Skip to content

Latest commit

 

History

History
160 lines (114 loc) · 7.4 KB

CONTRIBUTING.md

File metadata and controls

160 lines (114 loc) · 7.4 KB
Raw

Contributing to PyThaiNLP

Hi! Thanks for your interest in contributing to PyThaiNLP.

Please refer to our Contributor Covenant Code of Conduct.

Issue Report and Discussion

Code

Code Guidelines

  • Follow PEP8, use black with --line-length = 79;
  • Name identifiers (variables, classes, functions, module names) with meaningful and pronounceable names (x is always wrong);
    • Please follow this naming convention. For example, global constant variables must be in ALL_CAPS;
  • Write tests for your new features. The test suite is in tests/ directory. (see "Testing" section below);
  • Run all tests before pushing (just execute tox) so you will know if your changes broke something;
  • Commented out codes are dead codes;
  • All #TODO comments should be turned into issues in GitHub;
  • When appropriate, use f-string (use f"{a} = {b}", instead of "{} = {}".format(a, b) and "%s = %s' % (a, b)");
  • All text files, including source codes, must end with one empty line. This is to please git and to keep up with POSIX standard.

Version Control System

Commit Message

Pull Request

  • We use the famous gitflow to manage our branches.
  • When you create pull requests on GitHub, Github Actions and AppVeyor will run tests and several checks automatically. Click the "Details" link at the end of each check to see what needs to be fixed.

Documentation

  • We use Sphinx to generate API document automatically from "docstring" comments in source codes. This means the comment section in the source codes is important for the quality of documentation.
  • A docstring should start with one summary line, end with one line with a full stop (period), then be followed by a blank line before starting a new paragraph.
  • A commit to release branches (e.g. 2.2, 2.1) with a title "(build and deploy docs)" (without quotes) will trigger the system to rebuild the documentation files and upload them to the website https://pythainlp.github.io/docs.

Testing

We use standard Python unittest. The test suite is in tests/ directory.

To run unit tests locally together with code coverage test:

(from main pythainlp/ directory)

coverage run -m unittest discover

See code coverage test:

coverage report

Generate code coverage test in HTML (files will be available in htmlcov/ directory):

coverage html

Make sure the tests pass on both Github Actions and AppVeyor.

Releasing

  • We use semantic versioning: MAJOR.MINOR.PATCH, with development build suffix: MAJOR.MINOR.PATCH-devBUILD
  • We use bumpversion to manage versioning.
    • bumpversion [major|minor|patch|release|build]
    • Example:
    #current_version = 2.3.3-dev0

bumpversion build
#current_version = 2.3.3-dev1

bumpversion build
#current_version = 2.3.3-dev2

bumpversion release
#current_version = 2.3.3-beta0

bumpversion release
#current_version = 2.3.3

bumpversion patch
#current_version = 2.3.6-dev0

bumpversion minor
#current_version = 2.3.1-dev0

bumpversion build
#current_version = 2.3.1-dev1

bumpversion major
#current_version = 3.0.0-dev0

bumpversion release
#current_version = 3.0.0-beta0

bumpversion release
#current_version = 3.0.0

Credits

Thanks to all contributors. (Image made with contributors-img)

Development Leads

  • Wannaphong Phatthiyaphaibun wannaphong@yahoo.com - foundation, distribution and maintenance
  • Korakot Chaovavanich - initial tokenization and soundex codes
  • Charin Polpanumas - classification and benchmarking
  • Arthit Suriyawongkul - refactoring, packaging, distribution, and maintenance
  • Lalita Lowphansirikul - documentation
  • Pattarawat Chormai - benchmarking
  • Peerat Limkonchotiwat
  • Thanathip Suntorntip - nlpO3 maintenance, Rust Developer
  • Can Udomcharoenchaikit - documentation and codes

Maintainers

  • Arthit Suriyawongkul
  • Wannaphong Phatthiyaphaibun

Past

  • Peeradej Tanruangporn - documentation

References

  • [Maximum Matching] -- Manabu Sassano. Deterministic Word Segmentation Using Maximum Matching with Fully Lexicalized Rules. Retrieved from http://www.aclweb.org/anthology/E14-4016
  • [MetaSound] -- Snae & Brückner. (2009). Novel Phonetic Name Matching Algorithm with a Statistical Ontology for Analysing Names Given in Accordance with Thai Astrology. Retrieved from https://pdfs.semanticscholar.org/3983/963e87ddc6dfdbb291099aa3927a0e3e4ea6.pdf
  • [Thai Character Cluster] -- T. Teeramunkong, V. Sornlertlamvanich, T. Tanhermhong and W. Chinnan, “Character cluster based Thai information retrieval,” in IRAL '00 Proceedings of the fifth international workshop on on Information retrieval with Asian languages, 2000.
  • [Enhanced Thai Character Cluster] -- Jeeragone Inrut, Patiroop Yuanghirun, Sarayut Paludkong, Supot Nitsuwat, and Para Limmaneepraserth. “Thai word segmentation using combination of forward and backward longest matching techniques.” In International Symposium on Communications and Information Technology (ISCIT), pp. 37-40. 2001.
  • เพ็ญศิริ ลี้ตระกูล. การเลือกประโยคสำคัญในการสรุปความภาษาไทย โดยใช้แบบจำลองแบบลำดับชั้น (Selection of Important Sentences in Thai Text Summarization Using a Hierarchical Model). Retrieved from http://digi.library.tu.ac.th/thesis/st/0192/
  • [Thai Discourse Treebank] -- Ponrawee Prasertsom, Apiwat Jaroonpol, Attapol T. Rutherford; The Thai Discourse Treebank: Annotating and Classifying Thai Discourse Connectives. Transactions of the Association for Computational Linguistics 2024; 12 613–629. doi: https://doi.org/10.1162/tacl_a_00650