textattack.shared package

Shared TextAttack Functions

This package includes functions shared across packages.

• textattack.shared.utils package

Attacked Text Class

A helper class that represents a string that can be attacked.

class textattack.shared.attacked_text.AttackedText(text_input, attack_attrs=None)

Bases: object

A helper class that represents a string that can be attacked.

Models that take multiple sentences as input separate them by SPLIT_TOKEN. Attacks “see” the entire input, joined into one string, without the split token.

AttackedText instances that were perturbed from other AttackedText objects contain a pointer to the previous text (attack_attrs["previous_attacked_text"]), so that the full chain of perturbations might be reconstructed by using this key to form a linked list.

Parameters

• text (string) – The string that this AttackedText represents

• attack_attrs (dict) – Dictionary of various attributes stored during the course of an attack.

align_with_model_tokens(model_wrapper)

Align AttackedText’s words with target model’s tokenization scheme (e.g. word, character, subword). Specifically, we map each word to list of indices of tokens that compose the word (e.g. embedding –> [“em”, “##bed”, “##ding”])

Parameters

model_wrapper (textattack.models.wrappers.ModelWrapper) – ModelWrapper of the target model

Returns

Dictionary that maps i-th word to list of indices.

Return type

word2token_mapping (dict[int, list[int]])

all_words_diff(other_attacked_text)

Returns the set of indices for which this and other_attacked_text have different words.

convert_from_original_idxs(idxs)

Takes indices of words from original string and converts them to indices of the same words in the current string.

Uses information from self.attack_attrs['original_index_map'], which maps word indices from the original to perturbed text.

delete_word_at_index(index)

This code returns a new AttackedText object where the word at index is removed.

first_word_diff(other_attacked_text)

Returns the first word in self.words that differs from other_attacked_text.

Useful for word swap strategies.

first_word_diff_index(other_attacked_text)

Returns the index of the first word in self.words that differs from other_attacked_text.

Useful for word swap strategies.

free_memory()

Delete items that take up memory.

Can be called once the AttackedText is only needed to display.

generate_new_attacked_text(new_words)

Returns a new AttackedText object and replaces old list of words with a new list of words, but preserves the punctuation and spacing of the original message.

self.words is a list of the words in the current text with punctuation removed. However, each “word” in new_words could be an empty string, representing a word deletion, or a string with multiple space-separated words, representation an insertion of one or more words.

get_deletion_indices()

insert_text_after_word_index(index, text)

Inserts a string before word at index index and attempts to add appropriate spacing.

insert_text_before_word_index(index, text)

Inserts a string before word at index index and attempts to add appropriate spacing.

ith_word_diff(other_attacked_text, i)

Returns whether the word at index i differs from other_attacked_text.

ner_of_word_index(desired_word_idx, model_name='ner')

Returns the ner tag of the word at index word_idx.

Uses FLAIR ner tagger.

pos_of_word_index(desired_word_idx)

Returns the part-of-speech of the word at index word_idx.

Uses FLAIR part-of-speech tagger.

printable_text(key_color='bold', key_color_method=None)

Represents full text input. Adds field descriptions.

For example, entailment inputs look like:

` premise: ... hypothesis: ... `

replace_word_at_index(index, new_word)

This code returns a new AttackedText object where the word at index is replaced with a new word.

replace_words_at_indices(indices, new_words)

This code returns a new AttackedText object where the word at index is replaced with a new word.

text_after_word_index(i)

Returns the text after the end of word at index i.

text_until_word_index(i)

Returns the text before the beginning of word at index i.

text_window_around_index(index, window_size)

The text window of window_size words centered around index.

words_diff_num(other_attacked_text)

words_diff_ratio(x)

Get the ratio of words difference between current text and x.

Note that current text and x must have same number of words.

SPLIT_TOKEN = '<SPLIT>'

property column_labels

Returns the labels for this text’s columns.

For single-sequence inputs, this simply returns [‘text’].

property num_words

Returns the number of words in the sequence.

property text

Represents full text input.

Multiply inputs are joined with a line break.

property tokenizer_input

The tuple of inputs to be passed to the tokenizer.

property words

property words_per_input

Returns a list of lists of words corresponding to each input.

