Implement --wrap flag

Which causes long lines to be nicely wrapped in side-by-side mode. To be
used primarily with `-s -w0`.

cdiff.py

     'lightcyan'     : '\x1b[1;36m',
 }
 
-
 # Keys for revision control probe, diff and log with diff
 VCS_INFO = {
     'Git': {
     return COLORS[start_color] + text + COLORS[end_color]
 
 
+def strsplit(text, width):
+    r"""strsplit() splits a given string into two substrings, respecting the
+    escape sequences (in a global var COLORS).
+
+    It returns 3-tuple: (first string, second string, number of visible chars
+    in the first string).
+
+    If some color was active at the splitting point, then the first string is
+    appended with the resetting sequence, and the second string is prefixed
+    with all active colors.
+    """
+    first = ''
+    second = ''
+    found_colors = []
+    chars_cnt = 0
+    bytes_cnt = 0
+    while len(text) > 0:
+        # First of all, check if current string begins with any escape
+        # sequence.
+        append_len = 0
+        for color in COLORS:
+            if text.startswith(COLORS[color]):
+                if color == 'reset':
+                    found_colors = []
+                else:
+                    found_colors.append(color)
+                append_len = len(COLORS[color])
+                break
+
+        if append_len == 0:
+            # Current string does not start with any escape sequence, so,
+            # either add one more visible char to the "first" string, or
+            # break if that string is already large enough.
+            if chars_cnt >= width:
+                break
+            chars_cnt += 1
+            append_len = 1
+
+        first += text[:append_len]
+        text = text[append_len:]
+        bytes_cnt += append_len
+
+    second = text
+
+    # If the first string has some active colors at the splitting point,
+    # reset it and append the same colors to the second string
+    if len(found_colors) > 0:
+        first += COLORS['reset']
+        for color in found_colors:
+            second = COLORS[color] + second
+
+    return (first, second, chars_cnt)
+
+
+def strtrim(text, width, wrap_char, pad):
+    r"""strtrim() trims given string respecting the escape sequences (using
+    strsplit), so that if text is larger than width, it's trimmed to have
+    width-1 chars plus wrap_char. Additionally, if pad is True, short strings
+    are padded with space to have exactly needed width.
+
+    Returns resulting string.
+    """
+    text, _, tlen = strsplit(text, width + 1)
+    if tlen > width:
+        text, _, _ = strsplit(text, width - 1)
+
+        # Old code always added trailing 'reset' sequence, but strsplit is
+        # smarter and only adds it when there is something to reset. However,
+        # in order not to distract with changed test data, here's workaround
+        # which keeps output exactly the same. TODO: remove it; it doesn't add
+        # any practical value for the user.
+        if not text.endswith(COLORS['reset']):
+            text += COLORS['reset']
+
+        text += wrap_char
+    elif pad:
+        # The string is short enough, but it might need to be padded.
+        text = "%s%*s" % (text, width - tlen, '')
+    return text
+
+
 class Hunk(object):
 
     def __init__(self, hunk_headers, hunk_meta, old_addr, new_addr):
 
 class DiffMarker(object):
 
-    def markup(self, diffs, side_by_side=False, width=0, tab_width=8):
+    def markup(self, diffs, side_by_side=False, width=0, tab_width=8,
+               wrap=False):
         """Returns a generator"""
         if side_by_side:
             for diff in diffs:
-                for line in self._markup_side_by_side(diff, width, tab_width):
+                for line in self._markup_side_by_side(diff, width, tab_width,
+                                                      wrap):
                     yield line
         else:
             for diff in diffs:
                 else:
                     yield self._markup_common(' ' + old[1])
 
-    def _markup_side_by_side(self, diff, width, tab_width):
+    def _markup_side_by_side(self, diff, width, tab_width, wrap):
         """Returns a generator"""
-        wrap_char = colorize('>', 'lightmagenta')
 
         def _normalize(line):
             return line.replace(
                 '\t', ' ' * tab_width).replace('\n', '').replace('\r', '')
 
-        def _fit_with_marker(text, markup_fn, width, pad=False):
-            """Wrap or pad input pure text, then markup"""
-            if len(text) > width:
-                return markup_fn(text[:(width - 1)]) + wrap_char
-            elif pad:
-                pad_len = width - len(text)
-                return '%s%*s' % (markup_fn(text), pad_len, '')
-            else:
-                return markup_fn(text)
-
-        def _fit_with_marker_mix(text, base_color, width, pad=False):
-            """Wrap or pad input text which contains mdiff tags, markup at the
-            meantime, note only left side need to set `pad`
+        def _fit_with_marker_mix(text, base_color, width):
+            """Wrap input text which contains mdiff tags, markup at the
+            meantime
             """
             out = [COLORS[base_color]]
-            count = 0
             tag_re = re.compile(r'\x00[+^-]|\x01')
 
