Add a script for formatting the VerifySignedData() test data.

net/data/verify_signed_data_unittest/annotate_test_data.py

Index: net/data/verify_signed_data_unittest/annotate_test_data.py
diff --git a/net/data/verify_signed_data_unittest/annotate_test_data.py b/net/data/verify_signed_data_unittest/annotate_test_data.py
new file mode 100755
index 0000000000000000000000000000000000000000..7f4f7082fa3d724376362af90dfdaab9b7f2f9d4
--- /dev/null
+++ b/net/data/verify_signed_data_unittest/annotate_test_data.py
@@ -0,0 +1,166 @@
+#!/usr/bin/python
+# Copyright (c) 2015 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+"""This script is called without any arguments to re-format all of the *.pem
+files in the script's parent directory.
+
+The main formatting change is to run "openssl asn1parse" for each of the PEM
+block sections (except for DATA), and add that output to the comment.
+
+Refer to the README file for more information.
+"""
+
+import glob
+import os
+import re
+import base64
+import subprocess
+
+
+def Transform(file_data):
+  """Returns a transformed (formatted) version of file_data"""
+
+  result = ''
+
+  # Get the file's description (all the text before the first PEM block)
+  file_description = GetTextUntilNextPemBlock(file_data)
+
+  result += file_description + '\n'
+
+  for block in GetPemBlocks(file_data):
+    result += '\n\n\n'
+
+    result += MakePemBlockString(block.name, block.data)
+
+    # If there was a user comment (non-script-generated comment) associated
+    # with the block, output it immediately after the block.
+    user_comment = GetUserComment(block.comment)
+    if user_comment:
+      result += '\n' + user_comment + '\n'
+
+    # For every block except for DATA, try to pretty print the parsed ASN.1.
+    # DATA blocks likely would be DER in practice, but for the purposes of
+    # these tests seeing its structure doesn't clarify
+    # anything and is just a distraction.
+    if block.name != 'DATA':
+      generated_comment = GenerateCommentForBlock(block.name, block.data)
+      result += '\n' + generated_comment + '\n'
+
+  return result
+
+
+def GenerateCommentForBlock(block_name, block_data):
+  """Returns a string describing the ASN.1 structure of block_data"""
+
+  p = subprocess.Popen(['openssl', 'asn1parse', '-i', '-inform', 'DER'],
+                       stdout=subprocess.PIPE, stdin=subprocess.PIPE,
+                       stderr=subprocess.PIPE)
+  stdout_data, stderr_data = p.communicate(input=block_data)
+  generated_comment = '$ openssl asn1parse -i < [%s]\n%s' % (block_name,
+                                                             stdout_data)
+  return generated_comment.strip('\n')
+
+
+def GetTextUntilNextPemBlock(text):
+  return text.split('-----BEGIN ', 1)[0].strip('\n')
+
+
+def GetUserComment(comment):
+  """Removes any script-generated lines (everything after the $ openssl line)"""
+
+  # Consider everything after "$ openssl" to be a generated comment.
+  comment = comment.split('$ openssl asn1parse -i', 1)[0].strip('\n')
+  if IsEntirelyWhiteSpace(comment):
+    comment = ''
+  return comment
+
+
+def MakePemBlockString(name, data):
+  return ('-----BEGIN %s-----\n'
+          '%s'
+          '-----END %s-----\n') % (name, EncodeDataForPem(data), name)
+
+
+def GetPemFilePaths():
+  """Returns an iterable for all the paths to the PEM test files"""
+
+  base_dir = os.path.dirname(os.path.realpath(__file__))
+  return glob.iglob(os.path.join(base_dir, '*.pem'))
+
+
+def ReadFileToString(path):
+  with open(path, 'r') as f:
+    return f.read()
+
+
+def WrapTextToLineWidth(text, column_width):
+  result = ''
+  pos = 0
+  while pos < len(text):
+    result += text[pos : pos + column_width] + '\n'
+    pos += column_width
+  return result
+
+
+def EncodeDataForPem(data):
+  result = base64.b64encode(data)
+  return WrapTextToLineWidth(result, 75)
+
+
+class PemBlock(object):
+  def __init__(self):
+    self.name = None
+    self.data = None
+    self.comment = None
+
+
+def StripAllWhitespace(text):
+  pattern = re.compile(r'\s+')
+  return re.sub(pattern, '', text)
+
+
+def IsEntirelyWhiteSpace(text):
+  return len(StripAllWhitespace(text)) == 0
+
+
+def DecodePemBlockData(text):
+  text = StripAllWhitespace(text)
+  return base64.b64decode(text)
+
+
+def GetPemBlocks(data):
+  """Returns an iterable of PemBlock"""
+
+  regex = re.compile(r'-----BEGIN ([\w ]+)-----(.*)-----END \1-----', re.DOTALL)

[mattm, 2015/07/28 22:27:00]

probably want to use non-greedy match, in case the same block name appears
multiple times (add a question mark after the .*).

[eroman, 2015/07/29 01:07:55]

Done.

I also incorporated the comment into the regexp (rather than using
GetTextUntilNextPemBlock).

[mattm, 2015/07/29 01:17:47]

Hm, did you test that? I wasn't sure a non-greedy match at the end of a pattern
would work like that or just match nothing.

[eroman, 2015/07/29 01:38:47]

You are correct, that did not work, thanks!
Reverted that change.

[mattm, 2015/07/29 01:55:23]

I guess you could also use a lookahead like (.*?)(?=-----BEGIN |$)', but that
might be getting too clever. This is fine.

+
+  for match in regex.finditer(data):
+    block = PemBlock()
+
+    block.name = match.group(1)
+    block.data = DecodePemBlockData(match.group(2))
+
+    # Keep track of any non-PEM text between blocks
+    block.comment = GetTextUntilNextPemBlock(data[match.end():])
+
+    yield block
+
+
+def WriteStringToFile(data, path):
+  with open(path, "w") as f:
+    f.write(data)
+
+
+def main():
+  for path in GetPemFilePaths():
+    print "Processing %s ..." % (path)
+    original_data = ReadFileToString(path)
+    transformed_data = Transform(original_data)
+    if original_data != transformed_data:
+      WriteStringToFile(transformed_data, path)
+      print "Rewrote %s" % (path)
+
+
+if __name__ == "__main__":
+  main()