From 150d6762f77ab1a2d089a2baea43ac16e06d314c Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Mon, 6 Jun 2016 18:19:04 +0530 Subject: [PATCH] Amazon metadata download: Recognize common patterns of embedding the series name after the title and remove it automatically --- src/calibre/ebooks/metadata/sources/amazon.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/src/calibre/ebooks/metadata/sources/amazon.py b/src/calibre/ebooks/metadata/sources/amazon.py index 7615c57f75..76ab28b5cc 100644 --- a/src/calibre/ebooks/metadata/sources/amazon.py +++ b/src/calibre/ebooks/metadata/sources/amazon.py @@ -881,6 +881,10 @@ class Amazon(Source): mi.isbn = check_isbn(mi.isbn) if mi.series and docase: mi.series = fixcase(mi.series) + if mi.title and mi.series: + for pat in (r':\s*Book\s+\d+\s+of\s+%s$', r'\(%s\)$'): + pat = pat % re.escape(mi.series) + mi.title = re.sub(pat, '', mi.title, flags=re.I) def get_website_domain(self, domain): udomain = domain