Improve diff display

[Dirk Stöcker]

The diff functionality in trac has a very simple method to show changes in a line. In JOSM wiki we use this diff a lot for translation and I wanted a better display. Attached is a patch improving trac/versioncontrol/diff.py

diff.py

@@ -30 +30 @@
     return ' '.join(_whitespace_split(text))
 
 
-def get_change_extent(str1, str2):
-    """Determines the extent of differences between two strings.
+def get_change_extents(str1, str2):
+    """Determines the extents of differences between two strings.
 
-    Returns a pair containing the offset at which the changes start,
-    and the negative offset at which the changes end.
-
-    If the two strings have neither a common prefix nor a common
-    suffix, ``(0, 0)`` is returned.
+    Returns a list of typles containing the offsets at which the changes start,
+    and the offsets at which the changes end:
+    [(startstr1,endstr1,startstr2,endstr2),(...)].
+    The list is sorted from the last to the first.
     """
+    done = []
     start = 0
-    limit = min(len(str1), len(str2))
+    minlen = 7
+    l1 = len(str1)
+    l2 = len(str2)
+    limit = min(l1, l2)
     while start < limit and str1[start] == str2[start]:
         start += 1
     end = -1
     limit = limit - start
     while -end <= limit and str1[end] == str2[end]:
         end -= 1
-    return start, end + 1
+    new = [(start, l1 + end + 1, start, l2 + end + 1)]
+    matcher = difflib.SequenceMatcher(None, str1, str2)
+    while new:
+        n = []
+        for e in new:
+            if min(e[1]-e[0],e[3]-e[2]) < minlen+2:
+                done.append(e)
+            else:
+                (s1,s2,s) = matcher.find_longest_match(e[0], e[1], e[2], e[3])
+                if s < minlen:
+                    done.append(e)
+                else:
+                    n.append((e[0],s1,e[2],s2))
+                    n.append((s1+s,e[1],s2+s,e[3]))
+        new = n
+    return sorted(done, reverse=True)
 
 
 def get_filtered_hunks(fromlines, tolines, context=None,
@@ -207 +225 @@
             if tag == 'replace' and i2 - i1 == j2 - j1:
                 for i in range(i2 - i1):
                     fromline, toline = fromlines[i1 + i], tolines[j1 + i]
-                    (start, end) = get_change_extent(fromline, toline)
-                    if start != 0 or end != 0:
-                        last = end + len(fromline)
-                        fromlines[i1 + i] = (
-                            fromline[:start] + '\0' + fromline[start:last] +
-                            '\1' + fromline[last:])
-                        last = end+len(toline)
-                        tolines[j1 + i] = (
-                            toline[:start] + '\0' + toline[start:last] +
-                            '\1' + toline[last:])
+                    fromlen = len(fromline)
+                    tolen = len(toline)
+                    for fromstart, fromend, tostart, toend in get_change_extents(fromline, toline):
+                        if fromstart != fromend and (fromstart != 0 or fromend != fromlen):
+                            fromline = (
+                                fromline[:fromstart] + '\0' + fromline[fromstart:fromend] +
+                                '\1' + fromline[fromend:])
+                        if tostart != toend and (tostart != 0 or toend != tolen):
+                            toline = (
+                                toline[:tostart] + '\0' + toline[tostart:toend] +
+                                    '\1' + toline[toend:])
+                    fromlines[i1 + i] = fromline
+                    tolines[j1 + i] = toline
             yield tag, i1, i2, j1, j2
 
     changes = []

old:

new:

[Dirk Stöcker]

The diff as file

[Dirk Stöcker]

old state

[Dirk Stöcker]

new state

[Jun Omae]

That is the same of (or probably related) #6858. I'll try the patch.