From e4cf0e9b02570ebb29289fdffe87d25a72e56663 Mon Sep 17 00:00:00 2001
From: =?utf8?q?Jos=C3=A9=20Fonseca?= <jfonseca@vmware.com>
Date: Fri, 3 Jul 2009 13:10:37 +0100
Subject: [PATCH] Handle incomplete gzip files.

---
 xml2txt.py | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

diff --git a/xml2txt.py b/xml2txt.py
index 5c92aff..21f0fc4 100755
--- a/xml2txt.py
+++ b/xml2txt.py
@@ -22,6 +22,7 @@
 import sys
 import optparse
 import xml.parsers.expat
+import gzip
 
 
 ELEMENT_START, ELEMENT_END, CHARACTER_DATA, EOF = range(4)
@@ -99,6 +100,7 @@ class XmlTokenizer:
             self.index = 0
             data = self.fp.read(size)
             self.final = len(data) < size
+            data = data.rstrip('\0')
             try:
                 self.parser.Parse(data, self.final)
             except xml.parsers.expat.ExpatError, e:
@@ -176,6 +178,16 @@ class XmlParser:
         return data
 
 
+class GzipFile(gzip.GzipFile):
+
+    def _read_eof(self):
+        # Ignore incomplete files
+        try:
+            gzip.GzipFile._read_eof(self)
+        except IOError:
+            pass
+
+
 class Formatter:
     
     def function(self, name):
@@ -337,7 +349,6 @@ def main():
     if args:
         for arg in args:
             if arg.endswith('.gz'):
-                from gzip import GzipFile
                 stream = GzipFile(arg, 'rt')
             elif arg.endswith('.bz2'):
                 from bz2 import BZ2File
-- 
2.45.2