o
Zhi @ s@ d Z ddlZddlZddlZddlmZ ddlmZmZm Z ddl
Z
ddl m Z ddl
mZ eeZdd d
Zd d ZG d
d de Z dZdZededdddddf Ze
dde e
je
jB e
j B Z e
dZ e
ee
je
jB e
j B Z e
dZ d-d d Z!d.d"d#Z"G d$d% d%Z#d&d' Z$d(d) Z%d/d+d,Z&dgZ'dS )0z!Tokenization classes for BERTweet N)copyfile)ListOptionalTuple )PreTrainedTokenizer)loggingz vocab.txtz bpe.codes)
vocab_file merges_filec C s> t }| d }| dd D ] }|||f |}q
t |}|S )z
Return set of symbol pairs in a word.
Word is represented as tuple of symbols (symbols being variable-length strings).
r N)setadd)wordpairsZ prev_charchar r a/var/www/auris/lib/python3.10/site-packages/transformers/models/bertweet/tokenization_bertweet.py get_pairs& s r c
s0 e Zd ZdZeZ d- fdd Z
d.d ee d e ee d
ee fddZ
d/d ee d e ee de d
ee f fdd
Z
d.d ee d e ee d
ee fddZ
edd Zdd Zdd Zdd Zd d Zd d Zd!d" Zd#d$ Zd%d& Zd.d'ed(e e d
ee fd)d*Zd+d, Z Z S )0BertweetTokenizera
Constructs a BERTweet tokenizer, using Byte-Pair-Encoding.
This tokenizer inherits from [`PreTrainedTokenizer`] which contains most of the main methods. Users should refer to
this superclass for more information regarding those methods.
Args:
vocab_file (`str`):
Path to the vocabulary file.
merges_file (`str`):
Path to the merges file.
normalization (`bool`, *optional*, defaults to `False`):
Whether or not to apply a normalization preprocess.
bos_token (`str`, *optional*, defaults to `""`):
The beginning of sequence token that was used during pretraining. Can be used a sequence classifier token.
When building a sequence using special tokens, this is not the token that is used for the beginning of
sequence. The token used is the `cls_token`.
eos_token (`str`, *optional*, defaults to `""`):
The end of sequence token.
When building a sequence using special tokens, this is not the token that is used for the end of sequence.
The token used is the `sep_token`.
sep_token (`str`, *optional*, defaults to `""`):
The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for
sequence classification or for a text and a question for question answering. It is also used as the last
token of a sequence built with special tokens.
cls_token (`str`, *optional*, defaults to `""`):
The classifier token which is used when doing sequence classification (classification of the whole sequence
instead of per-token classification). It is the first token of the sequence when built with special tokens.
unk_token (`str`, *optional*, defaults to `""`):
The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this
token instead.
pad_token (`str`, *optional*, defaults to `""`):
The token used for padding, for example when batching sequences of different lengths.
mask_token (`str`, *optional*, defaults to `""`):
The token used for masking values. This is the token used when training this model with masked language
modeling. This is the token which the model will try to predict.
F X `
- pair of sequences: ` A B `
Args:
token_ids_0 (`List[int]`):
List of IDs to which the special tokens will be added.
token_ids_1 (`List[int]`, *optional*):
Optional second list of IDs for sequence pairs.
Returns:
`List[int]`: List of [input IDs](../glossary#input-ids) with the appropriate special tokens.
N) cls_token_id sep_token_id)rJ rN rO clssepr r r build_inputs_with_special_tokens s
z2BertweetTokenizer.build_inputs_with_special_tokensalready_has_special_tokensc sh |r t j||ddS |du r dgdgt| dg S dgdgt| ddg dgt| dg S )a
Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
special tokens using the tokenizer `prepare_for_model` method.
Args:
token_ids_0 (`List[int]`):
List of IDs.
token_ids_1 (`List[int]`, *optional*):
Optional second list of IDs for sequence pairs.
already_has_special_tokens (`bool`, *optional*, defaults to `False`):
Whether or not the token list is already formatted with special tokens for the model.
Returns:
`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
T)rN rO rV Nr r )rH get_special_tokens_maskrB )rJ rN rO rV rL r r rW s 0z)BertweetTokenizer.get_special_tokens_maskc C sP | j g}| jg}|du rt|| | dg S t|| | | | | dg S )a
Create a mask from the two sequences passed to be used in a sequence-pair classification task. BERTweet does
not make use of token type ids, therefore a list of zeros is returned.
Args:
token_ids_0 (`List[int]`):
List of IDs.
token_ids_1 (`List[int]`, *optional*):
Optional second list of IDs for sequence pairs.
Returns:
`List[int]`: List of zeros.
