Regexes truly are write-only.
\[(.+?)\]\(<?(.+?)\.md>?\) does a shockingly simple thing for how opaque it is to look at.
To explain: this takes a markdown link in the format [Text](address.md) and extracts 'Text' and 'address' as matching groups I plan to use in some code to rewrite the link into a different format.
\[ \] \( and \) are literal bracket/paren matches. (.+?) means "Take one or more characters that match here, but as few as possible to meet the condition (in this case literally 'all characters between [ and ]'). <? and >? mean "match < and > if they're there. If they're not, cool".
And I know this isn't optimal even. I should remove \] from the matching set in the match groups, but meh.
\[ \] \( and \) are literal bracket/paren matches. (.+?) means "Take one or more characters that match here, but as few as possible to meet the condition (in this case literally 'all characters between [ and ]'). <? and >? mean "match < and > if they're there. If they're not, cool".
And I know this isn't optimal even. I should remove \] from the matching set in the match groups, but meh.
@rlamacraft The former requires at least one character in the match. * can match 0.
@trysdyn Oh right, so the question mark doesn't mean 0 or 1 of these here, instead it makes the plus lazy thereby ensuring it doesn't consume the period before the "md"? You're not wrong, regexs are really complex in subtle ways
