Release 0.9.0

This represents possibly the largest minor release thus far. There is quite a lot of breaking changes involved in this version. Upgrade at your own risk.

Bugs

Changes

Features

Version 0.8.0

Bugs

• Make the TC-XML writer write validity properly.

Changes

• Add parsing/convenience module.
  • The module contains helper/convenience methods for parsing.
  • Added helper method for parsing a slash-separated phrasal string. These are used in for instance the Brown corpus.
  • Added helper method for parsing a bar-separated phrasal string. These are used by for instance LCC.
• Improve error reporting for bad input in Phrase#add_word.
• Add convenience methods words_to_phrase and word_pos_tuples_to_phrase. These are methods for quickly instantiating Phrase object from lists.
• Add batch utility method which batches iterables.
• Change PhraseValidity enums to be strings.

Version 0.7.0

Bug fixes

• Fix a bug where validity was not properly parsed to an enum
• Fix a bug in tests where basestring was used instead of six.string_types

Breaking changes

• Removed Python 3.3 and 3.5 official support

Changes

• Added methods for removing child objects (phrases in text, words in phrase and so on)
• Added methods for clearing all child objects
• Added methods for clearing all tags
• Added an integration towards NLTK
  • Added POS tagging using NLTK
  • Added (word and sentence) tokenization using NLTK
  • Added named entity recognition using NLTK

Version 0.6.0

Changes

• Fix a bug related to reading and writing phrase comments
• Rename phrase attribute 'free_translation' → 'translation'
• Fix a bug with a relative path import (breaking in python3)

Version 0.5.0

Changes

• Added globaltags support for the phrase model
• Added proper parsing and dumping of global tags

Version 0.4.3

Changes

• Fixed a bug with botched dependency in requirements_dev

v0.4.2

Changes

• Added initialiser-values to all constructors
• Added some more tests

Version 0.4.1

Changes

• Added UTF-8 encoding to writing of files.

Version 0.4.0

• Fixed a bug where unicode entries in any Text-tree would be incorrectly rendered upon writing xml
• Implemented add_iterable to all models
• Added the Word.from_text method, enabling us to instantiate Word objects from strings

Version 0.3.2

• Removed click as dependency.