Update from master

.flake8

@@ -6,9 +6,5 @@ exclude =
     .env,
     .git,
     __pycache__,
-    lemmatizer.py,
-    lookup.py,
     _tokenizer_exceptions_list.py,
-    spacy/lang/fr/lemmatizer,
-    spacy/lang/nb/lemmatizer
     spacy/__init__.py

.github/contributors/mihaigliga21.md

@@ -0,0 +1,106 @@
+# spaCy contributor agreement
+
+This spaCy Contributor Agreement (**"SCA"**) is based on the
+[Oracle Contributor Agreement](http://www.oracle.com/technetwork/oca-405177.pdf).
+The SCA applies to any contribution that you make to any product or project
+managed by us (the **"project"**), and sets out the intellectual property rights
+you grant to us in the contributed materials. The term **"us"** shall mean
+[ExplosionAI UG (haftungsbeschränkt)](https://explosion.ai/legal). The term
+**"you"** shall mean the person or entity identified below.
+
+If you agree to be bound by these terms, fill in the information requested
+below and include the filled-in version with your first pull request, under the
+folder [`.github/contributors/`](/.github/contributors/). The name of the file
+should be your GitHub username, with the extension `.md`. For example, the user
+example_user would create the file `.github/contributors/example_user.md`.
+
+Read this agreement carefully before signing. These terms and conditions
+constitute a binding legal agreement.
+
+## Contributor Agreement
+
+1. The term "contribution" or "contributed materials" means any source code,
+object code, patch, tool, sample, graphic, specification, manual,
+documentation, or any other material posted or submitted by you to the project.
+
+2. With respect to any worldwide copyrights, or copyright applications and
+registrations, in your contribution:
+
+    * you hereby assign to us joint ownership, and to the extent that such
+    assignment is or becomes invalid, ineffective or unenforceable, you hereby
+    grant to us a perpetual, irrevocable, non-exclusive, worldwide, no-charge,
+    royalty-free, unrestricted license to exercise all rights under those
+    copyrights. This includes, at our option, the right to sublicense these same
+    rights to third parties through multiple levels of sublicensees or other
+    licensing arrangements;
+
+    * you agree that each of us can do all things in relation to your
+    contribution as if each of us were the sole owners, and if one of us makes
+    a derivative work of your contribution, the one who makes the derivative
+    work (or has it made will be the sole owner of that derivative work;
+
+    * you agree that you will not assert any moral rights in your contribution
+    against us, our licensees or transferees;
+
+    * you agree that we may register a copyright in your contribution and
+    exercise all ownership rights associated with it; and
+
+    * you agree that neither of us has any duty to consult with, obtain the
+    consent of, pay or render an accounting to the other for any use or
+    distribution of your contribution.
+
+3. With respect to any patents you own, or that you can license without payment
+to any third party, you hereby grant to us a perpetual, irrevocable,
+non-exclusive, worldwide, no-charge, royalty-free license to:
+
+    * make, have made, use, sell, offer to sell, import, and otherwise transfer
+    your contribution in whole or in part, alone or in combination with or
+    included in any product, work or materials arising out of the project to
+    which your contribution was submitted, and
+
+    * at our option, to sublicense these same rights to third parties through
+    multiple levels of sublicensees or other licensing arrangements.
+
+4. Except as set out above, you keep all right, title, and interest in your
+contribution. The rights that you grant to us under these terms are effective
+on the date you first submitted a contribution to us, even if your submission
+took place before the date you sign these terms.
+
+5. You covenant, represent, warrant and agree that:
+
+    * Each contribution that you submit is and shall be an original work of
+    authorship and you can legally grant the rights set out in this SCA;
+
+    * to the best of your knowledge, each contribution will not violate any
+    third party's copyrights, trademarks, patents, or other intellectual
+    property rights; and
+
+    * each contribution shall be in compliance with U.S. export control laws and
+    other applicable export and import laws. You agree to notify us if you
+    become aware of any circumstance which would make any of the foregoing
+    representations inaccurate in any respect. We may publicly disclose your
+    participation in the project, including the fact that you have signed the SCA.
+
+6. This SCA is governed by the laws of the State of California and applicable
+U.S. Federal law. Any choice of law rules will not apply.
+
+7. Please place an “x” on one of the applicable statement below. Please do NOT
+mark both statements:
+
+    * [x] I am signing on behalf of myself as an individual and no other person
+    or entity, including my employer, has or will have rights with respect to my
+    contributions.
+
+    * [x] I am signing on behalf of my employer or a legal entity and I have the
+    actual authority to contractually bind that entity.
+
+## Contributor Details
+
+| Field                          | Entry                      |
+|------------------------------- | -------------------------- |
+| Name                           | Mihai Gliga                |
+| Company name (if applicable)   |                            |
+| Title or role (if applicable)  |                            |
+| Date                           | September 9, 2019          |
+| GitHub username                | mihaigliga21               |
+| Website (optional)             |                            |