-            while text and count < width:
+            while text:
                 if text.startswith('\x00-'):    # del
                     out.append(COLORS['reverse'] + COLORS[base_color])
                     text = text[2:]
                     out.append(COLORS['underline'] + COLORS[base_color])
                     text = text[2:]
                 elif text.startswith('\x01'):   # reset
+                    # TODO: Append resetting sequence if only there is some
+                    # text after that. That is, call out.append(...) if only
+                    # len(text) > 1.
                     out.append(COLORS['reset'] + COLORS[base_color])
                     text = text[1:]
                 else:
                     # this tool never put that kind of symbol in their code :-)
                     #
                     out.append(text[0])
-                    count += 1
                     text = text[1:]
 
-            if count == width and tag_re.sub('', text):
-                # Was stripped: output fulfil and still has normal char in text
-                out[-1] = COLORS['reset'] + wrap_char
-            elif count < width and pad:
-                pad_len = width - count
-                out.append('%s%*s' % (COLORS['reset'], pad_len, ''))
-            else:
-                out.append(COLORS['reset'])
+            out.append(COLORS['reset'])
 
             return ''.join(out)
 
 
                 if changed:
                     if not old[0]:
-                        left = '%*s' % (width, ' ')
+                        left = ''
                         right = right.rstrip('\x01')
                         if right.startswith('\x00+'):
                             right = right[2:]
-                        right = _fit_with_marker(
-                            right, self._markup_new, width)
+                        right = self._markup_new(right)
                     elif not new[0]:
                         left = left.rstrip('\x01')
                         if left.startswith('\x00-'):
                             left = left[2:]
-                        left = _fit_with_marker(left, self._markup_old, width)
+                        left = self._markup_old(left)
                         right = ''
                     else:
-                        left = _fit_with_marker_mix(left, 'red', width, 1)
+                        left = _fit_with_marker_mix(left, 'red', width)
                         right = _fit_with_marker_mix(right, 'green', width)
                 else:
-                    left = _fit_with_marker(
-                        left, self._markup_common, width, 1)
-                    right = _fit_with_marker(right, self._markup_common, width)
-                yield line_fmt % {
-                    'left_num': left_num,
-                    'left': left,
-                    'right_num': right_num,
-                    'right': right
-                }
+                    left = self._markup_common(left)
+                    right = self._markup_common(right)
+
+                if wrap:
+                    # Need to wrap long lines, so here we'll iterate,
+                    # shaving off `width` chars from both left and right
+                    # strings, until both are empty. Also, line number needs to
+                    # be printed only for the first part.
+                    lncur = left_num
+                    rncur = right_num
+                    while len(left) > 0 or len(right) > 0:
+                        # Split both left and right lines, preserving escaping
+                        # sequences correctly.
+                        lcur, left, llen = strsplit(left, width)
+                        rcur, right, rlen = strsplit(right, width)
+
+                        # Pad left line with spaces if needed
+                        if llen < width:
+                            lcur = "%s%*s" % (lcur, width - llen, '')
+
+                        yield line_fmt % {
+                            'left_num': lncur,
+                            'left': lcur,
+                            'right_num': rncur,
+                            'right': rcur
+                        }
+
+                        # Clean line numbers for further iterations
+                        lncur = ''
+                        rncur = ''
+                else:
+                    # Don't need to wrap long lines; instead, a trailing '>'
+                    # char needs to be appended.
+                    wrap_char = colorize('>', 'lightmagenta')
+                    left = strtrim(left, width, wrap_char, len(right) > 0)
+                    right = strtrim(right, width, wrap_char, False)
+
+                    yield line_fmt % {
+                        'left_num': left_num,
+                        'left': left,
+                        'right_num': right_num,
+                        'right': right
+                    }
 
     def _markup_header(self, line):
         return colorize(line, 'cyan')
     diffs = DiffParser(stream).get_diff_generator()
     marker = DiffMarker()
     color_diff = marker.markup(diffs, side_by_side=opts.side_by_side,
-                               width=opts.width, tab_width=opts.tab_width)
+                               width=opts.width, tab_width=opts.tab_width,
+                               wrap=opts.wrap)
 
     for line in color_diff:
         pager.stdin.write(line.encode('utf-8'))
     parser.add_option(
         '-t', '--tab-width', type='int', default=8, metavar='N',
         help="""convert tab characters to this many spcaes (default: 8)""")
+    parser.add_option(
+        '', '--wrap', action='store_true',
+        help='wrap long lines in side-by-side view')
 
