spaCy/spacy/vocab.pyx

544 lines
21 KiB
Cython
Raw Normal View History

# coding: utf8
# cython: profile=True
from __future__ import unicode_literals
2017-08-22 20:46:35 +03:00
import numpy
2018-12-06 20:53:16 +03:00
import srsly
2017-05-29 14:04:40 +03:00
from collections import OrderedDict
2017-10-30 19:59:43 +03:00
from thinc.neural.util import get_array_module
from .lexeme cimport EMPTY_LEXEME
from .lexeme cimport Lexeme
2015-01-31 08:38:58 +03:00
from .typedefs cimport attr_t
from .tokens.token cimport Token
from .attrs cimport PROB, LANG, ORTH, TAG, POS
from .structs cimport SerializedLexemeC
2017-10-27 20:45:19 +03:00
from .compat import copy_reg, basestring_
from .errors import Errors
from .lemmatizer import Lemmatizer
from .attrs import intify_attrs, NORM
2017-08-19 19:50:16 +03:00
from .vectors import Vectors
2017-09-22 17:38:22 +03:00
from ._ml import link_vectors_to_models
2017-10-27 22:07:59 +03:00
from . import util
2014-12-24 09:42:00 +03:00
cdef class Vocab:
"""A look-up table that allows you to access `Lexeme` objects. The `Vocab`
instance also provides access to the `StringStore`, and owns underlying
C-data that is shared between `Doc` objects.
"""
def __init__(self, lex_attr_getters=None, tag_map=None, lemmatizer=None,
strings=tuple(), oov_prob=-20., **deprecated_kwargs):
"""Create the vocabulary.
2017-10-27 20:45:19 +03:00
lex_attr_getters (dict): A dictionary mapping attribute IDs to
functions to compute them. Defaults to `None`.
tag_map (dict): Dictionary mapping fine-grained tags to coarse-grained
parts-of-speech, and optionally morphological attributes.
lemmatizer (object): A lemmatizer. Defaults to `None`.
strings (StringStore): StringStore that maps strings to integers, and
vice versa.
2017-10-27 20:45:19 +03:00
RETURNS (Vocab): The newly constructed object.
"""
lex_attr_getters = lex_attr_getters if lex_attr_getters is not None else {}
tag_map = tag_map if tag_map is not None else {}
if lemmatizer in (None, True, False):
lemmatizer = Lemmatizer({}, {}, {})
self.cfg = {'oov_prob': oov_prob}
self.mem = Pool()
self._by_orth = PreshMap()
self.strings = StringStore()
2017-06-02 11:57:25 +03:00
self.length = 0
if strings:
for string in strings:
_ = self[string]
self.lex_attr_getters = lex_attr_getters
self.morphology = Morphology(self.strings, tag_map, lemmatizer)
2017-10-31 20:25:08 +03:00
self.vectors = Vectors()
2016-12-21 20:04:41 +03:00
property lang:
def __get__(self):
langfunc = None
if self.lex_attr_getters:
langfunc = self.lex_attr_getters.get(LANG, None)
return langfunc('_') if langfunc else ''
def __len__(self):
"""The current number of lexemes stored.
RETURNS (int): The current number of lexemes stored.
"""
return self.length
def add_flag(self, flag_getter, int flag_id=-1):
"""Set a new boolean flag to words in the vocabulary.
2016-12-21 20:04:41 +03:00
The flag_getter function will be called over the words currently in the
2016-11-01 14:25:36 +03:00
vocab, and then applied to new words as they occur. You'll then be able
2017-10-27 20:45:19 +03:00
to access the flag value on each token using token.check_flag(flag_id).
See also: `Lexeme.set_flag`, `Lexeme.check_flag`, `Token.set_flag`,
`Token.check_flag`.
2017-10-27 20:45:19 +03:00
flag_getter (callable): A function `f(unicode) -> bool`, to get the
flag value.
flag_id (int): An integer between 1 and 63 (inclusive), specifying
the bit at which the flag will be stored. If -1, the lowest
available bit will be chosen.
RETURNS (int): The integer ID by which the flag value can be checked.
EXAMPLE:
2017-10-27 20:45:19 +03:00
>>> my_product_getter = lambda text: text in ['spaCy', 'dislaCy']
>>> MY_PRODUCT = nlp.vocab.add_flag(my_product_getter)
>>> doc = nlp(u'I like spaCy')
>>> assert doc[2].check_flag(MY_PRODUCT) == True
"""
if flag_id == -1:
for bit in range(1, 64):
if bit not in self.lex_attr_getters:
flag_id = bit
break
else:
raise ValueError(Errors.E062)
elif flag_id >= 64 or flag_id < 1:
raise ValueError(Errors.E063.format(value=flag_id))
for lex in self:
lex.set_flag(flag_id, flag_getter(lex.orth_))
self.lex_attr_getters[flag_id] = flag_getter
return flag_id
cdef const LexemeC* get(self, Pool mem, unicode string) except NULL:
2017-10-27 20:45:19 +03:00
"""Get a pointer to a `LexemeC` from the lexicon, creating a new
`Lexeme` if necessary using memory acquired from the given pool. If the
pool is the lexicon's own memory, the lexeme is saved in the lexicon.
