3 Star 6 Fork 7

tanloong / neosca

加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
克隆/下载
CHANGELOG.md 7.07 KB
一键复制 编辑 原始数据 按行查看 历史
tanloong 提交于 2024-03-24 00:57 . chore: version bump

Changelog

0.1.2 (24 March 2024)

Bug fixes

  • Fix crash on exporting table

0.1.1 (17 March 2024)

Improvements

  • Able to query long sentences

Now query constituency trees non-recursively to prevent reaching the maximum recursion depth limit (default 1000 in Python). If any, trees for sentences that have hundreds of words can be queried normally.

  • Support macOS 10

Migrate from PySide6 to PyQt5 for earlier macOS version support. The testing macOS 10 complains that "Qt requires macOS 11.0.0 or later" when importing PySide6/PyQt6. Fallback to Qt5.

0.1.0 (13 March 2024)

Breaking Changes

  • Rename --reserve-matched to --save-matches
  • Rename --reserve-parsed to --cache

Dependency Changes

New Features

  • Add GUI (nsca gui). NeoSCA borrows tons of code and design ideas to its GUI component from Wordless.

Deprecated Features

  • --no-querying
  • --expand
  • --max-length
  • --config

Bug fixes

  • Fix including Windows temporary files (~$\*) as input files

License change

Change the license from GPLv2-or-later to GPLv3. According to the license compatibility matrix, the combination of code under GPLv2-or-later, which is the current license, and code under LGPLv3, which is the license of the recently added dependency PySide6, should be licensed under GPLv3.

0.0.55 (2 October 2023)

Bug fixes

  • Fix exiting early when checking spaCy and model

0.0.54 (2 October 2023)

Dependency Changes

  • Update dependency on spaCy to 3.7.0

0.0.53 (25 September 2023)

Improvements

  • Add option to install spaCy from Chinese mirror site and en_core_web_sm from sourceforge.net

0.0.52 (1 September 2023)

New features

  • Add support of PTB POS tagset for nsca-lca
  • Add --text option for nsca-lca

Improvements

  • Add formulae in README for LCA measures

Bug fixes

  • Fix not correctly using Chinese JDK mirror

0.0.51 (23 August 2023)

Bug fixes

  • Don't check empty key/value when initializing Structure

0.0.50 (23 August 2023)

Bug fixes

  • Fix --reserve-matched not working since 0.0.48

0.0.49 (19 August 2023)

Bug fixes

  • Fix missing filename in the output of nsca-lca

0.0.48 (19 August 2023)

New features

  • Add lexical complexity analyzing feature (nsca-lca)

0.0.47 (18 August 2023)

Bug fixes

  • Re-calculate values of structures defined by value_source after '+' operation
  • Check empty key/value when loading custom config file

0.0.46 (14 August 2023)

Bug fixes

  • Fix division by zero error introduced in 0.0.44

Improvements

  • Update docs for custom structure definition
  • Add release.yml for workflow

0.0.45 (14 August 2023)

New features

  • Add Tregex command line interface

Bug fixes

  • Include structure_data.json in dist

0.0.44 (14 August 2023)

New features

  • Add nsca.json config file feature (--config)

Improvements

  • Skip unsupported file types instead of immediately exit
  • Try to parse a file if its extension is not among the extensions_to_exclude

Bug fixes

  • Fix precision issue of values on combined files caused by Python's floating-point inaccuracy
  • Fix Tregex matching for "ROOT" nodes

0.0.43 (7 July 2023)

Breaking changes

  • Should now use slashes (e.g., C/S) instead of underscores (e.g., C_S) for --select

New features

  • Add --no-parse
  • Add support for reading .odt files

Bug fixes

  • Fix urlopen error on macOS
  • Store Java dependencies at root direcotry if Windows username contains non-latin characters

0.0.42 (29 April 2023)

Bug fixes

  • Update dependents in setup.py

0.0.41 (29 April 2023)

Improvements

  • Try reusing previously detected encoding for txt files before detecting again

0.0.40 (29 April 2023)

New features

  • Add support for input directories, allowing users to process batches of files more easily.
  • Add support for reading docx files, in addition to the previously-supported txt files.
  • Add the ability to guess the encoding of txt files, making it easier to read files with unknown or non-utf8 encodings.

Bug fixes

  • Fix subsequent files cannot be processed if parsed results for preceding files are present (#20)
  • Fix incorrect column alignment in the csv output if filenames contain commas

0.0.39 (8 April 2023)

Bug fixes

  • Fix raising IndexError when running only on subfiles
  • Fix wrong column alignment in the csv output if filenames contain commas

0.0.38 (25 March 2023)

New features

  • Add --pretokenized
  • Add --quiet

Bug fixes

  • Fix not rasing error for typos in arguments when --text is specified
  • Fix zero frequency for some measures with --select (#12)

0.0.37 (10 March 2023)

Bug fixes

  • Fix not being able to start JVM if Java is installed but JAVA_HOME is not set.

0.0.36 (5 March 2023)

New features

  • Add --combine-subfiles
  • Add --expand-wildcards
  • Add --max-length
  • Add --newline-break
  • Add --select

Bug fixes

  • Fix raising PermissionError when installing dependencies on Windows
  • Fix missing subfiles when expanding wildcards
  • Fix the slightly inaccurate word counting due to changes in the output format of Stanford Parser
  • Fix raising an error if target_dir does not exist when installing dependencies
  • Fix unexpectedly adding blanklines at the end of each line within each matched subtree on Windows

Dependency changes

  • Add JPype1
1
https://gitee.com/tanloong/neosca.git
git@gitee.com:tanloong/neosca.git
tanloong
neosca
neosca
master

搜索帮助