Author usernames should not be translatable

[its-miroma]

Usernames of authors are currently translatable in Crowdin.

Example: https://github.com/FabricMC/fabric-docs/blob/main/develop/events.md?plain=1#L4-L15

[IMB11]

Not really much we can do about this, as we can't disable the translation of yml frontmatter without disabling the ability to translate the page title and description.

[its-miroma]

I'm not super familiar with Crowdin's configuration, but if I understand
correctly you can add custom rules to ignore some sections in files.
See https://support.crowdin.com/custom-segmentation

After experimenting on https://regex101.com, I think this may work:

<rule break="yes">
	<beforebreak>\A---\n(?:(?!---\n).*\n)*?authors:\n</beforebreak>
	<afterbreak> *+(?!- )</afterbreak>
</rule>

/regex/ explanation

These regexes rely on lookarounds, which Crowdin should support.

beforebreak

• \A: assert position at the beginning of the string (aka file)
• ---\n: make sure we're inside of the frontmatter
• (?:(?!---\n).*\n)*?: match all lines before authors: in the frontmatter
  • (?:...)*?: match zero or more lines lazily, to stop when reaching authors: below
  • (?!---\n): a negative lookahead to make sure we're still in the frontmatter. If this matches, then the file does not mention any authors
  • .*\n: match zero or more characters followed by a newline
• authors:\n: match the authors: line
• the break starts here

afterbreak

• the break ends if this matches
• *+: match zero or more spaces greedily (aka ignore indentation)
  • I'm pretty sure that YAML only supports spaces, but if that's not the case you can substitute this with [ \t]*+ to also match tabs
• (?!- ): negative lookahead for - (aka make sure there are no more authors)

[IMB11]

This is only for manual XML/HTML based uploads, we use a different system to upload translations that doesn't support segments.