bin/ud/ud_run_test.py

@@ -5,7 +5,6 @@
 from __future__ import unicode_literals
 
 import plac
-import tqdm
 from pathlib import Path
 import re
 import sys

bin/ud/ud_train.py

@@ -5,7 +5,6 @@
 from __future__ import unicode_literals
 
 import plac
-import tqdm
 from pathlib import Path
 import re
 import sys
@@ -486,6 +485,9 @@ def main(
     vectors_dir=None,
     use_oracle_segments=False,
 ):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     spacy.util.fix_random_seed()
     lang.zh.Chinese.Defaults.use_jieba = False
     lang.ja.Japanese.Defaults.use_janome = False

examples/training/conllu.py

@@ -3,11 +3,9 @@
 """
 from __future__ import unicode_literals
 import plac
-import tqdm
 import attr
 from pathlib import Path
 import re
-import sys
 import json
 
 import spacy
@@ -23,7 +21,7 @@ import itertools
 import random
 import numpy.random
 
-import conll17_ud_eval
+from bin.ud import conll17_ud_eval
 
 import spacy.lang.zh
 import spacy.lang.ja
@@ -394,6 +392,9 @@ class TreebankPaths(object):
     limit=("Size limit", "option", "n", int),
 )
 def main(ud_dir, parses_dir, config, corpus, limit=0):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     paths = TreebankPaths(ud_dir, corpus)
     if not (parses_dir / corpus).exists():
         (parses_dir / corpus).mkdir()

examples/training/pretrain_textcat.py

@@ -18,7 +18,6 @@ import random
 import spacy
 import thinc.extra.datasets
 from spacy.util import minibatch, use_gpu, compounding
-import tqdm
 from spacy._ml import Tok2Vec
 from spacy.pipeline import TextCategorizer
 import numpy
@@ -107,6 +106,9 @@ def create_pipeline(width, embed_size, vectors_model):
 
 
 def train_tensorizer(nlp, texts, dropout, n_iter):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     tensorizer = nlp.create_pipe("tensorizer")
     nlp.add_pipe(tensorizer)
     optimizer = nlp.begin_training()
@@ -120,6 +122,9 @@ def train_tensorizer(nlp, texts, dropout, n_iter):
 
 
 def train_textcat(nlp, n_texts, n_iter=10):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     textcat = nlp.get_pipe("textcat")
     tok2vec_weights = textcat.model.tok2vec.to_bytes()
     (train_texts, train_cats), (dev_texts, dev_cats) = load_textcat_data(limit=n_texts)

examples/vectors_tensorboard.py

@@ -13,7 +13,6 @@ import numpy
 import plac
 import spacy
 import tensorflow as tf
-import tqdm
 from tensorflow.contrib.tensorboard.plugins.projector import (
     visualize_embeddings,
     ProjectorConfig,
@@ -36,6 +35,9 @@ from tensorflow.contrib.tensorboard.plugins.projector import (
     ),
 )
 def main(vectors_loc, out_loc, name="spaCy_vectors"):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     meta_file = "{}.tsv".format(name)
     out_meta_file = path.join(out_loc, meta_file)
 

spacy/cli/init_model.py

@@ -3,7 +3,6 @@ from __future__ import unicode_literals
 
 import plac
 import math
-from tqdm import tqdm
 import numpy
 from ast import literal_eval
 from pathlib import Path
@@ -109,6 +108,9 @@ def open_file(loc):
 
 
 def read_attrs_from_deprecated(freqs_loc, clusters_loc):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    from tqdm import tqdm
+
     if freqs_loc is not None:
         with msg.loading("Counting frequencies..."):
             probs, _ = read_freqs(freqs_loc)
@@ -186,6 +188,9 @@ def add_vectors(nlp, vectors_loc, prune_vectors):
 
 
 def read_vectors(vectors_loc):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    from tqdm import tqdm
+
     f = open_file(vectors_loc)
     shape = tuple(int(size) for size in next(f).split())
     vectors_data = numpy.zeros(shape=shape, dtype="f")
@@ -202,6 +207,9 @@ def read_vectors(vectors_loc):
 
 
 def read_freqs(freqs_loc, max_length=100, min_doc_freq=5, min_freq=50):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    from tqdm import tqdm
+
     counts = PreshCounter()
     total = 0
     with freqs_loc.open() as f:
@@ -231,6 +239,9 @@ def read_freqs(freqs_loc, max_length=100, min_doc_freq=5, min_freq=50):
 