"""
2015-07-26 01:18:30 +03:00
if string == u'':
return &EMPTY_LEXEME
cdef LexemeC* lex
2018-08-16 00:43:34 +03:00
cdef hash_t key = self.strings[string]
lex = <LexemeC*>self._by_orth.get(key)
cdef size_t addr
if lex != NULL:
2018-08-16 00:43:34 +03:00
assert lex.orth in self.strings
if lex.orth != key:
raise KeyError(Errors.E064.format(string=lex.orth,
orth=key, orth_id=string))
return lex
else:
2015-08-22 23:04:34 +03:00
return self._new_lexeme(mem, string)
cdef const LexemeC* get_by_orth(self, Pool mem, attr_t orth) except NULL:
2017-10-27 20:45:19 +03:00
"""Get a pointer to a `LexemeC` from the lexicon, creating a new
`Lexeme` if necessary using memory acquired from the given pool. If the
pool is the lexicon's own memory, the lexeme is saved in the lexicon.
"""
2015-07-26 20:26:41 +03:00
if orth == 0:
2015-07-26 19:39:27 +03:00
return &EMPTY_LEXEME
cdef LexemeC* lex
lex = <LexemeC*>self._by_orth.get(orth)
if lex != NULL:
return lex
2015-08-22 23:04:34 +03:00
else:
return self._new_lexeme(mem, self.strings[orth])
2015-08-22 23:04:34 +03:00
cdef const LexemeC* _new_lexeme(self, Pool mem, unicode string) except NULL:
if len(string) < 3 or self.length < 10000:
mem = self.mem
cdef bint is_oov = mem is not self.mem
lex = <LexemeC*>mem.alloc(sizeof(LexemeC), 1)
lex.orth = self.strings.add(string)
2015-08-26 20:21:46 +03:00
lex.length = len(string)
2017-11-15 16:23:58 +03:00
if self.vectors is not None:
lex.id = self.vectors.key2row.get(lex.orth, 0)
else:
lex.id = 0
if self.lex_attr_getters is not None:
for attr, func in self.lex_attr_getters.items():
value = func(string)
if isinstance(value, unicode):
value = self.strings.add(value)
2015-08-26 20:21:46 +03:00
if attr == PROB:
lex.prob = value
elif value is not None:
2015-08-26 20:21:46 +03:00
Lexeme.set_struct_attr(lex, attr, value)
if not is_oov:
self._add_lex_to_vocab(lex.orth, lex)
if lex == NULL:
raise ValueError(Errors.E085.format(string=string))
return lex
cdef int _add_lex_to_vocab(self, hash_t key, const LexemeC* lex) except -1:
self._by_orth.set(lex.orth, <void*>lex)
self.length += 1
def __contains__(self, key):
"""Check whether the string or int key has an entry in the vocabulary.
2016-11-01 14:25:36 +03:00
string (unicode): The ID string.
RETURNS (bool) Whether the string has an entry in the vocabulary.
"""
cdef hash_t int_key
if isinstance(key, bytes):
2018-08-16 00:43:34 +03:00
int_key = self.strings[key.decode('utf8')]
elif isinstance(key, unicode):
2018-08-16 00:43:34 +03:00
int_key = self.strings[key]
else:
int_key = key
lex = self._by_orth.get(int_key)
2017-01-11 12:18:22 +03:00
return lex is not NULL
def __iter__(self):
"""Iterate over the lexemes in the vocabulary.
2016-11-01 14:25:36 +03:00
YIELDS (Lexeme): An entry in the vocabulary.
"""
2017-10-31 04:00:26 +03:00
cdef attr_t key
cdef size_t addr
2017-10-31 04:00:26 +03:00
for key, addr in self._by_orth.items():
lex = Lexeme(self, key)
yield lex
2017-10-27 20:45:19 +03:00
def __getitem__(self, id_or_string):
"""Retrieve a lexeme, given an int ID or a unicode string. If a
previously unseen unicode string is given, a new lexeme is created and
stored.
id_or_string (int or unicode): The integer ID of a word, or its unicode
string. If `int >= Lexicon.size`, `IndexError` is raised. If
`id_or_string` is neither an int nor a unicode string, `ValueError`
is raised.
RETURNS (Lexeme): The lexeme indicated by the given ID.