     # Hack: use OptionGroup text for extra help message after option list
     option_group = OptionGroup(

tests/context/out.w70.wrap

+--- a/Lib/test/test_minidom.py	Fri Sep 30 08:46:25 2011 +0300
++++ b/Lib/test/test_minidom.py	Sat Oct 01 21:02:49 2011 +0300
+@@ -467,6 +467,13 @@
+467         dom.unlink()                                                   467         dom.unlink()
+468         self.confirm(domstr == str.replace("\n", "\r\n"))              468         self.confirm(domstr == str.replace("\n", "\r\n"))
+                                                                           469 
+                                                                           470     def testPrettyTextNode(self):
+                                                                           471         str = '<A>B</A>'
+                                                                           472         dom = parseString(str)
+                                                                           473         dom2 = parseString(dom.toprettyxml())
+                                                                           474         self.confirm(dom.childNodes[0].childNodes[0].toxml()==
+                                                                           475                      dom2.childNodes[0].childNodes[0].toxml())
+469                                                                        476 
+470     def testProcessingInstruction(self):                               477     def testProcessingInstruction(self):
+471         dom = parseString('<e><?mypi \t\n data \t\n ?></e>')           478         dom = parseString('<e><?mypi \t\n data \t\n ?></e>')
+472         pi = dom.documentElement.firstChild                            479         pi = dom.documentElement.firstChild
+--- a/Lib/xml/dom/minidom.py	Fri Sep 30 08:46:25 2011 +0300
++++ b/Lib/xml/dom/minidom.py	Sat Oct 01 21:02:49 2011 +0300
+@@ -836,7 +836,9 @@
+ 836             _write_data(writer, attrs[a_name].value)                    836             _write_data(writer, attrs[a_name].value)
+ 837             writer.write("\"")                                          837             writer.write("\"")
+ 838         if self.childNodes:                                             838         if self.childNodes:
+ 839             writer.write(">%s"%(newl))                                  839             writer.write(">")
+                                                                             840             if self.childNodes[0].nodeType != Node.TEXT_NODE:     # St
+                                                                                 rict check
+                                                                             841                 writer.write(newl)
+ 840             for node in self.childNodes:                                842             for node in self.childNodes:
+ 841                 node.writexml(writer,indent+addindent,addindent,newl)   843                 node.writexml(writer,indent+addindent,addindent,newl)
+ 842             writer.write("%s</%s>%s" % (indent,self.tagName,newl))      844             writer.write("%s</%s>%s" % (indent,self.tagName,newl))
+@@ -1061,7 +1063,7 @@
+1061         return newText                                                 1063         return newText
+1062                                                                        1064 
+1063     def writexml(self, writer, indent="", addindent="", newl=""):      1065     def writexml(self, writer, indent="", addindent="", newl=""):
+1064         _write_data(writer, "%s%s%s"%(indent, self.data, newl))        1066         _write_data(writer, self.data)
+1065                                                                        1067 
+1066     # DOM Level 3 (WD 9 April 2002)                                    1068     # DOM Level 3 (WD 9 April 2002)
+1067                                                                        1069 

tests/crlf/out.w70.wrap

+Index: README.md
+===================================================================
+--- README.md	(revision 41)
++++ README.md	(working copy)
+@@ -3,7 +3,6 @@
+  3 ## About                                                                 3 ## About
+  4                                                                          4 
+  5 Coderev is a toolkit generates static side-by-side html pages for code   5 Coderev is a toolkit generates static side-by-side html pages for code
+     review.                                                                    review.
+  6 Typical use case is to generate diff pages for local modification in s     
+    vn/cvs                                                                     
+  7 workspace and send page link to teammates for code review.               6 workspace and send page link to teammates for code review.
+  8                                                                          7 
+  9 See [joyus.org/pub/coderev-demo](http://joyus.org/pub/coderev-demo) fo   8 See [joyus.org/pub/coderev-demo](http://joyus.org/pub/coderev-demo) fo
+    r a demo.                                                                  r a demo.
+@@ -18,9 +17,10 @@
+ 18                                                                         17 
+ 19 ## Usage of coderev.sh                                                  18 ## Usage of coderev.sh
+ 20                                                                         19 
+ 21 Just type `./coderev.sh -h` to see the usage.                           20 Just type `./coderev.sh --help` to see the usage.
+ 22                                                                         21 
+ 23     Usage:                                                              22     Usage:
+                                                                            23 
+ 24         coderev.sh [-r revision] [-w width] [-o outdir] [-y] [-d name]  24         coderev.sh [-r revision] [-w width] [-o outdir] [-y] [-d name]
+     \                                                                          \
+ 25                 [-F comment-file | -m 'comment...'] [file...]           25                 [-F comment-file | -m 'comment...'] [file...]
+ 26                                                                         26 
+@@ -145,3 +145,5 @@
+145                             specify column number where lines are brok 145                             specify column number where lines are brok
+    en and                                                                     en and
+146                             wrapped for sdiff, default is no line wrap 146                             wrapped for sdiff, default is no line wrap
+    ping                                                                       ping
+147     -y, --yes             do not prompt for overwriting                147     -y, --yes             do not prompt for overwriting
+                                                                           148 
+                                                                           149 # EOF

