From 90878e844b82c12b629a6f81210272c261357d99 Mon Sep 17 00:00:00 2001
From: ldolse <ldolse@yahoo.com>
Date: Mon, 7 Feb 2011 18:46:30 +0800
Subject: [PATCH 1/2] fix for text based horizontal rules in dehyphenate and
 scene break markup

---
 src/calibre/ebooks/conversion/preprocess.py | 12 ++++++------
 src/calibre/ebooks/conversion/utils.py      |  3 ++-
 2 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/src/calibre/ebooks/conversion/preprocess.py b/src/calibre/ebooks/conversion/preprocess.py
index 691aa307d7..6fafbb992e 100644
--- a/src/calibre/ebooks/conversion/preprocess.py
+++ b/src/calibre/ebooks/conversion/preprocess.py
@@ -245,17 +245,17 @@ class Dehyphenator(object):
         self.html = html
         self.format = format
         if format == 'html':
-            intextmatch = re.compile(u'(?<=.{%i})(?P<firstpart>[^\[\]\\\^\$\.\|\?\*\+\(\)“"\s>]+)(-|‐)\s*(?=<)(?P<wraptags>(</span>)?\s*(</[iubp]>\s*){1,2}(?P<up2threeblanks><(p|div)[^>]*>\s*(<p[^>]*>\s*</p>\s*)?</(p|div)>\s+){0,3}\s*(<[iubp][^>]*>\s*){1,2}(<span[^>]*>)?)\s*(?P<secondpart>[\w\d]+)' % length)
+            intextmatch = re.compile(u'(?<=.{%i})(?P<firstpart>[^\W\-]+)(-|‐)\s*(?=<)(?P<wraptags>(</span>)?\s*(</[iubp]>\s*){1,2}(?P<up2threeblanks><(p|div)[^>]*>\s*(<p[^>]*>\s*</p>\s*)?</(p|div)>\s+){0,3}\s*(<[iubp][^>]*>\s*){1,2}(<span[^>]*>)?)\s*(?P<secondpart>[\w\d]+)' % length)
         elif format == 'pdf':
-            intextmatch = re.compile(u'(?<=.{%i})(?P<firstpart>[^\[\]\\\^\$\.\|\?\*\+\(\)“"\s>]+)(-|‐)\s*(?P<wraptags><p>|</[iub]>\s*<p>\s*<[iub]>)\s*(?P<secondpart>[\w\d]+)'% length)
+            intextmatch = re.compile(u'(?<=.{%i})(?P<firstpart>[^\W\-]+)(-|‐)\s*(?P<wraptags><p>|</[iub]>\s*<p>\s*<[iub]>)\s*(?P<secondpart>[\w\d]+)'% length)
         elif format == 'txt':
-            intextmatch = re.compile(u'(?<=.{%i})(?P<firstpart>[^\[\]\\\^\$\.\|\?\*\+\(\)“"\s>]+)(-|‐)(\u0020|\u0009)*(?P<wraptags>(\n(\u0020|\u0009)*)+)(?P<secondpart>[\w\d]+)'% length)
+            intextmatch = re.compile(u'(?<=.{%i})(?P<firstpart>[^\W\-]+)(-|‐)(\u0020|\u0009)*(?P<wraptags>(\n(\u0020|\u0009)*)+)(?P<secondpart>[\w\d]+)'% length)
         elif format == 'individual_words':
-            intextmatch = re.compile(u'(?!<)(?P<firstpart>\w+)(-|‐)\s*(?P<secondpart>\w+)(?![^<]*?>)')
+            intextmatch = re.compile(u'(?!<)(?P<firstpart>[^\W\-]+)(-|‐)\s*(?P<secondpart>\w+)(?![^<]*?>)')
         elif format == 'html_cleanup':
-            intextmatch = re.compile(u'(?P<firstpart>[^\[\]\\\^\$\.\|\?\*\+\(\)“"\s>]+)(-|‐)\s*(?=<)(?P<wraptags></span>\s*(</[iubp]>\s*<[iubp][^>]*>\s*)?<span[^>]*>|</[iubp]>\s*<[iubp][^>]*>)?\s*(?P<secondpart>[\w\d]+)')
+            intextmatch = re.compile(u'(?P<firstpart>[^\W\-]+)(-|‐)\s*(?=<)(?P<wraptags></span>\s*(</[iubp]>\s*<[iubp][^>]*>\s*)?<span[^>]*>|</[iubp]>\s*<[iubp][^>]*>)?\s*(?P<secondpart>[\w\d]+)')
         elif format == 'txt_cleanup':
-            intextmatch = re.compile(u'(?P<firstpart>\w+)(-|‐)(?P<wraptags>\s+)(?P<secondpart>[\w\d]+)')
+            intextmatch = re.compile(u'(?P<firstpart>[^\W\-]+)(-|‐)(?P<wraptags>\s+)(?P<secondpart>[\w\d]+)')
 
