Tex pool reorg #2358
Tex pool reorg #2358
Conversation
…ML-specific support and a binding for plain.tex
There was a problem hiding this comment.
This looks surprisingly well-cataloged! Kudos to David Bausum for the thematic structure.
I generally quite like it. One note is that this kind of PR is the perfect place for subtle typos to sneak in, since git won't micromanage the large chunks of text moving from one file to another, and we can't really inspect it all manually. So I hope copy-paste was bug free :> The tests passing is a big encouragement.
One curious question: Is there an obvious performance footprint from reorganizing? Or are make test times comparable to master?
lib/LaTeXML/Util/Pathname.pm
Outdated
| @@ -433,6 +433,7 @@ sub build_kpse_cache { | |||
| foreach my $path (split(/$KPATHSEP/, $texpaths)) { | |||
| $path =~ s/^!!//; $path =~ s|//+$|/|; | |||
| push(@filters, $path) if -d $path; } | |||
| my $filterre = '(?:' . join('|', map { "\Q$_\E"; } @filters) . ')'; | |||
There was a problem hiding this comment.
Edge case note: I think there is a subtle behavior change here if @filters is empty. The new regex will then evaluate to (?:), which apparently matches against every string. So in that case $skip would be always false, rather than always true. I think?
Edit:
Particularly !grep{ ... } () is always true.
While $d !~ /(?:)/ is always false.
…ools; move and rename all pool-level modules
|
@dginev here's another take, per our discussions; give it another look. |
|
✅ I like the new naming scheme. Looks good to merge. |
Split the (huge)
TeX.poolinto more manageable, modular components:plain.tex.Future (or continuing) work will
\lx@plain.texandlatex.ltxdirectly for the bulk of definitions, after establishing what needs LaTeXML-specific binding, eg for semantics.