tests/diff-of-diff/out.w70.wrap

+Index: patch-Alias.xs
+===================================================================
+--- patch-Alias.xs	(revision 384635)
++++ patch-Alias.xs	(revision 384636)
+@@ -140,17 +140,21 @@
+140          tmp = kLISTOP->op_first;                                      140          tmp = kLISTOP->op_first;
+141          if (inside)                                                   141          if (inside)
+142                  op_null(tmp);                                         142                  op_null(tmp);
+143 @@ -2001,6 +2035,9 @@ STATIC OP *da_ck_entersub(pTHX_ OP *o) {         143 @@ -2001,6 +2035,13 @@ STATIC OP *da_ck_entersub(pTHX_ OP *o) {
+144          while (kid->op_sibling != last)                               144          while (kid->op_sibling != last)
+145                  kid = kid->op_sibling;                                145                  kid = kid->op_sibling;
+146          kid->op_sibling = Nullop;                                     146          kid->op_sibling = Nullop;
+147 +#ifdef op_sibling_splice                                              147 +#ifdef op_sibling_splice
+                                                                           148 +#if (PERL_COMBI_VERSION >= 5021011)
+                                                                           149 +        kid->op_moresib = 0;
+                                                                           150 +#else
+148 +        kid->op_lastsib = 1;                                          151 +        kid->op_lastsib = 1;
+                                                                           152 +#endif
+149 +#endif                                                                153 +#endif
+150          cLISTOPx(cUNOPo->op_first)->op_last = kid;                    154          cLISTOPx(cUNOPo->op_first)->op_last = kid;
+151          if (kid->op_type == OP_NULL && inside)                        155          if (kid->op_type == OP_NULL && inside)
+152                  kid->op_flags &= ~OPf_SPECIAL;                        156                  kid->op_flags &= ~OPf_SPECIAL;
+153 @@ -2008,6 +2045,14 @@ STATIC OP *da_ck_entersub(pTHX_ OP *o) {        157 @@ -2008,6 +2049,14 @@ STATIC OP *da_ck_entersub(pTHX_ OP *o) {
+154          return o;                                                     158          return o;
+155  }                                                                     159  }
+156                                                                        160  
+@@ -165,7 +169,7 @@
+165                                                                        169  
+166  MODULE = Data::Alias  PACKAGE = Data::Alias                           170  MODULE = Data::Alias  PACKAGE = Data::Alias
+167                                                                        171  
+168 @@ -2025,6 +2070,10 @@ BOOT:                                           172 @@ -2025,6 +2074,10 @@ BOOT:
+169                  PL_check[OP_RV2CV] = da_ck_rv2cv;                     173                  PL_check[OP_RV2CV] = da_ck_rv2cv;
+170                  da_old_ck_entersub = PL_check[OP_ENTERSUB];           174                  da_old_ck_entersub = PL_check[OP_ENTERSUB];
+171                  PL_check[OP_ENTERSUB] = da_ck_entersub;               175                  PL_check[OP_ENTERSUB] = da_ck_entersub;

tests/diff-ru-bin/out.w70.wrap

+diff -ru a/README b/README
+--- a/README	2013-02-23 19:51:41.000000000 +0800
++++ b/README	2013-02-23 19:52:06.000000000 +0800
+@@ -1,6 +1,6 @@
+1 # To generate expected output, chdir to a subdir and use following com 1 # To generate expected output, cd to a subdir and use following comman
+  mand, then                                                               d, then
+2 # review with `less -R`                                                2 # review with `less -R`
+3 #                                                                        
+4 ../../cdiff.py -c always < in.diff > out.normal                        3 ../../cdiff.py -c always < in.diff > out.normal
+5 ../../cdiff.py -c always -s < in.diff > out.side-by-side               4 ../../cdiff.py -c always -s < in.diff > out.side-by-side
+6 ../../cdiff.py -c always -s < in.diff -w70 > out.w70                   5 ../../cdiff.py -c always -s < in.diff -w70 > out.w70
+                                                                         6 # EOF
+Binary files a/foo.pdf and b/foo.pdf differ
+diff -ru a/foo.txt b/foo.txt
+--- a/foo.txt	2013-02-23 19:55:03.000000000 +0800
++++ b/foo.txt	2013-02-23 19:55:10.000000000 +0800
+@@ -1,2 +1,2 @@
+1 Hello                                                                  1 hella
+2 world                                                                  2 world

tests/diff-ru/out.w70.wrap

+diff -ru a/README b/README
+--- a/README	2013-02-22 20:26:43.000000000 +0800
++++ b/README	2013-02-22 20:27:51.000000000 +0800
+@@ -1,6 +1,6 @@
+1 # To generate expected output, chdir to a subdir and use following com 1 # To generate expected output, cd to a subdir and use following comman
+  mand, then                                                               d, then
+2 # review with `less -R`                                                2 # review with `less -R`
+3 #                                                                        
+4 ../../cdiff.py -c always in.diff > out.normal                          3 ../../cdiff.py -c always in.diff > out.normal 
+5 ../../cdiff.py -c always -s in.diff > out.side-by-side                 4 ../../cdiff.py -c always -s in.diff > out.side-by-side
+6 ../../cdiff.py -c always -s in.diff -w70 > out.w70                     5 ../../cdiff.py -c always -s in.diff -w70 > out.w70
+                                                                         6 # EOF
+Only in a: a1
+Only in b: b1
+diff -ru a/common/foo.txt b/common/foo.txt
+--- a/common/foo.txt	2013-02-22 20:28:32.000000000 +0800
++++ b/common/foo.txt	2013-02-22 20:30:18.000000000 +0800
+@@ -1 +1 @@
+1 Hello                                                                  1 world
+Only in b: date.txt
+Only in a: time.txt
+

