalj. nadom. nagl. nakl. namer. nan. naniz. nasl. nat. navt. nač. ned. nem. nik. nizoz. nm. nn. +no. nom. norv. notr. nov. novogr. ns. o. ob. obd. obj. oblač. obl. oblik. obr. obraz. obs. obst. obt. obč. oc. +oct. od. odd. odg. odn. odst. odv. oec. off. ok. okla. okr. ont. oo. op. opis. opp. opr. orch. ord. ore. oreg. +org. orient. orig. ork. ort. oseb. osn. ot. ozir. ošk. p. pag. par. para. parc. parl. part. past. pat. pdk. +pen. perf. pert. perz. pesn. pet. pev. pf. pfc. ph. pharm. phil. pis. pl. po. pod. podr. podaljš. pogl. pogoj. pojm. +pok. pokr. pol. poljed. poljub. polu. pom. pomen. pon. ponov. pop. por. port. pos. posl. posn. pov. pp. ppl. pr. +praet. prav. pravopis. pravosl. preb. pred. predl. predm. predp. preds. pref. pregib. prel. prem. premen. prep. +pres. pret. prev. pribl. prih. pril. primerj. primor. prip. pripor. prir. prist. priv. proc. prof. prog. proiz. +prom. pron. prop. prot. protest. prov. ps. pss. pt. publ. pz. q. qld. qu. quad. que. r. racc. rastl. razgl. +razl. razv. rd. red. ref. reg. rel. relig. rep. repr. rer. resp. rest. ret. rev. revol. rež. rim. rist. rkp. rm. +roj. rom. romun. rp. rr. rt. rud. ruš. ry. sal. samogl. san. sc. scen. sci. scr. sdv. seg. sek. sen. sept. ser. +sev. sg. sgt. sh. sig. sigg. sign. sim. sin. sing. sinh. skand. skl. sklad. sklanj. sklep. skr. sl. slik. slov. +slovak. slovn. sn. so. sob. soc. sociol. sod. sopomen. sopr. sor. sov. sovj. sp. spec. spl. spr. spreg. sq. sr. +sre. sred. sredoz. srh. ss. ssp. st. sta. stan. stanstar. stcsl. ste. stim. stol. stom. str. stroj. strok. stsl. +stud. sup. supl. suppl. svet. sz. t. tab. tech. ted. tehn. tehnol. tek. teks. tekst. tel. temp. ten. teol. ter. +term. test. th. theol. tim. tip. tisočl. tit. tl. tol. tolmač. tom. tor. tov. tr. trad. traj. trans. tren. +trib. tril. trop. trp. trž. ts. tt. tu. tur. turiz. tvor. tvorb. tč. u. ul. umet. un. univ. up. upr. ur. urad. +us. ust. utr. v. va. val. var. varn. ven. ver. verb. vest. vezal. vic. vis. viv. viz. viš. vod. vok. vol. vpr. +vrst. vrstil. vs. vv. vzd. vzg. vzh. vzor. w. wed. wg. wk. x. y. z. zah. zaim. zak. zap. zasl. zavar. zač. zb. +združ. zg. zn. znan. znanstv. zoot. zun. zv. zvd. á. é. ć. č. čas. čet. čl. člen. čustv. đ. ľ. ł. ş. ŠT. š. šir. +škofl. škot. šol. št. števil. štud. ů. ű. žen. žival. +""".split() + +for orth in abbrv: + _exc[orth] = [{ORTH: orth}] + + +TOKENIZER_EXCEPTIONS = update_exc(BASE_EXCEPTIONS, _exc) diff --git a/spacy/lang/uk/__init__.py b/spacy/lang/uk/__init__.py index 737243b66..bfea9ff69 100644 --- a/spacy/lang/uk/__init__.py +++ b/spacy/lang/uk/__init__.py @@ -29,7 +29,7 @@ class Ukrainian(Language): assigns=["token.lemma"], default_config={ "model": None, - "mode": "pymorphy2", + "mode": "pymorphy3", "overwrite": False, "scorer": {"@scorers": "spacy.lemmatizer_scorer.v1"}, }, diff --git a/spacy/lang/uk/lemmatizer.py b/spacy/lang/uk/lemmatizer.py index a8bc56057..8337e7328 100644 --- a/spacy/lang/uk/lemmatizer.py +++ b/spacy/lang/uk/lemmatizer.py @@ -14,11 +14,11 @@ class UkrainianLemmatizer(RussianLemmatizer): model: Optional[Model], name: str = "lemmatizer", *, - mode: str = "pymorphy2", + mode: str = "pymorphy3", overwrite: bool = False, scorer: Optional[Callable] = lemmatizer_score, ) -> None: - if mode == "pymorphy2": + if mode in {"pymorphy2", "pymorphy2_lookup"}: try: from pymorphy2 import MorphAnalyzer except ImportError: @@ -29,6 +29,17 @@ class UkrainianLemmatizer(RussianLemmatizer): ) from None if getattr(self, "_morph", None) is None: self._morph = MorphAnalyzer(lang="uk") + elif mode == "pymorphy3": + try: + from pymorphy3 import MorphAnalyzer + except ImportError: + raise ImportError( + "The Ukrainian lemmatizer mode 'pymorphy3' requires the " + "pymorphy3 library and dictionaries. Install them with: " + "pip install pymorphy3 pymorphy3-dicts-uk" + ) from None + if getattr(self, "_morph", None) is None: + self._morph = MorphAnalyzer(lang="uk") super().__init__( vocab, model, name, mode=mode, overwrite=overwrite, scorer=scorer ) diff --git a/spacy/language.py b/spacy/language.py index 34a06e576..d391f15ab 100644 --- a/spacy/language.py +++ b/spacy/language.py @@ -1,4 +1,4 @@ -from typing import Iterator, Optional, Any, Dict, Callable, Iterable, Collection +from typing import Iterator, Optional, Any, Dict, Callable, Iterable from typing import Union, Tuple, List, Set, Pattern, Sequence from typing import NoReturn, TYPE_CHECKING, TypeVar, cast, overload @@ -10,6 +10,7 @@ from contextlib import contextmanager from copy import deepcopy from pathlib import Path import warnings + from thinc.api import get_current_ops, Config, CupyOps, Optimizer import srsly import multiprocessing as mp @@ -24,7 +25,7 @@ from .pipe_analysis import validate_attrs, analyze_pipes, print_pipe_analysis from .training import Example, validate_examples from .training.initialize import init_vocab, init_tok2vec from .scorer import Scorer -from .util import registry, SimpleFrozenList, _pipe, raise_error +from .util import registry, SimpleFrozenList, _pipe, raise_error, _DEFAULT_EMPTY_PIPES from .util import SimpleFrozenDict, combine_score_weights, CONFIG_SECTION_ORDER from .util import warn_if_jupyter_cupy from .lang.tokenizer_exceptions import URL_MATCH, BASE_EXCEPTIONS @@ -1698,9 +1699,9 @@ class Language: config: Union[Dict[str, Any], Config] = {}, *, vocab: Union[Vocab, bool] = True, - disable: Union[str, Iterable[str]] = SimpleFrozenList(), - enable: Union[str, Iterable[str]] = SimpleFrozenList(), - exclude: Union[str, Iterable[str]] = SimpleFrozenList(), + disable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + enable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + exclude: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, meta: Dict[str, Any] = SimpleFrozenDict(), auto_fill: bool = True, validate: bool = True, @@ -1727,12 +1728,6 @@ class Language: DOCS: https://spacy.io/api/language#from_config """ - if isinstance(disable, str): - disable = [disable] - if isinstance(enable, str): - enable = [enable] - if isinstance(exclude, str): - exclude = [exclude] if auto_fill: config = Config( cls.default_config, section_order=CONFIG_SECTION_ORDER @@ -1877,9 +1872,38 @@ class Language: nlp.vocab.from_bytes(vocab_b) # Resolve disabled/enabled settings. + if isinstance(disable, str): + disable = [disable] + if isinstance(enable, str): + enable = [enable] + if isinstance(exclude, str): + exclude = [exclude] + + def fetch_pipes_status(value: Iterable[str], key: str) -> Iterable[str]: + """Fetch value for `enable` or `disable` w.r.t. the specified config and passed arguments passed to + .load(). If both arguments and config specified values for this field, the passed arguments take precedence + and a warning is printed. + value (Iterable[str]): Passed value for `enable` or `disable`. + key (str): Key for field in config (either "enabled" or "disabled"). + RETURN (Iterable[str]): + """ + # We assume that no argument was passed if the value is the specified default value. + if id(value) == id(_DEFAULT_EMPTY_PIPES): + return config["nlp"].get(key, []) + else: + if len(config["nlp"].get(key, [])): + warnings.warn( + Warnings.W123.format( + arg=key[:-1], + arg_value=value, + config_value=config["nlp"][key], + ) + ) + return value + disabled_pipes = cls._resolve_component_status( - [*config["nlp"]["disabled"], *disable], - [*config["nlp"].get("enabled", []), *enable], + fetch_pipes_status(disable, "disabled"), + fetch_pipes_status(enable, "enabled"), config["nlp"]["pipeline"], ) nlp._disabled = set(p for p in disabled_pipes if p not in exclude) @@ -2064,14 +2088,7 @@ class Language: pipe_name for pipe_name in pipe_names if pipe_name not in enable ] if disable and disable != to_disable: - raise ValueError( - Errors.E1042.format( - arg1="enable", - arg2="disable", - arg1_values=enable, - arg2_values=disable, - ) - ) + raise ValueError(Errors.E1042.format(enable=enable, disable=disable)) return tuple(to_disable) diff --git a/spacy/ml/models/entity_linker.py b/spacy/ml/models/entity_linker.py index d847342a3..4d18d216a 100644 --- a/spacy/ml/models/entity_linker.py +++ b/spacy/ml/models/entity_linker.py @@ -1,11 +1,12 @@ from pathlib import Path from typing import Optional, Callable, Iterable, List, Tuple from thinc.types import Floats2d -from thinc.api import chain, clone, list2ragged, reduce_mean, residual -from thinc.api import Model, Maxout, Linear, noop, tuplify, Ragged +from thinc.api import chain, list2ragged, reduce_mean, residual +from thinc.api import Model, Maxout, Linear, tuplify, Ragged from ...util import registry -from ...kb import KnowledgeBase, Candidate, get_candidates +from ...kb import KnowledgeBase, InMemoryLookupKB +from ...kb import Candidate, get_candidates, get_candidates_batch from ...vocab import Vocab from ...tokens import Span, Doc from ..extract_spans import extract_spans @@ -78,9 +79,11 @@ def span_maker_forward(model, docs: List[Doc], is_train) -> Tuple[Ragged, Callab @registry.misc("spacy.KBFromFile.v1") -def load_kb(kb_path: Path) -> Callable[[Vocab], KnowledgeBase]: - def kb_from_file(vocab): - kb = KnowledgeBase(vocab, entity_vector_length=1) +def load_kb( + kb_path: Path, +) -> Callable[[Vocab], KnowledgeBase]: + def kb_from_file(vocab: Vocab): + kb = InMemoryLookupKB(vocab, entity_vector_length=1) kb.from_disk(kb_path) return kb @@ -88,9 +91,11 @@ def load_kb(kb_path: Path) -> Callable[[Vocab], KnowledgeBase]: @registry.misc("spacy.EmptyKB.v1") -def empty_kb(entity_vector_length: int) -> Callable[[Vocab], KnowledgeBase]: - def empty_kb_factory(vocab): - return KnowledgeBase(vocab=vocab, entity_vector_length=entity_vector_length) +def empty_kb( + entity_vector_length: int, +) -> Callable[[Vocab], KnowledgeBase]: + def empty_kb_factory(vocab: Vocab): + return InMemoryLookupKB(vocab=vocab, entity_vector_length=entity_vector_length) return empty_kb_factory @@ -98,3 +103,10 @@ def empty_kb(entity_vector_length: int) -> Callable[[Vocab], KnowledgeBase]: @registry.misc("spacy.CandidateGenerator.v1") def create_candidates() -> Callable[[KnowledgeBase, Span], Iterable[Candidate]]: return get_candidates + + +@registry.misc("spacy.CandidateBatchGenerator.v1") +def create_candidates_batch() -> Callable[ + [KnowledgeBase, Iterable[Span]], Iterable[Iterable[Candidate]] +]: + return get_candidates_batch diff --git a/spacy/pipeline/edit_tree_lemmatizer.py b/spacy/pipeline/edit_tree_lemmatizer.py index b7d615f6d..12f9b73a3 100644 --- a/spacy/pipeline/edit_tree_lemmatizer.py +++ b/spacy/pipeline/edit_tree_lemmatizer.py @@ -1,7 +1,6 @@ from typing import cast, Any, Callable, Dict, Iterable, List, Optional -from typing import Sequence, Tuple, Union +from typing import Tuple from collections import Counter -from copy import deepcopy from itertools import islice import numpy as np @@ -149,9 +148,7 @@ class EditTreeLemmatizer(TrainablePipe): if not any(len(doc) for doc in docs): # Handle cases where there are no tokens in any docs. n_labels = len(self.cfg["labels"]) - guesses: List[Ints2d] = [ - self.model.ops.alloc((0, n_labels), dtype="i") for doc in docs - ] + guesses: List[Ints2d] = [self.model.ops.alloc2i(0, n_labels) for _ in docs] assert len(guesses) == n_docs return guesses scores = self.model.predict(docs) diff --git a/spacy/pipeline/entity_linker.py b/spacy/pipeline/entity_linker.py index 73a90b268..62845287b 100644 --- a/spacy/pipeline/entity_linker.py +++ b/spacy/pipeline/entity_linker.py @@ -53,9 +53,11 @@ DEFAULT_NEL_MODEL = Config().from_str(default_model_config)["model"] "incl_context": True, "entity_vector_length": 64, "get_candidates": {"@misc": "spacy.CandidateGenerator.v1"}, + "get_candidates_batch": {"@misc": "spacy.CandidateBatchGenerator.v1"}, "overwrite": True, "scorer": {"@scorers": "spacy.entity_linker_scorer.v1"}, "use_gold_ents": True, + "candidates_batch_size": 1, "threshold": None, }, default_score_weights={ @@ -75,9 +77,13 @@ def make_entity_linker( incl_context: bool, entity_vector_length: int, get_candidates: Callable[[KnowledgeBase, Span], Iterable[Candidate]], + get_candidates_batch: Callable[ + [KnowledgeBase, Iterable[Span]], Iterable[Iterable[Candidate]] + ], overwrite: bool, scorer: Optional[Callable], use_gold_ents: bool, + candidates_batch_size: int, threshold: Optional[float] = None, ): """Construct an EntityLinker component. @@ -90,17 +96,21 @@ def make_entity_linker( incl_prior (bool): Whether or not to include prior probabilities from the KB in the model. incl_context (bool): Whether or not to include the local context in the model. entity_vector_length (int): Size of encoding vectors in the KB. - get_candidates (Callable[[KnowledgeBase, "Span"], Iterable[Candidate]]): Function that + get_candidates (Callable[[KnowledgeBase, Span], Iterable[Candidate]]): Function that produces a list of candidates, given a certain knowledge base and a textual mention. + get_candidates_batch ( + Callable[[KnowledgeBase, Iterable[Span]], Iterable[Iterable[Candidate]]], Iterable[Candidate]] + ): Function that produces a list of candidates, given a certain knowledge base and several textual mentions. scorer (Optional[Callable]): The scoring method. use_gold_ents (bool): Whether to copy entities from gold docs or not. If false, another component must provide entity annotations. + candidates_batch_size (int): Size of batches for entity candidate generation. threshold (Optional[float]): Confidence threshold for entity predictions. If confidence is below the threshold, prediction is discarded. If None, predictions are not filtered by any threshold. """ if not model.attrs.get("include_span_maker", False): - # The only difference in arguments here is that use_gold_ents is not available + # The only difference in arguments here is that use_gold_ents and threshold aren't available. return EntityLinker_v1( nlp.vocab, model, @@ -124,9 +134,11 @@ def make_entity_linker( incl_context=incl_context, entity_vector_length=entity_vector_length, get_candidates=get_candidates, + get_candidates_batch=get_candidates_batch, overwrite=overwrite, scorer=scorer, use_gold_ents=use_gold_ents, + candidates_batch_size=candidates_batch_size, threshold=threshold, ) @@ -160,9 +172,13 @@ class EntityLinker(TrainablePipe): incl_context: bool, entity_vector_length: int, get_candidates: Callable[[KnowledgeBase, Span], Iterable[Candidate]], + get_candidates_batch: Callable[ + [KnowledgeBase, Iterable[Span]], Iterable[Iterable[Candidate]] + ], overwrite: bool = BACKWARD_OVERWRITE, scorer: Optional[Callable] = entity_linker_score, use_gold_ents: bool, + candidates_batch_size: int, threshold: Optional[float] = None, ) -> None: """Initialize an entity linker. @@ -178,10 +194,14 @@ class EntityLinker(TrainablePipe): entity_vector_length (int): Size of encoding vectors in the KB. get_candidates (Callable[[KnowledgeBase, Span], Iterable[Candidate]]): Function that produces a list of candidates, given a certain knowledge base and a textual mention. - scorer (Optional[Callable]): The scoring method. Defaults to - Scorer.score_links. + get_candidates_batch ( + Callable[[KnowledgeBase, Iterable[Span]], Iterable[Iterable[Candidate]]], + Iterable[Candidate]] + ): Function that produces a list of candidates, given a certain knowledge base and several textual mentions. + scorer (Optional[Callable]): The scoring method. Defaults to Scorer.score_links. use_gold_ents (bool): Whether to copy entities from gold docs or not. If false, another component must provide entity annotations. + candidates_batch_size (int): Size of batches for entity candidate generation. threshold (Optional[float]): Confidence threshold for entity predictions. If confidence is below the threshold, prediction is discarded. If None, predictions are not filtered by any threshold. DOCS: https://spacy.io/api/entitylinker#init @@ -204,22 +224,27 @@ class EntityLinker(TrainablePipe): self.incl_prior = incl_prior self.incl_context = incl_context self.get_candidates = get_candidates + self.get_candidates_batch = get_candidates_batch self.cfg: Dict[str, Any] = {"overwrite": overwrite} self.distance = CosineDistance(normalize=False) # how many neighbour sentences to take into account - # create an empty KB by default. If you want to load a predefined one, specify it in 'initialize'. + # create an empty KB by default self.kb = empty_kb(entity_vector_length)(self.vocab) self.scorer = scorer self.use_gold_ents = use_gold_ents + self.candidates_batch_size = candidates_batch_size self.threshold = threshold + if candidates_batch_size < 1: + raise ValueError(Errors.E1044) + def set_kb(self, kb_loader: Callable[[Vocab], KnowledgeBase]): """Define the KB of this pipe by providing a function that will create it using this object's vocab.""" if not callable(kb_loader): raise ValueError(Errors.E885.format(arg_type=type(kb_loader))) - self.kb = kb_loader(self.vocab) + self.kb = kb_loader(self.vocab) # type: ignore def validate_kb(self) -> None: # Raise an error if the knowledge base is not initialized. @@ -241,8 +266,8 @@ class EntityLinker(TrainablePipe): get_examples (Callable[[], Iterable[Example]]): Function that returns a representative sample of gold-standard Example objects. nlp (Language): The current nlp object the component is part of. - kb_loader (Callable[[Vocab], KnowledgeBase]): A function that creates a KnowledgeBase from a Vocab instance. - Note that providing this argument, will overwrite all data accumulated in the current KB. + kb_loader (Callable[[Vocab], KnowledgeBase]): A function that creates a KnowledgeBase from a Vocab + instance. Note that providing this argument will overwrite all data accumulated in the current KB. Use this only when loading a KB as-such from file. DOCS: https://spacy.io/api/entitylinker#initialize @@ -419,66 +444,93 @@ class EntityLinker(TrainablePipe): if len(doc) == 0: continue sentences = [s for s in doc.sents] - # Looping through each entity (TODO: rewrite) - for ent in doc.ents: - sent_index = sentences.index(ent.sent) - assert sent_index >= 0 - if self.incl_context: - # get n_neighbour sentences, clipped to the length of the document - start_sentence = max(0, sent_index - self.n_sents) - end_sentence = min(len(sentences) - 1, sent_index + self.n_sents) - start_token = sentences[start_sentence].start - end_token = sentences[end_sentence].end - sent_doc = doc[start_token:end_token].as_doc() - # currently, the context is the same for each entity in a sentence (should be refined) - sentence_encoding = self.model.predict([sent_doc])[0] - sentence_encoding_t = sentence_encoding.T - sentence_norm = xp.linalg.norm(sentence_encoding_t) - entity_count += 1 - if ent.label_ in self.labels_discard: - # ignoring this entity - setting to NIL - final_kb_ids.append(self.NIL) - else: - candidates = list(self.get_candidates(self.kb, ent)) - if not candidates: - # no prediction possible for this entity - setting to NIL - final_kb_ids.append(self.NIL) - elif len(candidates) == 1 and self.threshold is None: - # shortcut for efficiency reasons: take the 1 candidate - final_kb_ids.append(candidates[0].entity_) - else: - random.shuffle(candidates) - # set all prior probabilities to 0 if incl_prior=False - prior_probs = xp.asarray([c.prior_prob for c in candidates]) - if not self.incl_prior: - prior_probs = xp.asarray([0.0 for _ in candidates]) - scores = prior_probs - # add in similarity from the context - if self.incl_context: - entity_encodings = xp.asarray( - [c.entity_vector for c in candidates] - ) - entity_norm = xp.linalg.norm(entity_encodings, axis=1) - if len(entity_encodings) != len(prior_probs): - raise RuntimeError( - Errors.E147.format( - method="predict", - msg="vectors not of equal length", - ) - ) - # cosine similarity - sims = xp.dot(entity_encodings, sentence_encoding_t) / ( - sentence_norm * entity_norm - ) - if sims.shape != prior_probs.shape: - raise ValueError(Errors.E161) - scores = prior_probs + sims - (prior_probs * sims) - final_kb_ids.append( - candidates[scores.argmax().item()].entity_ - if self.threshold is None or scores.max() >= self.threshold - else EntityLinker.NIL + # Loop over entities in batches. + for ent_idx in range(0, len(doc.ents), self.candidates_batch_size): + ent_batch = doc.ents[ent_idx : ent_idx + self.candidates_batch_size] + + # Look up candidate entities. + valid_ent_idx = [ + idx + for idx in range(len(ent_batch)) + if ent_batch[idx].label_ not in self.labels_discard + ] + + batch_candidates = list( + self.get_candidates_batch( + self.kb, [ent_batch[idx] for idx in valid_ent_idx] + ) + if self.candidates_batch_size > 1 + else [ + self.get_candidates(self.kb, ent_batch[idx]) + for idx in valid_ent_idx + ] + ) + + # Looping through each entity in batch (TODO: rewrite) + for j, ent in enumerate(ent_batch): + sent_index = sentences.index(ent.sent) + assert sent_index >= 0 + + if self.incl_context: + # get n_neighbour sentences, clipped to the length of the document + start_sentence = max(0, sent_index - self.n_sents) + end_sentence = min( + len(sentences) - 1, sent_index + self.n_sents ) + start_token = sentences[start_sentence].start + end_token = sentences[end_sentence].end + sent_doc = doc[start_token:end_token].as_doc() + # currently, the context is the same for each entity in a sentence (should be refined) + sentence_encoding = self.model.predict([sent_doc])[0] + sentence_encoding_t = sentence_encoding.T + sentence_norm = xp.linalg.norm(sentence_encoding_t) + entity_count += 1 + if ent.label_ in self.labels_discard: + # ignoring this entity - setting