Skip to content

HISTORY.rst

Latest commit

 

History

History
174 lines (147 loc) · 9.67 KB

HISTORY.rst

File metadata and controls

174 lines (147 loc) · 9.67 KB

Changelog

0.9.0 (2014-09-15)

  • TextBlob now depends on NLTK 3. The vendorized version of NLTK has been removed.
  • Fix bug that raised a SyntaxError when translating text with non-ascii characters on Python 3.
  • Fix bug that showed "double-escaped" unicode characters in translator output (issue #56). Thanks Evan Dempsey.
  • Backwards-incompatible: Completely remove import text.blob. You should import textblob instead.
  • Backwards-incompatible: Completely remove PerceptronTagger. Install textblob-aptagger instead.
  • Backwards-incompatible: Rename TextBlobException to TextBlobError and MissingCorpusException to MissingCorpusError.
  • Backwards-incompatible: Format classes are passed a file object rather than a file path.
  • Backwards-incompatible: If training a classifier with data from a file, you must pass a file object (rather than a file path).
  • Updated English sentiment corpus.
  • Add feature_extractor parameter to NaiveBayesAnalyzer.
  • Add textblob.formats.get_registry() and textblob.formats.register() which allows users to register custom data source formats.
  • Change BaseClassifier.detect from a staticmethod to a classmethod.
  • Improved docs.
  • Tested on Python 3.4.

0.8.4 (2014-02-02)

  • Fix display (__repr__) of WordList slices on Python 3.
  • Add download_corpora module. Corpora must now be downloaded using python -m textblob.download_corpora.

0.8.3 (2013-12-29)

  • Sentiment analyzers return namedtuples, e.g. Sentiment(polarity=0.12, subjectivity=0.34).
  • Memory usage improvements to NaiveBayesAnalyzer and basic_extractor (default feature extractor for classifiers module).
  • Add textblob.tokenizers.sent_tokenize and textblob.tokenizers.word_tokenize convenience functions.
  • Add textblob.classifiers.MaxEntClassifer.
  • Improved NLTKTagger.

0.8.2 (2013-12-21)

  • Fix bug in spelling correction that stripped some punctuation (Issue #48).
  • Various improvements to spelling correction: preserves whitespace characters (Issue #12); handle contractions and punctuation between words. Thanks @davidnk.
  • Make TextBlob.words more memory-efficient.
  • Translator now sends POST instead of GET requests. This allows for larger bodies of text to be translated (Issue #49).
  • Update pattern tagger for better accuracy.

0.8.1 (2013-11-16)

  • Fix bug that caused ValueError upon sentence tokenization. This removes modifications made to the NLTK sentence tokenizer.
  • Add Word.lemmatize() method that allows passing in a part-of-speech argument.
  • Word.lemma returns correct part of speech for Word objects that have their pos attribute set. Thanks @RomanYankovsky.

0.8.0 (2013-10-23)

  • Backwards-incompatible: Renamed package to textblob. This avoids clashes with other namespaces called text. TextBlob should now be imported with from textblob import TextBlob.
  • Update pattern resources for improved parser accuracy.
  • Update NLTK.
  • Allow Translator to connect to proxy server.
  • PerceptronTagger completely deprecated. Install the textblob-aptagger extension instead.

0.7.1 (2013-09-30)

  • Bugfix updates.
  • Fix bug in feature extraction for NaiveBayesClassifier.
  • basic_extractor is now case-sensitive, e.g. contains(I) != contains(i)
  • Fix repr output when a TextBlob contains non-ascii characters.
  • Fix part-of-speech tagging with PatternTagger on Windows.
  • Suppress warning about not having scikit-learn installed.

0.7.0 (2013-09-25)

  • Wordnet integration. Word objects have synsets and definitions properties. The text.wordnet module allows you to create Synset and Lemma objects directly.
  • Move all English-specific code to its own module, text.en.
  • Basic extensions framework in place. TextBlob has been refactored to make it easier to develop extensions.
  • Add text.classifiers.PositiveNaiveBayesClassifier.
  • Update NLTK.
  • NLTKTagger now working on Python 3.
  • Fix __str__ behavior. print(blob) should now print non-ascii text correctly in both Python 2 and 3.
  • Backwards-incompatible: All abstract base classes have been moved to the text.base module.
  • Backwards-incompatible: PerceptronTagger will now be maintained as an extension, textblob-aptagger. Instantiating a text.taggers.PerceptronTagger() will raise a DeprecationWarning.