tests/evil-udiff/out.w70.wrap

+--- evil.orig   2013-06-01 16:36:29.676003599 -0500
++++ evil.new    2013-06-01 16:36:28.280056137 -0500
+@@ -1,12 +1,12 @@
+ 1 blah blah blah                                                          1 blah blah blah
+ 2 humbug                                                                  2 bah humbug
+ 3 one two three                                                           3 one two three
+ 4 four five six                                                           4 four five six
+ 5 -- vim73/src/ui.c      2013-06-01 14:48:45.012467754 -0500              5 ++ vim73/src/ui.c      2013-06-01 14:48:45.012467754 -0500
+ 6 @@ -1,6 +1,6 @@                                                         6 @@ -1,6 +1,6 @@
+ 7 blah blah blah                                                          7 blah blah blah
+ 8 humbug                                                                  8 humbug
+ 9 one two three                                                           9 one two three
+10 four five six                                                          10 four five six
+11 eight nine ten                                                         11 seven eight nine ten
+12 zero                                                                   12 zero

tests/git-bin/out.w70.wrap

+diff --git a/2pdf.sh b/2pdf.sh
+index 529d8a3..ad71911 100755
+--- a/2pdf.sh
++++ b/2pdf.sh
+@@ -13,3 +13,5 @@ INPUT=${1:-foo.html}
+13 OUTPUT=${INPUT%.html}.pdf                                              13 OUTPUT=${INPUT%.html}.pdf
+14                                                                        14 
+15 wkhtmltopdf --page-size A4 $INPUT $OUTPUT  # very very long comments e 15 wkhtmltopdf --page-size A4 $INPUT $OUTPUT  # very very long comments e
+   nds here                                                                  nds here
+                                                                          16 
+                                                                          17 # EOF
+diff --git a/example.pdf b/example.pdf
+index 1eacfd8..3696851 100644
+Binary files a/example.pdf and b/example.pdf differ
+diff --git a/foo.html b/foo.html
+index d2fd3fb..13afa6e 100644
+--- a/foo.html
++++ b/foo.html
+@@ -1,4 +1,4 @@
+1 <!doctype html>                                                        1 <!DOCTYPE html>
+2 <html>                                                                 2 <html>
+3                                                                        3 
+4 <head>                                                                 4 <head>
+diff --git a/foo.pdf b/foo.pdf
+index 0e90017..3c3b90d 100644
+Binary files a/foo.pdf and b/foo.pdf differ
+

tests/git-log/out.w70.wrap

+commit 65f33a326fb1d81e9b56cfa9dbe3af887ed91c8d
+Author: Steven Myint
+Date:   Fri Mar 22 14:59:34 2013 -0700
+
+    Remove unused imports
+
+diff --git a/libmodernize/fixes/__init__.py b/libmodernize/fixes/__init__.py
+index f8628f1..2e06052 100644
+--- a/libmodernize/fixes/__init__.py
++++ b/libmodernize/fixes/__init__.py
+@@ -1,7 +1,3 @@
+1 import sys                                                               
+2 from lib2to3 import refactor                                             
+3                                                                          
+4                                                                          
+5 lib2to3_fix_names = set([                                              1 lib2to3_fix_names = set([
+6     'lib2to3.fixes.fix_apply',                                         2     'lib2to3.fixes.fix_apply',
+7     'lib2to3.fixes.fix_except',                                        3     'lib2to3.fixes.fix_except',
+
+commit 2e80837bf815d0686138beec9a5bb94e3282204d
+Author: Steven Myint
+Date:   Fri Mar 22 14:33:55 2013 -0700
+
+    Format docstring
+
+diff --git a/libmodernize/main.py b/libmodernize/main.py
+index caf847d..b24db29 100644
+--- a/libmodernize/main.py
++++ b/libmodernize/main.py
+@@ -14,6 +14,7 @@ def main(args=None):
+14     """Main program.                                                   14     """Main program.
+15                                                                        15 
+16     Returns a suggested exit status (0, 1, 2).                         16     Returns a suggested exit status (0, 1, 2).
+                                                                          17 
+17     """                                                                18     """
+18     # Set up option parser                                             19     # Set up option parser
+19     parser = optparse.OptionParser(usage="modernize [options] file|dir 20     parser = optparse.OptionParser(usage="modernize [options] file|dir
+    ...")                                                                     ...")

tests/git-perm/out.w70.wrap

+diff --git a/cdiff b/cdiff
+old mode 100755
+new mode 100644
+diff --git a/cdiff.py b/cdiff.py
+old mode 100755
+new mode 100644
+

