As reported in T175725: Deploy HTML5 sections to WMF production, some characters are problematic when used in wikitext anchor-encoded with html5 set as primary mode in $wgFragentMode.
| Subject | Repo | Branch | Lines +/- | |
|---|---|---|---|---|
| Don't break when reference names contain [] | mediawiki/extensions/Cite | master | +42 -14 |
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | MaxSem | T75092 Anchors to section names for non-ASCII letters are encoded in the URL | |||
| Resolved | matmarex | T172493 mw.util.escapeId should log deprecation warning | |||
| Resolved | MaxSem | T152540 Migrate to HTML5 section ids | |||
| Resolved | MaxSem | T175725 Deploy HTML5 sections to WMF production | |||
| Resolved | MaxSem | T179544 Some characters in HTML5 fragments mess with wikitext |
T176176 might be related, since it also related to wikitext-significant characters showing up in html5 ids.
I have a silly question: why do we even use reference names in IDs? Because these IDs are also using reference numbers, they're not stable anyway.
Good question. I would certainly support removing the wikitext-generated part of the ID.
Please don't miss {{anchorencode}} part of the task. Quite a lot of templates in ruwiki using {{anchorencode}} / mw.uri.anchorEncode() to encode anchor links have broken in many cases due to use of []. en:Module:Footnotes, for one, will break in many cases if you try delpoy to enwiki.
Change 390356 had a related patch set uploaded (by MaxSem; owner: MaxSem):
[mediawiki/extensions/Cite@master] Don't break when reference names contain []
Change 390356 had a related patch set uploaded (by Legoktm; owner: MaxSem):
[mediawiki/extensions/Cite@master] Don't break when reference names contain []
I merged this patch, but I'd like to see a follow-up task to remove the name part of the ID entirely.
Change 390356 merged by jenkins-bot:
[mediawiki/extensions/Cite@master] Don't break when reference names contain []
I can come up with at least one application: if the reference name is generated somewhere deep in the code (for example, if it is a hash of the reference content retrieved from Wikidata), there is no other way to learn it to reuse in a different place in an article. Or is it? @cscott