 
 def read_clusters(clusters_loc):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    from tqdm import tqdm
+
     clusters = {}
     if ftfy is None:
         user_warning(Warnings.W004)

spacy/cli/profile.py

@@ -7,7 +7,6 @@ import srsly
 import cProfile
 import pstats
 import sys
-import tqdm
 import itertools
 import thinc.extra.datasets
 from wasabi import Printer
@@ -48,6 +47,9 @@ def profile(model, inputs=None, n_texts=10000):
 
 
 def parse_texts(nlp, texts):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     for doc in nlp.pipe(tqdm.tqdm(texts), batch_size=16):
         pass
 

spacy/cli/train.py

@@ -4,7 +4,6 @@ from __future__ import unicode_literals, division, print_function
 import plac
 import os
 from pathlib import Path
-import tqdm
 from thinc.neural._classes.model import Model
 from timeit import default_timer as timer
 import shutil
@@ -103,6 +102,10 @@ def train(
     JSON format. To convert data from other formats, use the `spacy convert`
     command.
     """
+
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     msg = Printer()
     util.fix_random_seed()
     util.set_env_log(verbose)
@@ -392,6 +395,9 @@ def _score_for_model(meta):
 
 @contextlib.contextmanager
 def _create_progress_bar(total):
+    # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
+    import tqdm
+
     if int(os.environ.get("LOG_FRIENDLY", 0)):
         yield
     else:

spacy/errors.py

@@ -452,6 +452,9 @@ class Errors(object):
             "Make sure that you're passing in absolute token indices, not "
             "relative token offsets.\nstart: {start}, end: {end}, label: "
             "{label}, direction: {dir}")
+    E158 = ("Can't add table '{name}' to lookups because it already exists.")
+    E159 = ("Can't find table '{name}' in lookups. Available tables: {tables}")
+    E160 = ("Can't find language data file: {path}")
 
 @add_codes
 class TempErrors(object):

spacy/lang/punctuation.py

@@ -3,7 +3,7 @@ from __future__ import unicode_literals
 
 from .char_classes import LIST_PUNCT, LIST_ELLIPSES, LIST_QUOTES, LIST_CURRENCY
 from .char_classes import LIST_ICONS, HYPHENS, CURRENCY, UNITS
-from .char_classes import CONCAT_QUOTES, ALPHA_LOWER, ALPHA_UPPER, ALPHA
+from .char_classes import CONCAT_QUOTES, ALPHA_LOWER, ALPHA_UPPER, ALPHA, PUNCT
 
 
 _prefixes = (
@@ -27,8 +27,8 @@ _suffixes = (
         r"(?<=°[FfCcKk])\.",
         r"(?<=[0-9])(?:{c})".format(c=CURRENCY),
         r"(?<=[0-9])(?:{u})".format(u=UNITS),
-        r"(?<=[0-9{al}{e}(?:{q})])\.".format(
-            al=ALPHA_LOWER, e=r"%²\-\+", q=CONCAT_QUOTES
+        r"(?<=[0-9{al}{e}{p}(?:{q})])\.".format(
+            al=ALPHA_LOWER, e=r"%²\-\+", q=CONCAT_QUOTES, p=PUNCT
         ),
         r"(?<=[{au}][{au}])\.".format(au=ALPHA_UPPER),
     ]

spacy/lang/ro/__init__.py

@@ -9,6 +9,7 @@ from ..norm_exceptions import BASE_NORMS
 from ...language import Language
 from ...attrs import LANG, NORM
 from ...util import update_exc, add_lookups
+from .tag_map import TAG_MAP
 
 # Lemma data note:
 # Original pairs downloaded from http://www.lexiconista.com/datasets/lemmatization/
@@ -24,6 +25,7 @@ class RomanianDefaults(Language.Defaults):
     tokenizer_exceptions = update_exc(BASE_EXCEPTIONS, TOKENIZER_EXCEPTIONS)
     stop_words = STOP_WORDS
     resources = {"lemma_lookup": "lemma_lookup.json"}
+    tag_map = TAG_MAP
 
 
 class Romanian(Language):

spacy/lang/uk/__init__.py

@@ -24,7 +24,7 @@ class UkrainianDefaults(Language.Defaults):
     stop_words = STOP_WORDS
 
     @classmethod
-    def create_lemmatizer(cls, nlp=None):
+    def create_lemmatizer(cls, nlp=None, **kwargs):
         return UkrainianLemmatizer()
 