EXAMPLE:
>>> apple = nlp.vocab.strings['apple']
>>> assert nlp.vocab[apple] == nlp.vocab[u'apple']
"""
cdef attr_t orth
2017-10-31 04:00:26 +03:00
if isinstance(id_or_string, unicode):
orth = self.strings.add(id_or_string)
else:
orth = id_or_string
return Lexeme(self, orth)
cdef const TokenC* make_fused_token(self, substrings) except NULL:
cdef int i
tokens = <TokenC*>self.mem.alloc(len(substrings) + 1, sizeof(TokenC))
for i, props in enumerate(substrings):
2017-10-27 20:45:19 +03:00
props = intify_attrs(props, strings_map=self.strings,
_do_deprecated=True)
token = &tokens[i]
# Set the special tokens up to have arbitrary attributes
2017-10-27 20:45:19 +03:00
lex = <LexemeC*>self.get_by_orth(self.mem, props[ORTH])
token.lex = lex
2017-10-27 20:45:19 +03:00
if TAG in props:
self.morphology.assign_tag(token, props[TAG])
elif POS in props:
# Don't allow POS to be set without TAG -- this causes problems,
# see #1773
props.pop(POS)
for attr_id, value in props.items():
Token.set_struct_attr(token, attr_id, value)
# NORM is the only one that overlaps between the two
# (which is maybe not great?)
if attr_id != NORM:
Lexeme.set_struct_attr(lex, attr_id, value)
return tokens
2016-12-21 20:04:41 +03:00
2017-05-31 00:34:50 +03:00
@property
def vectors_length(self):
2017-08-22 20:46:35 +03:00
return self.vectors.data.shape[1]
2017-05-31 00:34:50 +03:00
2017-10-31 20:25:08 +03:00
def reset_vectors(self, *, width=None, shape=None):
2017-05-31 00:34:50 +03:00
"""Drop the current vector table. Because all vectors must be the same
width, you have to call this to change the size of the vectors.
"""
2017-10-31 20:25:08 +03:00
if width is not None and shape is not None:
raise ValueError(Errors.E065.format(width=width, shape=shape))
2017-10-31 20:25:08 +03:00
elif shape is not None:
self.vectors = Vectors(shape=shape)
else:
width = width if width is not None else self.vectors.data.shape[1]
self.vectors = Vectors(shape=(self.vectors.shape[0], width))
2017-10-31 04:00:26 +03:00
2017-10-31 20:25:08 +03:00
def prune_vectors(self, nr_row, batch_size=1024):
2017-10-31 04:00:26 +03:00
"""Reduce the current vector table to `nr_row` unique entries. Words
mapped to the discarded vectors will be remapped to the closest vector
among those remaining.
For example, suppose the original table had vectors for the words:
['sat', 'cat', 'feline', 'reclined']. If we prune the vector table to,
two rows, we would discard the vectors for 'feline' and 'reclined'.
These words would then be remapped to the closest remaining vector
-- so "feline" would have the same vector as "cat", and "reclined"
would have the same vector as "sat".
The similarities are judged by cosine. The original vectors may
be large, so the cosines are calculated in minibatches, to reduce
memory usage.
nr_row (int): The number of rows to keep in the vector table.
batch_size (int): Batch of vectors for calculating the similarities.
Larger batch sizes might be faster, while temporarily requiring
more memory.
RETURNS (dict): A dictionary keyed by removed words mapped to
`(string, score)` tuples, where `string` is the entry the removed
word was mapped to, and `score` the similarity score between the
two words.
"""
xp = get_array_module(self.vectors.data)
2017-10-31 20:25:08 +03:00
# Make prob negative so it sorts by rank ascending
# (key2row contains the rank)
priority = [(-lex.prob, self.vectors.key2row[lex.orth], lex.orth)
for lex in self if lex.orth in self.vectors.key2row]
priority.sort()
indices = xp.asarray([i for (prob, i, key) in priority], dtype='i')
keys = xp.asarray([key for (prob, i, key) in priority], dtype='uint64')
2017-11-01 02:34:55 +03:00
2017-10-31 20:25:08 +03:00
keep = xp.ascontiguousarray(self.vectors.data[indices[:nr_row]])
toss = xp.ascontiguousarray(self.vectors.data[indices[nr_row:]])
self.vectors = Vectors(data=keep, keys=keys)
💫 Port master changes over to develop (#2979) * Create aryaprabhudesai.md (#2681) * Update _install.jade (#2688) Typo fix: "models" -> "model" * Add FAC to spacy.explain (resolves #2706) * Remove docstrings for deprecated arguments (see #2703) * When calling getoption() in conftest.py, pass a default option (#2709) * When calling getoption() in conftest.py, pass a default option This is necessary to allow testing an installed spacy by running: pytest --pyargs spacy * Add contributor agreement * update bengali token rules for hyphen and digits (#2731) * Less norm computations in token similarity (#2730) * Less norm computations in token similarity * Contributor agreement * Remove ')' for clarity (#2737) Sorry, don't mean to be nitpicky, I just noticed this when going through the CLI and thought it was a quick fix. That said, if this was intention than please let me know. * added contributor agreement for mbkupfer (#2738) * Basic support for Telugu language (#2751) * Lex _attrs for polish language (#2750) * Signed spaCy contributor agreement * Added polish version of english lex_attrs * Introduces a bulk merge function, in order to solve