tests/git/out.w70.wrap

+commit 15bfa564b9db08fb277a343a3d0a01d377800606
+Author: Matthew Wang <XXXXXXX@gmail.com>
+Date:   Thu Jan 31 15:27:17 2013 +0800
+
+    Default width is now 80
+
+diff --git a/src/cdiff.py b/src/cdiff.py
+index 13f725f..bf15ef1 100755
+--- a/src/cdiff.py
++++ b/src/cdiff.py
+@@ -128,9 +128,7 @@ class Diff(object):
+128                     yield self._markup_common(' ' + old[1])            128                     yield self._markup_common(' ' + old[1])
+129                                                                        129 
+130     def markup_side_by_side(self, width):                              130     def markup_side_by_side(self, width):
+131         """width of 0 means infinite width, None means auto detect. Re 131         """Returns a generator"""
+    turns a                                                                    
+132         generator                                                          
+133         """                                                                
+134         def _normalize(line):                                          132         def _normalize(line):
+135             return line.replace('\t', ' ' * 8).replace('\n', '')       133             return line.replace('\t', ' ' * 8).replace('\n', '')
+136                                                                        134 
+@@ -147,7 +145,8 @@ class Diff(object):
+147                 return markup                                          145                 return markup
+148                                                                        146 
+149         # Setup line width and number width                            147         # Setup line width and number width
+150         if not width: width = 80                                       148         if width <= 0:
+                                                                           149             width = 80
+151         (start, offset) = self._hunks[-1].get_old_addr()               150         (start, offset) = self._hunks[-1].get_old_addr()
+152         max1 = start + offset - 1                                      151         max1 = start + offset - 1
+153         (start, offset) = self._hunks[-1].get_new_addr()               152         (start, offset) = self._hunks[-1].get_new_addr()
+@@ -430,13 +429,10 @@ if __name__ == '__main__':
+430     parser = optparse.OptionParser(usage)                              429     parser = optparse.OptionParser(usage)
+431     parser.add_option('-s', '--side-by-side', action='store_true',     430     parser.add_option('-s', '--side-by-side', action='store_true',
+432             help=('show in side-by-side mode'))                        431             help=('show in side-by-side mode'))
+433     parser.add_option('-w', '--width', type='int', default=None,       432     parser.add_option('-w', '--width', type='int', default=80,
+434             help='set line width (side-by-side mode only)')            433             help='set line width (side-by-side mode only), default is 
+                                                                               80')
+435     opts, args = parser.parse_args()                                   434     opts, args = parser.parse_args()
+436                                                                            
+437     if opts.width and opts.width < 0:                                      
+438         opts.width = 0                                                     
+439                                                                        435 
+440     if len(args) >= 1:                                                 436     if len(args) >= 1:
+441         diff_hdl = open(args[0], 'r')                                  437         diff_hdl = open(args[0], 'r')
+442     elif sys.stdin.isatty():                                           438     elif sys.stdin.isatty():

tests/hg-log/out.w70.wrap

+# HG changeset patch
+# User Dan Kenigsberg <danken@redhat.com>
+# Date 1317492169 -10800
+# Node ID a9a87f0e7c509ec6768379c08a0cf56f43d71b4a
+# Parent  b0ef6a5a6dccab0089d287bf6b9bcb8132bdbd0d
+xml.dom.minidom toprettyxml: omit whitespace for Text nodes
+
+http://bugs.python.org/issue4147
+
+This patch was very lightly tested, but I think it is nicer than the former one,
+as Text.writexml() should better know not to wrap its data with whitespace.
+Ever.
+
+diff -r b0ef6a5a6dcc -r a9a87f0e7c50 Lib/test/test_minidom.py
+--- a/Lib/test/test_minidom.py	Fri Sep 30 08:46:25 2011 +0300
++++ b/Lib/test/test_minidom.py	Sat Oct 01 21:02:49 2011 +0300
+@@ -467,6 +467,13 @@
+467         dom.unlink()                                                   467         dom.unlink()
+468         self.confirm(domstr == str.replace("\n", "\r\n"))              468         self.confirm(domstr == str.replace("\n", "\r\n"))
+                                                                           469 
+                                                                           470     def testPrettyTextNode(self):
+                                                                           471         str = '<A>B</A>'
+                                                                           472         dom = parseString(str)
+                                                                           473         dom2 = parseString(dom.toprettyxml())
+                                                                           474         self.confirm(dom.childNodes[0].childNodes[0].toxml()==
+                                                                           475                      dom2.childNodes[0].childNodes[0].toxml())
+469                                                                        476 
+470     def testProcessingInstruction(self):                               477     def testProcessingInstruction(self):
+471         dom = parseString('<e><?mypi \t\n data \t\n ?></e>')           478         dom = parseString('<e><?mypi \t\n data \t\n ?></e>')
+472         pi = dom.documentElement.firstChild                            479         pi = dom.documentElement.firstChild
+diff -r b0ef6a5a6dcc -r a9a87f0e7c50 Lib/xml/dom/minidom.py
+--- a/Lib/xml/dom/minidom.py	Fri Sep 30 08:46:25 2011 +0300
++++ b/Lib/xml/dom/minidom.py	Sat Oct 01 21:02:49 2011 +0300
+@@ -836,7 +836,9 @@
+ 836             _write_data(writer, attrs[a_name].value)                    836             _write_data(writer, attrs[a_name].value)
+ 837             writer.write("\"")                                          837             writer.write("\"")
+ 838         if self.childNodes:                                             838         if self.childNodes:
+ 839             writer.write(">%s"%(newl))                                  839             writer.write(">")
+                                                                             840             if self.childNodes[0].nodeType != Node.TEXT_NODE:
+                                                                             841                 writer.write(newl)
+ 840             for node in self.childNodes:                                842             for node in self.childNodes:
+ 841                 node.writexml(writer,indent+addindent,addindent,newl)   843                 node.writexml(writer,indent+addindent,addindent,newl)
+ 842             writer.write("%s</%s>%s" % (indent,self.tagName,newl))      844             writer.write("%s</%s>%s" % (indent,self.tagName,newl))
+@@ -1061,7 +1063,7 @@
+1061         return newText                                                 1063         return newText
+1062                                                                        1064 
+1063     def writexml(self, writer, indent="", addindent="", newl=""):      1065     def writexml(self, writer, indent="", addindent="", newl=""):
+1064         _write_data(writer, "%s%s%s"%(indent, self.data, newl))        1066         _write_data(writer, self.data)
+1065                                                                        1067 
+1066     # DOM Level 3 (WD 9 April 2002)                                    1068     # DOM Level 3 (WD 9 April 2002)
+1067                                                                        1069 

