Wip/ritesh/resilient stripper (#215)

* Be more resilient in parsing the signature file

The assumption that the signature file is well crafted is wrong. Over
time, there have been numerous reports from many derivatives, that add
many extra bits to the output. Obviously, apt never guaranteed a machine
readable format for its output.

With this change, we are being more resilient in dealing with each line
in the signature file. Things like malforned file (cdrom lines -
historical ones) or empty new lines should all be handled.

apt-offline should treat the signature file dumb and very carefully and
resiliently parse the content.

Closes: #214

* Be more explicit

* Set extracted size value to type integer

Author: rickysarraf

apt_offline_core/AptOfflineCoreLib.py

@@ -860,41 +860,18 @@ def stripper(item):
         and returns them.'''
 
         log.verbose("Item before split is: %s\n" % (item))
-        try:
-            url, localFile, size, checksum = item.split(' ')
-        except ValueError:
-            #INFO: For apt metadata, no checksum is present
-            # For such case, set checksum to None
-            url, localFile, size = item.split(' ')
-            checksum = None
-        log.verbose("Item after split is: %s %s %s %s\n" % (url, localFile, size, checksum))
-
-        url = url.rstrip("'")
-        url = url.lstrip("'")
-        log.verbose("Stripped item URL is: %s\n" % url)
-
-        localFile = localFile.rstrip("'")
-        localFile = localFile.lstrip("'")
-        log.verbose("Stripped item FILE is: %s\n" % localFile)
-
-        try:
-                size = size.rstrip("'")
-                size = size.lstrip("'")
-                size = int(size)
-        except ValueError:
-                log.verbose("%s is malformed\n" % (" ".join(item) ) )
-                size = 0
-        log.verbose("Stripped item SIZE is: %d\n" % size)
+        SplitItem = item.split(' ')
+        url = SplitItem[0].strip("'").strip()
+        localFile = SplitItem[1].strip("'").strip()
+        size = SplitItem[2].strip("'").strip()
+        checksum = SplitItem[3].strip("'").strip()
+        log.verbose("Items after split is: %s\n" % (SplitItem))
 
-        #INFO: md5 ends up having '\n' with it.
-        # That needs to be stripped too.
-        if checksum:
-            checksum = checksum.rstrip("'")
-            checksum = checksum.lstrip("'")
-            checksum = checksum.rstrip("\n")
-        log.verbose("Stripped item CHECKSUM is: %s\n" % checksum)
+        # Convert size to integer
+        size = int(size) if size.isdecimal() else 0
+        log.verbose("size of size is: %s\n" % (size))
 
-        return url, localFile, size, checksum
+        return (url, localFile, size, checksum)
 
 
 def errfunc(errno, errormsg, filename):
@@ -1273,8 +1250,14 @@ def buildChangelog(self, pkgPath, installedVersion):
                                             log.verbose("Printing Release URL/Files\n")
                                             log.verbose("%s %s" % (ReleaseItemURL, ReleaseItemFile) )
                                         FetchData['Item'].append( item )
+                                except IndexError:
+                                    log.err("Failed parisng file: %s, while processing line: %s\n" % (Str_GetArg, item))
+                                    log.err(traceback.format_exc())
+                                    sys.exit(1)
                                 except ValueError:
-                                        log.err("Cannot parse line %s\n" % (item))
+                                    log.err("Cannot parse line: %s\n" % (item))
+                                    log.err(traceback.format_exc())
+                                    sys.exit(1)
 
         del raw_data_list