issue #653 (#2696) * Fix comment * Introduce bulk merge to increase performance on many span merges * Sign contributor agreement * Implement pull request suggestions * Describe converters more explicitly (see #2643) * Add multi-threading note to Language.pipe (resolves #2582) [ci skip] * Fix formatting * Fix dependency scheme docs (closes #2705) [ci skip] * Don't set stop word in example (closes #2657) [ci skip] * Add words to portuguese language _num_words (#2759) * Add words to portuguese language _num_words * Add words to portuguese language _num_words * Update Indonesian model (#2752) * adding e-KTP in tokenizer exceptions list * add exception token * removing lines with containing space as it won't matter since we use .split() method in the end, added new tokens in exception * add tokenizer exceptions list * combining base_norms with norm_exceptions * adding norm_exception * fix double key in lemmatizer * remove unused import on punctuation.py * reformat stop_words to reduce number of lines, improve readibility * updating tokenizer exception * implement is_currency for lang/id * adding orth_first_upper in tokenizer_exceptions * update the norm_exception list * remove bunch of abbreviations * adding contributors file * Fixed spaCy+Keras example (#2763) * bug fixes in keras example * created contributor agreement * Adding French hyphenated first name (#2786) * Fix typo (closes #2784) * Fix typo (#2795) [ci skip] Fixed typo on line 6 "regcognizer --> recognizer" * Adding basic support for Sinhala language. (#2788) * adding Sinhala language package, stop words, examples and lex_attrs. * Adding contributor agreement * Updating contributor agreement * Also include lowercase norm exceptions * Fix error (#2802) * Fix error ValueError: cannot resize an array that references or is referenced by another array in this way. Use the resize function * added spaCy Contributor Agreement * Add charlax's contributor agreement (#2805) * agreement of contributor, may I introduce a tiny pl languge contribution (#2799) * Contributors agreement * Contributors agreement * Contributors agreement * Add jupyter=True to displacy.render in documentation (#2806) * Revert "Also include lowercase norm exceptions" This reverts commit 70f4e8adf37cfcfab60be2b97d6deae949b30e9e. * Remove deprecated encoding argument to msgpack * Set up dependency tree pattern matching skeleton (#2732) * Fix bug when too many entity types. Fixes #2800 * Fix Python 2 test failure * Require older msgpack-numpy * Restore encoding arg on msgpack-numpy * Try to fix version pin for msgpack-numpy * Update Portuguese Language (#2790) * Add words to portuguese language _num_words * Add words to portuguese language _num_words * Portuguese - Add/remove stopwords, fix tokenizer, add currency symbols * Extended punctuation and norm_exceptions in the Portuguese language * Correct error in spacy universe docs concerning spacy-lookup (#2814) * Update Keras Example for (Parikh et al, 2016) implementation (#2803) * bug fixes in keras example * created contributor agreement * baseline for Parikh model * initial version of parikh 2016 implemented * tested asymmetric models * fixed grevious error in normalization * use standard SNLI test file * begin to rework parikh example * initial version of running example * start to document the new version * start to document the new version * Update Decompositional Attention.ipynb * fixed calls to similarity * updated the README * import sys package duh * simplified indexing on mapping word to IDs * stupid python indent error * added code from https://github.com/tensorflow/tensorflow/issues/3388 for tf bug workaround * Fix typo (closes #2815) [ci skip] * Update regex version dependency * Set version to 2.0.13.dev3 * Skip seemingly problematic test * Remove problematic test * Try previous version of regex * Revert "Remove problematic test" This reverts commit bdebbef45552d698d390aa430b527ee27830f11b. * Unskip test * Try older version of regex * 💫 Update training examples and use minibatching (#2830) <!--- Provide a general summary of your changes in the title. --> ## Description Update the training examples in `/examples/training` to show usage of spaCy's `minibatch` and `compounding` helpers ([see here](https://spacy.io/usage/training#tips-batch-size) for details). The lack of batching in the examples has caused some confusion in the past, especially for beginners who would copy-paste the examples, update them with large training sets and experienced slow and unsatisfying results. ### Types of change enhancements ## Checklist <!