tests/latin1/out.w70.wrap

+diff --git a/test/latin1.cc b/test/latin1.cc
+index a556e5c..d81bb0c 100644
+--- a/test/latin1.cc
++++ b/test/latin1.cc
+@@ -1,4 +1,3 @@
+1 // é                                                                     
+2 int latin1()                                                           1 int latin1()
+3 {                                                                      2 {
+4         static int x;                                                  3         static int x;

tests/strange/out.w70.wrap

+commit 57263369bd810ba3a2c2dfd32a905f1e7d59cc6d
+Author: Matthew Wang <XXXXXXX@gmail.com>
+Date:   Sun Jan 6 21:56:54 2013 +0800
+
+    simplify with seperate node.py
+
+diff --git a/.gitignore b/.gitignore
+new file mode 100644
+index 0000000..0d20b64
+--- /dev/null
++++ b/.gitignore
+@@ -0,0 +1 @@
+                                                                           1 *.pyc
+diff --git a/README b/README
+index 7e70850..b5eb369 100644
+--- a/README
++++ b/README
+@@ -1 +1,2 @@
+1 Sun Feb  3 13:57:05 CST 2013                                           1 Sun Feb  3 13:57:05 CST 2013
+                                                                         2 Sun Feb  3 13:57:15 CST 2013

tests/svn-log/out.w70.wrap

+------------------------------------------------------------------------
+r1235 | ymattw | 2011-09-01 17:00:02 +0800 (Thu, 01 Sep 2011) | 3 lines
+
+Interpreter fix
+
+
+Index: src/share/example.sh
+===================================================================
+--- src/share/example.sh	(revision 1234)
++++ src/share/example.sh	(revision 1235)
+@@ -1,3 +1,3 @@
+1 #!/bin/sh                                                              1 #!/bin/bash
+2 #                                                                      2 #
+3 # Test program, also try run me as yroot (sudo ./example.sh)           3 # Test program, also try run me as yroot (sudo ./example.sh)
+
+------------------------------------------------------------------------
+r1234 | ymattw | 2011-09-01 16:00:02 +0800 (Thu, 01 Sep 2011) | 3 lines
+
+Implement common tool to run command as headless account
+
+
+Index: ChangeLog
+===================================================================
+--- ChangeLog	(revision 1233)
++++ ChangeLog	(revision 1234)
+@@ -1,3 +1,6 @@
+                                                                         1 Version 0.0.4
+                                                                         2     * Add prototype of perl module
+                                                                         3 
+1 Version 0.0.3                                                          4 Version 0.0.3
+2     * Implement '-d' option                                            5     * Implement '-d' option
+3     * Add configfile to set global debug flag                          6     * Add configfile to set global debug flag
+Index: src/share/example.sh
+===================================================================
+--- src/share/example.sh	(revision 0)
++++ src/share/example.sh	(revision 1234)
+@@ -0,0 +1,4 @@
+                                                                           1 #!/bin/sh
+                                                                           2 #
+                                                                           3 # Test program, also try run me as yroot (sudo ./example.sh)
+                                                                           4 #
+
+Property changes on: src/share/example.sh
+___________________________________________________________________
+Added: svn:executable
+## -0,0 +1 ##
+                                                                           1 *
+Added: svn:keywords
+## -0,0 +1 ##
+                                                                           1 Id

tests/svn-merge/out.w70.wrap

+Index: .
+==================================================================
+--- .   (revision 10)
++++ .   (working copy)
+
+Property changes on: .
+___________________________________________________________________
+Modified: svn:mergeinfo
+   Merged /repository/foo/branches/foo-bar:r4-9
+   Merged /repository/foo/trunk/foo:r2-3
+

tests/svn-property/out.w70.wrap

+Index: foo
+===================================================================
+Cannot display: file marked as a binary type.
+svn:mime-type = application/octet-stream
+

