Module talk:Unicode data/scripts

Please update this module with the contents of Module:Unicode data/scripts/sandbox. It contains the Unicode 12.0.0 version of this data module, including newly assigned code points and new script codes. — Eru·tuon 22:07, 6 March 2019 (UTC) [reply ]

This module is used in 272,000 pages. I expect that your edits are correct, but can you please create some testcases that show the changes? Also pinging Trappist the monk, who works quite a bit with Unicode and {{lang }}, for another set of eyeballs. – Jonesey95 (talk) 18:58, 10 March 2019 (UTC) [reply ]

I have no reason to believe that the sandbox is flawed so done. I think that I would prefer that the script that makes this data module did not rely on external tools ... Perhaps a rewrite as Module:Unicode data/scripts/make?

Also, if this is v12 then the documentation needs updating?

—Trappist the monk (talk) 19:30, 10 March 2019 (UTC) [reply ]

Thanks, Trappist. – Jonesey95 (talk) 20:03, 10 March 2019 (UTC) [reply ]

Updated the documentation. The data was generated in the same way as the previous version, but from the Unicode 12.0.0 versions of Scripts.txt and PropertyValueAliases.txt from the archive at https://www.unicode.org/Public/UCD/latest/ucd/UCD.zip. I'll consider rewriting the script so that it can be run here on-wiki. The script that I used to generate wikt:User:Erutuon/scripts/UnicodeScriptRecognition.js would be easier to adapt than the one here because the part that processes Scripts.txt is simpler. — Eru·tuon 10:23, 11 March 2019 (UTC) [reply ]

Okay, created Module:Unicode data/scripts/make. It generates the top table of this module, the part that actually depends on the Unicode database. — Eru·tuon 22:18, 11 March 2019 (UTC) [reply ]

This module needs to be updated for Unicode 13.0.0 by replacing the data table with the new one generated by Module:Unicode data/scripts/make. — Eru·tuon 19:38, 11 March 2020 (UTC) [reply ]

 Done Let me know if anything is not working correctly. – Jonesey95 (talk) 23:37, 11 March 2020 (UTC) [reply ]

Please change:

Nkoo = "Nko",

into

Nkoo = "NKo",

(capital "K").

Formal script name is spelled "N'Ko", Unicode Alias is "NKo" (see ISO 15924); . Unicode defines these script names as indifferent wrt case (A=a), so we're free to adjust this.

Maybe the bot applying Module:Unicode data/scripts/make should incorporate this for future updates. DePiep (talk) 18:58, 8 June 2022 (UTC) [reply ]

Unicode formal definitions, rules & backgrounds are in TR24. 'Script names' are referred to as 'Script Designators' because 'name' is overloaded with meanings. Ping creator @Erutuon:. -DePiep (talk) 05:30, 10 June 2022 (UTC) [reply ]

This is updated on the /scripts page, so  done; however, the /make subpage is still in question. Note that editor Erutuon's Unicode subpage has the correct "NKo" designator. P.I. Ellsworth , ed. ^put'r there 09:58, 4 July 2022 (UTC) [reply ]

Thanks. For now: looks like after every Unicode version update (annually for main versions 14-15-16-..), this NKo requires manual fixing. List User:Erutuon/Unicode is not the best source (nor best blocklist): it mixes blocknames and enwiki pages, some are a Redirect or a DAB. More info in Template:Unicode blocks/overview. -DePiep (talk) 10:45, 4 July 2022 (UTC) [reply ]

Don't think it matters much, because the /make subpage seems like just a test page. Only two transclusions and still seems to rely on User:Erutuon/Unicode/PropertyValueAliases.txt, because when I changed it on that page beginning with this edit , its /doc page transclusion threw a line #31 error (No code for "Nko") until I changed it back. No biggee surmised. P.I. Ellsworth , ed. ^put'r there 13:10, 4 July 2022 (UTC) [reply ]

• See edit request #above: manual edit after automated version update. (@Drmccreedy:)

Please change:

Nkoo = "Nko",

into

Nkoo = "NKo",

(capital "K").

DePiep (talk) 10:35, 26 September 2022 (UTC) [reply ]

I see "Nko" (lowercase k) used at https://www.unicode.org/Public/UCD/latest/ucd/PropertyValueAliases.txt and https://www.unicode.org/iso15924/iso15924-codes.html while only the block is NKo (uppercase K). Unless I'm missing something I think it's correct as-is. DRMcCreedy (talk) 17:30, 26 September 2022 (UTC) [reply ]

TL;DR: not an error to fix. Both uc and lc are acceptable. Then "NKo" is conform RL English script name, so is reader-friendly.

@Drmccreedy: long: lc "Nko" is not an error to fix. Both lc and uc allowed per TR24 c.a. In RL, the script is spelled "N'Ko" (e.g, see N'Ko; ISO English Name). So, "NKo" (uc K) conforms RL writing, is recognisable for the reader.

Both forms are correct per Unicode definition TR24. So one can use uppercase and lowercase. Block names are spelled both ways, all correct. (incidentally, we can also legally replace the "_" underscores that are in PropertyValueAliases.txt)

It's the only block name with irregular titlecase casing (second letter), hence this manual exception. -DePiep (talk) 06:15, 28 September 2022 (UTC) [reply ]

I understand now. I've update the template to use NKo for the script name.DRMcCreedy (talk) 15:09, 28 September 2022 (UTC) [reply ]