Misc Checkpoints

The AttackCheckpoint class saves in-progress attacks and loads saved attacks from disk.

class textattack.shared.checkpoint.AttackCheckpoint(attack_args, attack_log_manager, worklist, worklist_candidates, chkpt_time=None)

Bases: object

An object that stores necessary information for saving and loading checkpoints.

Parameters

• attack_args (textattack.AttackArgs) – Arguments of the original attack

• attack_log_manager (textattack.loggers.AttackLogManager) – Object for storing attack results

• worklist (deque[int]) – List of examples that will be attacked. Examples are represented by their indicies within the dataset.

• worklist_candidates (int) – List of other available examples we can attack. Used to get the next dataset element when attack_n=True.

• chkpt_time (float) – epoch time representing when checkpoint was made

classmethod load(path)

save(quiet=False)

property dataset_offset

Calculate offset into the dataset to start from.

property datetime

property num_failed_attacks

property num_maximized_attacks

property num_remaining_attacks

property num_skipped_attacks

property num_successful_attacks

property results_count

Return number of attacks made so far.

Shared data fields

Lists of named entities: countries, nationalities, cities.

Lists of person names, first and last.

Misc Validators

Validators ensure compatibility between search methods, transformations, constraints, and goal functions.

textattack.shared.validators.transformation_consists_of(transformation, transformation_classes)

Determines if transformation is or consists only of instances of a class in transformation_classes

textattack.shared.validators.transformation_consists_of_word_swaps(transformation)

Determines if transformation is a word swap or consists of only word swaps.

textattack.shared.validators.transformation_consists_of_word_swaps_and_deletions(transformation)

Determines if transformation is a word swap or consists of only word swaps and deletions.

textattack.shared.validators.validate_model_goal_function_compatibility(goal_function_class, model_class)

Determines if model_class is task-compatible with goal_function_class.

For example, a text-generative model like one intended for translation or summarization would not be compatible with a goal function that requires probability scores, like the UntargetedGoalFunction.

textattack.shared.validators.validate_model_gradient_word_swap_compatibility(model)

Determines if model is task-compatible with GradientBasedWordSwap.

We can only take the gradient with respect to an individual word if the model uses a word-based tokenizer.

Shared loads word embeddings and related distances

class textattack.shared.word_embeddings.AbstractWordEmbedding

Bases: abc.ABC

Abstract class representing word embedding used by TextAttack.

This class specifies all the methods that is required to be defined so that it can be used for transformation and constraints. For custom word embedding not supported by TextAttack, please create a class that inherits this class and implement the required methods. However, please first check if you can use WordEmbedding class, which has a lot of internal methods implemented.

abstract get_cos_sim(a, b)

Return cosine similarity between vector for word a and vector for word b.

Since this is a metric, get_mse_dist(a,b) and get_mse_dist(b,a) should return the same value. :param a: Either word or integer presenting the id of the word :type a: Union[str|int] :param b: Either word or integer presenting the id of the word :type b: Union[str|int]

Returns

cosine similarity

Return type

distance (float)

abstract get_mse_dist(a, b)

Return MSE distance between vector for word a and vector for word b.

Since this is a metric, get_mse_dist(a,b) and get_mse_dist(b,a) should return the same value. :param a: Either word or integer presenting the id of the word :type a: Union[str|int] :param b: Either word or integer presenting the id of the word :type b: Union[str|int]

Returns

MSE (L2) distance

Return type

distance (float)

abstract index2word(index)

Convert index to corresponding word :param index: :type index: int

Returns

word (str)

abstract nearest_neighbours(index, topn)

Get top-N nearest neighbours for a word :param index: ID of the word for which we’re finding the nearest neighbours :type index: int :param topn: Used for specifying N nearest neighbours :type topn: int

Returns

List of indices of the nearest neighbours

Return type

neighbours (list[int])

abstract word2index(word)

Convert between word to id (i.e. index of word in embedding matrix) :param word: :type word: str

Returns

index (int)

class textattack.shared.word_embeddings.GensimWordEmbedding(keyed_vectors)

Bases: textattack.shared.word_embeddings.AbstractWordEmbedding

Wraps Gensim’s models.keyedvectors module (https://radimrehurek.com/gensim/models/keyedvectors.html)

get_cos_sim(a, b)

Return cosine similarity between vector for word a and vector for word b.

