mattcg / language-subtag-registry Goto Github PK

While I cannot cite chapter and verse of the spec for the particular flavor of record-jar format used (because that link hangs and times out on me)-:, it is apparent from reading the official repository file itself that a line that begins with two spaces is a “continuation line”, i.e. a continuation of the previous line.

I do not know whether the intended use is that

1. For any line that starts with a blank (U+0020), remove the blank and the preceding LF (or CR+LF or whatever), or
2. for any line that starts with two blanks, replace the two blanks and the preceding LF (or CR+LF or whatever) with a blank.

But in this particular file both algorithms will achieve the same result, as all lines that start with a blank start with two blanks.

This is actually an important parsing distinction to make, so as to avoid considering something like “Interlingua (International Auxiliary Language Association)” as two separate Description s, or “Minnan, Hokkien, Amoy, Taiwanese, Southern Min, Southern Fujian, Hoklo, Southern Fukien, Ho-lo” as two separate Comment s.

P.S. This came up while writing a program to convert your JSON to XML. Even if you cannot generate a single string for these, if you keep the two initial blanks (rather than discard them, which is what currently happens), I could execute the logic necessary to combine the appropriate adjacent Description s or Comment s in the XSLT. I should have that program (which gives you credit, of course) up on GitHub any day now.

Hey Matthew!

Thank you for the package 🐨

We want to use it in our projects as a transitive dependency coming from eslint-plugin-jsx-a11y:

[email protected] → [email protected] → [email protected]

But it is not possible due to ODC-By-1.0 license. It is kind of hard to make sure that this license works in our country, so we're a bit scared and can't use this package. Especially because ODC-By-1.0 is not marked as FSF Free by SPDX.

Is it possible to relicense it with MIT?

Just wondering why ODC-By-1.0 was chosen over MIT.

Hi!

I would like to suggest a very minor enhancement. Consider this when you get a chance :)

I noticed that this package uses the ODC-BY license in the licenses field of package.json. This is a fine choice. However, the officially registered SPDX license id is ODC-By-1.0. See https://spdx.org/licenses/. The reason why it is generally helpful to use an officially registered SPDX license id is that those of us who use an automated license checker (e.g., https://www.npmjs.com/package/license-checker) can automatically process this package without further whitelisting.

Thank you!