Page MenuHomePhabricator

Implement smarter wikilink serialization
Open, MediumPublic

Description

An editor attempted to change [[red]]ish to correct the spelling, but instead of producing [[red]]dish, this resulted in [[red|redd]]ish.

I have it on very good authority that this is poor syntax and am reporting it accordingly. :)


Version: unspecified
Severity: normal
See Also:

Details

Reference
bz52240

Event Timeline

bzimport raised the priority of this task from to Medium.Nov 22 2014, 2:01 AM
bzimport set Reference to bz52240.

Yeah, Parsoid's serialiser should pick one (either you do [[Foo]]bar or [[Foo|Foobar]] but not [[Foo|Foob]]ar.

For edited content, we should normalize wikilinks to accepted canonical forms.

I suggest that links tails should remain, either whenever possible or at least if the original link already had a tail.

In flective languages that heavily add affixes with no major modifications to the word root, canonical forms often bloat the link syntax very much.

Even in English you can get [[transcontinental railroad]]s which would become [[transcontinental railroad|transcontinental railroads]]. It is my impression that editors who edit in wikitext prefer to avoid such constructions, unlike most software.

I am currently working on software that converts wikitext to HTML, modifies it to some extent (translates the text nodes and some attributes such as link targets from Czech to Slovak) and then converts the result back to wikitext. If canonical forms are going to be forced, I am afraid of loosing the possibility to preserve link tails from the source article in the resulting translation (Czech and Slovak are so close to each other that often translation consists only of slight modification of affixes).

  • Bug 63199 has been marked as a duplicate of this bug. ***
Arlolra added a project: Parsoid.
Arlolra set Security to None.