Build the language-aware H2 BLUF-heading regex used by readBlufParagraph. The universal BLUF\b token is always accepted (preserved as an English acronym in roughly half of translated briefs); when lang is given, each entry from LANG_BLUF_SECTION_NAMES is added as an alternation, anchored to the heading line so it never matches an in-body mention.
Strips Markdown emphasis markers (*, _) from emoji-suffixed heading bodies via the [^\n]*? lookahead before the name, so ## ๐ฏ *Sammanfattning* still matches.
Build the language-aware H2 BLUF-heading regex used by readBlufParagraph. The universal
BLUF\btoken is always accepted (preserved as an English acronym in roughly half of translated briefs); whenlangis given, each entry from LANG_BLUF_SECTION_NAMES is added as an alternation, anchored to the heading line so it never matches an in-body mention.Strips Markdown emphasis markers (
*,_) from emoji-suffixed heading bodies via the[^\n]*?lookahead before the name, so## ๐ฏ *Sammanfattning*still matches.Pure function โ exported for tests.