tests/svn/out.w70.wrap

+Index: README.md
+===================================================================
+--- README.md	(revision 41)
++++ README.md	(working copy)
+@@ -3,7 +3,6 @@
+  3 ## About                                                                 3 ## About
+  4                                                                          4 
+  5 Coderev is a toolkit generates static side-by-side html pages for code   5 Coderev is a toolkit generates static side-by-side html pages for code
+     review.                                                                    review.
+  6 Typical use case is to generate diff pages for local modification in s     
+    vn/cvs                                                                     
+  7 workspace and send page link to teammates for code review.               6 workspace and send page link to teammates for code review.
+  8                                                                          7 
+  9 See [joyus.org/pub/coderev-demo](http://joyus.org/pub/coderev-demo) fo   8 See [joyus.org/pub/coderev-demo](http://joyus.org/pub/coderev-demo) fo
+    r a demo.                                                                  r a demo.
+@@ -18,9 +17,10 @@
+ 18                                                                         17 
+ 19 ## Usage of coderev.sh                                                  18 ## Usage of coderev.sh
+ 20                                                                         19 
+ 21 Just type `./coderev.sh -h` to see the usage.                           20 Just type `./coderev.sh --help` to see the usage.
+ 22                                                                         21 
+ 23     Usage:                                                              22     Usage:
+                                                                            23 
+ 24         coderev.sh [-r revision] [-w width] [-o outdir] [-y] [-d name]  24         coderev.sh [-r revision] [-w width] [-o outdir] [-y] [-d name]
+     \                                                                          \
+ 25                 [-F comment-file | -m 'comment...'] [file...]           25                 [-F comment-file | -m 'comment...'] [file...]
+ 26                                                                         26 
+@@ -145,3 +145,5 @@
+145                             specify column number where lines are brok 145                             specify column number where lines are brok
+    en and                                                                     en and
+146                             wrapped for sdiff, default is no line wrap 146                             wrapped for sdiff, default is no line wrap
+    ping                                                                       ping
+147     -y, --yes             do not prompt for overwriting                147     -y, --yes             do not prompt for overwriting
+                                                                           148 
+                                                                           149 # EOF

tests/test_cdiff.py

         diff = self._init_diff()
         marker = cdiff.DiffMarker()
 
-        out = list(marker._markup_side_by_side(diff, 7, 8))
+        out = list(marker._markup_side_by_side(diff, 7, 8, False))
         self.assertEqual(len(out), 11)
 
         sys.stdout.write('\n')
             '\x1b[33m5\x1b[0m '
             '\x1b[31m\x1b[7m\x1b[31m      \x1b[0m\x1b[1;35m>\x1b[0m '
             '\x1b[0m\x1b[33m5\x1b[0m '
-            '\x1b[32m\x1b[7m\x1b[32m spaced\x1b[0m\n')
+            '\x1b[32m\x1b[7m\x1b[32m spaced\x1b[0m\x1b[32m\x1b[0m\n')
 
     # This test is not valid anymore
     def __test_markup_side_by_side_neg_width(self):
         diff = self._init_diff()
         marker = cdiff.DiffMarker()
-        out = list(marker._markup_side_by_side(diff, -1, 8))
+        out = list(marker._markup_side_by_side(diff, -1, 8, False))
         self.assertEqual(len(out), 11)
 
         self.assertEqual(out[0], '\x1b[36mheader\n\x1b[0m')
     def test_markup_side_by_side_off_by_one(self):
         diff = self._init_diff()
         marker = cdiff.DiffMarker()
-        out = list(marker._markup_side_by_side(diff, 6, 8))
+        out = list(marker._markup_side_by_side(diff, 6, 8, False))
         self.assertEqual(len(out), 11)
 
         sys.stdout.write('\n')
         self.assertEqual(out[4], '\x1b[1;34m@@ -1,5 +1,5 @@\n\x1b[0m')
         self.assertEqual(
             out[5],
+
             '\x1b[33m1\x1b[0m '
             '\x1b[31m\x1b[7m\x1b[31mh\x1b[0m\x1b[31mhello\x1b[0m '
             '\x1b[0m\x1b[33m1\x1b[0m '
-            '\x1b[32mhello\x1b[7m\x1b[32mo\x1b[0m\n')
+            '\x1b[32mhello\x1b[7m\x1b[32mo\x1b[0m\x1b[32m\x1b[0m\n')
         self.assertEqual(
             out[6],
             '\x1b[33m \x1b[0m        '
     def test_markup_side_by_side_wrapped(self):
         diff = self._init_diff()
         marker = cdiff.DiffMarker()
-        out = list(marker._markup_side_by_side(diff, 5, 8))
+        out = list(marker._markup_side_by_side(diff, 5, 8, False))
         self.assertEqual(len(out), 11)
 
         sys.stdout.write('\n')
     def test_markup_side_by_side_tabbed(self):
         diff = self._init_diff()
         marker = cdiff.DiffMarker()
-        out = list(marker._markup_side_by_side(diff, 8, 2))
+        out = list(marker._markup_side_by_side(diff, 8, 2, False))
         self.assertEqual(len(out), 11)
 
         sys.stdout.write('\n')
         self.assertEqual(
             out[10],
             '\x1b[33m5\x1b[0m '
-            '\x1b[31m\x1b[7m\x1b[31m  tabbed\x1b[0m '
+            '\x1b[31m\x1b[7m\x1b[31m  tabbed\x1b[0m\x1b[31m\x1b[0m '
             '\x1b[0m\x1b[33m5\x1b[0m '
             '\x1b[32m\x1b[7m\x1b[32m spaced\x1b[0m\x1b[32m\x1b[0m\n')