to NIL + final_kb_ids.append(self.NIL) + else: + candidates = list(batch_candidates[j]) + if not candidates: + # no prediction possible for this entity - setting to NIL + final_kb_ids.append(self.NIL) + elif len(candidates) == 1 and self.threshold is None: + # shortcut for efficiency reasons: take the 1 candidate + final_kb_ids.append(candidates[0].entity_) + else: + random.shuffle(candidates) + # set all prior probabilities to 0 if incl_prior=False + prior_probs = xp.asarray([c.prior_prob for c in candidates]) + if not self.incl_prior: + prior_probs = xp.asarray([0.0 for _ in candidates]) + scores = prior_probs + # add in similarity from the context + if self.incl_context: + entity_encodings = xp.asarray( + [c.entity_vector for c in candidates] + ) + entity_norm = xp.linalg.norm(entity_encodings, axis=1) + if len(entity_encodings) != len(prior_probs): + raise RuntimeError( + Errors.E147.format( + method="predict", + msg="vectors not of equal length", + ) + ) + # cosine similarity + sims = xp.dot(entity_encodings, sentence_encoding_t) / ( + sentence_norm * entity_norm + ) + if sims.shape != prior_probs.shape: + raise ValueError(Errors.E161) + scores = prior_probs + sims - (prior_probs * sims) + final_kb_ids.append( + candidates[scores.argmax().item()].entity_ + if self.threshold is None + or scores.max() >= self.threshold + else EntityLinker.NIL + ) + if not (len(final_kb_ids) == entity_count): err = Errors.E147.format( method="predict", msg="result variables not of equal length" diff --git a/spacy/pipeline/entityruler.py b/spacy/pipeline/entityruler.py index 3cb1ca676..8154a077d 100644 --- a/spacy/pipeline/entityruler.py +++ b/spacy/pipeline/entityruler.py @@ -1,6 +1,5 @@ -import warnings from typing import Optional, Union, List, Dict, Tuple, Iterable, Any, Callable, Sequence -from typing import cast +import warnings from collections import defaultdict from pathlib import Path import srsly @@ -317,7 +316,7 @@ class EntityRuler(Pipe): phrase_pattern["id"] = ent_id phrase_patterns.append(phrase_pattern) for entry in token_patterns + phrase_patterns: # type: ignore[operator] - label = entry["label"] + label = entry["label"] # type: ignore if "id" in entry: ent_label = label label = self._create_label(label, entry["id"]) diff --git a/spacy/pipeline/legacy/entity_linker.py b/spacy/pipeline/legacy/entity_linker.py index 2f8a1f8ea..c14dfa1db 100644 --- a/spacy/pipeline/legacy/entity_linker.py +++ b/spacy/pipeline/legacy/entity_linker.py @@ -68,8 +68,7 @@ class EntityLinker_v1(TrainablePipe): entity_vector_length (int): Size of encoding vectors in the KB. get_candidates (Callable[[KnowledgeBase, Span], Iterable[Candidate]]): Function that produces a list of candidates, given a certain knowledge base and a textual mention. - scorer (Optional[Callable]): The scoring method. Defaults to - Scorer.score_links. + scorer (Optional[Callable]): The scoring method. Defaults to Scorer.score_links. DOCS: https://spacy.io/api/entitylinker#init """ self.vocab = vocab @@ -115,7 +114,7 @@ class EntityLinker_v1(TrainablePipe): get_examples (Callable[[], Iterable[Example]]): Function that returns a representative sample of gold-standard Example objects. nlp (Language): The current nlp object the component is part of. - kb_loader (Callable[[Vocab], KnowledgeBase]): A function that creates a KnowledgeBase from a Vocab instance. + kb_loader (Callable[[Vocab], KnowledgeBase]): A function that creates an InMemoryLookupKB from a Vocab instance. Note that providing this argument, will overwrite all data accumulated in the current KB. Use this only when loading a KB as-such from file. diff --git a/spacy/pipeline/spancat.py b/spacy/pipeline/spancat.py index 1b7a9eecb..956bbb72c 100644 --- a/spacy/pipeline/spancat.py +++ b/spacy/pipeline/spancat.py @@ -26,17 +26,17 @@ scorer = {"@layers": "spacy.LinearLogistic.v1"} hidden_size = 128 [model.tok2vec] -@architectures = "spacy.Tok2Vec.v1" +@architectures = "spacy.Tok2Vec.v2" [model.tok2vec.embed] -@architectures = "spacy.MultiHashEmbed.v1" +@architectures = "spacy.MultiHashEmbed.v2" width = 96 rows = [5000, 2000, 1000, 1000] attrs = ["ORTH", "PREFIX", "SUFFIX", "SHAPE"] include_static_vectors = false [model.tok2vec.encode] -@architectures = "spacy.MaxoutWindowEncoder.v1" +@architectures = "spacy.MaxoutWindowEncoder.v2" width = ${model.tok2vec.embed.width} window_size = 1 maxout_pieces = 3 @@ -133,6 +133,9 @@ def make_spancat( spans_key (str): Key of the doc.spans dict to save the spans under. During initialization and training, the component will look for spans on the reference document under the same key. + scorer (Optional[Callable]): The scoring method. Defaults to + Scorer.score_spans for the Doc.spans[spans_key] with overlapping + spans allowed. threshold (float): Minimum probability to consider a prediction positive. Spans with a positive prediction will be saved on the Doc. Defaults to 0.5. diff --git a/spacy/pipeline/textcat.py b/spacy/pipeline/textcat.py index c45f819fc..59549ad99 100644 --- a/spacy/pipeline/textcat.py +++ b/spacy/pipeline/textcat.py @@ -24,8 +24,8 @@ single_label_default_config = """ [model.tok2vec.embed] @architectures = "spacy.MultiHashEmbed.v2" width = 64 -rows = [2000, 2000, 1000, 1000, 1000, 1000] -attrs = ["ORTH", "LOWER", "PREFIX", "SUFFIX", "SHAPE", "ID"] +rows = [2000, 2000, 500, 1000, 500] +attrs = ["NORM", "LOWER", "PREFIX", "SUFFIX", "SHAPE"] include_static_vectors = false [model.tok2vec.encode] diff --git a/spacy/pipeline/textcat_multilabel.py b/spacy/pipeline/textcat_multilabel.py index e33a885f8..eb83d9cb7 100644 --- a/spacy/pipeline/textcat_multilabel.py +++ b/spacy/pipeline/textcat_multilabel.py @@ -19,17 +19,17 @@ multi_label_default_config = """ @architectures = "spacy.TextCatEnsemble.v2" [model.tok2vec] -@architectures = "spacy.Tok2Vec.v1" +@architectures = "spacy.Tok2Vec.v2" [model.tok2vec.embed] @architectures = "spacy.MultiHashEmbed.v2" width = 64 -rows = [2000, 2000, 1000, 1000, 1000, 1000] -attrs = ["ORTH", "LOWER", "PREFIX", "SUFFIX", "SHAPE", "ID"] +rows = [2000, 2000, 500, 1000, 500] +attrs = ["NORM", "LOWER", "PREFIX", "SUFFIX", "SHAPE"] include_static_vectors = false [model.tok2vec.encode] -@architectures = "spacy.MaxoutWindowEncoder.v1" +@architectures = "spacy.MaxoutWindowEncoder.v2" width = ${model.tok2vec.embed.width} window_size = 1 maxout_pieces = 3 @@ -96,8 +96,8 @@ def make_multilabel_textcat( model: Model[List[Doc], List[Floats2d]], threshold: float, scorer: Optional[Callable], -) -> "TextCategorizer": - """Create a TextCategorizer component. The text categorizer predicts categories +) -> "MultiLabel_TextCategorizer": + """Create a MultiLabel_TextCategorizer component. The text categorizer predicts categories over a whole document. It can learn one or more labels, and the labels are considered to be non-mutually exclusive, which means that there can be zero or more labels per doc). @@ -105,6 +105,7 @@ def make_multilabel_textcat( model (Model[List[Doc], List[Floats2d]]): A model instance that predicts scores for each category. threshold (float): Cutoff to consider a prediction "positive". + scorer (Optional[Callable]): The scoring method. """ return MultiLabel_TextCategorizer( nlp.vocab, model, name, threshold=threshold, scorer=scorer @@ -147,6 +148,7 @@ class MultiLabel_TextCategorizer(TextCategorizer): name (str): The component instance name, used to add entries to the losses during training. threshold (float): Cutoff to consider a prediction "positive". + scorer (Optional[Callable]): The scoring method. DOCS: https://spacy.io/api/textcategorizer#init """ diff --git a/spacy/pipeline/tok2vec.py b/spacy/pipeline/tok2vec.py index 2e3dde3cb..c742aaeaa 100644 --- a/spacy/pipeline/tok2vec.py +++ b/spacy/pipeline/tok2vec.py @@ -123,9 +123,6 @@ class Tok2Vec(TrainablePipe): width = self.model.get_dim("nO") return [self.model.ops.alloc((0, width)) for doc in docs] tokvecs = self.model.predict(docs) - batch_id = Tok2VecListener.get_batch_id(docs) - for listener in self.listeners: - listener.receive(batch_id, tokvecs, _empty_backprop) return tokvecs def set_annotations(self, docs: Sequence[Doc], tokvecses) -> None: @@ -286,8 +283,19 @@ class Tok2VecListener(Model): def forward(model: Tok2VecListener, inputs, is_train: bool): """Supply the outputs from the upstream Tok2Vec component.""" if is_train: - model.verify_inputs(inputs) - return model._outputs, model._backprop + # This might occur during training when the tok2vec layer is frozen / hasn't been updated. + # In that case, it should be set to "annotating" so we can retrieve the embeddings from the doc. + if model._batch_id is None: + outputs = [] + for doc in inputs: + if doc.tensor.size == 0: + raise ValueError(Errors.E203.format(name="tok2vec")) + else: + outputs.append(doc.tensor) + return outputs, _empty_backprop + else: + model.verify_inputs(inputs) + return model._outputs, model._backprop else: # This is pretty grim, but it's hard to do better :(. # It's hard to avoid relying on the doc.tensor attribute, because the @@ -306,7 +314,7 @@ def forward(model: Tok2VecListener, inputs, is_train: bool): outputs.append(model.ops.alloc2f(len(doc), width)) else: outputs.append(doc.tensor) - return outputs, lambda dX: [] + return outputs, _empty_backprop def _empty_backprop(dX): # for pickling diff --git a/spacy/schemas.py b/spacy/schemas.py index 048082134..c824d76b9 100644 --- a/spacy/schemas.py +++ b/spacy/schemas.py @@ -181,12 +181,12 @@ class TokenPatternNumber(BaseModel): IS_SUBSET: Optional[List[StrictInt]] = Field(None, alias="is_subset") IS_SUPERSET: Optional[List[StrictInt]] = Field(None, alias="is_superset") INTERSECTS: Optional[List[StrictInt]] = Field(None, alias="intersects") - EQ: Union[StrictInt, StrictFloat] = Field(None, alias="==") - NEQ: Union[StrictInt, StrictFloat] = Field(None, alias="!=") - GEQ: Union[StrictInt, StrictFloat] = Field(None, alias=">=") - LEQ: Union[StrictInt, StrictFloat] = Field(None, alias="<=") - GT: Union[StrictInt, StrictFloat] = Field(None, alias=">") - LT: Union[StrictInt, StrictFloat] = Field(None, alias="<") + EQ: Optional[Union[StrictInt, StrictFloat]] = Field(None, alias="==") + NEQ: Optional[Union[StrictInt, StrictFloat]] = Field(None, alias="!=") + GEQ: Optional[Union[StrictInt, StrictFloat]] = Field(None, alias=">=") + LEQ: Optional[Union[StrictInt, StrictFloat]] = Field(None, alias="<=") + GT: Optional[Union[StrictInt, StrictFloat]] = Field(None, alias=">") + LT: Optional[Union[StrictInt, StrictFloat]] = Field(None, alias="<") class Config: extra = "forbid" @@ -430,7 +430,7 @@ class ProjectConfigAssetURL(BaseModel): # fmt: off dest: StrictStr = Field(..., title="Destination of downloaded asset") url: Optional[StrictStr] = Field(None, title="URL of asset") - checksum: str = Field(None, title="MD5 hash of file", regex=r"([a-fA-F\d]{32})") + checksum: Optional[str] = Field(None, title="MD5 hash of file", regex=r"([a-fA-F\d]{32})") description: StrictStr = Field("", title="Description of asset") # fmt: on @@ -438,7 +438,7 @@ class ProjectConfigAssetURL(BaseModel): class ProjectConfigAssetGit(BaseModel): # fmt: off git: ProjectConfigAssetGitItem = Field(..., title="Git repo information") - checksum: str = Field(None, title="MD5 hash of file", regex=r"([a-fA-F\d]{32})") + checksum: Optional[str] = Field(None, title="MD5 hash of file", regex=r"([a-fA-F\d]{32})") description: Optional[StrictStr] = Field(None, title="Description of asset") # fmt: on @@ -508,9 +508,9 @@ class DocJSONSchema(BaseModel): None, title="Indices of sentences' start and end indices" ) text: StrictStr = Field(..., title="Document text") - spans: Dict[StrictStr, List[Dict[StrictStr, Union[StrictStr, StrictInt]]]] = Field( - None, title="Span information - end/start indices, label, KB ID" - ) + spans: Optional[ + Dict[StrictStr, List[Dict[StrictStr, Union[StrictStr, StrictInt]]]] + ] = Field(None, title="Span information - end/start indices, label, KB ID") tokens: List[Dict[StrictStr, Union[StrictStr, StrictInt]]] = Field( ..., title="Token information - ID, start, annotations" ) @@ -519,9 +519,9 @@ class DocJSONSchema(BaseModel): title="Any custom data stored in the document's _ attribute", alias="_", ) - underscore_token: Optional[Dict[StrictStr, Dict[StrictStr, Any]]] = Field( + underscore_token: Optional[Dict[StrictStr, List[Dict[StrictStr, Any]]]] = Field( None, title="Any custom data stored in the token's _ attribute" ) - underscore_span: Optional[Dict[StrictStr, Dict[StrictStr, Any]]] = Field( + underscore_span: Optional[Dict[StrictStr, List[Dict[StrictStr, Any]]]] = Field( None, title="Any custom data stored in the span's _ attribute" ) diff --git a/spacy/tests/conftest.py b/spacy/tests/conftest.py index 742bfcc6a..0fc74243d 100644 --- a/spacy/tests/conftest.py +++ b/spacy/tests/conftest.py @@ -333,16 +333,24 @@ def ro_tokenizer(): @pytest.fixture(scope="session") def ru_tokenizer(): - pytest.importorskip("pymorphy2") + pytest.importorskip("pymorphy3") return get_lang_class("ru")().tokenizer @pytest.fixture def ru_lemmatizer(): - pytest.importorskip("pymorphy2") + pytest.importorskip("pymorphy3") return get_lang_class("ru")().add_pipe("lemmatizer") +@pytest.fixture +def ru_lookup_lemmatizer(): + pytest.importorskip("pymorphy2") + return get_lang_class("ru")().add_pipe( + "lemmatizer", config={"mode": "pymorphy2_lookup"} + ) + + @pytest.fixture(scope="session") def sa_tokenizer(): return get_lang_class("sa")().tokenizer @@ -411,15 +419,24 @@ def ky_tokenizer(): @pytest.fixture(scope="session") def uk_tokenizer(): - pytest.importorskip("pymorphy2") + pytest.importorskip("pymorphy3") return get_lang_class("uk")().tokenizer @pytest.fixture def uk_lemmatizer(): + pytest.importorskip("pymorphy3") + pytest.importorskip("pymorphy3_dicts_uk") + return get_lang_class("uk")().add_pipe("lemmatizer") + + +@pytest.fixture +def uk_lookup_lemmatizer(): pytest.importorskip("pymorphy2") pytest.importorskip("pymorphy2_dicts_uk") - return get_lang_class("uk")().add_pipe("lemmatizer") + return get_lang_class("uk")().add_pipe( + "lemmatizer", config={"mode": "pymorphy2_lookup"} + ) @pytest.fixture(scope="session") diff --git a/spacy/tests/doc/test_doc_api.py b/spacy/tests/doc/test_doc_api.py index a64ab2ba8..38003dea9 100644 --- a/spacy/tests/doc/test_doc_api.py +++ b/spacy/tests/doc/test_doc_api.py @@ -82,6 +82,21 @@ def test_issue2396(en_vocab): assert (span.get_lca_matrix() == matrix).all() +@pytest.mark.issue(11499) +def test_init_args_unmodified(en_vocab): + words = ["A", "sentence"] + ents = ["B-TYPE1", ""] + sent_starts = [True, False] + Doc( + vocab=en_vocab, + words=words, + ents=ents, + sent_starts=sent_starts, + ) + assert ents == ["B-TYPE1", ""] + assert sent_starts == [True, False] + + @pytest.mark.parametrize("text", ["-0.23", "+123,456", "±1"]) @pytest.mark.parametrize("lang_cls", [English, MultiLanguage]) @pytest.mark.issue(2782) diff --git a/spacy/tests/doc/test_json_doc_conversion.py b/spacy/tests/doc/test_json_doc_conversion.py index 0d7c061c9..19698cfb2 100644 --- a/spacy/tests/doc/test_json_doc_conversion.py +++ b/spacy/tests/doc/test_json_doc_conversion.py @@ -128,7 +128,9 @@ def test_doc_to_json_with_token_span_attributes(doc): doc._.json_test1 = "hello world" doc._.json_test2 = [1, 2, 3] doc[0:1]._.span_test = "span_attribute" + doc[0:2]._.span_test = "span_attribute_2" doc[0]._.token_test = 117 + doc[1]._.token_test = 118 doc.spans["span_group"] = [doc[0:1]] json_doc = doc.to_json( underscore=["json_test1", "json_test2", "token_test", "span_test"] @@ -139,8 +141,10 @@ def test_doc_to_json_with_token_span_attributes(doc): assert json_doc["_"]["json_test2"] == [1, 2, 3] assert "underscore_token" in json_doc assert "underscore_span" in json_doc - assert json_doc["underscore_token"]["token_test"]["value"] == 117 - assert json_doc["underscore_span"]["span_test"]["value"] == "span_attribute" + assert json_doc["underscore_token"]["token_test"][0]["value"] == 117 + assert json_doc["underscore_token"]["token_test"][1]["value"] == 118 + assert json_doc["underscore_span"]["span_test"][0]["value"] == "span_attribute" + assert json_doc["underscore_span"]["span_test"][1]["value"] == "span_attribute_2" assert len(schemas.validate(schemas.DocJSONSchema, json_doc)) == 0 assert srsly.json_loads(srsly.json_dumps(json_doc)) == json_doc @@ -161,8 +165,8 @@ def test_doc_to_json_with_custom_user_data(doc): assert json_doc["_"]["json_test"] == "hello world" assert "underscore_token" in json_doc assert "underscore_span" in json_doc - assert json_doc["underscore_token"]["token_test"]["value"] == 117 - assert json_doc["underscore_span"]["span_test"]["value"] == "span_attribute" + assert json_doc["underscore_token"]["token_test"][0]["value"] == 117 + assert json_doc["underscore_span"]["span_test"][0]["value"] == "span_attribute" assert len(schemas.validate(schemas.DocJSONSchema, json_doc)) == 0 assert srsly.json_loads(srsly.json_dumps(json_doc)) == json_doc @@ -181,8 +185,8 @@ def test_doc_to_json_with_token_span_same_identifier(doc): assert json_doc["_"]["my_ext"] == "hello world" assert "underscore_token" in json_doc assert "underscore_span" in json_doc - assert json_doc["underscore_token"]["my_ext"]["value"] == 117 - assert json_doc["underscore_span"]["my_ext"]["value"] == "span_attribute" + assert json_doc["underscore_token"]["my_ext"][0]["value"] == 117 + assert json_doc["underscore_span"]["my_ext"][0]["value"] == "span_attribute" assert len(schemas.validate(schemas.DocJSONSchema, json_doc)) == 0 assert srsly.json_loads(srsly.json_dumps(json_doc)) == json_doc @@ -195,10 +199,9 @@ def test_doc_to_json_with_token_attributes_missing(doc): doc[0]._.token_test = 117 json_doc = doc.to_json(underscore=["span_test"]) - assert "underscore_token" in json_doc assert "underscore_span" in json_doc - assert json_doc["underscore_span"]["span_test"]["value"] == "span_attribute" - assert "token_test" not in json_doc["underscore_token"] + assert json_doc["underscore_span"]["span_test"][0]["value"] == "span_attribute" + assert "underscore_token" not in json_doc assert len(schemas.validate(schemas.DocJSONSchema, json_doc)) == 0 @@ -283,7 +286,9 @@ def test_json_to_doc_with_token_span_attributes(doc): doc._.json_test1 = "hello world" doc._.json_test2 = [1, 2, 3] doc[0:1]._.span_test = "span_attribute" + doc[0:2]._.span_test = "span_attribute_2" doc[0]._.token_test = 117 + doc[1]._.token_test = 118 json_doc = doc.to_json( underscore=["json_test1", "json_test2", "token_test", "span_test"] @@ -295,7 +300,9 @@ def test_json_to_doc_with_token_span_attributes(doc): assert new_doc._.json_test1 == "hello world" assert new_doc._.json_test2 == [1, 2, 3] assert new_doc[0]._.token_test == 117 + assert new_doc[1]._.token_test == 118 assert new_doc[0:1]._.span_test == "span_attribute" + assert new_doc[0:2]._.span_test == "span_attribute_2" assert new_doc.user_data == doc.user_data assert new_doc.to_bytes(exclude=["user_data"]) == doc.to_bytes( exclude=["user_data"] diff --git a/spacy/tests/lang/grc/test_tokenizer.py b/spacy/tests/lang/grc/test_tokenizer.py new file mode 100644 index 000000000..3df5b546b --- /dev/null +++ b/spacy/tests/lang/grc/test_tokenizer.py @@ -0,0 +1,18 @@ +import pytest + + +# fmt: off +GRC_TOKEN_EXCEPTION_TESTS = [ + ("τὸ 〈τῆς〉 φιλοσοφίας ἔργον ἔνιοί φασιν ἀπὸ ⟦βαρβάρων⟧ ἄρξαι.", ["τὸ", "〈", "τῆς", "〉", "φιλοσοφίας", "ἔργον", "ἔνιοί", "φασιν", "ἀπὸ", "⟦", "βαρβάρων", "⟧", "ἄρξαι", "."]), + ("τὴν δὲ τῶν Αἰγυπτίων φιλοσοφίαν εἶναι τοιαύτην περί τε †θεῶν† καὶ ὑπὲρ δικαιοσύνης.", ["τὴν", "δὲ", "τῶν", "Αἰγυπτίων", "φιλοσοφίαν", "εἶναι", "τοιαύτην", "περί", "τε", "†", "θεῶν", "†", "καὶ", "ὑπὲρ", "δικαιοσύνης", "."]), + ("⸏πόσις δ' Ἐρεχθεύς ἐστί μοι σεσωσμένος⸏", ["⸏", "πόσις", "δ'", "Ἐρεχθεύς", "ἐστί", "μοι", "σεσωσμένος", "⸏"]), + ("⸏ὔπνον ἴδωμεν⸎", ["⸏", "ὔπνον", "ἴδωμεν", "⸎"]), +] +# fmt: on + + +@pytest.mark.parametrize("text,expected_tokens", GRC_TOKEN_EXCEPTION_TESTS) +def test_grc_tokenizer(grc_tokenizer, text, expected_tokens): + tokens = grc_tokenizer(text) + token_list = [token.text for token in tokens if not token.is_space] + assert expected_tokens == token_list diff --git a/spacy/tests/lang/ru/test_lemmatizer.py b/spacy/tests/lang/ru/test_lemmatizer.py index 9ca7f441b..e82fd4f8c 100644 --- a/spacy/tests/lang/ru/test_lemmatizer.py +++ b/spacy/tests/lang/ru/test_lemmatizer.py @@ -78,3 +78,17 @@ def test_ru_lemmatizer_punct(ru_lemmatizer): assert ru_lemmatizer.pymorphy2_lemmatize(doc[0]) == ['"'] doc = Doc(ru_lemmatizer.vocab, words=["»"], pos=["PUNCT"]) assert ru_lemmatizer.pymorphy2_lemmatize(doc[0]) == ['"'] + + +def test_ru_doc_lookup_lemmatization(ru_lookup_lemmatizer): + words = ["мама", "мыла", "раму"] + pos = ["NOUN", "VERB", "NOUN"] + morphs = [ + "Animacy=Anim|Case=Nom|Gender=Fem|Number=Sing", + "Aspect=Imp|Gender=Fem|Mood=Ind|Number=Sing|Tense=Past|VerbForm=Fin|Voice=Act", + "Animacy=Anim|Case=Acc|Gender=Fem|Number=Sing", + ] + doc = Doc(ru_lookup_lemmatizer.vocab, words=words, pos=pos, morphs=morphs) + doc = ru_lookup_lemmatizer(doc) + lemmas = [token.lemma_ for token in doc] + assert lemmas == ["мама", "мыла", "раму"] diff --git a/spacy/tests/lang/sl/test_text.py b/spacy/tests/lang/sl/test_text.py index ddc5b6b5d..a2a932077 100644 --- a/spacy/tests/lang/sl/test_text.py +++ b/spacy/tests/lang/sl/test_text.py @@ -20,7 +20,6 @@ od katerih so te svoboščine odvisne, assert len(tokens) == 116 -@pytest.mark.xfail def test_ordinal_number(sl_tokenizer): text = "10. decembra 1948" tokens = sl_tokenizer(text) diff --git a/spacy/tests/lang/uk/test_lemmatizer.py b/spacy/tests/lang/uk/test_lemmatizer.py index 57dd4198a..788744aa1 100644 --- a/spacy/tests/lang/uk/test_lemmatizer.py +++ b/spacy/tests/lang/uk/test_lemmatizer.py @@ -9,3 +9,11 @@ def test_uk_lemmatizer(uk_lemmatizer): """Check that the default uk lemmatizer runs.""" doc = Doc(uk_lemmatizer.vocab, words=["a", "b", "c"]) uk_lemmatizer(doc) + assert [token.lemma for token in doc] + + +def test_uk_lookup_lemmatizer(uk_lookup_lemmatizer): + """Check that the lookup uk lemmatizer runs.""" + doc = Doc(uk_lookup_lemmatizer.vocab, words=["a", "b", "c"]) + uk_lookup_lemmatizer(doc) + assert [token.lemma for token in doc] diff --git a/spacy/tests/pipeline/test_entity_linker.py b/spacy/tests/pipeline/test_entity_linker.py index 82bc976bb..4d683acc5 100644 --- a/spacy/tests/pipeline/test_entity_linker.py +++ b/spacy/tests/pipeline/test_entity_linker.py @@ -6,7 +6,7 @@ from numpy.testing import assert_equal from spacy import registry, util from spacy.attrs import ENT_KB_ID from spacy.compat import pickle -from spacy.kb import Candidate, KnowledgeBase, get_candidates +from spacy.kb import Candidate, InMemoryLookupKB, get_candidates, KnowledgeBase from spacy.lang.en import English from spacy.ml import load_kb from spacy.pipeline import EntityLinker @@ -34,7 +34,7 @@ def assert_almost_equal(a, b): def test_issue4674(): """Test that setting entities with overlapping identifiers does not mess up IO""" nlp = English() - kb = KnowledgeBase(nlp.vocab, entity_vector_length=3) + kb = InMemoryLookupKB(nlp.vocab, entity_vector_length=3) vector1 = [0.9, 1.1, 1.01] vector2 = [1.8, 2.25, 2.01] with pytest.warns(UserWarning): @@ -51,7 +51,7 @@ def test_issue4674(): dir_path.mkdir() file_path = dir_path / "kb" kb.to_disk(str(file_path)) - kb2 = KnowledgeBase(nlp.vocab, entity_vector_length=3) + kb2 = InMemoryLookupKB(nlp.vocab, entity_vector_length=3) kb2.from_disk(str(file_path)) assert kb2.get_size_entities() == 1 @@ -59,9 +59,9 @@ def test_issue4674(): @pytest.mark.issue(6730) def test_issue6730(en_vocab): """Ensure that the KB does not accept empty strings, but otherwise IO works fine.""" - from spacy.kb import KnowledgeBase + from spacy.kb.kb_in_memory import InMemoryLookupKB - kb = KnowledgeBase(en_vocab, entity_vector_length=3) + kb = InMemoryLookupKB(en_vocab, entity_vector_length=3) kb.add_entity(entity="1", freq=148, entity_vector=[1, 2, 3]) with pytest.raises(ValueError): @@ -127,7 +127,7 @@ def test_issue7065_b(): def create_kb(vocab): # create artificial KB - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q270853", freq=12, entity_vector=[9, 1, -7]) mykb.add_alias( alias="No. 8", @@ -190,7 +190,7 @@ def test_no_entities(): def create_kb(vocab): # create artificial KB - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) mykb.add_alias("Russ Cochran", ["Q2146908"], [0.9]) return mykb @@ -231,7 +231,7 @@ def test_partial_links(): def create_kb(vocab): # create artificial KB - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) mykb.add_alias("Russ Cochran", ["Q2146908"], [0.9]) return mykb @@ -263,7 +263,7 @@ def test_partial_links(): def test_kb_valid_entities(nlp): """Test the valid construction of a KB with 3 entities and two aliases""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=3) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=3) # adding entities mykb.add_entity(entity="Q1", freq=19, entity_vector=[8, 4, 3]) @@ -292,7 +292,7 @@ def test_kb_valid_entities(nlp): def test_kb_invalid_entities(nlp): """Test the invalid construction of a KB with an alias linked to a non-existing entity""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) # adding entities mykb.add_entity(entity="Q1", freq=19, entity_vector=[1]) @@ -308,7 +308,7 @@ def test_kb_invalid_entities(nlp): def test_kb_invalid_probabilities(nlp): """Test the invalid construction of a KB with wrong prior probabilities""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) # adding entities mykb.add_entity(entity="Q1", freq=19, entity_vector=[1]) @@ -322,7 +322,7 @@ def test_kb_invalid_probabilities(nlp): def test_kb_invalid_combination(nlp): """Test the invalid construction of a KB with non-matching entity and probability lists""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) # adding entities mykb.add_entity(entity="Q1", freq=19, entity_vector=[1]) @@ -338,7 +338,7 @@ def test_kb_invalid_combination(nlp): def test_kb_invalid_entity_vector(nlp): """Test the invalid construction of a KB with non-matching entity vector lengths""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=3) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=3) # adding entities mykb.add_entity(entity="Q1", freq=19, entity_vector=[1, 2, 3]) @@ -376,7 +376,7 @@ def test_kb_initialize_empty(nlp): def test_kb_serialize(nlp): """Test serialization of the KB""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) with make_tempdir() as d: # normal read-write behaviour mykb.to_disk(d / "kb") @@ -393,12 +393,12 @@ def test_kb_serialize(nlp): @pytest.mark.issue(9137) def test_kb_serialize_2(nlp): v = [5, 6, 7, 8] - kb1 = KnowledgeBase(vocab=nlp.vocab, entity_vector_length=4) + kb1 = InMemoryLookupKB(vocab=nlp.vocab, entity_vector_length=4) kb1.set_entities(["E1"], [1], [v]) assert kb1.get_vector("E1") == v with make_tempdir() as d: kb1.to_disk(d / "kb") - kb2 = KnowledgeBase(vocab=nlp.vocab, entity_vector_length=4) + kb2 = InMemoryLookupKB(vocab=nlp.vocab, entity_vector_length=4) kb2.from_disk(d / "kb") assert kb2.get_vector("E1") == v @@ -408,7 +408,7 @@ def test_kb_set_entities(nlp): v = [5, 6, 7, 8] v1 = [1, 1, 1, 0] v2 = [2, 2, 2, 3] - kb1 = KnowledgeBase(vocab=nlp.vocab, entity_vector_length=4) + kb1 = InMemoryLookupKB(vocab=nlp.vocab, entity_vector_length=4) kb1.set_entities(["E0"], [1], [v]) assert kb1.get_entity_strings() == ["E0"] kb1.set_entities(["E1", "E2"], [1, 9], [v1, v2]) @@ -417,7 +417,7 @@ def test_kb_set_entities(nlp): assert kb1.get_vector("E2") == v2 with make_tempdir() as d: kb1.to_disk(d / "kb") - kb2 = KnowledgeBase(vocab=nlp.vocab, entity_vector_length=4) + kb2 = InMemoryLookupKB(vocab=nlp.vocab, entity_vector_length=4) kb2.from_disk(d / "kb") assert set(kb2.get_entity_strings()) == {"E1", "E2"} assert kb2.get_vector("E1") == v1 @@ -428,7 +428,7 @@ def test_kb_serialize_vocab(nlp): """Test serialization of the KB and custom strings""" entity = "MyFunnyID" assert entity not in nlp.vocab.strings - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) assert not mykb.contains_entity(entity) mykb.add_entity(entity, freq=342, entity_vector=[3]) assert mykb.contains_entity(entity) @@ -436,14 +436,14 @@ def test_kb_serialize_vocab(nlp): with make_tempdir() as d: # normal read-write behaviour mykb.to_disk(d / "kb") - mykb_new = KnowledgeBase(Vocab(), entity_vector_length=1) + mykb_new = InMemoryLookupKB(Vocab(), entity_vector_length=1) mykb_new.from_disk(d / "kb") assert entity in mykb_new.vocab.strings def test_candidate_generation(nlp): """Test correct candidate generation""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) doc = nlp("douglas adam Adam shrubbery") douglas_ent = doc[0:1] @@ -481,7 +481,7 @@ def test_el_pipe_configuration(nlp): ruler.add_patterns([pattern]) def create_kb(vocab): - kb = KnowledgeBase(vocab, entity_vector_length=1) + kb = InMemoryLookupKB(vocab, entity_vector_length=1) kb.add_entity(entity="Q2", freq=12, entity_vector=[2]) kb.add_entity(entity="Q3", freq=5, entity_vector=[3]) kb.add_alias(alias="douglas", entities=["Q2", "Q3"], probabilities=[0.8, 0.1]) @@ -500,10 +500,21 @@ def test_el_pipe_configuration(nlp): def get_lowercased_candidates(kb, span): return kb.get_alias_candidates(span.text.lower()) + def get_lowercased_candidates_batch(kb, spans): + return [get_lowercased_candidates(kb, span) for span in spans] + @registry.misc("spacy.LowercaseCandidateGenerator.v1") - def create_candidates() -> Callable[[KnowledgeBase, "Span"], Iterable[Candidate]]: + def create_candidates() -> Callable[ + [InMemoryLookupKB, "Span"], Iterable[Candidate] + ]: return get_lowercased_candidates + @registry.misc("spacy.LowercaseCandidateBatchGenerator.v1") + def create_candidates_batch() -> Callable[ + [InMemoryLookupKB, Iterable["Span"]], Iterable[Iterable[Candidate]] + ]: + return get_lowercased_candidates_batch + # replace the pipe with a new one with with a different candidate generator entity_linker = nlp.replace_pipe( "entity_linker", @@ -511,6 +522,9 @@ def test_el_pipe_configuration(nlp): config={ "incl_context": False, "get_candidates": {"@misc": "spacy.LowercaseCandidateGenerator.v1"}, + "get_candidates_batch": { + "@misc": "spacy.LowercaseCandidateBatchGenerator.v1" + }, }, ) entity_linker.set_kb(create_kb) @@ -532,7 +546,7 @@ def test_nel_nsents(nlp): def test_vocab_serialization(nlp): """Test that string information is retained across storage""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) # adding entities mykb.add_entity(entity="Q1", freq=27, entity_vector=[1]) @@ -552,7 +566,7 @@ def test_vocab_serialization(nlp): with make_tempdir() as d: mykb.to_disk(d / "kb") - kb_new_vocab = KnowledgeBase(Vocab(), entity_vector_length=1) + kb_new_vocab = InMemoryLookupKB(Vocab(), entity_vector_length=1) kb_new_vocab.from_disk(d / "kb") candidates = kb_new_vocab.get_alias_candidates("adam") @@ -568,7 +582,7 @@ def test_vocab_serialization(nlp): def test_append_alias(nlp): """Test that we can append additional alias-entity pairs""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) # adding entities mykb.add_entity(entity="Q1", freq=27, entity_vector=[1]) @@ -599,7 +613,7 @@ def test_append_alias(nlp): @pytest.mark.filterwarnings("ignore:\\[W036") def test_append_invalid_alias(nlp): """Test that append an alias will throw an error if prior probs are exceeding 1""" - mykb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + mykb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) # adding entities mykb.add_entity(entity="Q1", freq=27, entity_vector=[1]) @@ -621,7 +635,7 @@ def test_preserving_links_asdoc(nlp): vector_length = 1 def create_kb(vocab): - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) # adding entities mykb.add_entity(entity="Q1", freq=19, entity_vector=[1]) mykb.add_entity(entity="Q2", freq=8, entity_vector=[1]) @@ -723,7 +737,7 @@ def test_overfitting_IO(): # create artificial KB - assign same prior weight to the two russ cochran's # Q2146908 (Russ Cochran): American golfer # Q7381115 (Russ Cochran): publisher - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) mykb.add_entity(entity="Q7381115", freq=12, entity_vector=[9, 1, -7]) mykb.add_alias( @@ -805,7 +819,7 @@ def test_kb_serialization(): kb_dir = tmp_dir / "kb" nlp1 = English() assert "Q2146908" not in nlp1.vocab.strings - mykb = KnowledgeBase(nlp1.vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(nlp1.vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) mykb.add_alias(alias="Russ Cochran", entities=["Q2146908"], probabilities=[0.8]) assert "Q2146908" in nlp1.vocab.strings @@ -828,7 +842,7 @@ def test_kb_serialization(): def test_kb_pickle(): # Test that the KB can be pickled nlp = English() - kb_1 = KnowledgeBase(nlp.vocab, entity_vector_length=3) + kb_1 = InMemoryLookupKB(nlp.vocab, entity_vector_length=3) kb_1.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) assert not kb_1.contains_alias("Russ Cochran") kb_1.add_alias(alias="Russ Cochran", entities=["Q2146908"], probabilities=[0.8]) @@ -842,7 +856,7 @@ def test_kb_pickle(): def test_nel_pickle(): # Test that a pipeline with an EL component can be pickled def create_kb(vocab): - kb = KnowledgeBase(vocab, entity_vector_length=3) + kb = InMemoryLookupKB(vocab, entity_vector_length=3) kb.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) kb.add_alias(alias="Russ Cochran", entities=["Q2146908"], probabilities=[0.8]) return kb @@ -864,7 +878,7 @@ def test_nel_pickle(): def test_kb_to_bytes(): # Test that the KB's to_bytes method works correctly nlp = English() - kb_1 = KnowledgeBase(nlp.vocab, entity_vector_length=3) + kb_1 = InMemoryLookupKB(nlp.vocab, entity_vector_length=3) kb_1.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) kb_1.add_entity(entity="Q66", freq=9, entity_vector=[1, 2, 3]) kb_1.add_alias(alias="Russ Cochran", entities=["Q2146908"], probabilities=[0.8]) @@ -874,7 +888,7 @@ def test_kb_to_bytes(): ) assert kb_1.contains_alias("Russ Cochran") kb_bytes = kb_1.to_bytes() - kb_2 = KnowledgeBase(nlp.vocab, entity_vector_length=3) + kb_2 = InMemoryLookupKB(nlp.vocab, entity_vector_length=3) assert not kb_2.contains_alias("Russ Cochran") kb_2 = kb_2.from_bytes(kb_bytes) # check that both KBs are exactly the same @@ -897,7 +911,7 @@ def test_kb_to_bytes(): def test_nel_to_bytes(): # Test that a pipeline with an EL component can be converted to bytes def create_kb(vocab): - kb = KnowledgeBase(vocab, entity_vector_length=3) + kb = InMemoryLookupKB(vocab, entity_vector_length=3) kb.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) kb.add_alias(alias="Russ Cochran", entities=["Q2146908"], probabilities=[0.8]) return kb @@ -987,7 +1001,7 @@ def test_legacy_architectures(name, config): train_examples.append(Example.from_dict(doc, annotation)) def create_kb(vocab): - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q2146908", freq=12, entity_vector=[6, -4, 3]) mykb.add_entity(entity="Q7381115", freq=12, entity_vector=[9, 1, -7]) mykb.add_alias( @@ -1054,7 +1068,7 @@ def test_no_gold_ents(patterns): def create_kb(vocab): # create artificial KB - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q613241", freq=12, entity_vector=[6, -4, 3]) mykb.add_alias("Kirby", ["Q613241"], [0.9]) # Placeholder @@ -1104,7 +1118,7 @@ def test_tokenization_mismatch(): def create_kb(vocab): # create placeholder KB - mykb = KnowledgeBase(vocab, entity_vector_length=vector_length) + mykb = InMemoryLookupKB(vocab, entity_vector_length=vector_length) mykb.add_entity(entity="Q613241", freq=12, entity_vector=[6, -4, 3]) mykb.add_alias("Kirby", ["Q613241"], [0.9]) return mykb @@ -1121,6 +1135,12 @@ def test_tokenization_mismatch(): nlp.evaluate(train_examples) +def test_abstract_kb_instantiation(): + """Test whether instantiation of abstract KB base class fails.""" + with pytest.raises(TypeError): + KnowledgeBase(None, 3) + + # fmt: off @pytest.mark.parametrize( "meet_threshold,config", @@ -1151,7 +1171,7 @@ def test_threshold(meet_threshold: bool, config: Dict[str, Any]): def create_kb(vocab): # create artificial KB - mykb = KnowledgeBase(vocab, entity_vector_length=3) + mykb = InMemoryLookupKB(vocab, entity_vector_length=3) mykb.add_entity(entity=entity_id, freq=12, entity_vector=[6, -4, 3]) mykb.add_alias( alias="Mahler", diff --git a/spacy/tests/pipeline/test_pipe_methods.py b/spacy/tests/pipeline/test_pipe_methods.py index b946061f6..14a7a36e5 100644 --- a/spacy/tests/pipeline/test_pipe_methods.py +++ b/spacy/tests/pipeline/test_pipe_methods.py @@ -605,10 +605,35 @@ def test_update_with_annotates(): assert results[component] == "" -def test_load_disable_enable() -> None: - """ - Tests spacy.load() with dis-/enabling components. - """ +@pytest.mark.issue(11443) +def test_enable_disable_conflict_with_config(): + """Test conflict between enable/disable w.r.t. `nlp.disabled` set in the config.""" + nlp = English() + nlp.add_pipe("tagger") + nlp.add_pipe("senter") + nlp.add_pipe("sentencizer") + + with make_tempdir() as tmp_dir: + nlp.to_disk(tmp_dir) + # Expected to fail, as config and arguments conflict. + with pytest.raises(ValueError): + spacy.load( + tmp_dir, enable=["tagger"], config={"nlp": {"disabled": ["senter"]}} + ) + # Expected to succeed without warning due to the lack of a conflicting config option. + spacy.load(tmp_dir, enable=["tagger"]) + # Expected to succeed with a warning, as disable=[] should override the config setting. + with pytest.warns(UserWarning): + spacy.load( + tmp_dir, + enable=["tagger"], + disable=[], + config={"nlp": {"disabled": ["senter"]}}, + ) + + +def test_load_disable_enable(): + """Tests spacy.load() with dis-/enabling components.""" base_nlp = English() for pipe in ("sentencizer", "tagger", "parser"): diff --git a/spacy/tests/pipeline/test_tok2vec.py b/spacy/tests/pipeline/test_tok2vec.py index 64faf133d..e423d9a19 100644 --- a/spacy/tests/pipeline/test_tok2vec.py +++ b/spacy/tests/pipeline/test_tok2vec.py @@ -230,6 +230,97 @@ def test_tok2vec_listener_callback(): assert get_dX(Y) is not None +def test_tok2vec_listener_overfitting(): + """Test that a pipeline with a listener properly overfits, even if 'tok2vec' is in the annotating components""" + orig_config = Config().from_str(cfg_string) + nlp = util.load_model_from_config(orig_config, auto_fill=True, validate=True) + train_examples = [] + for t in TRAIN_DATA: + train_examples.append(Example.from_dict(nlp.make_doc(t[0]), t[1])) + optimizer = nlp.initialize(get_examples=lambda: train_examples) + + for i in range(50): + losses = {} + nlp.update(train_examples, sgd=optimizer, losses=losses, annotates=["tok2vec"]) + assert losses["tagger"] < 0.00001 + + # test the trained model + test_text = "I like blue eggs" + doc = nlp(test_text) + assert doc[0].tag_ == "N" + assert doc[1].tag_ == "V" + assert doc[2].tag_ == "J" + assert doc[3].tag_ == "N" + + # Also test the results are still the same after IO + with make_tempdir() as tmp_dir: + nlp.to_disk(tmp_dir) + nlp2 = util.load_model_from_path(tmp_dir) + doc2 = nlp2(test_text) + assert doc2[0].tag_ == "N" + assert doc2[1].tag_ == "V" + assert doc2[2].tag_ == "J" + assert doc2[3].tag_ == "N" + + +def test_tok2vec_frozen_not_annotating(): + """Test that a pipeline with a frozen tok2vec raises an error when the tok2vec is not annotating""" + orig_config = Config().from_str(cfg_string) + nlp = util.load_model_from_config(orig_config, auto_fill=True, validate=True) + train_examples = [] + for t in TRAIN_DATA: + train_examples.append(Example.from_dict(nlp.make_doc(t[0]), t[1])) + optimizer = nlp.initialize(get_examples=lambda: train_examples) + + for i in range(2): + losses = {} + with pytest.raises( + ValueError, match=r"the tok2vec embedding layer is not updated" + ): + nlp.update( + train_examples, sgd=optimizer, losses=losses, exclude=["tok2vec"] + ) + + +def test_tok2vec_frozen_overfitting(): + """Test that a pipeline with a frozen & annotating tok2vec can still overfit""" + orig_config = Config().from_str(cfg_string) + nlp = util.load_model_from_config(orig_config, auto_fill=True, validate=True) + train_examples = [] + for t in TRAIN_DATA: + train_examples.append(Example.from_dict(nlp.make_doc(t[0]), t[1])) + optimizer = nlp.initialize(get_examples=lambda: train_examples) + + for i in range(100): + losses = {} + nlp.update( + train_examples, + sgd=optimizer, + losses=losses, + exclude=["tok2vec"], + annotates=["tok2vec"], + ) + assert losses["tagger"] < 0.0001 + + # test the trained model + test_text = "I like blue eggs" + doc = nlp(test_text) + assert doc[0].tag_ == "N" + assert doc[1].tag_ == "V" + assert doc[2].tag_ == "J" + assert doc[3].tag_ == "N" + + # Also test the results are still the same after IO + with make_tempdir() as tmp_dir: + nlp.to_disk(tmp_dir) + nlp2 = util.load_model_from_path(tmp_dir) + doc2 = nlp2(test_text) + assert doc2[0].tag_ == "N" + assert doc2[1].tag_ == "V" + assert doc2[2].tag_ == "J" + assert doc2[3].tag_ == "N" + + def test_replace_listeners(): orig_config = Config().from_str(cfg_string) nlp = util.load_model_from_config(orig_config, auto_fill=True, validate=True) diff --git a/spacy/tests/serialize/test_resource_warning.py b/spacy/tests/serialize/test_resource_warning.py index a00b2a688..38701c6d9 100644 --- a/spacy/tests/serialize/test_resource_warning.py +++ b/spacy/tests/serialize/test_resource_warning.py @@ -3,7 +3,7 @@ from unittest import TestCase import pytest import srsly from numpy import zeros -from spacy.kb import KnowledgeBase, Writer +from spacy.kb.kb_in_memory import InMemoryLookupKB, Writer from spacy.vectors import Vectors from spacy.language import Language from spacy.pipeline import TrainablePipe @@ -71,7 +71,7 @@ def entity_linker(): nlp = Language() def create_kb(vocab): - kb = KnowledgeBase(vocab, entity_vector_length=1) + kb = InMemoryLookupKB(vocab, entity_vector_length=1) kb.add_entity("test", 0.0, zeros((1, 1), dtype="f")) return kb @@ -120,7 +120,7 @@ def test_writer_with_path_py35(): def test_save_and_load_knowledge_base(): nlp = Language() - kb = KnowledgeBase(nlp.vocab, entity_vector_length=1) + kb = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) with make_tempdir() as d: path = d / "kb" try: @@ -129,7 +129,7 @@ def test_save_and_load_knowledge_base(): pytest.fail(str(e)) try: - kb_loaded = KnowledgeBase(nlp.vocab, entity_vector_length=1) + kb_loaded = InMemoryLookupKB(nlp.vocab, entity_vector_length=1) kb_loaded.from_disk(path) except Exception as e: pytest.fail(str(e)) diff --git a/spacy/tests/serialize/test_serialize_kb.py b/spacy/tests/serialize/test_serialize_kb.py index 1e0ae3c76..8d3653ab1 100644 --- a/spacy/tests/serialize/test_serialize_kb.py +++ b/spacy/tests/serialize/test_serialize_kb.py @@ -2,7 +2,7 @@ from typing import Callable from spacy import util from spacy.util import ensure_path, registry, load_model_from_config -from spacy.kb import KnowledgeBase +from spacy.kb.kb_in_memory import InMemoryLookupKB from spacy.vocab import Vocab from thinc.api import Config @@ -22,7 +22,7 @@ def test_serialize_kb_disk(en_vocab): dir_path.mkdir() file_path = dir_path / "kb" kb1.to_disk(str(file_path)) - kb2 = KnowledgeBase(vocab=en_vocab, entity_vector_length=3) + kb2 = InMemoryLookupKB(vocab=en_vocab, entity_vector_length=3) kb2.from_disk(str(file_path)) # final assertions @@ -30,7 +30,7 @@ def test_serialize_kb_disk(en_vocab): def _get_dummy_kb(vocab): - kb = KnowledgeBase(vocab, entity_vector_length=3) + kb = InMemoryLookupKB(vocab, entity_vector_length=3) kb.add_entity(entity="Q53", freq=33, entity_vector=[0, 5, 3]) kb.add_entity(entity="Q17", freq=2, entity_vector=[7, 1, 0]) kb.add_entity(entity="Q007", freq=7, entity_vector=[0, 0, 7]) @@ -104,7 +104,7 @@ def test_serialize_subclassed_kb(): custom_field = 666 """ - class SubKnowledgeBase(KnowledgeBase): + class SubInMemoryLookupKB(InMemoryLookupKB): def __init__(self, vocab, entity_vector_length, custom_field): super().