 

spacy/lookups.py

@@ -1,52 +1,157 @@
 # coding: utf8
 from __future__ import unicode_literals
 
-from .util import SimpleFrozenDict
+import srsly
+from collections import OrderedDict
+
+from .errors import Errors
+from .util import SimpleFrozenDict, ensure_path
 
 
 class Lookups(object):
+    """Container for large lookup tables and dictionaries, e.g. lemmatization
+    data or tokenizer exception lists. Lookups are available via vocab.lookups,
+    so they can be accessed before the pipeline components are applied (e.g.
+    in the tokenizer and lemmatizer), as well as within the pipeline components
+    via doc.vocab.lookups.
+
+    Important note: At the moment, this class only performs a very basic
+    dictionary lookup. We're planning to replace this with a more efficient
+    implementation. See #3971 for details.
+    """
+
     def __init__(self):
-        self._tables = {}
+        """Initialize the Lookups object.
+
+        RETURNS (Lookups): The newly created object.
+        """
+        self._tables = OrderedDict()
 
     def __contains__(self, name):
+        """Check if the lookups contain a table of a given name. Delegates to
+        Lookups.has_table.
+
+        name (unicode): Name of the table.
+        RETURNS (bool): Whether a table of that name exists.
+        """
         return self.has_table(name)
 
+    def __len__(self):
+        """RETURNS (int): The number of tables in the lookups."""
+        return len(self._tables)
+
     @property
     def tables(self):
+        """RETURNS (list): Names of all tables in the lookups."""
         return list(self._tables.keys())
 
     def add_table(self, name, data=SimpleFrozenDict()):
+        """Add a new table to the lookups. Raises an error if the table exists.
+
+        name (unicode): Unique name of table.
+        data (dict): Optional data to add to the table.
+        RETURNS (Table): The newly added table.
+        """
         if name in self.tables:
-            raise ValueError("Table '{}' already exists".format(name))
+            raise ValueError(Errors.E158.format(name=name))
         table = Table(name=name)
         table.update(data)
         self._tables[name] = table
         return table
 
     def get_table(self, name):
+        """Get a table. Raises an error if the table doesn't exist.
+
+        name (unicode): Name of the table.
+        RETURNS (Table): The table.
+        """
         if name not in self._tables:
-            raise KeyError("Can't find table '{}'".format(name))
+            raise KeyError(Errors.E159.format(name=name, tables=self.tables))
         return self._tables[name]
 
+    def remove_table(self, name):
+        """Remove a table. Raises an error if the table doesn't exist.
+
+        name (unicode): The name to remove.
+        RETURNS (Table): The removed table.
+        """
+        if name not in self._tables:
+            raise KeyError(Errors.E159.format(name=name, tables=self.tables))
+        return self._tables.pop(name)
+
     def has_table(self, name):
+        """Check if the lookups contain a table of a given name.
+
+        name (unicode): Name of the table.
+        RETURNS (bool): Whether a table of that name exists.
+        """
         return name in self._tables
 
     def to_bytes(self, exclude=tuple(), **kwargs):
-        raise NotImplementedError
+        """Serialize the lookups to a bytestring.
+
+        exclude (list): String names of serialization fields to exclude.
+        RETURNS (bytes): The serialized Lookups.
+        """
+        return srsly.msgpack_dumps(self._tables)
 
     def from_bytes(self, bytes_data, exclude=tuple(), **kwargs):
-        raise NotImplementedError
+        """Load the lookups from a bytestring.
 
-    def to_disk(self, path, exclude=tuple(), **kwargs):
-        raise NotImplementedError
+        exclude (list): String names of serialization fields to exclude.
+        RETURNS (bytes): The loaded Lookups.
+        """
+        self._tables = OrderedDict()
+        msg = srsly.msgpack_loads(bytes_data)
+        for key, value in msg.items():
+            self._tables[key] = Table.from_dict(value)
+        return self
 
-    def from_disk(self, path, exclude=tuple(), **kwargs):
-        raise NotImplementedError
+    def to_disk(self, path, **kwargs):
+        """Save the lookups to a directory as lookups.bin.
+
+        path (unicode / Path): The file path.
+        """
+        if len(self._tables):
+            path = ensure_path(path)
+            filepath = path / "lookups.bin"
+            with filepath.open("wb") as file_:
+                file_.write(self.to_bytes())
+
+    def from_disk(self, path, **kwargs):
+        """Load lookups from a directory containing a lookups.bin.
+
+        path (unicode / Path): The file path.
+        RETURNS (Lookups): The loaded lookups.
+        """
+        path = ensure_path(path)
+        filepath = path / "lookups.bin"
+        if filepath.exists():
+            with filepath.open("rb") as file_:
+                data = file_.read()
+            return self.from_bytes(data)
+        return self
 
