Add 'Sept' for 'September'

[wulmer]

Due to an error with my Google Sheets table, I found that Google (and DIN 5008) use Sept as abbreviation for September, but dateparser only understood Sep. It looks like both Sept and Sep are valid.

This PR adds Sept as variant of September.

[serhii73]

Please, see https://dateparser.readthedocs.io/en/latest/contributing.html#guidelines-for-editing-translation-data

[wulmer]

@serhii73 : thanks a lot for your response. I have updated the JSON data now and also included a test.

[Gallaecio]

I believe the CLDR data JSON file is not meant to be updated manually, that you should be updating the YAML file instead.

[wulmer]

I believe the CLDR data JSON file is not meant to be updated manually, that you should be updating the YAML file instead.

I don't think so: the script in dateparser_scripts/write_complete_data.py (lines 49-54) first reads all data from the two directories (with the JSON and YAML files) and then combines the data before writing it to the Python files:

    if language in cldr_languages:
        with open(cldr_date_directory + language + '.json') as f:
            cldr_data = json.load(f, object_pairs_hook=OrderedDict)
    if language in supplementary_languages:
        with open(supplementary_date_directory + language + '.yaml') as g:
            supplementary_data = OrderedDict(RoundTripLoader(g).get_data())
    complete_data = combine_dicts(cldr_data, supplementary_data)

At least, when running the script, the Python files get updated based on the changes in the JSON or YAML files.

[Gallaecio]

Yes, and the CLDR file is meant to be kept in line with the upstream CLDR file, without customizations of any kind. Customizations are managed on the YAML files. Otherwise, your changes get removed the next time we run https://github.com/scrapinghub/dateparser/blob/master/dateparser_scripts/get_cldr_data.py.

[wulmer]

@Gallaecio : the upstream files in the given repos will not change as the repos are currently archived, see the referenced repos from get_raw_data at https://github.com/unicode-cldr/cldr-dates-full, https://github.com/unicode-cldr/cldr-core, and https://github.com/unicode-cldr/cldr-rbnf.

There is however a newer still maintained repo at https://github.com/unicode-org/cldr-json/ , and they seem to have Sept as abbreviation for September.

How should I proceed here now? Next year, in September, my scripts will crash again... :-(

[serhii73]

I added a fix in the last commit, and the PR looks good.
Thank you very much.

[codecov]

Codecov Report

Base: 98.23% // Head: 98.23% // No change to project coverage 👍

Coverage data is based on head (f5f6fca) compared to base (6cdd6b5).
Patch has no changes to coverable lines.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1071   +/-   ##
=======================================
  Coverage   98.23%   98.23%           
=======================================
  Files         232      232           
  Lines        2605     2605           
=======================================
  Hits         2559     2559           
  Misses         46       46           

Impacted Files	Coverage Δ
dateparser/data/date_translation_data/de.py	100.00% <ø> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[Gallaecio]

@wulmer Abotu CLDR data being outdated, that is it seems a known issue, and definitely one we would like to address.

[RadhiFadlillah]

Apologize for necroing this PR, but I just want to notify that it seems the fix by @serhii73 is not merged in this PR.

I've checked the master branch, and the sept keyword is still applied directly in de.json file instead of in de.yaml.