__init__(vocab, entity_vector_length) self.custom_field = custom_field @@ -112,9 +112,9 @@ def test_serialize_subclassed_kb(): @registry.misc("spacy.CustomKB.v1") def custom_kb( entity_vector_length: int, custom_field: int - ) -> Callable[[Vocab], KnowledgeBase]: + ) -> Callable[[Vocab], InMemoryLookupKB]: def custom_kb_factory(vocab): - kb = SubKnowledgeBase( + kb = SubInMemoryLookupKB( vocab=vocab, entity_vector_length=entity_vector_length, custom_field=custom_field, @@ -129,7 +129,7 @@ def test_serialize_subclassed_kb(): nlp.initialize() entity_linker = nlp.get_pipe("entity_linker") - assert type(entity_linker.kb) == SubKnowledgeBase + assert type(entity_linker.kb) == SubInMemoryLookupKB assert entity_linker.kb.entity_vector_length == 342 assert entity_linker.kb.custom_field == 666 @@ -139,6 +139,6 @@ def test_serialize_subclassed_kb(): nlp2 = util.load_model_from_path(tmp_dir) entity_linker2 = nlp2.get_pipe("entity_linker") # After IO, the KB is the standard one - assert type(entity_linker2.kb) == KnowledgeBase + assert type(entity_linker2.kb) == InMemoryLookupKB assert entity_linker2.kb.entity_vector_length == 342 assert not hasattr(entity_linker2.kb, "custom_field") diff --git a/spacy/tests/serialize/test_serialize_pipeline.py b/spacy/tests/serialize/test_serialize_pipeline.py index 9fcf18e2d..b948bb76c 100644 --- a/spacy/tests/serialize/test_serialize_pipeline.py +++ b/spacy/tests/serialize/test_serialize_pipeline.py @@ -404,10 +404,11 @@ def test_serialize_pipeline_disable_enable(): assert nlp3.component_names == ["ner", "tagger"] with make_tempdir() as d: nlp3.to_disk(d) - nlp4 = spacy.load(d, disable=["ner"]) - assert nlp4.pipe_names == [] + with pytest.warns(UserWarning): + nlp4 = spacy.load(d, disable=["ner"]) + assert nlp4.pipe_names == ["tagger"] assert nlp4.component_names == ["ner", "tagger"] - assert nlp4.disabled == ["ner", "tagger"] + assert nlp4.disabled == ["ner"] with make_tempdir() as d: nlp.to_disk(d) nlp5 = spacy.load(d, exclude=["tagger"]) diff --git a/spacy/tests/test_models.py b/spacy/tests/test_models.py index 2306cabb7..d91ed1201 100644 --- a/spacy/tests/test_models.py +++ b/spacy/tests/test_models.py @@ -23,7 +23,7 @@ def get_textcat_bow_kwargs(): def get_textcat_cnn_kwargs(): - return {"tok2vec": test_tok2vec(), "exclusive_classes": False, "nO": 13} + return {"tok2vec": make_test_tok2vec(), "exclusive_classes": False, "nO": 13} def get_all_params(model): @@ -65,7 +65,7 @@ def get_tok2vec_kwargs(): } -def test_tok2vec(): +def make_test_tok2vec(): return build_Tok2Vec_model(**get_tok2vec_kwargs()) diff --git a/spacy/tests/training/test_augmenters.py b/spacy/tests/training/test_augmenters.py index e3639c5da..35860a199 100644 --- a/spacy/tests/training/test_augmenters.py +++ b/spacy/tests/training/test_augmenters.py @@ -31,7 +31,7 @@ def doc(nlp): words = ["Sarah", "'s", "sister", "flew", "to", "Silicon", "Valley", "via", "London", "."] tags = ["NNP", "POS", "NN", "VBD", "IN", "NNP", "NNP", "IN", "NNP", "."] pos = ["PROPN", "PART", "NOUN", "VERB", "ADP", "PROPN", "PROPN", "ADP", "PROPN", "PUNCT"] - ents = ["B-PERSON", "I-PERSON", "O", "O", "O", "B-LOC", "I-LOC", "O", "B-GPE", "O"] + ents = ["B-PERSON", "I-PERSON", "O", "", "O", "B-LOC", "I-LOC", "O", "B-GPE", "O"] cats = {"TRAVEL": 1.0, "BAKING": 0.0} # fmt: on doc = Doc(nlp.vocab, words=words, tags=tags, pos=pos, ents=ents) @@ -106,6 +106,7 @@ def test_lowercase_augmenter(nlp, doc): assert [(e.start, e.end, e.label) for e in eg.reference.ents] == ents for ref_ent, orig_ent in zip(eg.reference.ents, doc.ents): assert ref_ent.text == orig_ent.text.lower() + assert [t.ent_iob for t in doc] == [t.ent_iob for t in eg.reference] assert [t.pos_ for t in eg.reference] == [t.pos_ for t in doc] # check that augmentation works when lowercasing leads to different @@ -166,7 +167,7 @@ def test_make_whitespace_variant(nlp): lemmas = ["they", "fly", "to", "New", "York", "City", ".", "\n", "then", "they", "drive", "to", "Washington", ",", "D.C."] heads = [1, 1, 1, 4, 5, 2, 1, 10, 10, 10, 10, 10, 11, 12, 12] deps = ["nsubj", "ROOT", "prep", "compound", "compound", "pobj", "punct", "dep", "advmod", "nsubj", "ROOT", "prep", "pobj", "punct", "appos"] - ents = ["O", "O", "O", "B-GPE", "I-GPE", "I-GPE", "O", "O", "O", "O", "O", "O", "B-GPE", "O", "B-GPE"] + ents = ["O", "", "O", "B-GPE", "I-GPE", "I-GPE", "O", "O", "O", "O", "O", "O", "B-GPE", "O", "B-GPE"] # fmt: on doc = Doc( nlp.vocab, @@ -215,6 +216,8 @@ def test_make_whitespace_variant(nlp): assert mod_ex2.reference[j].head.i == j - 1 # entities are well-formed assert len(doc.ents) == len(mod_ex.reference.ents) + # there is one token with missing entity information + assert any(t.ent_iob == 0 for t in mod_ex.reference) for ent in mod_ex.reference.ents: assert not ent[0].is_space assert not ent[-1].is_space diff --git a/spacy/tokens/_dict_proxies.py b/spacy/tokens/_dict_proxies.py index 9630da261..6edcce13d 100644 --- a/spacy/tokens/_dict_proxies.py +++ b/spacy/tokens/_dict_proxies.py @@ -42,7 +42,8 @@ class SpanGroups(UserDict): def copy(self, doc: Optional["Doc"] = None) -> "SpanGroups": if doc is None: doc = self._ensure_doc() - return SpanGroups(doc).from_bytes(self.to_bytes()) + data_copy = ((k, v.copy(doc=doc)) for k, v in self.items()) + return SpanGroups(doc, items=data_copy) def setdefault(self, key, default=None): if not isinstance(default, SpanGroup): diff --git a/spacy/tokens/doc.pyi b/spacy/tokens/doc.pyi index a40fa74aa..f0cdaee87 100644 --- a/spacy/tokens/doc.pyi +++ b/spacy/tokens/doc.pyi @@ -72,7 +72,7 @@ class Doc: lemmas: Optional[List[str]] = ..., heads: Optional[List[int]] = ..., deps: Optional[List[str]] = ..., - sent_starts: Optional[List[Union[bool, None]]] = ..., + sent_starts: Optional[List[Union[bool, int, None]]] = ..., ents: Optional[List[str]] = ..., ) -> None: ... @property diff --git a/spacy/tokens/doc.pyx b/spacy/tokens/doc.pyx index 7ba9a3341..295f91c28 100644 --- a/spacy/tokens/doc.pyx +++ b/spacy/tokens/doc.pyx @@ -217,9 +217,9 @@ cdef class Doc: head in the doc. Defaults to None. deps (Optional[List[str]]): A list of unicode strings, of the same length as words, to assign as token.dep. Defaults to None. - sent_starts (Optional[List[Union[bool, None]]]): A list of values, of - the same length as words, to assign as token.is_sent_start. Will be - overridden by heads if heads is provided. Defaults to None. + sent_starts (Optional[List[Union[bool, int, None]]]): A list of values, + of the same length as words, to assign as token.is_sent_start. Will + be overridden by heads if heads is provided. Defaults to None. ents (Optional[List[str]]): A list of unicode strings, of the same length as words, as IOB tags to assign as token.ent_iob and token.ent_type. Defaults to None. @@ -285,6 +285,7 @@ cdef class Doc: heads = [0] * len(deps) if heads and not deps: raise ValueError(Errors.E1017) + sent_starts = list(sent_starts) if sent_starts is not None else None if sent_starts is not None: for i in range(len(sent_starts)): if sent_starts[i] is True: @@ -300,12 +301,11 @@ cdef class Doc: ent_iobs = None ent_types = None if ents is not None: + ents = [ent if ent != "" else None for ent in ents] iob_strings = Token.iob_strings() # make valid IOB2 out of IOB1 or IOB2 for i, ent in enumerate(ents): - if ent is "": - ents[i] = None - elif ent is not None and not isinstance(ent, str): + if ent is not None and not isinstance(ent, str): raise ValueError(Errors.E177.format(tag=ent)) if i < len(ents) - 1: # OI -> OB @@ -1608,24 +1608,20 @@ cdef class Doc: Doc.set_extension(attr) self._.set(attr, doc_json["_"][attr]) - if doc_json.get("underscore_token", {}): - for token_attr in doc_json["underscore_token"]: - token_start = doc_json["underscore_token"][token_attr]["token_start"] - value = doc_json["underscore_token"][token_attr]["value"] - - if not Token.has_extension(token_attr): - Token.set_extension(token_attr) - self[token_start]._.set(token_attr, value) + for token_attr in doc_json.get("underscore_token", {}): + if not Token.has_extension(token_attr): + Token.set_extension(token_attr) + for token_data in doc_json["underscore_token"][token_attr]: + start = token_by_char(self.c, self.length, token_data["start"]) + value = token_data["value"] + self[start]._.set(token_attr, value) - if doc_json.get("underscore_span", {}): - for span_attr in doc_json["underscore_span"]: - token_start = doc_json["underscore_span"][span_attr]["token_start"] - token_end = doc_json["underscore_span"][span_attr]["token_end"] - value = doc_json["underscore_span"][span_attr]["value"] - - if not Span.has_extension(span_attr): - Span.set_extension(span_attr) - self[token_start:token_end]._.set(span_attr, value) + for span_attr in doc_json.get("underscore_span", {}): + if not Span.has_extension(span_attr): + Span.set_extension(span_attr) + for span_data in doc_json["underscore_span"][span_attr]: + value = span_data["value"] + self.char_span(span_data["start"], span_data["end"])._.set(span_attr, value) return self def to_json(self, underscore=None): @@ -1673,30 +1669,34 @@ cdef class Doc: if underscore: user_keys = set() if self.user_data: - data["_"] = {} - data["underscore_token"] = {} - data["underscore_span"] = {} - for data_key in self.user_data: + for data_key, value in self.user_data.copy().items(): if type(data_key) == tuple and len(data_key) >= 4 and data_key[0] == "._.": attr = data_key[1] start = data_key[2] end = data_key[3] if attr in underscore: user_keys.add(attr) - value = self.user_data[data_key] if not srsly.is_json_serializable(value): raise ValueError(Errors.E107.format(attr=attr, value=repr(value))) # Check if doc attribute if start is None: + if "_" not in data: + data["_"] = {} data["_"][attr] = value # Check if token attribute elif end is None: + if "underscore_token" not in data: + data["underscore_token"] = {} if attr not in data["underscore_token"]: - data["underscore_token"][attr] = {"token_start": start, "value": value} + data["underscore_token"][attr] = [] + data["underscore_token"][attr].append({"start": start, "value": value}) # Else span attribute else: + if "underscore_span" not in data: + data["underscore_span"] = {} if attr not in data["underscore_span"]: - data["underscore_span"][attr] = {"token_start": start, "token_end": end, "value": value} + data["underscore_span"][attr] = [] + data["underscore_span"][attr].append({"start": start, "end": end, "value": value}) for attr in underscore: if attr not in user_keys: diff --git a/spacy/tokens/span_group.pyi b/spacy/tokens/span_group.pyi index 245eb4dbe..21cd124ab 100644 --- a/spacy/tokens/span_group.pyi +++ b/spacy/tokens/span_group.pyi @@ -1,4 +1,4 @@ -from typing import Any, Dict, Iterable +from typing import Any, Dict, Iterable, Optional from .doc import Doc from .span import Span @@ -24,4 +24,4 @@ class SpanGroup: def __getitem__(self, i: int) -> Span: ... def to_bytes(self) -> bytes: ... def from_bytes(self, bytes_data: bytes) -> SpanGroup: ... - def copy(self) -> SpanGroup: ... + def copy(self, doc: Optional[Doc] = ...) -> SpanGroup: ... diff --git a/spacy/tokens/span_group.pyx b/spacy/tokens/span_group.pyx index bb0fab24f..1aa3c0bc8 100644 --- a/spacy/tokens/span_group.pyx +++ b/spacy/tokens/span_group.pyx @@ -241,15 +241,18 @@ cdef class SpanGroup: cdef void push_back(self, SpanC span) nogil: self.c.push_back(span) - def copy(self) -> SpanGroup: + def copy(self, doc: Optional["Doc"] = None) -> SpanGroup: """Clones the span group. + doc (Doc): New reference document to which the copy is bound. RETURNS (SpanGroup): A copy of the span group. DOCS: https://spacy.io/api/spangroup#copy """ + if doc is None: + doc = self.doc return SpanGroup( - self.doc, + doc, name=self.name, attrs=deepcopy(self.attrs), spans=list(self), diff --git a/spacy/training/augment.py b/spacy/training/augment.py index 55d780ba4..2fe8c24fb 100644 --- a/spacy/training/augment.py +++ b/spacy/training/augment.py @@ -6,7 +6,7 @@ from functools import partial from ..util import registry from .example import Example -from .iob_utils import split_bilu_label +from .iob_utils import split_bilu_label, _doc_to_biluo_tags_with_partial if TYPE_CHECKING: from ..language import Language # noqa: F401 @@ -62,6 +62,9 @@ def combined_augmenter( if orth_variants and random.random() < orth_level: raw_text = example.text orig_dict = example.to_dict() + orig_dict["doc_annotation"]["entities"] = _doc_to_biluo_tags_with_partial( + example.reference + ) variant_text, variant_token_annot = make_orth_variants( nlp, raw_text, @@ -128,6 +131,9 @@ def lower_casing_augmenter( def make_lowercase_variant(nlp: "Language", example: Example): example_dict = example.to_dict() + example_dict["doc_annotation"]["entities"] = _doc_to_biluo_tags_with_partial( + example.reference + ) doc = nlp.make_doc(example.text.lower()) example_dict["token_annotation"]["ORTH"] = [t.lower_ for t in example.reference] return example.from_dict(doc, example_dict) @@ -146,6 +152,9 @@ def orth_variants_augmenter( else: raw_text = example.text orig_dict = example.to_dict() + orig_dict["doc_annotation"]["entities"] = _doc_to_biluo_tags_with_partial( + example.reference + ) variant_text, variant_token_annot = make_orth_variants( nlp, raw_text, @@ -248,6 +257,9 @@ def make_whitespace_variant( RETURNS (Example): Example with one additional space token. """ example_dict = example.to_dict() + example_dict["doc_annotation"]["entities"] = _doc_to_biluo_tags_with_partial( + example.reference + ) doc_dict = example_dict.get("doc_annotation", {}) token_dict = example_dict.get("token_annotation", {}) # returned unmodified if: diff --git a/spacy/training/iob_utils.py b/spacy/training/iob_utils.py index 61f83a1c3..0d4d246b0 100644 --- a/spacy/training/iob_utils.py +++ b/spacy/training/iob_utils.py @@ -60,6 +60,14 @@ def doc_to_biluo_tags(doc: Doc, missing: str = "O"): ) +def _doc_to_biluo_tags_with_partial(doc: Doc) -> List[str]: + ents = doc_to_biluo_tags(doc, missing="-") + for i, token in enumerate(doc): + if token.ent_iob == 2: + ents[i] = "O" + return ents + + def offsets_to_biluo_tags( doc: Doc, entities: Iterable[Tuple[int, int, Union[str, int]]], missing: str = "O" ) -> List[str]: diff --git a/spacy/util.py b/spacy/util.py index 4e1a62d05..3034808ba 100644 --- a/spacy/util.py +++ b/spacy/util.py @@ -67,7 +67,6 @@ LEXEME_NORM_LANGS = ["cs", "da", "de", "el", "en", "id", "lb", "mk", "pt", "ru", CONFIG_SECTION_ORDER = ["paths", "variables", "system", "nlp", "components", "corpora", "training", "pretraining", "initialize"] # fmt: on - logger = logging.getLogger("spacy") logger_stream_handler = logging.StreamHandler() logger_stream_handler.setFormatter( @@ -394,13 +393,17 @@ def get_module_path(module: ModuleType) -> Path: return file_path.parent +# Default value for passed enable/disable values. +_DEFAULT_EMPTY_PIPES = SimpleFrozenList() + + def load_model( name: Union[str, Path], *, vocab: Union["Vocab", bool] = True, - disable: Union[str, Iterable[str]] = SimpleFrozenList(), - enable: Union[str, Iterable[str]] = SimpleFrozenList(), - exclude: Union[str, Iterable[str]] = SimpleFrozenList(), + disable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + enable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + exclude: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, config: Union[Dict[str, Any], Config] = SimpleFrozenDict(), ) -> "Language": """Load a model from a package or data path. @@ -470,9 +473,9 @@ def load_model_from_path( *, meta: Optional[Dict[str, Any]] = None, vocab: Union["Vocab", bool] = True, - disable: Union[str, Iterable[str]] = SimpleFrozenList(), - enable: Union[str, Iterable[str]] = SimpleFrozenList(), - exclude: Union[str, Iterable[str]] = SimpleFrozenList(), + disable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + enable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + exclude: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, config: Union[Dict[str, Any], Config] = SimpleFrozenDict(), ) -> "Language": """Load a model from a data directory path. Creates Language class with @@ -516,9 +519,9 @@ def load_model_from_config( *, meta: Dict[str, Any] = SimpleFrozenDict(), vocab: Union["Vocab", bool] = True, - disable: Union[str, Iterable[str]] = SimpleFrozenList(), - enable: Union[str, Iterable[str]] = SimpleFrozenList(), - exclude: Union[str, Iterable[str]] = SimpleFrozenList(), + disable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + enable: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, + exclude: Union[str, Iterable[str]] = _DEFAULT_EMPTY_PIPES, auto_fill: bool = False, validate: bool = True, ) -> "Language": diff --git a/website/docs/api/architectures.md b/website/docs/api/architectures.md index 2537faff6..4c5447f75 100644 --- a/website/docs/api/architectures.md +++ b/website/docs/api/architectures.md @@ -11,6 +11,7 @@ menu: - ['Text Classification', 'textcat'] - ['Span Classification', 'spancat'] - ['Entity Linking', 'entitylinker'] + - ['Coreference', 'coref-architectures'] --- A **model architecture** is a function that wires up a @@ -587,8 +588,8 @@ consists of either two or three subnetworks: run once for each batch. - **lower**: Construct a feature-specific vector for each `(token, feature)` pair. This is also run once for each batch. Constructing the state - representation is then a matter of summing the component features and - applying the non-linearity. + representation is then a matter of summing the component features and applying + the non-linearity. - **upper** (optional): A feed-forward network that predicts scores from the state representation. If not present, the output from the lower model is used as action scores directly. @@ -628,8 +629,8 @@ same signature, but the `use_upper` argument was `True` by default. > ``` Build a tagger model, using a provided token-to-vector component. The tagger -model adds a linear layer with softmax activation to predict scores given -the token vectors. +model adds a linear layer with softmax activation to predict scores given the +token vectors. | Name | Description | | ----------- | ------------------------------------------------------------------------------------------ | @@ -920,5 +921,84 @@ A function that reads an existing `KnowledgeBase` from file. A function that takes as input a [`KnowledgeBase`](/api/kb) and a [`Span`](/api/span) object denoting a named entity, and returns a list of plausible [`Candidate`](/api/kb/#candidate) objects. The default -`CandidateGenerator` uses the text of a mention to find its potential -aliases in the `KnowledgeBase`. Note that this function is case-dependent. +`CandidateGenerator` uses the text of a mention to find its potential aliases in +the `KnowledgeBase`. Note that this function is case-dependent. + +## Coreference {#coref-architectures tag="experimental"} + +A [`CoreferenceResolver`](/api/coref) component identifies tokens that refer to +the same entity. A [`SpanResolver`](/api/span-resolver) component infers spans +from single tokens. Together these components can be used to reproduce +traditional coreference models. You can also omit the `SpanResolver` if working +with only token-level clusters is acceptable. + +### spacy-experimental.Coref.v1 {#Coref tag="experimental"} + +> #### Example Config +> +> ```ini +> +> [model] +> @architectures = "spacy-experimental.Coref.v1" +> distance_embedding_size = 20 +> dropout = 0.3 +> hidden_size = 1024 +> depth = 2 +> antecedent_limit = 50 +> antecedent_batch_size = 512 +> +> [model.tok2vec] +> @architectures = "spacy-transformers.TransformerListener.v1" +> grad_factor = 1.0 +> upstream = "transformer" +> pooling = {"@layers":"reduce_mean.v1"} +> ``` + +The `Coref` model architecture is a Thinc `Model`. + +| Name | Description | +| ------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| `tok2vec` | The [`tok2vec`](#tok2vec) layer of the model. ~~Model~~ | +| `distance_embedding_size` | A representation of the distance between candidates. ~~int~~ | +| `dropout` | The dropout to use internally. Unlike some Thinc models, this has separate dropout for the internal PyTorch layers. ~~float~~ | +| `hidden_size` | Size of the main internal layers. ~~int~~ | +| `depth` | Depth of the internal network. ~~int~~ | +| `antecedent_limit` | How many candidate antecedents to keep after rough scoring. This has a significant effect on memory usage. Typical values would be 50 to 200, or higher for very long documents. ~~int~~ | +| `antecedent_batch_size` | Internal batch size. ~~int~~ | +| **CREATES** | The model using the architecture. ~~Model[List[Doc], Floats2d]~~ | + +### spacy-experimental.SpanResolver.v1 {#SpanResolver tag="experimental"} + +> #### Example Config +> +> ```ini +> +> [model] +> @architectures = "spacy-experimental.SpanResolver.v1" +> hidden_size = 1024 +> distance_embedding_size = 64 +> conv_channels = 4 +> window_size = 1 +> max_distance = 128 +> prefix = "coref_head_clusters" +> +> [model.tok2vec] +> @architectures = "spacy-transformers.TransformerListener.v1" +> grad_factor = 1.0 +> upstream = "transformer" +> pooling = {"@layers":"reduce_mean.v1"} +> ``` + +The `SpanResolver` model architecture is a Thinc `Model`. Note that +`MentionClusters` is `List[List[Tuple[int, int]]]`. + +| Name | Description | +| ------------------------- | -------------------------------------------------------------------------------------------------------------------- | +| `tok2vec` | The [`tok2vec`](#tok2vec) layer of the model. ~~Model~~ | +| `hidden_size` | Size of the main internal layers. ~~int~~ | +| `distance_embedding_size` | A representation of the distance between two candidates. ~~int~~ | +| `conv_channels` | The number of channels in the internal CNN. ~~int~~ | +| `window_size` | The number of neighboring tokens to consider in the internal CNN. `1` means consider one token on each side. ~~int~~ | +| `max_distance` | The longest possible length of a predicted span. ~~int~~ | +| `prefix` | The prefix that indicates spans to use for input data. ~~string~~ | +| **CREATES** | The model using the architecture. ~~Model[List[Doc], List[MentionClusters]]~~ | diff --git a/website/docs/api/cli.md b/website/docs/api/cli.md index e5cd3089b..fc2c46022 100644 --- a/website/docs/api/cli.md +++ b/website/docs/api/cli.md @@ -1482,7 +1482,7 @@ You'll also need to add the assets you want to track with ```cli -$ python -m spacy project dvc [project_dir] [workflow] [--force] [--verbose] +$ python -m spacy project dvc [project_dir] [workflow] [--force] [--verbose] [--quiet] ``` > #### Example @@ -1499,6 +1499,7 @@ $ python -m spacy project dvc [project_dir] [workflow] [--force] [--verbose] | `workflow` | Name of workflow defined in `project.yml`. Defaults to first workflow if not set. ~~Optional[str] \(option)~~ | | `--force`, `-F` | Force-updating config file. ~~bool (flag)~~ | | `--verbose`, `-V` | Print more output generated by DVC. ~~bool (flag)~~ | +| `--quiet`, `-q` | Print no output generated by DVC. ~~bool (flag)~~ | | `--help`, `-h` | Show help message and available arguments. ~~bool (flag)~~ | | **CREATES** | A `dvc.yaml` file in the project directory, based on the steps defined in the given workflow. | diff --git a/website/docs/api/coref.md b/website/docs/api/coref.md new file mode 100644 index 000000000..8f54422d6 --- /dev/null +++ b/website/docs/api/coref.md @@ -0,0 +1,353 @@ +--- +title: CoreferenceResolver +tag: class,experimental +source: spacy-experimental/coref/coref_component.py +teaser: 'Pipeline component for word-level coreference resolution' +api_base_class: /api/pipe +api_string_name: coref +api_trainable: true +--- + +> #### Installation +> +> ```bash +> $ pip install -U spacy-experimental +> ``` + + + +This component is not yet integrated into spaCy core, and is available via the +extension package +[`spacy-experimental`](https://github.com/explosion/spacy-experimental) starting +in version 0.6.0. It exposes the component via +[entry points](/usage/saving-loading/#entry-points), so if you have the package +installed, using `factory = "experimental_coref"` in your +[training config](/usage/training#config) or +`nlp.add_pipe("experimental_coref")` will work out-of-the-box. + + + +A `CoreferenceResolver` component groups tokens into clusters that refer to the +same thing. Clusters are represented as SpanGroups that start with a prefix +(`coref_clusters` by default). + +A `CoreferenceResolver` component can be paired with a +[`SpanResolver`](/api/span-resolver) to expand single tokens to spans. + +## Assigned Attributes {#assigned-attributes} + +Predictions will be saved to `Doc.spans` as a [`SpanGroup`](/api/spangroup). The +span key will be a prefix plus a serial number referring to the coreference +cluster, starting from zero. + +The span key prefix defaults to `"coref_clusters"`, but can be passed as a +parameter. + +| Location | Value | +| ------------------------------------------ | ------------------------------------------------------------------------------------------------------- | +| `Doc.spans[prefix + "_" + cluster_number]` | One coreference cluster, represented as single-token spans. Cluster numbers start from 1. ~~SpanGroup~~ | + +## Config and implementation {#config} + +The default config is defined by the pipeline component factory and describes +how the component should be configured. You can override its settings via the +`config` argument on [`nlp.add_pipe`](/api/language#add_pipe) or in your +[`config.cfg` for training](/usage/training#config). See the +[model architectures](/api/architectures#coref-architectures) documentation for +details on the architectures and their arguments and hyperparameters. + +> #### Example +> +> ```python +> from spacy_experimental.coref.coref_component import DEFAULT_COREF_MODEL +> from spacy_experimental.coref.coref_util import DEFAULT_CLUSTER_PREFIX +> config={ +> "model": DEFAULT_COREF_MODEL, +> "span_cluster_prefix": DEFAULT_CLUSTER_PREFIX, +> }, +> nlp.add_pipe("experimental_coref", config=config) +> ``` + +| Setting | Description | +| --------------------- | ---------------------------------------------------------------------------------------------------------------------------------------- | +| `model` | The [`Model`](https://thinc.ai/docs/api-model) powering the pipeline component. Defaults to [Coref](/api/architectures#Coref). ~~Model~~ | +| `span_cluster_prefix` | The prefix for the keys for clusters saved to `doc.spans`. Defaults to `coref_clusters`. ~~str~~ | + +## CoreferenceResolver.\_\_init\_\_ {#init tag="method"} + +> #### Example +> +> ```python +> # Construction via add_pipe with default model +> coref = nlp.add_pipe("experimental_coref") +> +> # Construction via add_pipe with custom model +> config = {"model": {"@architectures": "my_coref.v1"}} +> coref = nlp.add_pipe("experimental_coref", config=config) +> +> # Construction from class +> from spacy_experimental.coref.coref_component import CoreferenceResolver +> coref = CoreferenceResolver(nlp.vocab, model) +> ``` + +Create a new pipeline instance. In your application, you would normally use a +shortcut for this and instantiate the component using its string name and +[`nlp.add_pipe`](/api/language#add_pipe). + +| Name | Description | +| --------------------- | --------------------------------------------------------------------------------------------------- | +| `vocab` | The shared vocabulary. ~~Vocab~~ | +| `model` | The [`Model`](https://thinc.ai/docs/api-model) powering the pipeline component. ~~Model~~ | +| `name` | String name of the component instance. Used to add entries to the `losses` during training. ~~str~~ | +| _keyword-only_ | | +| `span_cluster_prefix` | The prefix for the key for saving clusters of spans. ~~bool~~ | + +## CoreferenceResolver.\_\_call\_\_ {#call tag="method"} + +Apply the pipe to one document. The document is modified in place and returned. +This usually happens under the hood when the `nlp` object is called on a text +and all pipeline components are applied to the `Doc` in order. Both +[`__call__`](/api/coref#call) and [`pipe`](/api/coref#pipe) delegate to the +[`predict`](/api/coref#predict) and +[`set_annotations`](/api/coref#set_annotations) methods. + +> #### Example +> +> ```python +> doc = nlp("This is a sentence.") +> coref = nlp.add_pipe("experimental_coref") +> # This usually happens under the hood +> processed = coref(doc) +> ``` + +| Name | Description | +| ----------- | -------------------------------- | +| `doc` | The document to process. ~~Doc~~ | +| **RETURNS** | The processed document. ~~Doc~~ | + +## CoreferenceResolver.pipe {#pipe tag="method"} + +Apply the pipe to a stream of documents. This usually happens under the hood +when the `nlp` object is called on a text and all pipeline components are +applied to the `Doc` in order. Both [`__call__`](/api/coref#call) and +[`pipe`](/api/coref#pipe) delegate to the [`predict`](/api/coref#predict) and +[`set_annotations`](/api/coref#set_annotations) methods. + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> for doc in coref.pipe(docs, batch_size=50): +> pass +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------- | +| `stream` | A stream of documents. ~~Iterable[Doc]~~ | +| _keyword-only_ | | +| `batch_size` | The number of documents to buffer. Defaults to `128`. ~~int~~ | +| **YIELDS** | The processed documents in order. ~~Doc~~ | + +## CoreferenceResolver.initialize {#initialize tag="method"} + +Initialize the component for training. `get_examples` should be a function that +returns an iterable of [`Example`](/api/example) objects. **At least one example +should be supplied.** The data examples are used to **initialize the model** of +the component and can either be the full training data or a representative +sample. Initialization includes validating the network, +[inferring missing shapes](https://thinc.ai/docs/usage-models#validation) and +setting up the label scheme based on the data. This method is typically called +by [`Language.initialize`](/api/language#initialize). + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> coref.initialize(lambda: examples, nlp=nlp) +> ``` + +| Name | Description | +| -------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| `get_examples` | Function that returns gold-standard annotations in the form of [`Example`](/api/example) objects. Must contain at least one `Example`. ~~Callable[[], Iterable[Example]]~~ | +| _keyword-only_ | | +| `nlp` | The current `nlp` object. Defaults to `None`. ~~Optional[Language]~~ | + +## CoreferenceResolver.predict {#predict tag="method"} + +Apply the component's model to a batch of [`Doc`](/api/doc) objects, without +modifying them. Clusters are returned as a list of `MentionClusters`, one for +each input `Doc`. A `MentionClusters` instance is just a list of lists of pairs +of `int`s, where each item corresponds to a cluster, and the `int`s correspond +to token indices. + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> clusters = coref.predict([doc1, doc2]) +> ``` + +| Name | Description | +| ----------- | ---------------------------------------------------------------------------- | +| `docs` | The documents to predict. ~~Iterable[Doc]~~ | +| **RETURNS** | The predicted coreference clusters for the `docs`. ~~List[MentionClusters]~~ | + +## CoreferenceResolver.set_annotations {#set_annotations tag="method"} + +Modify a batch of documents, saving coreference clusters in `Doc.spans`. + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> clusters = coref.predict([doc1, doc2]) +> coref.set_annotations([doc1, doc2], clusters) +> ``` + +| Name | Description | +| ---------- | ---------------------------------------------------------------------------- | +| `docs` | The documents to modify. ~~Iterable[Doc]~~ | +| `clusters` | The predicted coreference clusters for the `docs`. ~~List[MentionClusters]~~ | + +## CoreferenceResolver.update {#update tag="method"} + +Learn from a batch of [`Example`](/api/example) objects. Delegates to +[`predict`](/api/coref#predict). + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> optimizer = nlp.initialize() +> losses = coref.update(examples, sgd=optimizer) +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------------------------------------ | +| `examples` | A batch of [`Example`](/api/example) objects to learn from. ~~Iterable[Example]~~ | +| _keyword-only_ | | +| `drop` | The dropout rate. ~~float~~ | +| `sgd` | An optimizer. Will be created via [`create_optimizer`](#create_optimizer) if not set. ~~Optional[Optimizer]~~ | +| `losses` | Optional record of the loss during training. Updated using the component name as the key. ~~Optional[Dict[str, float]]~~ | +| **RETURNS** | The updated `losses` dictionary. ~~Dict[str, float]~~ | + +## CoreferenceResolver.create_optimizer {#create_optimizer tag="method"} + +Create an optimizer for the pipeline component. + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> optimizer = coref.create_optimizer() +> ``` + +| Name | Description | +| ----------- | ---------------------------- | +| **RETURNS** | The optimizer. ~~Optimizer~~ | + +## CoreferenceResolver.use_params {#use_params tag="method, contextmanager"} + +Modify the pipe's model, to use the given parameter values. At the end of the +context, the original parameters are restored. + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> with coref.use_params(optimizer.averages): +> coref.to_disk("/best_model") +> ``` + +| Name | Description | +| -------- | -------------------------------------------------- | +| `params` | The parameter values to use in the model. ~~dict~~ | + +## CoreferenceResolver.to_disk {#to_disk tag="method"} + +Serialize the pipe to disk. + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> coref.to_disk("/path/to/coref") +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------ | +| `path` | A path to a directory, which will be created if it doesn't exist. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | + +## CoreferenceResolver.from_disk {#from_disk tag="method"} + +Load the pipe from disk. Modifies the object in place and returns it. + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> coref.from_disk("/path/to/coref") +> ``` + +| Name | Description | +| -------------- | ----------------------------------------------------------------------------------------------- | +| `path` | A path to a directory. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | +| **RETURNS** | The modified `CoreferenceResolver` object. ~~CoreferenceResolver~~ | + +## CoreferenceResolver.to_bytes {#to_bytes tag="method"} + +> #### Example +> +> ```python +> coref = nlp.add_pipe("experimental_coref") +> coref_bytes = coref.to_bytes() +> ``` + +Serialize the pipe to a bytestring, including the `KnowledgeBase`. + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------- | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | +| **RETURNS** | The serialized form of the `CoreferenceResolver` object. ~~bytes~~ | + +## CoreferenceResolver.from_bytes {#from_bytes tag="method"} + +Load the pipe from a bytestring. Modifies the object in place and returns it. + +> #### Example +> +> ```python +> coref_bytes = coref.to_bytes() +> coref = nlp.add_pipe("experimental_coref") +> coref.from_bytes(coref_bytes) +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------- | +| `bytes_data` | The data to load from. ~~bytes~~ | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | +| **RETURNS** | The `CoreferenceResolver` object. ~~CoreferenceResolver~~ | + +## Serialization fields {#serialization-fields} + +During serialization, spaCy will export several data fields used to restore +different aspects of the object. If needed, you can exclude them from +serialization by passing in the string names via the `exclude` argument. + +> #### Example +> +> ```python +> data = coref.to_disk("/path", exclude=["vocab"]) +> ``` + +| Name | Description | +| ------- | -------------------------------------------------------------- | +| `vocab` | The shared [`Vocab`](/api/vocab). | +| `cfg` | The config file. You usually don't want to exclude this. | +| `model` | The binary model data. You usually don't want to exclude this. | diff --git a/website/docs/api/doc.md b/website/docs/api/doc.md index f97f4ad83..f97ed4547 100644 --- a/website/docs/api/doc.md +++ b/website/docs/api/doc.md @@ -31,21 +31,21 @@ Construct a `Doc` object. The most common way to get a `Doc` object is via the > doc = Doc(nlp.vocab, words=words, spaces=spaces) > ``` -| Name | Description | -| ---------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -| `vocab` | A storage container for lexical types. ~~Vocab~~ | -| `words` | A list of strings or integer hash values to add to the document as words. ~~Optional[List[Union[str,int]]]~~ | -| `spaces` | A list of boolean values indicating whether each word has a subsequent space. Must have the same length as `words`, if specified. Defaults to a sequence of `True`. ~~Optional[List[bool]]~~ | -| _keyword-only_ | | -| `user\_data` | Optional extra data to attach to the Doc. ~~Dict~~ | -| `tags` 3 | A list of strings, of the same length as `words`, to assign as `token.tag` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | -| `pos` 3 | A list of strings, of the same length as `words`, to assign as `token.pos` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | -| `morphs` 3 | A list of strings, of the same length as `words`, to assign as `token.morph` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | -| `lemmas` 3 | A list of strings, of the same length as `words`, to assign as `token.lemma` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | -| `heads` 3 | A list of values, of the same length as `words`, to assign as the head for each word. Head indices are the absolute position of the head in the `Doc`. Defaults to `None`. ~~Optional[List[int]]~~ | -| `deps` 3 | A list of strings, of the same length as `words`, to assign as `token.dep` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | -| `sent_starts` 3 | A list of values, of the same length as `words`, to assign as `token.is_sent_start`. Will be overridden by heads if `heads` is provided. Defaults to `None`. ~~Optional[List[Optional[bool]]]~~ | -| `ents` 3 | A list of strings, of the same length of `words`, to assign the token-based IOB tag. Defaults to `None`. ~~Optional[List[str]]~~ | +| Name | Description | +| ---------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| `vocab` | A storage container for lexical types. ~~Vocab~~ | +| `words` | A list of strings or integer hash values to add to the document as words. ~~Optional[List[Union[str,int]]]~~ | +| `spaces` | A list of boolean values indicating whether each word has a subsequent space. Must have the same length as `words`, if specified. Defaults to a sequence of `True`. ~~Optional[List[bool]]~~ | +| _keyword-only_ | | +| `user\_data` | Optional extra data to attach to the Doc. ~~Dict~~ | +| `tags` 3 | A list of strings, of the same length as `words`, to assign as `token.tag` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | +| `pos` 3 | A list of strings, of the same length as `words`, to assign as `token.pos` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | +| `morphs` 3 | A list of strings, of the same length as `words`, to assign as `token.morph` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | +| `lemmas` 3 | A list of strings, of the same length as `words`, to assign as `token.lemma` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | +| `heads` 3 | A list of values, of the same length as `words`, to assign as the head for each word. Head indices are the absolute position of the head in the `Doc`. Defaults to `None`. ~~Optional[List[int]]~~ | +| `deps` 3 | A list of strings, of the same length as `words`, to assign as `token.dep` for each word. Defaults to `None`. ~~Optional[List[str]]~~ | +| `sent_starts` 3 | A list of values, of the same length as `words`, to assign as `token.is_sent_start`. Will be overridden by heads if `heads` is provided. Defaults to `None`. ~~Optional[List[Union[bool, int, None]]]~~ | +| `ents` 3 | A list of strings, of the same length of `words`, to assign the token-based IOB tag. Defaults to `None`. ~~Optional[List[str]]~~ | ## Doc.\_\_getitem\_\_ {#getitem tag="method"} diff --git a/website/docs/api/entitylinker.md b/website/docs/api/entitylinker.md index 43e08a39c..40ec8afb5 100644 --- a/website/docs/api/entitylinker.md +++ b/website/docs/api/entitylinker.md @@ -14,7 +14,8 @@ entities) to unique identifiers, grounding the named entities into the "real world". It requires a `KnowledgeBase`, as well as a function to generate plausible candidates from that `KnowledgeBase` given a certain textual mention, and a machine learning model to pick the right candidate, given the local -context of the mention. +context of the mention. `EntityLinker` defaults to using the +[`InMemoryLookupKB`](/api/kb_in_memory) implementation. ## Assigned Attributes {#assigned-attributes} @@ -170,7 +171,7 @@ with the current vocab. > > ```python > def create_kb(vocab): -> kb = KnowledgeBase(vocab, entity_vector_length=128) +> kb = InMemoryLookupKB(vocab, entity_vector_length=128) > kb.add_entity(...) > kb.add_alias(...) > return kb diff --git a/website/docs/api/kb.md b/website/docs/api/kb.md index e7a8fcd6f..b217a1678 100644 --- a/website/docs/api/kb.md +++ b/website/docs/api/kb.md @@ -4,27 +4,45 @@ teaser: A storage class for entities and aliases of a specific knowledge base (ontology) tag: class -source: spacy/kb.pyx +source: spacy/kb/kb.pyx new: 2.2 --- -The `KnowledgeBase` object provides a method to generate -[`Candidate`](/api/kb/#candidate) objects, which are plausible external +The `KnowledgeBase` object is an abstract class providing a method to generate +[`Candidate`](/api/kb#candidate) objects, which are plausible external identifiers given a certain textual mention. Each such `Candidate` holds information from the relevant KB entities, such as its frequency in text and possible aliases. Each entity in the knowledge base also has a pretrained entity vector of a fixed size. +Beyond that, `KnowledgeBase` classes have to implement a number of utility +functions called by the [`EntityLinker`](/api/entitylinker) component. + + + +This class was not abstract up to spaCy version 3.5. The `KnowledgeBase` +implementation up to that point is available as `InMemoryLookupKB` from 3.5 +onwards. + + + ## KnowledgeBase.