Improve fuzzer

[tyler92]

The current fuzzer target covered only basic things and didn't cover the most interesting parts of HTTP request parsing. I've extended fuzzer with more targets and added a corpus. As a result, coverage has been increased significantly and several memory issues have been found; I will report them separately to avoid mixing them with this MR.

Notes:

1. Due to the reuse policy, I had to add ".license" files for each file in the corpus. I'm unsure if it's a good idea and would appreciate your suggestions.
2. The "router" target can't be covered via public interface because the Router class requires dependencies that send messages over a network. So Router is covered via the SegmentTreeNode class.
3. The 'removeRoute' method is excluded from the fuzzer due to crashes (there are no null pointer checks). Probably we can enable it later when it's fixed

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 78.09%. Comparing base (7cc8732) to head (5dba79a).
Report is 5 commits behind head on master.

❗ Current head 5dba79a differs from pull request most recent head 89cf458. Consider uploading reports for the commit 89cf458 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1191      +/-   ##
==========================================
- Coverage   78.09%   78.09%   -0.01%     
==========================================
  Files          53       53              
  Lines        7082     7077       -5     
==========================================
- Hits         5531     5527       -4     
+ Misses       1551     1550       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[kiplingw]

Good work @tyler92. We'll get back to you as soon as we've had a chance to review.

[Tachi107]

Hi @tyler92, thanks for the patch! Fuzzing is definitely an interesting area where we could do better.

Due to the reuse policy, I had to add ".license" files for each file in the corpus. I'm unsure if it's a good idea and would appreciate your suggestions.

What you can do instead is add a .reuse/dep5 file in the project root which contains licensing information for all those small files, for example:

Format: https://www.debian.org/doc/packaging-manuals/copyright-format/1.0/

Files: tests/fuzzers/corpus/http-*.txt
Copyright: 2024 Mikhail Khachayants
License: Apache-2.0

[tyler92]

@Tachi107 Thanks for the hint with .reuse/dep5, it works!

[tyler92]

@kiplingw Just in case - there are no more changes expected on my part, the MR is ready for review. Please let me know if something has to be fixed or improved.