 
-class Table(dict):
+class Table(OrderedDict):
+    """A table in the lookups. Subclass of builtin dict that implements a
+    slightly more consistent and unified API.
+    """
+    @classmethod
+    def from_dict(cls, data, name=None):
+        self = cls(name=name)
+        self.update(data)
+        return self
+
     def __init__(self, name=None):
+        """Initialize a new table.
+
+        name (unicode): Optional table name for reference.
+        RETURNS (Table): The newly created object.
+        """
+        OrderedDict.__init__(self)
         self.name = name
 
     def set(self, key, value):
+        """Set new key/value pair. Same as table[key] = value."""
         self[key] = value

spacy/tests/lang/en/test_prefix_suffix_infix.py

@@ -133,3 +133,9 @@ def test_en_tokenizer_splits_em_dash_infix(en_tokenizer):
     assert tokens[6].text == "Puddleton"
     assert tokens[7].text == "?"
     assert tokens[8].text == "\u2014"
+
+
+@pytest.mark.parametrize("text,length", [("_MATH_", 3), ("_MATH_.", 4)])
+def test_final_period(en_tokenizer, text, length):
+    tokens = en_tokenizer(text)
+    assert len(tokens) == length

spacy/tests/regression/test_issue1001-1500.py

@@ -13,7 +13,6 @@ from spacy.lemmatizer import Lemmatizer
 from spacy.symbols import ORTH, LEMMA, POS, VERB, VerbForm_part
 
 
-@pytest.mark.xfail
 def test_issue1061():
     '''Test special-case works after tokenizing. Was caching problem.'''
     text = 'I like _MATH_ even _MATH_ when _MATH_, except when _MATH_ is _MATH_! but not _MATH_.'

spacy/tests/serialize/test_serialize_pipeline.py

@@ -41,8 +41,8 @@ def test_serialize_parser_roundtrip_bytes(en_vocab, Parser):
     parser.model, _ = parser.Model(10)
     new_parser = Parser(en_vocab)
     new_parser.model, _ = new_parser.Model(10)
-    new_parser = new_parser.from_bytes(parser.to_bytes())
-    assert new_parser.to_bytes() == parser.to_bytes()
+    new_parser = new_parser.from_bytes(parser.to_bytes(exclude=["vocab"]))
+    assert new_parser.to_bytes(exclude=["vocab"]) == parser.to_bytes(exclude=["vocab"])
 
 
 @pytest.mark.parametrize("Parser", test_parsers)
@@ -55,8 +55,8 @@ def test_serialize_parser_roundtrip_disk(en_vocab, Parser):
         parser_d = Parser(en_vocab)
         parser_d.model, _ = parser_d.Model(0)
         parser_d = parser_d.from_disk(file_path)
-        parser_bytes = parser.to_bytes(exclude=["model"])
-        parser_d_bytes = parser_d.to_bytes(exclude=["model"])
+        parser_bytes = parser.to_bytes(exclude=["model", "vocab"])
+        parser_d_bytes = parser_d.to_bytes(exclude=["model", "vocab"])
         assert parser_bytes == parser_d_bytes
 
 
@@ -64,7 +64,7 @@ def test_to_from_bytes(parser, blank_parser):
     assert parser.model is not True
     assert blank_parser.model is True
     assert blank_parser.moves.n_moves != parser.moves.n_moves
-    bytes_data = parser.to_bytes()
+    bytes_data = parser.to_bytes(exclude=["vocab"])
     blank_parser.from_bytes(bytes_data)
     assert blank_parser.model is not True
     assert blank_parser.moves.n_moves == parser.moves.n_moves
@@ -97,9 +97,9 @@ def test_serialize_tagger_roundtrip_disk(en_vocab, taggers):
 def test_serialize_tensorizer_roundtrip_bytes(en_vocab):
     tensorizer = Tensorizer(en_vocab)
     tensorizer.model = tensorizer.Model()
-    tensorizer_b = tensorizer.to_bytes()
+    tensorizer_b = tensorizer.to_bytes(exclude=["vocab"])
     new_tensorizer = Tensorizer(en_vocab).from_bytes(tensorizer_b)
-    assert new_tensorizer.to_bytes() == tensorizer_b
+    assert new_tensorizer.to_bytes(exclude=["vocab"]) == tensorizer_b
 
 
 def test_serialize_tensorizer_roundtrip_disk(en_vocab):
@@ -109,13 +109,15 @@ def test_serialize_tensorizer_roundtrip_disk(en_vocab):
         file_path = d / "tensorizer"
         tensorizer.to_disk(file_path)
         tensorizer_d = Tensorizer(en_vocab).from_disk(file_path)
-        assert tensorizer.to_bytes() == tensorizer_d.to_bytes()
+        assert tensorizer.to_bytes(exclude=["vocab"]) == tensorizer_d.to_bytes(
+            exclude=["vocab"]
+        )
 
 
 def test_serialize_textcat_empty(en_vocab):
     # See issue #1105
     textcat = TextCategorizer(en_vocab, labels=["ENTITY", "ACTION", "MODIFIER"])
-    textcat.to_bytes()
+    textcat.to_bytes(exclude=["vocab"])
 