\_\_init\_\_ {#init tag="method"} -Create the knowledge base. +`KnowledgeBase` is an abstract class and cannot be instantiated. Its child +classes should call `__init__()` to set up some necessary attributes. > #### Example > > ```python > from spacy.kb import KnowledgeBase +> from spacy.vocab import Vocab +> +> class FullyImplementedKB(KnowledgeBase): +> def __init__(self, vocab: Vocab, entity_vector_length: int): +> super().__init__(vocab, entity_vector_length) +> ... > vocab = nlp.vocab -> kb = KnowledgeBase(vocab=vocab, entity_vector_length=64) +> kb = FullyImplementedKB(vocab=vocab, entity_vector_length=64) > ``` | Name | Description | @@ -40,133 +58,66 @@ The length of the fixed-size entity vectors in the knowledge base. | ----------- | ------------------------------------------------ | | **RETURNS** | Length of the fixed-size entity vectors. ~~int~~ | -## KnowledgeBase.add_entity {#add_entity tag="method"} +## KnowledgeBase.get_candidates {#get_candidates tag="method"} -Add an entity to the knowledge base, specifying its corpus frequency and entity -vector, which should be of length -[`entity_vector_length`](/api/kb#entity_vector_length). +Given a certain textual mention as input, retrieve a list of candidate entities +of type [`Candidate`](/api/kb#candidate). > #### Example > > ```python -> kb.add_entity(entity="Q42", freq=32, entity_vector=vector1) -> kb.add_entity(entity="Q463035", freq=111, entity_vector=vector2) +> from spacy.lang.en import English +> nlp = English() +> doc = nlp("Douglas Adams wrote 'The Hitchhiker's Guide to the Galaxy'.") +> candidates = kb.get_candidates(doc[0:2]) > ``` -| Name | Description | -| --------------- | ---------------------------------------------------------- | -| `entity` | The unique entity identifier. ~~str~~ | -| `freq` | The frequency of the entity in a typical corpus. ~~float~~ | -| `entity_vector` | The pretrained vector of the entity. ~~numpy.ndarray~~ | +| Name | Description | +| ----------- | -------------------------------------------------------------------- | +| `mention` | The textual mention or alias. ~~Span~~ | +| **RETURNS** | An iterable of relevant `Candidate` objects. ~~Iterable[Candidate]~~ | -## KnowledgeBase.set_entities {#set_entities tag="method"} +## KnowledgeBase.get_candidates_batch {#get_candidates_batch tag="method"} -Define the full list of entities in the knowledge base, specifying the corpus -frequency and entity vector for each entity. +Same as [`get_candidates()`](/api/kb#get_candidates), but for an arbitrary +number of mentions. The [`EntityLinker`](/api/entitylinker) component will call +`get_candidates_batch()` instead of `get_candidates()`, if the config parameter +`candidates_batch_size` is greater or equal than 1. + +The default implementation of `get_candidates_batch()` executes +`get_candidates()` in a loop. We recommend implementing a more efficient way to +retrieve candidates for multiple mentions at once, if performance is of concern +to you. > #### Example > > ```python -> kb.set_entities(entity_list=["Q42", "Q463035"], freq_list=[32, 111], vector_list=[vector1, vector2]) +> from spacy.lang.en import English +> nlp = English() +> doc = nlp("Douglas Adams wrote 'The Hitchhiker's Guide to the Galaxy'.") +> candidates = kb.get_candidates((doc[0:2], doc[3:])) > ``` -| Name | Description | -| ------------- | ---------------------------------------------------------------- | -| `entity_list` | List of unique entity identifiers. ~~Iterable[Union[str, int]]~~ | -| `freq_list` | List of entity frequencies. ~~Iterable[int]~~ | -| `vector_list` | List of entity vectors. ~~Iterable[numpy.ndarray]~~ | - -## KnowledgeBase.add_alias {#add_alias tag="method"} - -Add an alias or mention to the knowledge base, specifying its potential KB -identifiers and their prior probabilities. The entity identifiers should refer -to entities previously added with [`add_entity`](/api/kb#add_entity) or -[`set_entities`](/api/kb#set_entities). The sum of the prior probabilities -should not exceed 1. Note that an empty string can not be used as alias. - -> #### Example -> -> ```python -> kb.add_alias(alias="Douglas", entities=["Q42", "Q463035"], probabilities=[0.6, 0.3]) -> ``` - -| Name | Description | -| --------------- | --------------------------------------------------------------------------------- | -| `alias` | The textual mention or alias. Can not be the empty string. ~~str~~ | -| `entities` | The potential entities that the alias may refer to. ~~Iterable[Union[str, int]]~~ | -| `probabilities` | The prior probabilities of each entity. ~~Iterable[float]~~ | - -## KnowledgeBase.\_\_len\_\_ {#len tag="method"} - -Get the total number of entities in the knowledge base. - -> #### Example -> -> ```python -> total_entities = len(kb) -> ``` - -| Name | Description | -| ----------- | ----------------------------------------------------- | -| **RETURNS** | The number of entities in the knowledge base. ~~int~~ | - -## KnowledgeBase.get_entity_strings {#get_entity_strings tag="method"} - -Get a list of all entity IDs in the knowledge base. - -> #### Example -> -> ```python -> all_entities = kb.get_entity_strings() -> ``` - -| Name | Description | -| ----------- | --------------------------------------------------------- | -| **RETURNS** | The list of entities in the knowledge base. ~~List[str]~~ | - -## KnowledgeBase.get_size_aliases {#get_size_aliases tag="method"} - -Get the total number of aliases in the knowledge base. - -> #### Example -> -> ```python -> total_aliases = kb.get_size_aliases() -> ``` - -| Name | Description | -| ----------- | ---------------------------------------------------- | -| **RETURNS** | The number of aliases in the knowledge base. ~~int~~ | - -## KnowledgeBase.get_alias_strings {#get_alias_strings tag="method"} - -Get a list of all aliases in the knowledge base. - -> #### Example -> -> ```python -> all_aliases = kb.get_alias_strings() -> ``` - -| Name | Description | -| ----------- | -------------------------------------------------------- | -| **RETURNS** | The list of aliases in the knowledge base. ~~List[str]~~ | +| Name | Description | +| ----------- | -------------------------------------------------------------------------------------------- | +| `mentions` | The textual mention or alias. ~~Iterable[Span]~~ | +| **RETURNS** | An iterable of iterable with relevant `Candidate` objects. ~~Iterable[Iterable[Candidate]]~~ | ## KnowledgeBase.get_alias_candidates {#get_alias_candidates tag="method"} -Given a certain textual mention as input, retrieve a list of candidate entities -of type [`Candidate`](/api/kb/#candidate). + +This method is _not_ available from spaCy 3.5 onwards. + -> #### Example -> -> ```python -> candidates = kb.get_alias_candidates("Douglas") -> ``` - -| Name | Description | -| ----------- | ------------------------------------------------------------- | -| `alias` | The textual mention or alias. ~~str~~ | -| **RETURNS** | The list of relevant `Candidate` objects. ~~List[Candidate]~~ | +From spaCy 3.5 on `KnowledgeBase` is an abstract class (with +[`InMemoryLookupKB`](/api/kb_in_memory) being a drop-in replacement) to allow +more flexibility in customizing knowledge bases. Some of its methods were moved +to [`InMemoryLookupKB`](/api/kb_in_memory) during this refactoring, one of those +being `get_alias_candidates()`. This method is now available as +[`InMemoryLookupKB.get_alias_candidates()`](/api/kb_in_memory#get_alias_candidates). +Note: [`InMemoryLookupKB.get_candidates()`](/api/kb_in_memory#get_candidates) +defaults to +[`InMemoryLookupKB.get_alias_candidates()`](/api/kb_in_memory#get_alias_candidates). ## KnowledgeBase.get_vector {#get_vector tag="method"} @@ -178,27 +129,30 @@ Given a certain entity ID, retrieve its pretrained entity vector. > vector = kb.get_vector("Q42") > ``` -| Name | Description | -| ----------- | ------------------------------------ | -| `entity` | The entity ID. ~~str~~ | -| **RETURNS** | The entity vector. ~~numpy.ndarray~~ | +| Name | Description | +| ----------- | -------------------------------------- | +| `entity` | The entity ID. ~~str~~ | +| **RETURNS** | The entity vector. ~~Iterable[float]~~ | -## KnowledgeBase.get_prior_prob {#get_prior_prob tag="method"} +## KnowledgeBase.get_vectors {#get_vectors tag="method"} -Given a certain entity ID and a certain textual mention, retrieve the prior -probability of the fact that the mention links to the entity ID. +Same as [`get_vector()`](/api/kb#get_vector), but for an arbitrary number of +entity IDs. + +The default implementation of `get_vectors()` executes `get_vector()` in a loop. +We recommend implementing a more efficient way to retrieve vectors for multiple +entities at once, if performance is of concern to you. > #### Example > > ```python -> probability = kb.get_prior_prob("Q42", "Douglas") +> vectors = kb.get_vectors(("Q42", "Q3107329")) > ``` -| Name | Description | -| ----------- | ------------------------------------------------------------------------- | -| `entity` | The entity ID. ~~str~~ | -| `alias` | The textual mention or alias. ~~str~~ | -| **RETURNS** | The prior probability of the `alias` referring to the `entity`. ~~float~~ | +| Name | Description | +| ----------- | --------------------------------------------------------- | +| `entities` | The entity IDs. ~~Iterable[str]~~ | +| **RETURNS** | The entity vectors. ~~Iterable[Iterable[numpy.ndarray]]~~ | ## KnowledgeBase.to_disk {#to_disk tag="method"} @@ -207,12 +161,13 @@ Save the current state of the knowledge base to a directory. > #### Example > > ```python -> kb.to_disk(loc) +> kb.to_disk(path) > ``` -| Name | Description | -| ----- | ------------------------------------------------------------------------------------------------------------------------------------------ | -| `loc` | A path to a directory, which will be created if it doesn't exist. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| Name | Description | +| --------- | ------------------------------------------------------------------------------------------------------------------------------------------ | +| `path` | A path to a directory, which will be created if it doesn't exist. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| `exclude` | List of components to exclude. ~~Iterable[str]~~ | ## KnowledgeBase.from_disk {#from_disk tag="method"} @@ -222,16 +177,16 @@ Restore the state of the knowledge base from a given directory. Note that the > #### Example > > ```python -> from spacy.kb import KnowledgeBase > from spacy.vocab import Vocab > vocab = Vocab().from_disk("/path/to/vocab") -> kb = KnowledgeBase(vocab=vocab, entity_vector_length=64) +> kb = FullyImplementedKB(vocab=vocab, entity_vector_length=64) > kb.from_disk("/path/to/kb") > ``` | Name | Description | | ----------- | ----------------------------------------------------------------------------------------------- | | `loc` | A path to a directory. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| `exclude` | List of components to exclude. ~~Iterable[str]~~ | | **RETURNS** | The modified `KnowledgeBase` object. ~~KnowledgeBase~~ | ## Candidate {#candidate tag="class"} diff --git a/website/docs/api/kb_in_memory.md b/website/docs/api/kb_in_memory.md new file mode 100644 index 000000000..9e3279e6a --- /dev/null +++ b/website/docs/api/kb_in_memory.md @@ -0,0 +1,302 @@ +--- +title: InMemoryLookupKB +teaser: + The default implementation of the KnowledgeBase interface. Stores all + information in-memory. +tag: class +source: spacy/kb/kb_in_memory.pyx +new: 3.5 +--- + +The `InMemoryLookupKB` class inherits from [`KnowledgeBase`](/api/kb) and +implements all of its methods. It stores all KB data in-memory and generates +[`Candidate`](/api/kb#candidate) objects by exactly matching mentions with +entity names. It's highly optimized for both a low memory footprint and speed of +retrieval. + +## InMemoryLookupKB.\_\_init\_\_ {#init tag="method"} + +Create the knowledge base. + +> #### Example +> +> ```python +> from spacy.kb import InMemoryLookupKB +> vocab = nlp.vocab +> kb = InMemoryLookupKB(vocab=vocab, entity_vector_length=64) +> ``` + +| Name | Description | +| ---------------------- | ------------------------------------------------ | +| `vocab` | The shared vocabulary. ~~Vocab~~ | +| `entity_vector_length` | Length of the fixed-size entity vectors. ~~int~~ | + +## InMemoryLookupKB.entity_vector_length {#entity_vector_length tag="property"} + +The length of the fixed-size entity vectors in the knowledge base. + +| Name | Description | +| ----------- | ------------------------------------------------ | +| **RETURNS** | Length of the fixed-size entity vectors. ~~int~~ | + +## InMemoryLookupKB.add_entity {#add_entity tag="method"} + +Add an entity to the knowledge base, specifying its corpus frequency and entity +vector, which should be of length +[`entity_vector_length`](/api/kb_in_memory#entity_vector_length). + +> #### Example +> +> ```python +> kb.add_entity(entity="Q42", freq=32, entity_vector=vector1) +> kb.add_entity(entity="Q463035", freq=111, entity_vector=vector2) +> ``` + +| Name | Description | +| --------------- | ---------------------------------------------------------- | +| `entity` | The unique entity identifier. ~~str~~ | +| `freq` | The frequency of the entity in a typical corpus. ~~float~~ | +| `entity_vector` | The pretrained vector of the entity. ~~numpy.ndarray~~ | + +## InMemoryLookupKB.set_entities {#set_entities tag="method"} + +Define the full list of entities in the knowledge base, specifying the corpus +frequency and entity vector for each entity. + +> #### Example +> +> ```python +> kb.set_entities(entity_list=["Q42", "Q463035"], freq_list=[32, 111], vector_list=[vector1, vector2]) +> ``` + +| Name | Description | +| ------------- | ---------------------------------------------------------------- | +| `entity_list` | List of unique entity identifiers. ~~Iterable[Union[str, int]]~~ | +| `freq_list` | List of entity frequencies. ~~Iterable[int]~~ | +| `vector_list` | List of entity vectors. ~~Iterable[numpy.ndarray]~~ | + +## InMemoryLookupKB.add_alias {#add_alias tag="method"} + +Add an alias or mention to the knowledge base, specifying its potential KB +identifiers and their prior probabilities. The entity identifiers should refer +to entities previously added with [`add_entity`](/api/kb_in_memory#add_entity) +or [`set_entities`](/api/kb_in_memory#set_entities). The sum of the prior +probabilities should not exceed 1. Note that an empty string can not be used as +alias. + +> #### Example +> +> ```python +> kb.add_alias(alias="Douglas", entities=["Q42", "Q463035"], probabilities=[0.6, 0.3]) +> ``` + +| Name | Description | +| --------------- | --------------------------------------------------------------------------------- | +| `alias` | The textual mention or alias. Can not be the empty string. ~~str~~ | +| `entities` | The potential entities that the alias may refer to. ~~Iterable[Union[str, int]]~~ | +| `probabilities` | The prior probabilities of each entity. ~~Iterable[float]~~ | + +## InMemoryLookupKB.\_\_len\_\_ {#len tag="method"} + +Get the total number of entities in the knowledge base. + +> #### Example +> +> ```python +> total_entities = len(kb) +> ``` + +| Name | Description | +| ----------- | ----------------------------------------------------- | +| **RETURNS** | The number of entities in the knowledge base. ~~int~~ | + +## InMemoryLookupKB.get_entity_strings {#get_entity_strings tag="method"} + +Get a list of all entity IDs in the knowledge base. + +> #### Example +> +> ```python +> all_entities = kb.get_entity_strings() +> ``` + +| Name | Description | +| ----------- | --------------------------------------------------------- | +| **RETURNS** | The list of entities in the knowledge base. ~~List[str]~~ | + +## InMemoryLookupKB.get_size_aliases {#get_size_aliases tag="method"} + +Get the total number of aliases in the knowledge base. + +> #### Example +> +> ```python +> total_aliases = kb.get_size_aliases() +> ``` + +| Name | Description | +| ----------- | ---------------------------------------------------- | +| **RETURNS** | The number of aliases in the knowledge base. ~~int~~ | + +## InMemoryLookupKB.get_alias_strings {#get_alias_strings tag="method"} + +Get a list of all aliases in the knowledge base. + +> #### Example +> +> ```python +> all_aliases = kb.get_alias_strings() +> ``` + +| Name | Description | +| ----------- | -------------------------------------------------------- | +| **RETURNS** | The list of aliases in the knowledge base. ~~List[str]~~ | + +## InMemoryLookupKB.get_candidates {#get_candidates tag="method"} + +Given a certain textual mention as input, retrieve a list of candidate entities +of type [`Candidate`](/api/kb#candidate). Wraps +[`get_alias_candidates()`](/api/kb_in_memory#get_alias_candidates). + +> #### Example +> +> ```python +> from spacy.lang.en import English +> nlp = English() +> doc = nlp("Douglas Adams wrote 'The Hitchhiker's Guide to the Galaxy'.") +> candidates = kb.get_candidates(doc[0:2]) +> ``` + +| Name | Description | +| ----------- | -------------------------------------------------------------------- | +| `mention` | The textual mention or alias. ~~Span~~ | +| **RETURNS** | An iterable of relevant `Candidate` objects. ~~Iterable[Candidate]~~ | + +## InMemoryLookupKB.get_candidates_batch {#get_candidates_batch tag="method"} + +Same as [`get_candidates()`](/api/kb_in_memory#get_candidates), but for an +arbitrary number of mentions. The [`EntityLinker`](/api/entitylinker) component +will call `get_candidates_batch()` instead of `get_candidates()`, if the config +parameter `candidates_batch_size` is greater or equal than 1. + +The default implementation of `get_candidates_batch()` executes +`get_candidates()` in a loop. We recommend implementing a more efficient way to +retrieve candidates for multiple mentions at once, if performance is of concern +to you. + +> #### Example +> +> ```python +> from spacy.lang.en import English +> nlp = English() +> doc = nlp("Douglas Adams wrote 'The Hitchhiker's Guide to the Galaxy'.") +> candidates = kb.get_candidates((doc[0:2], doc[3:])) +> ``` + +| Name | Description | +| ----------- | -------------------------------------------------------------------------------------------- | +| `mentions` | The textual mention or alias. ~~Iterable[Span]~~ | +| **RETURNS** | An iterable of iterable with relevant `Candidate` objects. ~~Iterable[Iterable[Candidate]]~~ | + +## InMemoryLookupKB.get_alias_candidates {#get_alias_candidates tag="method"} + +Given a certain textual mention as input, retrieve a list of candidate entities +of type [`Candidate`](/api/kb#candidate). + +> #### Example +> +> ```python +> candidates = kb.get_alias_candidates("Douglas") +> ``` + +| Name | Description | +| ----------- | ------------------------------------------------------------- | +| `alias` | The textual mention or alias. ~~str~~ | +| **RETURNS** | The list of relevant `Candidate` objects. ~~List[Candidate]~~ | + +## InMemoryLookupKB.get_vector {#get_vector tag="method"} + +Given a certain entity ID, retrieve its pretrained entity vector. + +> #### Example +> +> ```python +> vector = kb.get_vector("Q42") +> ``` + +| Name | Description | +| ----------- | ------------------------------------ | +| `entity` | The entity ID. ~~str~~ | +| **RETURNS** | The entity vector. ~~numpy.ndarray~~ | + +## InMemoryLookupKB.get_vectors {#get_vectors tag="method"} + +Same as [`get_vector()`](/api/kb_in_memory#get_vector), but for an arbitrary +number of entity IDs. + +The default implementation of `get_vectors()` executes `get_vector()` in a loop. +We recommend implementing a more efficient way to retrieve vectors for multiple +entities at once, if performance is of concern to you. + +> #### Example +> +> ```python +> vectors = kb.get_vectors(("Q42", "Q3107329")) +> ``` + +| Name | Description | +| ----------- | --------------------------------------------------------- | +| `entities` | The entity IDs. ~~Iterable[str]~~ | +| **RETURNS** | The entity vectors. ~~Iterable[Iterable[numpy.ndarray]]~~ | + +## InMemoryLookupKB.get_prior_prob {#get_prior_prob tag="method"} + +Given a certain entity ID and a certain textual mention, retrieve the prior +probability of the fact that the mention links to the entity ID. + +> #### Example +> +> ```python +> probability = kb.get_prior_prob("Q42", "Douglas") +> ``` + +| Name | Description | +| ----------- | ------------------------------------------------------------------------- | +| `entity` | The entity ID. ~~str~~ | +| `alias` | The textual mention or alias. ~~str~~ | +| **RETURNS** | The prior probability of the `alias` referring to the `entity`. ~~float~~ | + +## InMemoryLookupKB.to_disk {#to_disk tag="method"} + +Save the current state of the knowledge base to a directory. + +> #### Example +> +> ```python +> kb.to_disk(path) +> ``` + +| Name | Description | +| --------- | ------------------------------------------------------------------------------------------------------------------------------------------ | +| `path` | A path to a directory, which will be created if it doesn't exist. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| `exclude` | List of components to exclude. ~~Iterable[str]~~ | + +## InMemoryLookupKB.from_disk {#from_disk tag="method"} + +Restore the state of the knowledge base from a given directory. Note that the +[`Vocab`](/api/vocab) should also be the same as the one used to create the KB. + +> #### Example +> +> ```python +> from spacy.vocab import Vocab +> vocab = Vocab().from_disk("/path/to/vocab") +> kb = FullyImplementedKB(vocab=vocab, entity_vector_length=64) +> kb.from_disk("/path/to/kb") +> ``` + +| Name | Description | +| ----------- | ----------------------------------------------------------------------------------------------- | +| `loc` | A path to a directory. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| `exclude` | List of components to exclude. ~~Iterable[str]~~ | +| **RETURNS** | The modified `KnowledgeBase` object. ~~KnowledgeBase~~ | diff --git a/website/docs/api/language.md b/website/docs/api/language.md index ed763e36a..767a7450a 100644 --- a/website/docs/api/language.md +++ b/website/docs/api/language.md @@ -164,6 +164,9 @@ examples, see the Apply the pipeline to some text. The text can span multiple sentences, and can contain arbitrary whitespace. Alignment into the original string is preserved. +Instead of text, a `Doc` can be passed as input, in which case tokenization is +skipped, but the rest of the pipeline is run. + > #### Example > > ```python @@ -173,7 +176,7 @@ contain arbitrary whitespace. Alignment into the original string is preserved. | Name | Description | | --------------- | ---------------------------------------------------------------------------------------------------------------------------------------------- | -| `text` | The text to be processed. ~~str~~ | +| `text` | The text to be processed, or a Doc. ~~Union[str, Doc]~~ | | _keyword-only_ | | | `disable` | Names of pipeline components to [disable](/usage/processing-pipelines#disabling). ~~List[str]~~ | | `component_cfg` | Optional dictionary of keyword arguments for components, keyed by component names. Defaults to `None`. ~~Optional[Dict[str, Dict[str, Any]]]~~ | @@ -184,6 +187,9 @@ contain arbitrary whitespace. Alignment into the original string is preserved. Process texts as a stream, and yield `Doc` objects in order. This is usually more efficient than processing texts one-by-one. +Instead of text, a `Doc` object can be passed as input. In this case +tokenization is skipped but the rest of the pipeline is run. + > #### Example > > ```python @@ -194,7 +200,7 @@ more efficient than processing texts one-by-one. | Name | Description | | ------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -| `texts` | A sequence of strings. ~~Iterable[str]~~ | +| `texts` | A sequence of strings (or `Doc` objects). ~~Iterable[Union[str, Doc]]~~ | | _keyword-only_ | | | `as_tuples` | If set to `True`, inputs should be a sequence of `(text, context)` tuples. Output will then be a sequence of `(doc, context)` tuples. Defaults to `False`. ~~bool~~ | | `batch_size` | The number of texts to buffer. ~~Optional[int]~~ | diff --git a/website/docs/api/lemmatizer.md b/website/docs/api/lemmatizer.md index 422f34040..905096338 100644 --- a/website/docs/api/lemmatizer.md +++ b/website/docs/api/lemmatizer.md @@ -70,7 +70,7 @@ lemmatizer is available. The lemmatizer modes `rule` and `pos_lookup` require [`token.pos`](/api/token) from a previous pipeline component (see example pipeline configurations in the [pretrained pipeline design details](/models#design-cnn)) or rely on third-party -libraries (`pymorphy2`). +libraries (`pymorphy3`). | Language | Default Mode | | -------- | ------------ | @@ -86,9 +86,9 @@ libraries (`pymorphy2`). | `nb` | `rule` | | `nl` | `rule` | | `pl` | `pos_lookup` | -| `ru` | `pymorphy2` | +| `ru` | `pymorphy3` | | `sv` | `rule` | -| `uk` | `pymorphy2` | +| `uk` | `pymorphy3` | ```python %%GITHUB_SPACY/spacy/pipeline/lemmatizer.py diff --git a/website/docs/api/pipeline-functions.md b/website/docs/api/pipeline-functions.md index 1b7017ca7..070292782 100644 --- a/website/docs/api/pipeline-functions.md +++ b/website/docs/api/pipeline-functions.md @@ -153,3 +153,36 @@ whole pipeline has run. | `attrs` | A dict of the `Doc` attributes and the values to set them to. Defaults to `{"tensor": None, "_.trf_data": None}` to clean up after `tok2vec` and `transformer` components. ~~dict~~ | | `silent` | If `False`, show warnings if attributes aren't found or can't be set. Defaults to `True`. ~~bool~~ | | **RETURNS** | The modified `Doc` with the modified attributes. ~~Doc~~ | + +## span_cleaner {#span_cleaner tag="function,experimental"} + +Remove `SpanGroup`s from `doc.spans` based on a key prefix. This is used to +clean up after the [`CoreferenceResolver`](/api/coref) when it's paired with a +[`SpanResolver`](/api/span-resolver). + + + +This pipeline function is not yet integrated into spaCy core, and is available +via the extension package +[`spacy-experimental`](https://github.com/explosion/spacy-experimental) starting +in version 0.6.0. It exposes the component via +[entry points](/usage/saving-loading/#entry-points), so if you have the package +installed, using `factory = "span_cleaner"` in your +[training config](/usage/training#config) or `nlp.add_pipe("span_cleaner")` will +work out-of-the-box. + + + +> #### Example +> +> ```python +> config = {"prefix": "coref_head_clusters"} +> nlp.add_pipe("span_cleaner", config=config) +> doc = nlp("text") +> assert "coref_head_clusters_1" not in doc.spans +> ``` + +| Setting | Description | +| ----------- | ------------------------------------------------------------------------------------------------------------------------- | +| `prefix` | A prefix to check `SpanGroup` keys for. Any matching groups will be removed. Defaults to `"coref_head_clusters"`. ~~str~~ | +| **RETURNS** | The modified `Doc` with any matching spans removed. ~~Doc~~ | diff --git a/website/docs/api/scorer.md b/website/docs/api/scorer.md index 8dbe3b276..ca3462aa9 100644 --- a/website/docs/api/scorer.md +++ b/website/docs/api/scorer.md @@ -270,3 +270,62 @@ Compute micro-PRF and per-entity PRF scores. | Name | Description | | ---------- | ------------------------------------------------------------------------------------------------------------------- | | `examples` | The `Example` objects holding both the predictions and the correct gold-standard annotations. ~~Iterable[Example]~~ | + +## score_coref_clusters {#score_coref_clusters tag="experimental"} + +Returns LEA ([Moosavi and Strube, 2016](https://aclanthology.org/P16-1060/)) PRF +scores for coreference clusters. + + + +Note this scoring function is not yet included in spaCy core - for details, see +the [CoreferenceResolver](/api/coref) docs. + + + +> #### Example +> +> ```python +> scores = score_coref_clusters( +> examples, +> span_cluster_prefix="coref_clusters", +> ) +> print(scores["coref_f"]) +> ``` + +| Name | Description | +| --------------------- | ------------------------------------------------------------------------------------------------------------------- | +| `examples` | The `Example` objects holding both the predictions and the correct gold-standard annotations. ~~Iterable[Example]~~ | +| _keyword-only_ | | +| `span_cluster_prefix` | The prefix used for spans representing coreference clusters. ~~str~~ | +| **RETURNS** | A dictionary containing the scores. ~~Dict[str, Optional[float]]~~ | + +## score_span_predictions {#score_span_predictions tag="experimental"} + +Return accuracy for reconstructions of spans from single tokens. Only exactly +correct predictions are counted as correct, there is no partial credit for near +answers. Used by the [SpanResolver](/api/span-resolver). + + + +Note this scoring function is not yet included in spaCy core - for details, see +the [SpanResolver](/api/span-resolver) docs. + + + +> #### Example +> +> ```python +> scores = score_span_predictions( +> examples, +> output_prefix="coref_clusters", +> ) +> print(scores["span_coref_clusters_accuracy"]) +> ``` + +| Name | Description | +| --------------- | ------------------------------------------------------------------------------------------------------------------- | +| `examples` | The `Example` objects holding both the predictions and the correct gold-standard annotations. ~~Iterable[Example]~~ | +| _keyword-only_ | | +| `output_prefix` | The prefix used for spans representing the final predicted spans. ~~str~~ | +| **RETURNS** | A dictionary containing the scores. ~~Dict[str, Optional[float]]~~ | diff --git a/website/docs/api/span-resolver.md b/website/docs/api/span-resolver.md new file mode 100644 index 000000000..3e992cd03 --- /dev/null +++ b/website/docs/api/span-resolver.md @@ -0,0 +1,356 @@ +--- +title: SpanResolver +tag: class,experimental +source: spacy-experimental/coref/span_resolver_component.py +teaser: 'Pipeline component for resolving tokens into spans' +api_base_class: /api/pipe +api_string_name: span_resolver +api_trainable: true +--- + +> #### Installation +> +> ```bash +> $ pip install -U spacy-experimental +> ``` + + + +This component not yet integrated into spaCy core, and is available via the +extension package +[`spacy-experimental`](https://github.com/explosion/spacy-experimental) starting +in version 0.6.0. It exposes the component via +[entry points](/usage/saving-loading/#entry-points), so if you have the package +installed, using `factory = "experimental_span_resolver"` in your +[training config](/usage/training#config) or +`nlp.add_pipe("experimental_span_resolver")` will work out-of-the-box. + + + +A `SpanResolver` component takes in tokens (represented as `Span` objects of +length 1) and resolves them into `Span` objects of arbitrary length. The initial +use case is as a post-processing step on word-level +[coreference resolution](/api/coref). The input and output keys used to store +`Span` objects are configurable. + +## Assigned Attributes {#assigned-attributes} + +Predictions will be saved to `Doc.spans` as [`SpanGroup`s](/api/spangroup). + +Input token spans will be read in using an input prefix, by default +`"coref_head_clusters"`, and output spans will be saved using an output prefix +(default `"coref_clusters"`) plus a serial number starting from one. The +prefixes are configurable. + +| Location | Value | +| ------------------------------------------------- | ------------------------------------------------------------------------- | +| `Doc.spans[output_prefix + "_" + cluster_number]` | One group of predicted spans. Cluster number starts from 1. ~~SpanGroup~~ | + +## Config and implementation {#config} + +The default config is defined by the pipeline component factory and describes +how the component should be configured. You can override its settings via the +`config` argument on [`nlp.add_pipe`](/api/language#add_pipe) or in your +[`config.cfg` for training](/usage/training#config). See the +[model architectures](/api/architectures#coref-architectures) documentation for +details on the architectures and their arguments and hyperparameters. + +> #### Example +> +> ```python +> from spacy_experimental.coref.span_resolver_component import DEFAULT_SPAN_RESOLVER_MODEL +> from spacy_experimental.coref.coref_util import DEFAULT_CLUSTER_PREFIX, DEFAULT_CLUSTER_HEAD_PREFIX +> config={ +> "model": DEFAULT_SPAN_RESOLVER_MODEL, +> "input_prefix": DEFAULT_CLUSTER_HEAD_PREFIX, +> "output_prefix": DEFAULT_CLUSTER_PREFIX, +> }, +> nlp.add_pipe("experimental_span_resolver", config=config) +> ``` + +| Setting | Description | +| --------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------ | +| `model` | The [`Model`](https://thinc.ai/docs/api-model) powering the pipeline component. Defaults to [SpanResolver](/api/architectures#SpanResolver). ~~Model~~ | +| `input_prefix` | The prefix to use for input `SpanGroup`s. Defaults to `coref_head_clusters`. ~~str~~ | +| `output_prefix` | The prefix for predicted `SpanGroup`s. Defaults to `coref_clusters`. ~~str~~ | + +## SpanResolver.\_\_init\_\_ {#init tag="method"} + +> #### Example +> +> ```python +> # Construction via add_pipe with default model +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> +> # Construction via add_pipe with custom model +> config = {"model": {"@architectures": "my_span_resolver.v1"}} +> span_resolver = nlp.add_pipe("experimental_span_resolver", config=config) +> +> # Construction from class +> from spacy_experimental.coref.span_resolver_component import SpanResolver +> span_resolver = SpanResolver(nlp.vocab, model) +> ``` + +Create a new pipeline instance. In your application, you would normally use a +shortcut for this and instantiate the component using its string name and +[`nlp.add_pipe`](/api/language#add_pipe). + +| Name | Description | +| --------------- | --------------------------------------------------------------------------------------------------- | +| `vocab` | The shared vocabulary. ~~Vocab~~ | +| `model` | The [`Model`](https://thinc.ai/docs/api-model) powering the pipeline component. ~~Model~~ | +| `name` | String name of the component instance. Used to add entries to the `losses` during training. ~~str~~ | +| _keyword-only_ | | +| `input_prefix` | The prefix to use for input `SpanGroup`s. Defaults to `coref_head_clusters`. ~~str~~ | +| `output_prefix` | The prefix for predicted `SpanGroup`s. Defaults to `coref_clusters`. ~~str~~ | + +## SpanResolver.\_\_call\_\_ {#call tag="method"} + +Apply the pipe to one document. The document is modified in place and returned. +This usually happens under the hood when the `nlp` object is called on a text +and all pipeline components are applied to the `Doc` in order. Both +[`__call__`](#call) and [`pipe`](#pipe) delegate to the [`predict`](#predict) +and [`set_annotations`](#set_annotations) methods. + +> #### Example +> +> ```python +> doc = nlp("This is a sentence.") +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> # This usually happens under the hood +> processed = span_resolver(doc) +> ``` + +| Name | Description | +| ----------- | -------------------------------- | +| `doc` | The document to process. ~~Doc~~ | +| **RETURNS** | The processed document. ~~Doc~~ | + +## SpanResolver.pipe {#pipe tag="method"} + +Apply the pipe to a stream of documents. This usually happens under the hood +when the `nlp` object is called on a text and all pipeline components are +applied to the `Doc` in order. Both [`__call__`](/api/span-resolver#call) and +[`pipe`](/api/span-resolver#pipe) delegate to the +[`predict`](/api/span-resolver#predict) and +[`set_annotations`](/api/span-resolver#set_annotations) methods. + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> for doc in span_resolver.pipe(docs, batch_size=50): +> pass +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------- | +| `stream` | A stream of documents. ~~Iterable[Doc]~~ | +| _keyword-only_ | | +| `batch_size` | The number of documents to buffer. Defaults to `128`. ~~int~~ | +| **YIELDS** | The processed documents in order. ~~Doc~~ | + +## SpanResolver.initialize {#initialize tag="method"} + +Initialize the component for training. `get_examples` should be a function that +returns an iterable of [`Example`](/api/example) objects. **At least one example +should be supplied.** The data examples are used to **initialize the model** of +the component and can either be the full training data or a representative +sample. Initialization includes validating the network, +[inferring missing shapes](https://thinc.ai/docs/usage-models#validation) and +setting up the label scheme based on the data. This method is typically called +by [`Language.initialize`](/api/language#initialize). + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> span_resolver.initialize(lambda: examples, nlp=nlp) +> ``` + +| Name | Description | +| -------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| `get_examples` | Function that returns gold-standard annotations in the form of [`Example`](/api/example) objects. Must contain at least one `Example`. ~~Callable[[], Iterable[Example]]~~ | +| _keyword-only_ | | +| `nlp` | The current `nlp` object. Defaults to `None`. ~~Optional[Language]~~ | + +## SpanResolver.predict {#predict tag="method"} + +Apply the component's model to a batch of [`Doc`](/api/doc) objects, without +modifying them. Predictions are returned as a list of `MentionClusters`, one for +each input `Doc`. A `MentionClusters` instance is just a list of lists of pairs +of `int`s, where each item corresponds to an input `SpanGroup`, and the `int`s +correspond to token indices. + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> spans = span_resolver.predict([doc1, doc2]) +> ``` + +| Name | Description | +| ----------- | ------------------------------------------------------------- | +| `docs` | The documents to predict. ~~Iterable[Doc]~~ | +| **RETURNS** | The predicted spans for the `Doc`s. ~~List[MentionClusters]~~ | + +## SpanResolver.set_annotations {#set_annotations tag="method"} + +Modify a batch of documents, saving predictions using the output prefix in +`Doc.spans`. + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> spans = span_resolver.predict([doc1, doc2]) +> span_resolver.set_annotations([doc1, doc2], spans) +> ``` + +| Name | Description | +| ------- | ------------------------------------------------------------- | +| `docs` | The documents to modify. ~~Iterable[Doc]~~ | +| `spans` | The predicted spans for the `docs`. ~~List[MentionClusters]~~ | + +## SpanResolver.update {#update tag="method"} + +Learn from a batch of [`Example`](/api/example) objects. Delegates to +[`predict`](/api/span-resolver#predict). + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> optimizer = nlp.initialize() +> losses = span_resolver.update(examples, sgd=optimizer) +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------------------------------------ | +| `examples` | A batch of [`Example`](/api/example) objects to learn from. ~~Iterable[Example]~~ | +| _keyword-only_ | | +| `drop` | The dropout rate. ~~float~~ | +| `sgd` | An optimizer. Will be created via [`create_optimizer`](#create_optimizer) if not set. ~~Optional[Optimizer]~~ | +| `losses` | Optional record of the loss during training. Updated using the component name as the key. ~~Optional[Dict[str, float]]~~ | +| **RETURNS** | The updated `losses` dictionary. ~~Dict[str, float]~~ | + +## SpanResolver.create_optimizer {#create_optimizer tag="method"} + +Create an optimizer for the pipeline component. + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> optimizer = span_resolver.create_optimizer() +> ``` + +| Name | Description | +| ----------- | ---------------------------- | +| **RETURNS** | The optimizer. ~~Optimizer~~ | + +## SpanResolver.use_params {#use_params tag="method, contextmanager"} + +Modify the pipe's model, to use the given parameter values. At the end of the +context, the original parameters are restored. + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> with span_resolver.use_params(optimizer.averages): +> span_resolver.to_disk("/best_model") +> ``` + +| Name | Description | +| -------- | -------------------------------------------------- | +| `params` | The parameter values to use in the model. ~~dict~~ | + +## SpanResolver.to_disk {#to_disk tag="method"} + +Serialize the pipe to disk. + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> span_resolver.to_disk("/path/to/span_resolver") +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------ | +| `path` | A path to a directory, which will be created if it doesn't exist. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | + +## SpanResolver.from_disk {#from_disk tag="method"} + +Load the pipe from disk. Modifies the object in place and returns it. + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> span_resolver.from_disk("/path/to/span_resolver") +> ``` + +| Name | Description | +| -------------- | ----------------------------------------------------------------------------------------------- | +| `path` | A path to a directory. Paths may be either strings or `Path`-like objects. ~~Union[str, Path]~~ | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | +| **RETURNS** | The modified `SpanResolver` object. ~~SpanResolver~~ | + +## SpanResolver.to_bytes {#to_bytes tag="method"} + +> #### Example +> +> ```python +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> span_resolver_bytes = span_resolver.to_bytes() +> ``` + +Serialize the pipe to a bytestring. + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------- | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | +| **RETURNS** | The serialized form of the `SpanResolver` object. ~~bytes~~ | + +## SpanResolver.from_bytes {#from_bytes tag="method"} + +Load the pipe from a bytestring. Modifies the object in place and returns it. + +> #### Example +> +> ```python +> span_resolver_bytes = span_resolver.to_bytes() +> span_resolver = nlp.add_pipe("experimental_span_resolver") +> span_resolver.from_bytes(span_resolver_bytes) +> ``` + +| Name | Description | +| -------------- | ------------------------------------------------------------------------------------------- | +| `bytes_data` | The data to load from. ~~bytes~~ | +| _keyword-only_ | | +| `exclude` | String names of [serialization fields](#serialization-fields) to exclude. ~~Iterable[str]~~ | +| **RETURNS** | The `SpanResolver` object. ~~SpanResolver~~ | + +## Serialization fields {#serialization-fields} + +During serialization, spaCy will export several data fields used to restore +different aspects of the object. If needed, you can exclude them from +serialization by passing in the string names via the `exclude` argument. + +> #### Example +> +> ```python +> data = span_resolver.to_disk("/path", exclude=["vocab"]) +> ``` + +| Name | Description | +| ------- | -------------------------------------------------------------- | +| `vocab` | The shared [`Vocab`](/api/vocab). | +| `cfg` | The config file. You usually don't want to exclude this. | +| `model` | The binary model data. You usually don't want to exclude this. | diff --git a/website/docs/api/spangroup.md b/website/docs/api/spangroup.md index 8dbdefc01..2d1cf73c4 100644 --- a/website/docs/api/spangroup.md +++ b/website/docs/api/spangroup.md @@ -255,9 +255,10 @@ Return a copy of the span group. > new_group = doc.spans["errors"].copy() > ``` -| Name | Description | -| ----------- | ----------------------------------------------- | -| **RETURNS** | A copy of the `SpanGroup` object. ~~SpanGroup~~ | +| Name | Description | +| ----------- | -------------------------------------------------------------------------------------------------- | +| `doc` | The document to which the copy is bound. Defaults to `None` for the current doc. ~~Optional[Doc]~~ | +| **RETURNS** | A copy of the `SpanGroup` object. ~~SpanGroup~~ | ## SpanGroup.to_bytes {#to_bytes tag="method"} diff --git a/website/docs/usage/101/_architecture.md b/website/docs/usage/101/_architecture.md index 22e2b961e..4ebca2756 100644 --- a/website/docs/usage/101/_architecture.md +++ b/website/docs/usage/101/_architecture.md @@ -78,7 +78,9 @@ operates on a `Doc` and gives you access to the matched tokens **in context**. | Name | Description | | ------------------------------------------------ | -------------------------------------------------------------------------------------------------- | | [`Corpus`](/api/corpus) | Class for managing annotated corpora for training and evaluation data. | -| [`KnowledgeBase`](/api/kb) | Storage for entities and aliases of a knowledge base for entity linking. | +| [`KnowledgeBase`](/api/kb) | Abstract base class for storage and retrieval of data for entity linking. | +| [`InMemoryLookupKB`](/api/kb_in_memory) | Implementation of `KnowledgeBase` storing all data in memory. | +| [`Candidate`](/api/kb#candidate) | Object associating a textual mention with a specific entity contained in a `KnowledgeBase`. | | [`Lookups`](/api/lookups) | Container for convenient access to large lookup tables and dictionaries. | | [`MorphAnalysis`](/api/morphology#morphanalysis) | A morphological analysis. | | [`Morphology`](/api/morphology) | Store morphological analyses and map them to and from hash values. | diff --git a/website/docs/usage/projects.md b/website/docs/usage/projects.md index 35150035a..90b612358 100644 --- a/website/docs/usage/projects.md +++ b/website/docs/usage/projects.md @@ -148,6 +148,13 @@ skipped. You can also set `--force` to force re-running a command, or `--dry` to perform a "dry run" and see what would happen (without actually running the script). +Since spaCy v3.4.2, `spacy projects run` checks your installed dependencies to +verify that your environment is properly set up and aligns with the project's +`requirements.txt`, if there is one. If missing or conflicting dependencies are +detected, a corresponding warning is displayed. If you'd like to disable the +dependency check, set `check_requirements: false` in your project's +`project.yml`. + ### 4. Run a workflow {#run-workfow} > #### project.yml @@ -226,26 +233,49 @@ pipelines. ```yaml %%GITHUB_PROJECTS/pipelines/tagger_parser_ud/project.yml ``` + > #### Tip: Overriding variables on the CLI > -> If you want to override one or more variables on the CLI and are not already specifying a -> project directory, you need to add `.` as a placeholder: +> If you want to override one or more variables on the CLI and are not already +> specifying a project directory, you need to add `.` as a placeholder: > > ``` > python -m spacy project run test . --vars.foo bar > ``` -| Section | Description | -| --------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -| `title` | An optional project title used in `--help` message and [auto-generated docs](#custom-docs). | -| `description` | An optional project description used in [auto-generated docs](#custom-docs). | -| `vars` | A dictionary of variables that can be referenced in paths, URLs and scripts and overriden on the CLI, just like [`config.cfg` variables](/usage/training#config-interpolation). For example, `${vars.name}` will use the value of the variable `name`. Variables need to be defined in the section `vars`, but can be a nested dict, so you're able to reference `${vars.model.name}`. | -| `env` | A dictionary of variables, mapped to the names of environment variables that will be read in when running the project. For example, `${env.name}` will use the value of the environment variable defined as `name`. | -| `directories` | An optional list of [directories](#project-files) that should be created in the project for assets, training outputs, metrics etc. spaCy will make sure that these directories always exist. | -| `assets` | A list of assets that can be fetched with the [`project assets`](/api/cli#project-assets) command. `url` defines a URL or local path, `dest` is the destination file relative to the project directory, and an optional `checksum` ensures that an error is raised if the file's checksum doesn't match. Instead of `url`, you can also provide a `git` block with the keys `repo`, `branch` and `path`, to download from a Git repo. | -| `workflows` | A dictionary of workflow names, mapped to a list of command names, to execute in order. Workflows can be run with the [`project run`](/api/cli#project-run) command. | -| `commands` | A list of named commands. A command can define an optional help message (shown in the CLI when the user adds `--help`) and the `script`, a list of commands to run. The `deps` and `outputs` let you define the created file the command depends on and produces, respectively. This lets spaCy determine whether a command needs to be re-run because its dependencies or outputs changed. Commands can be run as part of a workflow, or separately with the [`project run`](/api/cli#project-run) command. | -| `spacy_version` | Optional spaCy version range like `>=3.0.0,<3.1.0` that the project is compatible with. If it's loaded with an incompatible version, an error is raised when the project is loaded. | +> #### Tip: Environment Variables +> +> Commands in a project file are not executed in a shell, so they don't have +> direct access to environment variables. But you can insert environment +> variables using the `env` dictionary to make values available for +> interpolation, just like values in `vars`. Here's an example `env` dict that +> makes `$PATH` available as `ENV_PATH`: +> +> ```yaml +> env: +> ENV_PATH: PATH +> ``` +> +> This can be used in a project command like so: +> +> ```yaml +> - name: "echo-path" +> script: +> - "echo ${env.ENV_PATH}" +> ``` + +| Section | Description | +| --------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | +| `title` | An optional project title used in `--help` message and [auto-generated docs](#custom-docs). | +| `description` | An optional project description used in [auto-generated docs](#custom-docs). | +| `vars` | A dictionary of variables that can be referenced in paths, URLs and scripts and overriden on the CLI, just like [`config.cfg` variables](/usage/training#config-interpolation). For example, `${vars.name}` will use the value of the variable `name`. Variables need to be defined in the section `vars`, but can be a nested dict, so you're able to reference `${vars.model.name}`. | +| `env` | A dictionary of variables, mapped to the names of environment variables that will be read in when running the project. For example, `${env.name}` will use the value of the environment variable defined as `name`. | +| `directories` | An optional list of [directories](#project-files) that should be created in the project for assets, training outputs, metrics etc. spaCy will make sure that these directories always exist. | +| `assets` | A list of assets that can be fetched with the [`project assets`](/api/cli#project-assets) command. `url` defines a URL or local path, `dest` is the destination file relative to the project directory, and an optional `checksum` ensures that an error is raised if the file's checksum doesn't match. Instead of `url`, you can also provide a `git` block with the keys `repo`, `branch` and `path`, to download from a Git repo. | +| `workflows` | A dictionary of workflow names, mapped to a list of command names, to execute in order. Workflows can be run with the [`project run`](/api/cli#project-run) command. | +| `commands` | A list of named commands. A command can define an optional help message (shown in the CLI when the user adds `--help`) and the `script`, a list of commands to run. The `deps` and `outputs` let you define the created file the command depends on and produces, respectively. This lets spaCy determine whether a command needs to be re-run because its dependencies or outputs changed. Commands can be run as part of a workflow, or separately with the [`project run`](/api/cli#project-run) command. | +| `spacy_version` | Optional spaCy version range like `>=3.0.0,<3.1.0` that the project is compatible with. If it's loaded with an incompatible version, an error is raised when the project is loaded. | +| `check_requirements` 3.4.2 | A flag determining whether to verify that the installed dependencies align with the project's `requirements.txt`. Defaults to `true`. | ### Data assets {#data-assets} diff --git a/website/docs/usage/training.md b/website/docs/usage/training.md index 5e064b269..27a8bbca7 100644 --- a/website/docs/usage/training.md +++ b/website/docs/usage/training.md @@ -480,7 +480,7 @@ as-is. They are also excluded when calling > parse. So the evaluation results should always reflect what your pipeline will > produce at runtime. If you want a frozen component to run (without updating) > during training as well, so that downstream components can use its -> **predictions**, you can add it to the list of +> **predictions**, you should add it to the list of > [`annotating_components`](/usage/training#annotating-components). ```ini diff --git a/website/docs/usage/v3-4.md b/website/docs/usage/v3-4.md index 7cc4570d5..597fc3cc8 100644 --- a/website/docs/usage/v3-4.md +++ b/website/docs/usage/v3-4.md @@ -65,10 +65,10 @@ The English CNN pipelines have new word vectors: | Package | Model Version | TAG | Parser LAS | NER F | | ----------------------------------------------- | ------------- | ---: | ---------: | ----: | -| [`en_core_news_md`](/models/en#en_core_news_md) | v3.3.0 | 97.3 | 90.1 | 84.6 | -| [`en_core_news_md`](/models/en#en_core_news_lg) | v3.4.0 | 97.2 | 90.3 | 85.5 | -| [`en_core_news_lg`](/models/en#en_core_news_md) | v3.3.0 | 97.4 | 90.1 | 85.3 | -| [`en_core_news_lg`](/models/en#en_core_news_lg) | v3.4.0 | 97.3 | 90.2 | 85.6 | +| [`en_core_web_md`](/models/en#en_core_web_md) | v3.3.0 | 97.3 | 90.1 | 84.6 | +| [`en_core_web_md`](/models/en#en_core_web_lg) | v3.4.0 | 97.2 | 90.3 | 85.5 | +| [`en_core_web_lg`](/models/en#en_core_web_md) | v3.3.0 | 97.4 | 90.1 | 85.3 | +| [`en_core_web_lg`](/models/en#en_core_web_lg) | v3.4.0 | 97.3 | 90.2 | 85.6 | ## Notes about upgrading from v3.3 {#upgrading} diff --git a/website/meta/languages.json b/website/meta/languages.json index 79e1fc5d5..bd1535c90 100644 --- a/website/meta/languages.json +++ b/website/meta/languages.json @@ -4,12 +4,22 @@ "code": "af", "name": "Afrikaans" }, + { + "code": "am", + "name": "Amharic", + "has_examples": true + }, { "code": "ar", "name": "Arabic", "example": "هذه جملة", "has_examples": true }, + { + "code": "az", + "name": "Azerbaijani", + "has_examples": true + }, { "code": "bg", "name": "Bulgarian", @@ -65,7 +75,7 @@ { "code": "dsb", "name": "Lower Sorbian", - "has_examples": true + "has_examples": true }, { "code": "el", @@ -142,6 +152,11 @@ "code": "ga", "name": "Irish" }, + { + "code": "grc", + "name": "Ancient Greek", + "has_examples": true + }, { "code": "gu", "name": "Gujarati", @@ -172,7 +187,7 @@ { "code": "hsb", "name": "Upper Sorbian", - "has_examples": true + "has_examples": true }, { "code": "hu", @@ -260,6 +275,10 @@ "example": "Адамга эң кыйыны — күн сайын адам болуу", "has_examples": true }, + { + "code": "la", + "name": "Latin" + }, { "code": "lb", "name": "Luxembourgish", @@ -374,8 +393,8 @@ "has_examples": true, "dependencies": [ { - "name": "pymorphy2", - "url": "https://github.com/kmike/pymorphy2" + "name": "pymorphy3", + "url": "https://github.com/no-plagiarism/pymorphy3" } ], "models": [ @@ -448,6 +467,11 @@ "example": "นี่คือประโยค", "has_examples": true }, + { + "code": "ti", + "name": "Tigrinya", + "has_examples": true + }, { "code": "tl", "name": "Tagalog" @@ -480,12 +504,12 @@ ], "dependencies": [ { - "name": "pymorphy2", - "url": "https://github.com/kmike/pymorphy2" + "name": "pymorphy3", + "url": "https://github.com/no-plagiarism/pymorphy3" }, { - "name": "pymorphy2-dicts-uk", - "url": "https://github.com/kmike/pymorphy2-dicts/" + "name": "pymorphy3-dicts-uk", + "url": "https://github.com/no-plagiarism/pymorphy3-dicts" } ] }, diff --git a/website/meta/sidebars.json b/website/meta/sidebars.json index 1b743636c..2d8745d77 100644 --- a/website/meta/sidebars.json +++ b/website/meta/sidebars.json @@ -12,7 +12,6 @@ { "text": "New in v3.0", "url": "/usage/v3" }, { "text": "New in v3.1", "url": "/usage/v3-1" }, { "text": "New in v3.2", "url": "/usage/v3-2" }, - { "text": "New in v3.2", "url": "/usage/v3-2" }, { "text": "New in v3.3", "url": "/usage/v3-3" }, { "text": "New in v3.4", "url": "/usage/v3-4" } ] @@ -95,6 +94,7 @@ "label": "Pipeline", "items": [ { "text": "AttributeRuler", "url": "/api/attributeruler" }, + { "text": "CoreferenceResolver", "url": "/api/coref" }, { "text": "DependencyParser", "url": "/api/dependencyparser" }, { "text": "EditTreeLemmatizer", "url": "/api/edittreelemmatizer" }, { "text": "EntityLinker", "url": "/api/entitylinker" }, @@ -105,6 +105,7 @@ { "text": "SentenceRecognizer", "url": "/api/sentencerecognizer" }, { "text": "Sentencizer", "url": "/api/sentencizer" }, { "text": "SpanCategorizer", "url": "/api/spancategorizer" }, + { "text": "SpanResolver", "url": "/api/span-resolver" }, { "text": "SpanRuler", "url": "/api/spanruler" }, { "text": "Tagger", "url": "/api/tagger" }, { "text": "TextCategorizer", "url": "/api/textcategorizer" }, diff --git a/website/meta/universe.json b/website/meta/universe.json index 9145855c6..d7c99956b 100644 --- a/website/meta/universe.json +++ b/website/meta/universe.json @@ -1,5 +1,103 @@ { "resources": [ + { + "id": "spacy-cleaner", + "title": "spacy-cleaner", + "slogan": "Easily clean text with spaCy!", + "description": "**spacy-cleaner** utilises spaCy `Language` models to replace, remove, and \n mutate spaCy tokens. Cleaning actions available are:\n\n* Remove/replace stopwords.\n* Remove/replace punctuation.\n* Remove/replace numbers.\n* Remove/replace emails.\n* Remove/replace URLs.\n* Perform lemmatisation.\n\nSee our [docs](https://ce11an.github.io/spacy-cleaner/) for more information.", + "github": "Ce11an/spacy-cleaner", + "pip": "spacy-cleaner", + "code_example": [ + "import spacy", + "import spacy_cleaner", + "from spacy_cleaner.processing import removers, replacers, mutators", + "", + "model = spacy.load(\"en_core_web_sm\")", + "pipeline = spacy_cleaner.Pipeline(", + " model,", + " removers.remove_stopword_token,", + " replacers.replace_punctuation_token,", + " mutators.mutate_lemma_token,", + ")", + "", + "texts = [\"Hello, my name is Cellan! I love to swim!\"]", + "", + "pipeline.clean(texts)", + "# ['hello _IS_PUNCT_ Cellan _IS_PUNCT_ love swim _IS_PUNCT_']" + ], + "code_language": "python", + "url": "https://ce11an.github.io/spacy-cleaner/", + "image": "https://raw.githubusercontent.com/Ce11an/spacy-cleaner/main/docs/assets/images/spacemen.png", + "author": "Cellan Hall", + "author_links": { + "twitter": "Ce11an", + "github": "Ce11an", + "website": "https://www.linkedin.com/in/cellan-hall/" + }, + "category": [ + "extension" + ], + "tags": [ + "text-processing" + ] + }, + { + "id": "Zshot", + "title": "Zshot", + "slogan": "Zero and Few shot named entity & relationships recognition", + "github": "ibm/zshot", + "pip": "zshot", + "code_example": [ + "import spacy", + "from zshot import PipelineConfig, displacy", + "from zshot.linker import LinkerRegen", + "from zshot.mentions_extractor import MentionsExtractorSpacy", + "from zshot.utils.data_models import Entity", + "", + "nlp = spacy.load('en_core_web_sm')", + "# zero shot definition of entities", + "nlp_config = PipelineConfig(", + " mentions_extractor=MentionsExtractorSpacy(),", + " linker=LinkerRegen(),", + " entities=[", + " Entity(name='Paris',", + " description='Paris is located in northern central France, in a north-bending arc of the river Seine'),", + " Entity(name='IBM',", + " description='International Business Machines Corporation (IBM) is an American multinational technology corporation headquartered in Armonk, New York'),", + " Entity(name='New York', description='New York is a city in U.S. state'),", + " Entity(name='Florida', description='southeasternmost U.S. state'),", + " Entity(name='American',", + " description='American, something of, from, or related to the United States of America, commonly known as the United States or America'),", + " Entity(name='Chemical formula',", + " description='In chemistry, a chemical formula is a way of presenting information about the chemical proportions of atoms that constitute a particular chemical compound or molecul'),", + " Entity(name='Acetamide',", + " description='Acetamide (systematic name: ethanamide) is an organic compound with the formula CH3CONH2. It is the simplest amide derived from acetic acid. It finds some use as a plasticizer and as an industrial solvent.'),", + " Entity(name='Armonk',", + " description='Armonk is a hamlet and census-designated place (CDP) in the town of North Castle, located in Westchester County, New York, United States.'),", + " Entity(name='Acetic Acid',", + " description='Acetic acid, systematically named ethanoic acid, is an acidic, colourless liquid and organic compound with the chemical formula CH3COOH'),", + " Entity(name='Industrial solvent',", + " description='Acetamide (systematic name: ethanamide) is an organic compound with the formula CH3CONH2. It is the simplest amide derived from acetic acid. It finds some use as a plasticizer and as an industrial solvent.'),", + " ]", + ")", + "nlp.add_pipe('zshot', config=nlp_config, last=True)", + "", + "text = 'International Business Machines Corporation (IBM) is an American multinational technology corporation' \\", + " ' headquartered in Armonk, New York, with operations in over 171 countries.'", + "", + "doc = nlp(text)", + "displacy.serve(doc, style='ent')" + ], + "thumb": "https://ibm.github.io/zshot/img/graph.png", + "url": "https://ibm.github.io/zshot/", + "author": "IBM Research", + "author_links": { + "github": "ibm", + "twitter": "IBMResearch", + "website": "https://research.ibm.com/labs/ireland/" + }, + "category": ["scientific", "models", "research"] + }, { "id": "concepcy", "title": "concepCy", @@ -2403,20 +2501,20 @@ "import spacy", "from spacy_wordnet.wordnet_annotator import WordnetAnnotator ", "", - "# Load an spacy model (supported models are \"es\" and \"en\") ", - "nlp = spacy.load('en')", - "# Spacy 3.x", - "nlp.add_pipe(\"spacy_wordnet\", after='tagger', config={'lang': nlp.lang})", - "# Spacy 2.x", + "# Load a spaCy model (supported languages are \"es\" and \"en\") ", + "nlp = spacy.load('en_core_web_sm')", + "# spaCy 3.x", + "nlp.add_pipe(\"spacy_wordnet\", after='tagger')", + "# spaCy 2.x", "# nlp.add_pipe(WordnetAnnotator(nlp.lang), after='tagger')", "token = nlp('prices')[0]", "", - "# wordnet object link spacy token with nltk wordnet interface by giving acces to", + "# WordNet object links spaCy token with NLTK WordNet interface by giving access to", "# synsets and lemmas ", "token._.wordnet.synsets()", "token._.wordnet.lemmas()", "", - "# And automatically tags with wordnet domains", + "# And automatically add info about WordNet domains", "token._.wordnet.wordnet_domains()" ], "author": "recognai", @@ -3984,7 +4082,21 @@ }, "category": ["pipeline"], "tags": ["interpretation", "ja"] + }, + { + "id": "spacy-partial-tagger", + "title": "spaCy - Partial Tagger", + "slogan": "Sequence Tagger for Partially Annotated Dataset in spaCy", + "description": "This is a library to build a CRF tagger with a partially annotated dataset in spaCy. You can build your own tagger only from dictionary.", + "github": "doccano/spacy-partial-tagger", + "pip": "spacy-partial-tagger", + "category": ["pipeline", "training"], + "author": "Yasufumi Taniguchi", + "author_links": { + "github": "yasufumy" + } } + ], "categories": [