--- Before you submit the PR, go over this checklist and make sure you can tick off all the boxes. [] -> [x] --> - [x] I have submitted the spaCy Contributor Agreement. - [x] I ran the tests, and all new and existing tests passed. - [x] My changes don't require a change to the documentation, or if they do, I've added all required information. * Visual C++ link updated (#2842) (closes #2841) [ci skip] * New landing page * Add contribution agreement * Correcting lang/ru/examples.py (#2845) * Correct some grammatical inaccuracies in lang\ru\examples.py; filled Contributor Agreement * Correct some grammatical inaccuracies in lang\ru\examples.py * Move contributor agreement to separate file * Set version to 2.0.13.dev4 * Add Persian(Farsi) language support (#2797) * Also include lowercase norm exceptions * Remove in favour of https://github.com/explosion/spaCy/graphs/contributors * Rule-based French Lemmatizer (#2818) <!--- Provide a general summary of your changes in the title. --> ## Description <!--- Use this section to describe your changes. If your changes required testing, include information about the testing environment and the tests you ran. If your test fixes a bug reported in an issue, don't forget to include the issue number. If your PR is still a work in progress, that's totally fine – just include a note to let us know. --> Add a rule-based French Lemmatizer following the english one and the excellent PR for [greek language optimizations](https://github.com/explosion/spaCy/pull/2558) to adapt the Lemmatizer class. ### Types of change <!-- What type of change does your PR cover? Is it a bug fix, an enhancement or new feature, or a change to the documentation? --> - Lemma dictionary used can be found [here](http://infolingu.univ-mlv.fr/DonneesLinguistiques/Dictionnaires/telechargement.html), I used the XML version. - Add several files containing exhaustive list of words for each part of speech - Add some lemma rules - Add POS that are not checked in the standard Lemmatizer, i.e PRON, DET, ADV and AUX - Modify the Lemmatizer class to check in lookup table as a last resort if POS not mentionned - Modify the lemmatize function to check in lookup table as a last resort - Init files are updated so the model can support all the functionalities mentioned above - Add words to tokenizer_exceptions_list.py in respect to regex used in tokenizer_exceptions.py ## Checklist <!--- Before you submit the PR, go over this checklist and make sure you can tick off all the boxes. [] -> [x] --> - [X] I have submitted the spaCy Contributor Agreement. - [X] I ran the tests, and all new and existing tests passed. - [X] My changes don't require a change to the documentation, or if they do, I've added all required information. * Set version to 2.0.13 * Fix formatting and consistency * Update docs for new version [ci skip] * Increment version [ci skip] * Add info on wheels [ci skip] * Adding "This is a sentence" example to Sinhala (#2846) * Add wheels badge * Update badge [ci skip] * Update README.rst [ci skip] * Update murmurhash pin * Increment version to 2.0.14.dev0 * Update GPU docs for v2.0.14 * Add wheel to setup_requires * Import prefer_gpu and require_gpu functions from Thinc * Add tests for prefer_gpu() and require_gpu() * Update requirements and setup.py * Workaround bug in thinc require_gpu * Set version to v2.0.14 * Update push-tag script * Unhack prefer_gpu * Require thinc 6.10.6 * Update prefer_gpu and require_gpu docs [ci skip] * Fix specifiers for GPU * Set version to 2.0.14.dev1 * Set version to 2.0.14 * Update Thinc version pin * Increment version * Fix msgpack-numpy version pin * Increment version * Update version to 2.0.16 * Update version [ci skip] * Redundant ')' in the Stop words' example (#2856) <!--- Provide a general summary of your changes in the title. --> ## Description <!--- Use this section to describe your changes. If your changes required testing, include information about the testing environment and the tests you ran. If your test fixes a bug reported in an issue, don't forget to include the issue number. If your PR is still a work in progress, that's totally fine – just include a note to let us know. --> ### Types of change <!-- What type of change does your PR cover? Is it a bug fix, an enhancement or new feature, or a change to the documentation? --> ## Checklist <!--- Before you submit the PR, go over this checklist and make sure you can tick off all the boxes. [] -> [x] --> - [ ] I have submitted the spaCy Contributor Agreement. - [ ] I ran the tests, and all new and existing tests passed. - [ ] My changes don't require a change to the documentation, or if they do, I've added all required information. * Documentation improvement regarding joblib and SO (#2867) Some documentation improvements ## Description 1. Fixed the dead URL to joblib 2. Fixed Stack Overflow brand name (with space) ### Types of change Documentation ## Checklist <!