ossfuzz: moving towards the ideal integration

- Start with the basic code from the ossfuzz project.
- Rewrite fuzz corpora to be binary files full of Type-Length-Value
  data, and write a glue layer in the fuzzing function to convert
  corpora into CURL options.
- Have supporting functions to generate corpora from existing tests
- Integrate with Makefile.am

1 files changed, 13 insertions, 8 deletions

diff --git a/tests/curl_test_data.py b/tests/curl_test_data.py
index bfe1287d8..21747407d 100755
--- a/tests/curl_test_data.py
+++ b/tests/curl_test_data.py
@@ -24,12 +24,15 @@
 from __future__ import (absolute_import, division, print_function,
                         unicode_literals)
 import os
-import xml.etree.ElementTree as ET
+import re
 import logging
 
 log = logging.getLogger(__name__)
 
 
+REPLY_DATA = re.compile("<reply>\s*<data>(.*?)</data>", re.MULTILINE | re.DOTALL)
+
+
 class TestData(object):
     def __init__(self, data_folder):
         self.data_folder = data_folder
@@ -39,15 +42,17 @@ class TestData(object):
         filename = os.path.join(self.data_folder,
                                 "test{0}".format(test_number))
 
-        # The user should handle the exception from failing to find the file.
-        tree = ET.parse(filename)
+        log.debug("Parsing file %s", filename)
+
+        with open(filename, "rb") as f:
+            contents = f.read().decode("utf-8")
 
-        # We need the <reply><data> text.
-        reply = tree.find("reply")
-        data = reply.find("data")
+        m = REPLY_DATA.search(contents)
+        if not m:
+            raise Exception("Couldn't find a <reply><data> section")
 
-        # Return the text contents of the data
-        return data.text
+        # Left-strip the data so we don't get a newline before our data.
+        return m.group(1).lstrip()
 
 
 if __name__ == '__main__':