Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

tweaked chapter thresholds

Browse Source
This commit is contained in:
ldolse 2011-01-18 12:51:55 +08:00
parent 21ff2be472
commit 539f24213d
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/calibre/ebooks/conversion/utils.py
View File

@ -21,7 +21,6 @@ class HeuristicProcessor(object):
self.deleted_nbsps = False
self.totalwords = 0
self.min_chapters = 1
self.max_chapters = 150
self.chapters_no_title = 0
self.chapters_with_title = 0
self.blanks_deleted = False
@ -169,9 +168,12 @@ class HeuristicProcessor(object):
# Typical chapters are between 2000 and 7000 words, use the larger number to decide the
# minimum of chapters to search for. A max limit is calculated to prevent things like OCR
# or pdf page numbers from being treated as TOC markers
max_chapters = 150
typical_chapters = 7000.
if wordcount > 7000:
self.min_chapters = int(ceil(wordcount / 15000.))
self.max_chapters = int(ceil(wordcount / 1200.))
if wordcount > 200000:
typical_chapters = 15000.
self.min_chapters = int(ceil(wordcount / typical_chapters))
print "minimum chapters required are: "+str(self.min_chapters)
heading = re.compile('<h[1-3][^>]*>', re.IGNORECASE)
self.html_preprocess_sections = len(heading.findall(html))