--- Before you submit the PR, go over this checklist and make sure you can tick off all the boxes. [] -> [x] --> - [x] I have submitted the spaCy Contributor Agreement. - [x] I ran the tests, and all new and existing tests passed. - [x] My changes don't require a change to the documentation, or if they do, I've added all required information. * raise error when setting overlapping entities as doc.ents (#2880) * Fix out-of-bounds access in NER training The helper method state.B(1) gets the index of the first token of the buffer, or -1 if no such token exists. Normally this is safe because we pass this to functions like state.safe_get(), which returns an empty token. Here we used it directly as an array index, which is not okay! This error may have been the cause of out-of-bounds access errors during training. Similar errors may still be around, so much be hunted down. Hunting this one down took a long time...I printed out values across training runs and diffed, looking for points of divergence between runs, when no randomness should be allowed. * Change PyThaiNLP Url (#2876) * Fix missing comma * Add example showing a fix-up rule for space entities * Set version to 2.0.17.dev0 * Update regex version * Revert "Update regex version" This reverts commit 62358dd867d15bc6a475942dff34effba69dd70a. * Try setting older regex version, to align with conda * Set version to 2.0.17 * Add spacy-js to universe [ci-skip] * Add spacy-raspberry to universe (closes #2889) * Add script to validate universe json [ci skip] * Removed space in docs + added contributor indo (#2909) * - removed unneeded space in documentation * - added contributor info * Allow input text of length up to max_length, inclusive (#2922) * Include universe spec for spacy-wordnet component (#2919) * feat: include universe spec for spacy-wordnet component * chore: include spaCy contributor agreement * Minor formatting changes [ci skip] * Fix image [ci skip] Twitter URL doesn't work on live site * Check if the word is in one of the regular lists specific to each POS (#2886) * 💫 Create random IDs for SVGs to prevent ID clashes (#2927) Resolves #2924. ## Description Fixes problem where multiple visualizations in Jupyter notebooks would have clashing arc IDs, resulting in weirdly positioned arc labels. Generating a random ID prefix so even identical parses won't receive the same IDs for consistency (even if effect of ID clash isn't noticable here.) ### Types of change bug fix ## Checklist <!--- Before you submit the PR, go over this checklist and make sure you can tick off all the boxes. [] -> [x] --> - [x] I have submitted the spaCy Contributor Agreement. - [x] I ran the tests, and all new and existing tests passed. - [x] My changes don't require a change to the documentation, or if they do, I've added all required information. * Fix typo [ci skip] * fixes symbolic link on py3 and windows (#2949) * fixes symbolic link on py3 and windows during setup of spacy using command python -m spacy link en_core_web_sm en closes #2948 * Update spacy/compat.py Co-Authored-By: cicorias <cicorias@users.noreply.github.com> * Fix formatting * Update universe [ci skip] * Catalan Language Support (#2940) * Catalan language Support * Ddding Catalan to documentation * Sort languages alphabetically [ci skip] * Update tests for pytest 4.x (#2965) <!--- Provide a general summary of your changes in the title. --> ## Description - [x] Replace marks in params for pytest 4.0 compat ([see here](https://docs.pytest.org/en/latest/deprecations.html#marks-in-pytest-mark-parametrize)) - [x] Un-xfail passing tests (some fixes in a recent update resolved a bunch of issues, but tests were apparently never updated here) ### Types of change <!-- What type of change does your PR cover? Is it a bug fix, an enhancement or new feature, or a change to the documentation? --> ## Checklist <!--- Before you submit the PR, go over this checklist and make sure you can tick off all the boxes. [] -> [x] --> - [x] I have submitted the spaCy Contributor Agreement. - [x] I ran the tests, and all new and existing tests passed. - [x] My changes don't require a change to the documentation, or if they do, I've added all required information. * Fix regex pin to harmonize with conda (#2964) * Update README.rst * Fix bug where Vocab.prune_vector did not use 'batch_size' (#2977) Fixes #2976 * Fix typo * Fix typo * Remove duplicate file * Require thinc 7.0.0.dev2 Fixes bug in gpu_ops that would use cupy instead of numpy on CPU * Add missing import * Fix error IDs * Fix tests
2018-11-29 18:30:29 +03:00
syn_keys, syn_rows, scores = self.vectors.most_similar(toss, batch_size=batch_size)
2017-10-31 20:25:08 +03:00
2017-10-31 13:40:46 +03:00
remap = {}
2017-10-31 20:25:08 +03:00
for i, key in enumerate(keys[nr_row:]):
self.vectors.add(key, row=syn_rows[i])
word = self.strings[key]
synonym = self.strings[syn_keys[i]]
score = scores[i]
remap[word] = (synonym, score)
2017-10-31 04:00:26 +03:00
link_vectors_to_models(self)
return remap
2017-05-31 00:34:50 +03:00
def get_vector(self, orth, minn=None, maxn=None):
2017-10-27 20:45:19 +03:00
"""Retrieve a vector for a word in the vocabulary. Words can be looked
up by string or int ID. If no vectors data is loaded, ValueError is
raised.
2017-05-28 12:45:32 +03:00
2017-10-27 20:45:19 +03:00
RETURNS (numpy.ndarray): A word vector. Size
and shape determined by the `vocab.vectors` instance. Usually, a
numpy ndarray of shape (300,) and dtype float32.
