|
Lastest versions
Romanian Metadata 0.4.74 (vs 0.4.72)
Added _elefant_blocked flag — after the first 403 from Elefant, all subsequent Bing RSS and Google queries for that source are skipped in the same session, saving ~15 seconds of wasted requests. Added RAO fail-fast: if the first 2 products score below 80 (completely wrong results), the source aborts immediately instead of opening all remaining pages. Fixed TargulCartii UI noise: values like "Idei de cadou" in the author field are now filtered out, with the correct author recovered from the product URL instead.
Romanian Metadata Carturesti 0.1.25 (vs 0.1.22)
Brought in sync with the main plugin's title-cleaning pipeline: strip_pack_prefix (strips V12 / Nr.7 / C05- pack prefixes), strip_bracket_prefix ([Series] 04 Title → Title), strip_trailing_volume_suffix (Title Vol. 2 → Title), and clean_input_authors (removes duplicate authors and trailing series suffixes from author strings). These were already present in the main plugin and are now consistently applied before Carturesti direct search and Bing RSS fallback queries.
Romanian Metadata eMAG 0.0.18 (vs 0.0.15)
Added roman numeral volume normalization — Calibre titles with arabic volume numbers are automatically converted to roman variants for eMAG search queries and vice versa, matching eMAG's inconsistent naming. Added subtitle-first search fallback and conservative base-title extraction to handle eMAG edition suffixes. Added strip_author_suffix_from_query_title to clean author-like trailing text from Calibre titles before querying. Fixed Licia/Lucia Troisi-style author typos via a small equivalence table.
|