warc load: make http response/request protocol/verb validation optional

enabled for replay, disabled by default for cdx-indexing, though can
be enabled with -v option #99

pywb/utils/statusandheaders.py

@@ -133,8 +133,9 @@ class StatusAndHeadersParser(object):
     Parser which consumes a stream support readline() to read
     status and headers and return a StatusAndHeaders object
     """
-    def __init__(self, statuslist):
+    def __init__(self, statuslist, verify=True):
         self.statuslist = statuslist
+        self.verify = verify
 
     def parse(self, stream, full_statusline=None):
         """
@@ -160,12 +161,16 @@ class StatusAndHeadersParser(object):
                                     protocol='',
                                     total_len=total_read)
 
+        # validate only if verify is set
+        if self.verify:
             protocol_status = self.split_prefix(statusline, self.statuslist)
 
             if not protocol_status:
                 msg = 'Expected Status Line starting with {0} - Found: {1}'
                 msg = msg.format(self.statuslist, statusline)
                 raise StatusAndHeadersParserException(msg, full_statusline)
+        else:
+            protocol_status = statusline.split(' ', 1)
 
         line, total_read = _strip_count(stream.readline(), total_read)
         while line:

pywb/warc/archiveiterator.py

@@ -45,10 +45,11 @@ class ArchiveIterator(object):
     warc2warc -Z myfile.{0} > myfile.{0}.gz
     """
 
-    def __init__(self, fileobj, no_record_parse=False):
+    def __init__(self, fileobj, no_record_parse=False,
+                 verify_http=False):
         self.fh = fileobj
 
-        self.loader = ArcWarcRecordLoader()
+        self.loader = ArcWarcRecordLoader(verify_http=verify_http)
         self.reader = None
 
         self.offset = 0
@@ -445,7 +446,8 @@ class DefaultRecordIter(object):
         return entry
 
     def __call__(self, fh):
-        aiter = ArchiveIterator(fh, self.options.get('minimal', False))
+        aiter = ArchiveIterator(fh, self.options.get('minimal', False),
+                                    self.options.get('verify_http', False))
 
         entry_iter = self.create_record_iter(aiter)
 

pywb/warc/cdxindexer.py

@@ -311,6 +311,12 @@ Sort the output to each file before writing to create a total ordering
 Convert SURT (Sort-friendly URI Reordering Transform) back to regular
 urls for the cdx key. Default is to use SURT keys.
 Not-recommended for new cdx, use only for backwards-compatibility.
+"""
+
+    verify_help = """
+Verify HTTP protocol (1.0/1.1) status in response records and http verb
+on request records, ensuring the protocol or verb matches the expected list.
+Raise an exception on failure. (This was previously the default behavior).
 """
 
     cdx09_help = """
@@ -391,6 +397,10 @@ instead of current working directory
                         action='store_true',
                         help=unsurt_help)
 
+    parser.add_argument('-v', '--verify',
+                        action='store_true',
+                        help=verify_help)
+
     group = parser.add_mutually_exclusive_group()
     group.add_argument('-9', '--cdx09',
                         action='store_true',
@@ -416,6 +426,7 @@ instead of current working directory
                           append_post=cmd.postappend,
                           recurse=cmd.recurse,
                           rel_root=cmd.dir_root,
+                          verify_http=cmd.verify,
                           cdx09=cmd.cdx09,
                           cdxj=cmd.cdxj,
                           minimal=cmd.minimal_cdxj)

pywb/warc/recordloader.py

@@ -49,7 +49,8 @@ class ArcWarcRecordLoader:
 
     NON_HTTP_RECORDS = ('warcinfo', 'arc_header', 'metadata', 'resource')
 
-    def __init__(self, loader=None, cookie_maker=None, block_size=8192):
+    def __init__(self, loader=None, cookie_maker=None, block_size=8192,
+                 verify_http=True):
         if not loader:
             loader = BlockLoader(cookie_maker)
 
@@ -59,9 +60,9 @@ class ArcWarcRecordLoader:
         self.arc_parser = ARCHeadersParser(self.ARC_HEADERS)
 
         self.warc_parser = StatusAndHeadersParser(self.WARC_TYPES)
-        self.http_parser = StatusAndHeadersParser(self.HTTP_TYPES)
+        self.http_parser = StatusAndHeadersParser(self.HTTP_TYPES, verify_http)
 
-        self.http_req_parser = StatusAndHeadersParser(self.HTTP_VERBS)
+        self.http_req_parser = StatusAndHeadersParser(self.HTTP_VERBS, verify_http)
 
     def load(self, url, offset, length):
         """ Load a single record from given url at offset with length