2017-05-28 12:45:32 +03:00
"""
2017-08-19 20:52:25 +03:00
if isinstance(orth, basestring_):
orth = self.strings.add(orth)
word = self[orth].orth_
2017-08-22 20:46:35 +03:00
if orth in self.vectors.key2row:
return self.vectors[orth]
# Assign default ngram limits to minn and maxn which is the length of the word.
if minn is None:
minn = len(word)
if maxn is None:
maxn = len(word)
vectors = numpy.zeros((self.vectors_length,), dtype='f')
# Fasttext's ngram computation taken from https://github.com/facebookresearch/fastText
ngrams_size = 0;
for i in range(len(word)):
ngram = ""
if (word[i] and 0xC0) == 0x80:
continue
n = 1
j = i
while (j < len(word) and n <= maxn):
if n > maxn:
break
ngram += word[j]
j = j + 1
while (j < len(word) and (word[j] and 0xC0) == 0x80):
ngram += word[j]
j = j + 1
if (n >= minn and not (n == 1 and (i == 0 or j == len(word)))):
if self.strings[ngram] in self.vectors.key2row:
vectors = numpy.add(self.vectors[self.strings[ngram]],vectors)
ngrams_size += 1
n = n + 1
if ngrams_size > 0:
vectors = vectors * (1.0/ngrams_size)
return vectors
2017-05-28 12:45:32 +03:00
2017-05-31 00:34:50 +03:00
def set_vector(self, orth, vector):
2017-10-27 20:45:19 +03:00
"""Set a vector for a word in the vocabulary. Words can be referenced
by string or int ID.
2017-05-31 00:34:50 +03:00
"""
2017-10-31 20:25:08 +03:00
if isinstance(orth, basestring_):
orth = self.strings.add(orth)
if self.vectors.is_full and orth not in self.vectors:
new_rows = max(100, int(self.vectors.shape[0]*1.3))
if self.vectors.shape[1] == 0:
width = vector.size
else:
width = self.vectors.shape[1]
self.vectors.resize((new_rows, width))
lex = self[orth] # Adds worse to vocab
2017-10-31 20:25:08 +03:00
self.vectors.add(orth, vector=vector)
2017-08-19 21:35:33 +03:00
self.vectors.add(orth, vector=vector)
2017-05-31 00:34:50 +03:00
2017-05-28 12:45:32 +03:00
def has_vector(self, orth):
2017-10-27 20:45:19 +03:00
"""Check whether a word has a vector. Returns False if no vectors have
been loaded. Words can be looked up by string or int ID."""
2017-08-19 20:52:25 +03:00
if isinstance(orth, basestring_):
orth = self.strings.add(orth)
return orth in self.vectors
2017-05-28 12:45:32 +03:00
2017-08-18 21:46:56 +03:00
def to_disk(self, path, **exclude):
"""Save the current state to a directory.
path (unicode or Path): A path to a directory, which will be created if
2017-10-27 20:45:19 +03:00
it doesn't exist. Paths may be either strings or Path-like objects.
"""
path = util.ensure_path(path)
if not path.exists():
path.mkdir()
2017-05-29 00:34:12 +03:00
self.strings.to_disk(path / 'strings.json')
with (path / 'lexemes.bin').open('wb') as file_:
file_.write(self.lexemes_to_bytes())
2017-08-18 21:46:56 +03:00
if self.vectors is not None:
2017-08-19 22:27:35 +03:00
self.vectors.to_disk(path)
2017-08-18 21:46:56 +03:00
def from_disk(self, path, **exclude):
"""Loads state from a directory. Modifies the object in place and
returns it.
path (unicode or Path): A path to a directory. Paths may be either
strings or `Path`-like objects.
RETURNS (Vocab): The modified `Vocab` object.
"""
path = util.ensure_path(path)
2017-05-29 00:34:12 +03:00
self.strings.from_disk(path / 'strings.json')
with (path / 'lexemes.bin').open('rb') as file_:
self.lexemes_from_bytes(file_.read())
2017-08-18 21:46:56 +03:00
if self.vectors is not None:
2017-08-19 22:27:35 +03:00
self.vectors.from_disk(path, exclude='strings.json')
if self.vectors.name is not None:
link_vectors_to_models(self)
2017-05-29 00:34:12 +03:00
return self
2016-11-01 14:25:36 +03:00
def to_bytes(self, **exclude):
"""Serialize the current state to a binary string.
**exclude: Named attributes to prevent from being serialized.
RETURNS (bytes): The serialized form of the `Vocab` object.
"""
2017-08-18 21:46:56 +03:00
def deserialize_vectors():
if self.vectors is None:
return None
else:
return self.vectors.to_bytes()
2017-05-30 01:52:36 +03:00
getters = OrderedDict((
('strings', lambda: self.strings.to_bytes()),
('lexemes', lambda: self.lexemes_to_bytes()),
2017-08-18 21:46:56 +03:00
('vectors', deserialize_vectors)
2017-05-30 01:52:36 +03:00
))
2017-05-29 11:14:20 +03:00
return util.to_bytes(getters, exclude)
2017-05-21 15:18:46 +03:00
def from_bytes(self, bytes_data, **exclude):
"""Load state from a binary string.
bytes_data (bytes): The data to load from.