0.6.3 (2013-09-15)

  • Word tokenization fix: Words that stem from a contraction will still have an apostrophe, e.g. "Let's" => ["Let", "'s"].
  • Fix bug with comparing blobs to strings.
  • Add text.taggers.PerceptronTagger, a fast and accurate POS tagger. Thanks @syllog1sm.
  • Note for Python 3 users: You may need to update your corpora, since NLTK master has reorganized its corpus system. Just run curl https://raw.github.com/sloria/TextBlob/master/download_corpora.py | python again.
  • Add download_corpora_lite.py script for getting the minimum corpora requirements for TextBlob's basic features.

0.6.2 (2013-09-05)

  • Fix bug that resulted in a UnicodeEncodeError when tagging text with non-ascii characters.
  • Add DecisionTreeClassifier.
  • Add labels() and train() methods to classifiers.

0.6.1 (2013-09-01)

  • Classifiers can be trained and tested on CSV, JSON, or TSV data.
  • Add basic WordNet lemmatization via the Word.lemma property.
  • WordList.pluralize() and WordList.singularize() methods return WordList objects.

0.6.0 (2013-08-25)

  • Add Naive Bayes classification. New text.classifiers module, TextBlob.classify(), and Sentence.classify() methods.
  • Add parsing functionality via the TextBlob.parse() method. The text.parsers module currently has one implementation (PatternParser).
  • Add spelling correction. This includes the TextBlob.correct() and Word.spellcheck() methods.
  • Update NLTK.
  • Backwards incompatible: clean_html has been deprecated, just as it has in NLTK. Use Beautiful Soup's soup.get_text() method for HTML-cleaning instead.
  • Slight API change to language translation: if from_lang isn't specified, attempts to detect the language.
  • Add itokenize() method to tokenizers that returns a generator instead of a list of tokens.

0.5.3 (2013-08-21)

  • Unicode fixes: This fixes a bug that sometimes raised a UnicodeEncodeError upon creating accessing sentences for TextBlobs with non-ascii characters.
  • Update NLTK

0.5.2 (2013-08-14)

  • Important patch update for NLTK users: Fix bug with importing TextBlob if local NLTK is installed.
  • Fix bug with computing start and end indices of sentences.

0.5.1 (2013-08-13)

  • Fix bug that disallowed display of non-ascii characters in the Python REPL.
  • Backwards incompatible: Restore blob.json property for backwards compatibility with textblob<=0.3.10. Add a to_json() method that takes the same arguments as json.dumps.
  • Add WordList.append and WordList.extend methods that append Word objects.

0.5.0 (2013-08-10)

  • Language translation and detection API!
  • Add text.sentiments module. Contains the PatternAnalyzer (default implementation) as well as a NaiveBayesAnalyzer.
  • Part-of-speech tags can be accessed via TextBlob.tags or TextBlob.pos_tags.
  • Add polarity and subjectivity helper properties.

0.4.0 (2013-08-05)

  • New text.tokenizers module with WordTokenizer and SentenceTokenizer. Tokenizer instances (from either textblob itself or NLTK) can be passed to TextBlob's constructor. Tokens are accessed through the new tokens property.
  • New Blobber class for creating TextBlobs that share the same tagger, tokenizer, and np_extractor.
  • Add ngrams method.
  • Backwards-incompatible: TextBlob.json() is now a method, not a property. This allows you to pass arguments (the same that you would pass to json.dumps()).
  • New home for documentation: https://textblob.readthedocs.org/
  • Add parameter for cleaning HTML markup from text.
  • Minor improvement to word tokenization.
  • Updated NLTK.
  • Fix bug with adding blobs to bytestrings.

0.3.10 (2013-08-02)

  • Bundled NLTK no longer overrides local installation.
  • Fix sentiment analysis of text with non-ascii characters.

0.3.9 (2013-07-31)

  • Updated nltk.
  • ConllExtractor is now Python 3-compatible.
  • Improved sentiment analysis.
  • Blobs are equal (with ==) to their string counterparts.
  • Added instructions to install textblob without nltk bundled.
  • Dropping official 3.1 and 3.2 support.

0.3.8 (2013-07-30)

  • Importing TextBlob is now much faster. This is because the noun phrase parsers are trained only on the first call to noun_phrases (instead of training them every time you import TextBlob).
  • Add text.taggers module which allows user to change which POS tagger implementation to use. Currently supports PatternTagger and NLTKTagger (NLTKTagger only works with Python 2).
  • NPExtractor and Tagger objects can be passed to TextBlob's constructor.
  • Fix bug with POS-tagger not tagging one-letter words.
  • Rename text/np_extractor.py -> text/np_extractors.py
  • Add run_tests.py script.

0.3.7 (2013-07-28)

  • Every word in a Blob or Sentence is a Word instance which has methods for inflection, e.g word.pluralize() and word.singularize().
  • Updated the np_extractor module. Now has an new implementation, ConllExtractor that uses the Conll2000 chunking corpus. Only works on Py2.