Support context diff via filterdiff, fixed #15

cdiff.py

@@ -26,6 +26,8 @@ if sys.hexversion < 0x02050000:
 import re
 import subprocess
 import errno
+import fcntl
+import os
 import difflib
 
 
@@ -429,10 +431,6 @@ class UnifiedDiff(Diff):
         return re.match('^Binary files .* differ$', line.rstrip())
 
 
-class ContextDiff(Diff):
-    pass
-
-
 class PatchStream(object):
 
     def __init__(self, diff_hdl):
@@ -469,13 +467,54 @@ class PatchStream(object):
             yield line
 
 
+class PatchStreamForwarder(object):
+    """A non-block stream forwarder.  Note input stream is non-seekable, and
+    upstream has eaten some lines.
+    """
+    def __init__(self, istream, translator):
+        assert isinstance(istream, PatchStream)
+        self._istream = istream
+        self._translator = translator
+
+        self._istream_open = True
+        self._set_non_block(self._translator.stdin)
+        self._set_non_block(self._translator.stdout)
+
+        self._forward_until_block()
+
+    def _set_non_block(self, hdl):
+        fd = hdl.fileno()
+        fl = fcntl.fcntl(fd, fcntl.F_GETFL)
+        fcntl.fcntl(fd, fcntl.F_SETFL, fl | os.O_NONBLOCK)
+
+    def _forward_until_block(self):
+        for line in self._istream:
+            try:
+                self._translator.stdin.write(line.encode('utf-8'))
+            except IOError:
+                break       # EAGAIN
+        else:
+            self._translator.stdin.close()
+            self._istream_open = False
+
+    def __iter__(self):
+        while True:
+            try:
+                line = self._translator.stdout.readline()
+                if not line:
+                    return
+                yield line
+            except IOError:
+                if self._istream_open:
+                    self._forward_until_block()
+                continue    # EAGAIN
+
+
 class DiffParser(object):
 
     def __init__(self, stream):
-        self._stream = stream
 
-        header = [decode(line) for line in
-                  self._stream.read_stream_header(100)]
+        header = [decode(line) for line in stream.read_stream_header(100)]
         size = len(header)
 
         if size >= 4 and (header[0].startswith('*** ') and
@@ -483,34 +522,37 @@ class DiffParser(object):
                           header[2].rstrip() == '***************' and
                           header[3].startswith('*** ') and
                           header[3].rstrip().endswith(' ****')):
-            self._type = 'context'
-
             # For context diff, try use `filterdiff` to translate it to unified
             # format and provide a new stream
             #
-            # TODO
-
+            self._type = 'context'
+            try:
+                self._translator = subprocess.Popen(
+                    ['filterdiff', '--format=unified'], stdin=subprocess.PIPE,
+                    stdout=subprocess.PIPE)
+            except OSError:
+                raise SystemExit('*** Context diff support depends on '
+                                 'filterdiff')
+            self._stream = PatchStreamForwarder(stream, self._translator)
             return
 
         for n in range(size):
             if header[n].startswith('--- ') and (n < size - 1) and \
                     header[n+1].startswith('+++ '):
                 self._type = 'unified'
+                self._stream = stream
                 break
         else:
-            # `filterdiff` translate unknown diff to nothing, fall through to
+            # `filterdiff` translates unknown diff to nothing, fall through to
             # unified diff give cdiff a chance to show everything as headers
             #
             sys.stderr.write("*** unknown format, fall through to 'unified'\n")
             self._type = 'unified'
+            self._stream = stream
 
     def get_diff_generator(self):
         """parse all diff lines, construct a list of Diff objects"""
-        if self._type == 'unified':
-            difflet = UnifiedDiff(None, None, None, None)
-        else:
-            raise RuntimeError('unsupported diff format')
-
+        difflet = UnifiedDiff(None, None, None, None)
         diff = Diff([], None, None, [])
         headers = []
 

tests/context/in.diff

@@ -0,0 +1,56 @@
+*** 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
+***************
+*** 467,472 ****
+--- 467,479 ----
+          dom.unlink()
+          self.confirm(domstr == str.replace("\n", "\r\n"))
+  
++     def testPrettyTextNode(self):
++         str = '<A>B</A>'
++         dom = parseString(str)
++         dom2 = parseString(dom.toprettyxml())
++         self.confirm(dom.childNodes[0].childNodes[0].toxml()==
++                      dom2.childNodes[0].childNodes[0].toxml())
++ 
+      def testProcessingInstruction(self):
+          dom = parseString('<e><?mypi \t\n data \t\n ?></e>')
+          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
+***************
+*** 836,842 ****
+              _write_data(writer, attrs[a_name].value)
+              writer.write("\"")
+          if self.childNodes:
+!             writer.write(">%s"%(newl))
+              for node in self.childNodes:
+                  node.writexml(writer,indent+addindent,addindent,newl)
+              writer.write("%s</%s>%s" % (indent,self.tagName,newl))
+--- 836,844 ----
+              _write_data(writer, attrs[a_name].value)
+              writer.write("\"")
+          if self.childNodes:
+!             writer.write(">")
+!             if self.childNodes[0].nodeType != Node.TEXT_NODE:     # Strict check
+!                 writer.write(newl)
+              for node in self.childNodes:
+                  node.writexml(writer,indent+addindent,addindent,newl)
+              writer.write("%s</%s>%s" % (indent,self.tagName,newl))
+***************
+*** 1061,1067 ****
+          return newText
+  
+      def writexml(self, writer, indent="", addindent="", newl=""):
+!         _write_data(writer, "%s%s%s"%(indent, self.data, newl))
+  
+      # DOM Level 3 (WD 9 April 2002)
+  
+--- 1063,1069 ----
+          return newText
+  
+      def writexml(self, writer, indent="", addindent="", newl=""):
+!         _write_data(writer, self.data)
+  
+      # DOM Level 3 (WD 9 April 2002)
+  

tests/context/out.normal

@@ -0,0 +1,39 @@
+--- 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 @@
+         dom.unlink()
+         self.confirm(domstr == str.replace("\n", "\r\n"))
++
++    def testPrettyTextNode(self):
++        str = '<A>B</A>'
++        dom = parseString(str)
++        dom2 = parseString(dom.toprettyxml())
++        self.confirm(dom.childNodes[0].childNodes[0].toxml()==
++                     dom2.childNodes[0].childNodes[0].toxml())
+ 
+     def testProcessingInstruction(self):
+         dom = parseString('<e><?mypi \t\n data \t\n ?></e>')
+         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 @@
+             _write_data(writer, attrs[a_name].value)
+             writer.write("\"")
+         if self.childNodes:
+-            writer.write(">%s"%(newl))
++            writer.write(">")
++            if self.childNodes[0].nodeType != Node.TEXT_NODE:     # Strict check
++                writer.write(newl)
+             for node in self.childNodes:
+                 node.writexml(writer,indent+addindent,addindent,newl)
+             writer.write("%s</%s>%s" % (indent,self.tagName,newl))
+@@ -1061,7 +1063,7 @@
+         return newText
+ 
+     def writexml(self, writer, indent="", addindent="", newl=""):
+-        _write_data(writer, "%s%s%s"%(indent, self.data, newl))
++        _write_data(writer, self.data)
+ 
+     # DOM Level 3 (WD 9 April 2002)
+ 
+

tests/context/out.side-by-side

@@ -0,0 +1,36 @@
+--- 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:     # Strict 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/context/out.w70

@@ -0,0 +1,36 @@
+--- 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:     # S>
+                                                                             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