Since this is a metric, get_mse_dist(a,b) and get_mse_dist(b,a) should return the same value. :param a: Either word or integer presenting the id of the word :type a: Union[str|int] :param b: Either word or integer presenting the id of the word :type b: Union[str|int]

Returns

cosine similarity

Return type

distance (float)

get_mse_dist(a, b)

Return MSE distance between vector for word a and vector for word b.

Since this is a metric, get_mse_dist(a,b) and get_mse_dist(b,a) should return the same value. :param a: Either word or integer presenting the id of the word :type a: Union[str|int] :param b: Either word or integer presenting the id of the word :type b: Union[str|int]

Returns

MSE (L2) distance

Return type

distance (float)

index2word(index)

Convert index to corresponding word :param index: :type index: int

Returns

word (str)

nearest_neighbours(index, topn, return_words=True)

Get top-N nearest neighbours for a word :param index: ID of the word for which we’re finding the nearest neighbours :type index: int :param topn: Used for specifying N nearest neighbours :type topn: int

Returns

List of indices of the nearest neighbours

Return type

neighbours (list[int])

word2index(word)

Convert between word to id (i.e. index of word in embedding matrix) :param word: :type word: str

Returns

index (int)

class textattack.shared.word_embeddings.WordEmbedding(embedding_matrix, word2index, index2word, nn_matrix=None)

Bases: textattack.shared.word_embeddings.AbstractWordEmbedding

Object for loading word embeddings and related distances for TextAttack. This class has a lot of internal components (e.g. get consine similarity) implemented. Consider using this class if you can provide the appropriate input data to create the object.

Parameters

• emedding_matrix (ndarray) – 2-D array of shape N x D where N represents size of vocab and D is the dimension of embedding vectors.

• word2index (Union[dict|object]) – dictionary (or a similar object) that maps word to its index with in the embedding matrix.

• index2word (Union[dict|object]) – dictionary (or a similar object) that maps index to its word.

• nn_matrix (ndarray) – Matrix for precomputed nearest neighbours. It should be a 2-D integer array of shape N x K where N represents size of vocab and K is the top-K nearest neighbours. If this is set to None, we have to compute nearest neighbours on the fly for nearest_neighbours method, which is costly.

static counterfitted_GLOVE_embedding()

Returns a prebuilt counter-fitted GLOVE word embedding proposed by “Counter-fitting Word Vectors to Linguistic Constraints” (Mrkšić et al., 2016)

get_cos_sim(a, b)

Return cosine similarity between vector for word a and vector for word b.

Since this is a metric, get_mse_dist(a,b) and get_mse_dist(b,a) should return the same value. :param a: Either word or integer presenting the id of the word :type a: Union[str|int] :param b: Either word or integer presenting the id of the word :type b: Union[str|int]

Returns

cosine similarity

Return type

distance (float)

get_mse_dist(a, b)

Return MSE distance between vector for word a and vector for word b.

Since this is a metric, get_mse_dist(a,b) and get_mse_dist(b,a) should return the same value. :param a: Either word or integer presenting the id of the word :type a: Union[str|int] :param b: Either word or integer presenting the id of the word :type b: Union[str|int]

Returns

MSE (L2) distance

Return type

distance (float)

index2word(index)

Convert index to corresponding word :param index: :type index: int

Returns

word (str)

nearest_neighbours(index, topn)

Get top-N nearest neighbours for a word :param index: ID of the word for which we’re finding the nearest neighbours :type index: int :param topn: Used for specifying N nearest neighbours :type topn: int

Returns

List of indices of the nearest neighbours

Return type

neighbours (list[int])

word2index(word)

Convert between word to id (i.e. index of word in embedding matrix) :param word: :type word: str

Returns

index (int)

PATH = 'word_embeddings'