 
 @pytest.mark.parametrize("Parser", test_parsers)
@@ -128,13 +130,17 @@ def test_serialize_pipe_exclude(en_vocab, Parser):
     parser = Parser(en_vocab)
     parser.model, _ = parser.Model(0)
     parser.cfg["foo"] = "bar"
-    new_parser = get_new_parser().from_bytes(parser.to_bytes())
+    new_parser = get_new_parser().from_bytes(parser.to_bytes(exclude=["vocab"]))
     assert "foo" in new_parser.cfg
-    new_parser = get_new_parser().from_bytes(parser.to_bytes(), exclude=["cfg"])
+    new_parser = get_new_parser().from_bytes(
+        parser.to_bytes(exclude=["vocab"]), exclude=["cfg"]
+    )
     assert "foo" not in new_parser.cfg
-    new_parser = get_new_parser().from_bytes(parser.to_bytes(exclude=["cfg"]))
+    new_parser = get_new_parser().from_bytes(
+        parser.to_bytes(exclude=["cfg"]), exclude=["vocab"]
+    )
     assert "foo" not in new_parser.cfg
     with pytest.raises(ValueError):
-        parser.to_bytes(cfg=False)
+        parser.to_bytes(cfg=False, exclude=["vocab"])
     with pytest.raises(ValueError):
-        get_new_parser().from_bytes(parser.to_bytes(), cfg=False)
+        get_new_parser().from_bytes(parser.to_bytes(exclude=["vocab"]), cfg=False)

spacy/tests/serialize/test_serialize_vocab_strings.py

@@ -12,12 +12,14 @@ test_strings = [([], []), (["rats", "are", "cute"], ["i", "like", "rats"])]
 test_strings_attrs = [(["rats", "are", "cute"], "Hello")]
 
 
+@pytest.mark.xfail
 @pytest.mark.parametrize("text", ["rat"])
 def test_serialize_vocab(en_vocab, text):
     text_hash = en_vocab.strings.add(text)
-    vocab_bytes = en_vocab.to_bytes()
+    vocab_bytes = en_vocab.to_bytes(exclude=["lookups"])
     new_vocab = Vocab().from_bytes(vocab_bytes)
     assert new_vocab.strings[text_hash] == text
+    assert new_vocab.to_bytes(exclude=["lookups"]) == vocab_bytes
 
 
 @pytest.mark.parametrize("strings1,strings2", test_strings)

spacy/tests/vocab_vectors/test_lookups.py

@@ -3,6 +3,9 @@ from __future__ import unicode_literals
 
 import pytest
 from spacy.lookups import Lookups
+from spacy.vocab import Vocab
+
+from ..util import make_tempdir
 