Nr )rR rQ rB )rJ rN rO rT rS r r r $create_token_type_ids_from_sequences s
"z6BertweetTokenizer.create_token_type_ids_from_sequencesc C s
t | jS N)rB r8 rJ r r r
vocab_size s
z BertweetTokenizer.vocab_sizec C s t | jfi | jS rY )r? r8 Zadded_tokens_encoderrZ r r r get_vocab s zBertweetTokenizer.get_vocabc
s | j v r
j | S t|}tt|d d |d d g }t|}|s'|S t| fddd}| jvr8ny|\}}g }d}|t|k rz|||} W n ty` | ||d Y n?w | ||| | }|| |kr|t|d k r||d |kr|
|| |d 7 }n |
|| |d7 }|t|k sFt|}|}t|dkrnt|}q(d
|}|d d }| j |< |S ) Nr% zTc s j | tdS )Ninf)rC getfloat)pairrZ r r s z'BertweetTokenizer.bpe..)keyr r r @@ ) rD r& listr minrC rB index
ValueErrorextendappendjoin)
rJ tokenr r ZbigramfirstsecondZnew_wordijr rZ r bpe sN
"
,
zBertweetTokenizer.bpec C sH | j r| |}g }td|}|D ]}|t| |d q|S )zTokenize a string.z\S+\n? )r+ normalizeTweetrefindallri re rq r' )rJ textZ split_tokenswordsrl r r r _tokenize( s
zBertweetTokenizer._tokenizec s j D ] }|| j | }q j|}d fdd|D }|ddddddd d
d d }|d
ddddddddddd}|dddd d d d d }d| S )!z'
Normalize a raw Tweet
rr c s g | ]} |qS r )normalizeToken)r rl rZ r r r) ; r z4BertweetTokenizer.normalizeTweet..zcannot zcan not zn't z n't zn 't zca n'tzcan'tzai n'tzain'tz'm z 'm z're z 're z's z 's z'll z 'll z'd z 'd z've z 've z p . m .z p.m.z p . m z p.m z a . m .z a.m.z a . m z a.m )rG replacerF tokenizerk r' )rJ ZtweetpuncttokensZ normTweetr rZ r rs 3 s.
z BertweetTokenizer.normalizeTweetc C sj | }|dr dS |ds|drdS t|dkr3|| jv r'| j| S | jdur1| |S |S |S )z-
Normalize tokens in a Tweet
@z@USERhttpwwwZHTTPURLr N)lower
startswithrB rG r4 )rJ rl Zlowercased_tokenr r r ry U s
z BertweetTokenizer.normalizeTokenc C s | j || j | jS )z0Converts a token (str) in an id using the vocab.)r8 r^ r0 )rJ rl r r r _convert_token_to_idh s z&BertweetTokenizer._convert_token_to_idc C s | j || jS )z=Converts an index (integer) in a token (str) using the vocab.)r< r^ r0 )rJ rg r r r _convert_id_to_tokenl s z&BertweetTokenizer._convert_id_to_tokenc C s d |dd }|S )z:Converts a sequence of tokens (string) in a single string.rr rc )rk rz strip)rJ r} Z
out_stringr r r convert_tokens_to_stringp s z*BertweetTokenizer.convert_tokens_to_stringsave_directoryfilename_prefixc C s t j|std| d d S t j||r|d ndtd }t j||r,|d ndtd }t j| jt j|krNt j | jrNt
| j| n&t j | jstt |d}| j
}|| W d n1 sow Y t j| jt j|krt
| j| ||fS )NzVocabulary path (z) should be a directory-r r r
wb)ospathisdirr6 errorrk VOCAB_FILES_NAMESabspathr isfiler r= Zsp_modelZserialized_model_protowriter
)rJ r r Zout_vocab_fileZout_merge_filefiZcontent_spiece_modelr r r save_vocabularyu s& (
z!BertweetTokenizer.save_vocabularyc
C s t |trCz!t|ddd}| | W d W dS 1 s w Y W dS ty4 } z|d}~w tyB td| dw | }|D ]!}| }| d}|dkr\t
d |d| }t | j | j |< qIdS )
zi
Loads a pre-existing dictionary from a text file and adds its symbols to this instance.
rr! r" Nz Incorrect encoding detected in z , please rebuild the datasetrr r% z5Incorrect dictionary format, expected ' ')
isinstancer9 r= r: FileNotFoundError UnicodeError Exception readlinesr rfindrh rB r8 ) rJ ffdZfnfelinesZlineTmplineidxr r r r r: s.
z BertweetTokenizer.add_from_file)Fr r r r r r r rY )NF) __name__
__module__ __qualname____doc__r Zvocab_files_namesrI r intr rU boolrW rX propertyr[ r\ rq rx rs ry r r r r9 r r r:
__classcell__r r rL r r 6 sd 2=
, " r ac
(?:
[<>]?
[:;=8] # eyes
[\-o\*\']? # optional nose
[\)\]\(\[dDpP/\:\}\{@\|\\] # mouth
|
[\)\]\(\[dDpP/\:\}\{@\|\\] # mouth
[\-o\*\']? # optional nose
[:;=8] # eyes
[<>]?
|
<3 # heart
)u # Capture 1: entire matched URL
(?:
https?: # URL protocol and colon
(?:
/{1,3} # 1-3 slashes
| # or
[a-z0-9%] # Single letter or digit or '%'
# (Trying not to match e.g. "URI::Escape")
)
| # or
# looks like domain name followed by a slash:
[a-z0-9.\-]+[.]