**exclude: Named attributes to prevent from being loaded.
RETURNS (Vocab): The `Vocab` object.
"""
2017-08-18 21:46:56 +03:00
def serialize_vectors(b):
if self.vectors is None:
return None
else:
return self.vectors.from_bytes(b)
2017-05-29 14:04:40 +03:00
setters = OrderedDict((
('strings', lambda b: self.strings.from_bytes(b)),
2017-05-30 01:52:36 +03:00
('lexemes', lambda b: self.lexemes_from_bytes(b)),
2017-08-18 21:46:56 +03:00
('vectors', lambda b: serialize_vectors(b))
2017-05-29 14:04:40 +03:00
))
2017-06-02 11:56:40 +03:00
util.from_bytes(bytes_data, setters, exclude)
if self.vectors.name is not None:
link_vectors_to_models(self)
2017-06-02 11:56:40 +03:00
return self
2017-05-29 00:34:12 +03:00
def lexemes_to_bytes(self):
cdef hash_t key
cdef size_t addr
2017-03-11 21:43:09 +03:00
cdef LexemeC* lexeme = NULL
cdef SerializedLexemeC lex_data
cdef int size = 0
for key, addr in self._by_orth.items():
if addr == 0:
continue
size += sizeof(lex_data.data)
byte_string = b'\0' * size
byte_ptr = <unsigned char*>byte_string
cdef int j
cdef int i = 0
for key, addr in self._by_orth.items():
if addr == 0:
continue
lexeme = <LexemeC*>addr
lex_data = Lexeme.c_to_bytes(lexeme)
for j in range(sizeof(lex_data.data)):
byte_ptr[i] = lex_data.data[j]
i += 1
return byte_string
2016-11-01 14:25:36 +03:00
def lexemes_from_bytes(self, bytes bytes_data):
"""Load the binary vocabulary data from the given string."""
cdef LexemeC* lexeme
2017-03-07 22:25:12 +03:00
cdef hash_t key
cdef unicode py_str
cdef int i = 0
cdef int j = 0
cdef SerializedLexemeC lex_data
chunk_size = sizeof(lex_data.data)
cdef void* ptr
cdef unsigned char* bytes_ptr = bytes_data
for i in range(0, len(bytes_data), chunk_size):
lexeme = <LexemeC*>self.mem.alloc(1, sizeof(LexemeC))
for j in range(sizeof(lex_data.data)):
lex_data.data[j] = bytes_ptr[i+j]
Lexeme.c_from_bytes(lexeme, lex_data)
2017-03-07 22:25:12 +03:00
ptr = self.strings._map.get(lexeme.orth)
if ptr == NULL:
continue
2017-03-07 22:25:12 +03:00
py_str = self.strings[lexeme.orth]
if self.strings[py_str] != lexeme.orth:
raise ValueError(Errors.E086.format(string=py_str,
orth_id=lexeme.orth,
hash_id=self.strings[py_str]))
2017-03-07 22:25:12 +03:00
self._by_orth.set(lexeme.orth, lexeme)
self.length += 1
2017-11-14 21:15:04 +03:00
def _reset_cache(self, keys, strings):
# I'm not sure this made sense. Disable it for now.
raise NotImplementedError
2017-11-14 21:15:04 +03:00
2017-03-07 22:25:12 +03:00
def pickle_vocab(vocab):
sstore = vocab.strings
2017-11-23 19:19:50 +03:00
vectors = vocab.vectors
2017-03-07 22:25:12 +03:00
morph = vocab.morphology
length = vocab.length
data_dir = vocab.data_dir
2018-12-06 20:53:16 +03:00
lex_attr_getters = srsly.pickle_dumps(vocab.lex_attr_getters)
lexemes_data = vocab.lexemes_to_bytes()
2017-03-07 22:25:12 +03:00
return (unpickle_vocab,
2017-11-23 19:19:50 +03:00
(sstore, vectors, morph, data_dir, lex_attr_getters, lexemes_data, length))
2017-03-07 22:25:12 +03:00
2017-11-23 19:19:50 +03:00
def unpickle_vocab(sstore, vectors, morphology, data_dir,
2017-10-27 20:45:19 +03:00
lex_attr_getters, bytes lexemes_data, int length):
2017-03-07 22:25:12 +03:00
cdef Vocab vocab = Vocab()
vocab.length = length
2017-11-23 19:19:50 +03:00
vocab.vectors = vectors
2017-03-07 22:25:12 +03:00
vocab.strings = sstore
vocab.morphology = morphology
vocab.data_dir = data_dir
2018-12-06 20:53:16 +03:00
vocab.lex_attr_getters = srsly.pickle_loads(lex_attr_getters)
vocab.lexemes_from_bytes(lexemes_data)
2017-03-07 22:25:12 +03:00
vocab.length = length
return vocab
copy_reg.pickle(Vocab, pickle_vocab, unpickle_vocab)