 
 def test_lookups_api():
@@ -10,6 +13,7 @@ def test_lookups_api():
     data = {"foo": "bar", "hello": "world"}
     lookups = Lookups()
     lookups.add_table(table_name, data)
+    assert len(lookups) == 1
     assert table_name in lookups
     assert lookups.has_table(table_name)
     table = lookups.get_table(table_name)
@@ -22,5 +26,89 @@ def test_lookups_api():
     assert len(table) == 3
     with pytest.raises(KeyError):
         lookups.get_table("xyz")
-    # with pytest.raises(ValueError):
-    #     lookups.add_table(table_name)
+    with pytest.raises(ValueError):
+        lookups.add_table(table_name)
+    table = lookups.remove_table(table_name)
+    assert table.name == table_name
+    assert len(lookups) == 0
+    assert table_name not in lookups
+    with pytest.raises(KeyError):
+        lookups.get_table(table_name)
+
+
+# This fails on Python 3.5
+@pytest.mark.xfail
+def test_lookups_to_from_bytes():
+    lookups = Lookups()
+    lookups.add_table("table1", {"foo": "bar", "hello": "world"})
+    lookups.add_table("table2", {"a": 1, "b": 2, "c": 3})
+    lookups_bytes = lookups.to_bytes()
+    new_lookups = Lookups()
+    new_lookups.from_bytes(lookups_bytes)
+    assert len(new_lookups) == 2
+    assert "table1" in new_lookups
+    assert "table2" in new_lookups
+    table1 = new_lookups.get_table("table1")
+    assert len(table1) == 2
+    assert table1.get("foo") == "bar"
+    table2 = new_lookups.get_table("table2")
+    assert len(table2) == 3
+    assert table2.get("b") == 2
+    assert new_lookups.to_bytes() == lookups_bytes
+
+# This fails on Python 3.5
+@pytest.mark.xfail
+def test_lookups_to_from_disk():
+    lookups = Lookups()
+    lookups.add_table("table1", {"foo": "bar", "hello": "world"})
+    lookups.add_table("table2", {"a": 1, "b": 2, "c": 3})
+    with make_tempdir() as tmpdir:
+        lookups.to_disk(tmpdir)
+        new_lookups = Lookups()
+        new_lookups.from_disk(tmpdir)
+    assert len(new_lookups) == 2
+    assert "table1" in new_lookups
+    assert "table2" in new_lookups
+    table1 = new_lookups.get_table("table1")
+    assert len(table1) == 2
+    assert table1.get("foo") == "bar"
+    table2 = new_lookups.get_table("table2")
+    assert len(table2) == 3
+    assert table2.get("b") == 2
+
+# This fails on Python 3.5
+@pytest.mark.xfail
+def test_lookups_to_from_bytes_via_vocab():
+    table_name = "test"
+    vocab = Vocab()
+    vocab.lookups.add_table(table_name, {"foo": "bar", "hello": "world"})
+    assert len(vocab.lookups) == 1
+    assert table_name in vocab.lookups
+    vocab_bytes = vocab.to_bytes()
+    new_vocab = Vocab()
+    new_vocab.from_bytes(vocab_bytes)
+    assert len(new_vocab.lookups) == 1
+    assert table_name in new_vocab.lookups
+    table = new_vocab.lookups.get_table(table_name)
+    assert len(table) == 2
+    assert table.get("hello") == "world"
+    assert new_vocab.to_bytes() == vocab_bytes
+
+
+# This fails on Python 3.5
+@pytest.mark.xfail
+def test_lookups_to_from_disk_via_vocab():
+    table_name = "test"
+    vocab = Vocab()
+    vocab.lookups.add_table(table_name, {"foo": "bar", "hello": "world"})
+    assert len(vocab.lookups) == 1
+    assert table_name in vocab.lookups
+    with make_tempdir() as tmpdir:
+        vocab.to_disk(tmpdir)
+        new_vocab = Vocab()
+        new_vocab.from_disk(tmpdir)
+    assert len(new_vocab.lookups) == 1
+    assert table_name in new_vocab.lookups
+    table = new_vocab.lookups.get_table(table_name)
+    assert len(table) == 2
+    assert table.get("hello") == "world"

spacy/tokenizer.pxd

@@ -16,10 +16,10 @@ cdef class Tokenizer:
     cdef PreshMap _specials
     cpdef readonly Vocab vocab
 
-    cdef public object token_match
-    cdef public object prefix_search
-    cdef public object suffix_search
-    cdef public object infix_finditer
+    cdef object _token_match
+    cdef object _prefix_search
+    cdef object _suffix_search
+    cdef object _infix_finditer
     cdef object _rules
 
     cpdef Doc tokens_from_list(self, list strings)

spacy/tokenizer.pyx

@@ -61,6 +61,38 @@ cdef class Tokenizer:
             for chunk, substrings in sorted(rules.items()):
                 self.add_special_case(chunk, substrings)
 
+    property token_match:
+        def __get__(self):
+            return self._token_match
+
+        def __set__(self, token_match):
+            self._token_match = token_match
+            self._flush_cache()
+
+    property prefix_search:
+        def __get__(self):
+            return self._prefix_search
+
+        def __set__(self, prefix_search):
+            self._prefix_search = prefix_search
+            self._flush_cache()
+
+    property suffix_search:
+        def __get__(self):
+            return self._suffix_search
+
+        def __set__(self, suffix_search):
+            self._suffix_search = suffix_search
+            self._flush_cache()
+
+    property infix_finditer:
+        def __get__(self):
+            return self._infix_finditer
+
+        def __set__(self, infix_finditer):
+            self._infix_finditer = infix_finditer
+            self._flush_cache()
+
     def __reduce__(self):
         args = (self.vocab,
                 self._rules,
@@ -141,9 +173,23 @@ cdef class Tokenizer:
         for text in texts:
             yield self(text)
 