(?:[a-z]{2,13})
/
)
(?: # One or more:
[^\s()<>{}\[\]]+ # Run of non-space, non-()<>{}[]
| # or
\([^\s()]*?\([^\s()]+\)[^\s()]*?\) # balanced parens, one level deep: (...(...)...)
|
\([^\s]+?\) # balanced parens, non-recursive: (...)
)+
(?: # End with:
\([^\s()]*?\([^\s()]+\)[^\s()]*?\) # balanced parens, one level deep: (...(...)...)
|
\([^\s]+?\) # balanced parens, non-recursive: (...)
| # or
[^\s`!()\[\]{};:'".,<>?«»“”‘’] # not a space or one of these punct chars
)
| # OR, the following to match naked domains:
(?:
(?\s]+>z
[\-]+>|<[\-]+z (?:@[\w_]+)z (?:\#+[\w_]+[\w\'_\-]*[\w_]+)z#[\w.+-]+@[\w-]+\.(?:[\w-]\.?)+[\w-]a
(?:[^\W\d_](?:[^\W\d_]|['\-_])+[^\W\d_]) # Words with apostrophes or dashes.
|
(?:[+\-]?\d+[,/.:-]\d+[+\-]?) # Numbers, including fractions, decimals.
|
(?:[\w_]+) # Words without apostrophes or dashes.
|
(?:\.(?:\s*\.){1,}) # Ellipsis dots.
|
(?:\S) # Everything else that isn't whitespace.
z(%s)|z([^a-zA-Z0-9])\1{3,}z&(#?(x?))([^&;\s]+);strictc C s& |d u rd}t | tr| ||S | S )Nr! )r bytesdecode)rv r# errorsr r r _str_to_unicode\ s
r r Tr! c s fdd}t |t| |S )u
Remove entities from text by converting them to their corresponding unicode character.
Args:
text:
A unicode string or a byte string encoded in the given *encoding* (which defaults to 'utf-8').
keep (list):
List of entity names which should not be replaced. This supports both numeric entities (`&#nnnn;` and
`&#hhhh;`) and named entities (such as ` ` or `>`).
remove_illegal (bool):
If `True`, entities that can't be converted are removed. Otherwise, entities that can't be converted are
kept "as is".
Returns: A unicode string with the entities removed.
See https://github.com/scrapy/w3lib/blob/master/w3lib/html.py
Examples:
```python
>>> from nltk.tokenize.casual import _replace_html_entities
>>> _replace_html_entities(b"Price: £100")
'Price: \xa3100'
>>> print(_replace_html_entities(b"Price: £100"))
Price: £100
```c s | d}| dr=z'| drt|d}nt|d}d| kr%dkr0n n t|fdW S W n ty< d }Y nw | v rF| d S tjj|}|d urbzt |W S tt
fya Y nw rfd
S | d S ) Nr r r
cp1252r r ) groupr r r rh htmlentitiesname2codepointr^ chr
OverflowError)matchZ entity_bodynumberkeepremove_illegalr r _convert_entity s,
z/_replace_html_entities.._convert_entity)ENT_REsubr )rv r r r# r r r r _replace_html_entitiesd s r c @ s" e Zd ZdZd ddZdd ZdS )
rE a
Examples:
```python
>>> # Tokenizer for tweets.
>>> from nltk.tokenize import TweetTokenizer
>>> tknzr = TweetTokenizer()
>>> s0 = "This is a cooool #dummysmiley: :-) :-P <3 and some arrows < > -> <--"
>>> tknzr.tokenize(s0)
['This', 'is', 'a', 'cooool', '#dummysmiley', ':', ':-)', ':-P', '<3', 'and', 'some', 'arrows', '<', '>', '->', '<--']
>>> # Examples using *strip_handles* and *reduce_len parameters*:
>>> tknzr = TweetTokenizer(strip_handles=True, reduce_len=True)
>>> s1 = "@remy: This is waaaaayyyy too much for you!!!!!!"
>>> tknzr.tokenize(s1)
[':', 'This', 'is', 'waaayyy', 'too', 'much', 'for', 'you', '!', '!', '!']
```TFc C s || _ || _|| _d S rY
preserve_case
reduce_len
strip_handles)rJ r r r r r r rI s
zTweetTokenizer.__init__c C sR t |}| jr t|}| jrt|}td|}t|}| j s'dd |D }|S )z
Args:
text: str
Returns: list(str) A tokenized list of strings; concatenating this list returns the original string if
`preserve_case=False`
\1\1\1c S s" g | ]
}t |r |n| qS r ) EMOTICON_REsearchr )r xr r r r) s " z+TweetTokenizer.tokenize..)
r r remove_handlesr reduce_lengtheningHANG_REr WORD_REru r )rJ rv Z safe_textrw r r r r{ s
zTweetTokenizer.tokenizeNTFF)r r r r rI r{ r r r r rE s
rE c C t d}|d| S )za
Replace repeated character sequences of length 3 or greater with sequences of length 3.
z (.)\1{2,}r regexcompiler rv patternr r r r s
r c C r )z4
Remove Twitter username handles from text.
zv(? sT
&.$0
A8