 
         html = intextmatch.sub(self.dehyphenate, html)
diff --git a/src/calibre/ebooks/conversion/utils.py b/src/calibre/ebooks/conversion/utils.py
index c0c2ee8978..6583c258bf 100644
--- a/src/calibre/ebooks/conversion/utils.py
+++ b/src/calibre/ebooks/conversion/utils.py
@@ -34,6 +34,7 @@ class HeuristicProcessor(object):
         self.line_close = "(</(?P=inner3)>)?\s*(</(?P=inner2)>)?\s*(</(?P=inner1)>)?\s*</(?P=outer)>"
         self.single_blank = re.compile(r'(\s*<p[^>]*>\s*</p>)', re.IGNORECASE)
         self.scene_break_open = '<p class="scenebreak" style="text-align:center; text-indent:0%; margin-top:1em; margin-bottom:1em; page-break-before:avoid">'
+        self.common_in_text_endings = u'[\"\'—’”,\.!\?\…)\w]'
 
     def is_pdftohtml(self, src):
         return '<!-- created by calibre\'s pdftohtml -->' in src[:1000]
@@ -638,7 +639,7 @@ class HeuristicProcessor(object):
             blanks_count = len(self.any_multi_blank.findall(html))
             if blanks_count >= 1:
                 html = self.merge_blanks(html, blanks_count)
-            scene_break_regex = self.line_open+'(?![\w\'\"])(?P<break>((?P<break_char>((?!\s)\W))\s*(?P=break_char)?)+)\s*'+self.line_close
+            scene_break_regex = self.line_open+'(?!([\w\'\"]|.*?'+self.common_in_text_endings+'<))(?P<break>((?P<break_char>((?!\s)\W))\s*(?P=break_char)?)+)\s*'+self.line_close
             scene_break = re.compile(r'%s' % scene_break_regex, re.IGNORECASE|re.UNICODE)
             # If the user has enabled scene break replacement, then either softbreaks
             # or 'hard' scene breaks are replaced, depending on which is in use

From cdcfde662562105ebb8948828d9ddf37e3dccbf1 Mon Sep 17 00:00:00 2001
From: ldolse <ldolse@yahoo.com>
Date: Mon, 7 Feb 2011 19:11:47 +0800
Subject: [PATCH 2/2] added more line beginnings

---
 src/calibre/ebooks/conversion/utils.py | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/src/calibre/ebooks/conversion/utils.py b/src/calibre/ebooks/conversion/utils.py
index 6583c258bf..d075390e8e 100644
--- a/src/calibre/ebooks/conversion/utils.py
+++ b/src/calibre/ebooks/conversion/utils.py
@@ -34,7 +34,8 @@ class HeuristicProcessor(object):
         self.line_close = "(</(?P=inner3)>)?\s*(</(?P=inner2)>)?\s*(</(?P=inner1)>)?\s*</(?P=outer)>"
         self.single_blank = re.compile(r'(\s*<p[^>]*>\s*</p>)', re.IGNORECASE)
         self.scene_break_open = '<p class="scenebreak" style="text-align:center; text-indent:0%; margin-top:1em; margin-bottom:1em; page-break-before:avoid">'
-        self.common_in_text_endings = u'[\"\'—’”,\.!\?\…)\w]'
+        self.common_in_text_endings = u'[\"\'—’”,\.!\?\…\)„\w]'
+        self.common_in_text_beginnings = u'[\w\'\"“‘‛]'
 
     def is_pdftohtml(self, src):
         return '<!-- created by calibre\'s pdftohtml -->' in src[:1000]
@@ -639,7 +640,7 @@ class HeuristicProcessor(object):
             blanks_count = len(self.any_multi_blank.findall(html))
             if blanks_count >= 1:
                 html = self.merge_blanks(html, blanks_count)
-            scene_break_regex = self.line_open+'(?!([\w\'\"]|.*?'+self.common_in_text_endings+'<))(?P<break>((?P<break_char>((?!\s)\W))\s*(?P=break_char)?)+)\s*'+self.line_close
+            scene_break_regex = self.line_open+'(?!('+self.common_in_text_beginnings+'|.*?'+self.common_in_text_endings+'<))(?P<break>((?P<break_char>((?!\s)\W))\s*(?P=break_char)?)+)\s*'+self.line_close
             scene_break = re.compile(r'%s' % scene_break_regex, re.IGNORECASE|re.UNICODE)
             # If the user has enabled scene break replacement, then either softbreaks
             # or 'hard' scene breaks are replaced, depending on which is in use