+    def _flush_cache(self):
+        self._reset_cache([key for key in self._cache if not key in self._specials])
+
     def _reset_cache(self, keys):
         for k in keys:
             del self._cache[k]
+            if not k in self._specials:
+                cached = <_Cached*>self._cache.get(k)
+                if cached is not NULL:
+                    self.mem.free(cached)
+
+    def _reset_specials(self):
+        for k in self._specials:
+            cached = <_Cached*>self._specials.get(k)
+            del self._specials[k]
+            if cached is not NULL:
+                self.mem.free(cached)
 
     cdef int _try_cache(self, hash_t key, Doc tokens) except -1:
         cached = <_Cached*>self._cache.get(key)
@@ -183,6 +229,9 @@ cdef class Tokenizer:
         while string and len(string) != last_size:
             if self.token_match and self.token_match(string):
                 break
+            if self._specials.get(hash_string(string)) != NULL:
+                has_special[0] = 1
+                break
             last_size = len(string)
             pre_len = self.find_prefix(string)
             if pre_len != 0:
@@ -360,8 +409,15 @@ cdef class Tokenizer:
         cached.is_lex = False
         cached.data.tokens = self.vocab.make_fused_token(substrings)
         key = hash_string(string)
+        stale_special = <_Cached*>self._specials.get(key)
+        stale_cached = <_Cached*>self._cache.get(key)
+        self._flush_cache()
         self._specials.set(key, cached)
         self._cache.set(key, cached)
+        if stale_special is not NULL:
+            self.mem.free(stale_special)
+        if stale_special != stale_cached and stale_cached is not NULL:
+            self.mem.free(stale_cached)
         self._rules[string] = substrings
 
     def to_disk(self, path, **kwargs):
@@ -444,7 +500,10 @@ cdef class Tokenizer:
         if data.get("rules"):
             # make sure to hard reset the cache to remove data from the default exceptions
             self._rules = {}
+            self._reset_cache([key for key in self._cache])
+            self._reset_specials()
             self._cache = PreshMap()
+            self._specials = PreshMap()
             for string, substrings in data.get("rules", {}).items():
                 self.add_special_case(string, substrings)
 

spacy/util.py

@@ -131,8 +131,7 @@ def load_language_data(path):
     path = path.with_suffix(path.suffix + ".gz")
     if path.exists():
         return srsly.read_gzip_json(path)
-    # TODO: move to spacy.errors
-    raise ValueError("Can't find language data file: {}".format(path2str(path)))
+    raise ValueError(Errors.E160.format(path=path2str(path)))
 
 
 def get_module_path(module):
@@ -458,6 +457,14 @@ def expand_exc(excs, search, replace):
 
 
 def get_lemma_tables(lookups):
+    """Load lemmatizer data from lookups table. Mostly used via
+    Language.Defaults.create_lemmatizer, but available as helper so it can be
+    reused in language classes that implement custom lemmatizers.
+
+    lookups (Lookups): The lookups table.
+    RETURNS (tuple): A (lemma_rules, lemma_index, lemma_exc, lemma_lookup)
+        tuple that can be used to initialize a Lemmatizer.
+    """
     lemma_rules = {}
     lemma_index = {}
     lemma_exc = {}

spacy/vocab.pyx

@@ -43,6 +43,7 @@ cdef class Vocab:
         lemmatizer (object): A lemmatizer. Defaults to `None`.
         strings (StringStore): StringStore that maps strings to integers, and
             vice versa.
+        lookups (Lookups): Container for large lookup tables and dictionaries.
         RETURNS (Vocab): The newly constructed object.
         """
         lex_attr_getters = lex_attr_getters if lex_attr_getters is not None else {}
@@ -433,6 +434,8 @@ cdef class Vocab:
                 file_.write(self.lexemes_to_bytes())
         if "vectors" not in "exclude" and self.vectors is not None:
             self.vectors.to_disk(path)
+        if "lookups" not in "exclude" and self.lookups is not None:
+            self.lookups.to_disk(path)
 
     def from_disk(self, path, exclude=tuple(), **kwargs):
         """Loads state from a directory. Modifies the object in place and
@@ -457,6 +460,8 @@ cdef class Vocab:
                 self.vectors.from_disk(path, exclude=["strings"])
             if self.vectors.name is not None:
                 link_vectors_to_models(self)
+        if "lookups" not in exclude:
+            self.lookups.from_disk(path)
         return self
 
     def to_bytes(self, exclude=tuple(), **kwargs):
@@ -477,6 +482,7 @@ cdef class Vocab:
             ("strings", lambda: self.strings.to_bytes()),
             ("lexemes", lambda: self.lexemes_to_bytes()),
             ("vectors", deserialize_vectors),
+            ("lookups", lambda: self.lookups.to_bytes())
         ))
         exclude = util.get_serialization_exclude(getters, exclude, kwargs)
         return util.to_bytes(getters, exclude)
@@ -500,6 +506,7 @@ cdef class Vocab:
             ("strings", lambda b: self.strings.from_bytes(b)),
             ("lexemes", lambda b: self.lexemes_from_bytes(b)),
             ("vectors", lambda b: serialize_vectors(b)),
+            ("lookups", lambda b: self.lookups.from_bytes(b))
         ))
         exclude = util.get_serialization_exclude(setters, exclude, kwargs)
         util.from_bytes(bytes_data, setters, exclude)