Huang Xin 370a516620 feat(reader): glue non-breaking spaces after short Russian words (#4769) (#4798) ...

Russian typography requires short function words (prepositions,
conjunctions, particles) to never hang at the end of a line. Add an
`nbsp` content transformer that inserts U+00A0 after such words so they
stick to the following word. The source file is never modified.

The transformer is language-driven via an NBSP_LANGUAGES registry keyed
by language code (only `ru` is configured today), so adding another
language is a single entry. It runs only for matching books and rewrites
text between tags with a regex, leaving tags, attributes, and the XML
declaration intact. Runs after whitespace normalization so the inserted
spaces are not stripped under the override-layout setting.

The space-to-NBSP swap is length-preserving (both are single UTF-16 code
units), so DOM character offsets and CFIs stay valid for every word
before and after the transform; tests enforce this invariant.

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-06-26 04:52:32 +02:00

..

readest-app

feat(reader): glue non-breaking spaces after short Russian words (#4769) (#4798)

2026-06-26 04:52:32 +02:00

readest.koplugin

feat(koplugin): bulk download all cloud books from Library view (#4751) (#4765)

2026-06-24 17:42:31 +02:00