Mirror/zeek
1
0
Fork 0
mirror of https://github.com/zeek/zeek.git synced 2025-10-02 14:48:21 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

Merge remote-tracking branch 'origin/topic/timw/cleaner-utf8'

Browse source 
* origin/topic/timw/cleaner-utf8:
  GHI-486: Switch over to using LLVM utf8-checking code to better validate characters

I addressed a buffer over-read during the merge and added test-cases for
it.
This commit is contained in:
Johanna Amann 2019-07-29 09:12:52 -07:00
commit 486bf1e713
3 changed files with 57 additions and 53 deletions
Download patch file Download diff file Expand all files Collapse all files

14
testing/btest/Baseline/scripts.base.frameworks.logging.ascii-json-utf8/ssh.log
View file

@ -2,11 +2,23 @@
{"s":"\b\f\n\r\t\\x00\\x15"}
{"s":"ñ"}
{"s":"\\xc3("}
{"s":"\\xc0\\x81"}
{"s":"\\xc1\\x81"}
{"s":"\\xc2\\xcf"}
{"s":"\\xa0\\xa1"}
{"s":"₡"}
{"s":"࣡"}
{"s":"\\xe0\\x80\\xa1"}
{"s":"\\xe2(\\xa1"}
{"s":"\\xed\\xa0\\xa1"}
{"s":"\\xe2\\x82("}
{"s":"𐌼"}
{"s":"\\xf0(\\x8c\\xbc"}
{"s":"񀌼"}
{"s":"􀌼"}
{"s":"\\xf0\\x80\\x8c\\xbc"}
{"s":"\\xf2(\\x8c\\xbc"}
{"s":"\\xf4\\x90\\x8c\\xbc"}
{"s":"\\xf0\\x90(\\xbc"}
{"s":"\\xf0(\\x8c("}
{"s":"\\xf4\\x80\\x8c"}
{"s":"\\xf0"}

37
testing/btest/scripts/base/frameworks/logging/ascii-json-utf8.zeek
View file

@ -27,33 +27,50 @@ event zeek_init()
Log::write(SSH::LOG, [$s="a"]);
Log::write(SSH::LOG, [$s="\b\f\n\r\t\x00\x15"]);
# Table 3-7 in https://www.unicode.org/versions/Unicode12.0.0/ch03.pdf describes what is
# valid and invalid for the tests below
# Valid 2 Octet Sequence
Log::write(SSH::LOG, [$s="\xc3\xb1"]);
# Invalid 2 Octet Sequence
Log::write(SSH::LOG, [$s="\xc3\x28"]);
Log::write(SSH::LOG, [$s="\xc0\x81"]);
Log::write(SSH::LOG, [$s="\xc1\x81"]);
Log::write(SSH::LOG, [$s="\xc2\xcf"]);
# Invalid Sequence Identifier
Log::write(SSH::LOG, [$s="\xa0\xa1"]);
# Valid 3 Octet Sequence
Log::write(SSH::LOG, [$s="\xe2\x82\xa1"]);
Log::write(SSH::LOG, [$s="\xe0\xa3\xa1"]);
# Invalid 3 Octet Sequence (in 2nd Octet)
Log::write(SSH::LOG, [$s="\xe0\x80\xa1"]);
Log::write(SSH::LOG, [$s="\xe2\x28\xa1"]);
Log::write(SSH::LOG, [$s="\xed\xa0\xa1"]);
# Invalid 3 Octet Sequence (in 3rd Octet)
Log::write(SSH::LOG, [$s="\xe2\x82\x28"]);
# Valid 4 Octet Sequence
Log::write(SSH::LOG, [$s="\xf0\x90\x8c\xbc"]);
Log::write(SSH::LOG, [$s="\xf1\x80\x8c\xbc"]);
Log::write(SSH::LOG, [$s="\xf4\x80\x8c\xbc"]);
# Invalid 4 Octet Sequence (in 2nd Octet)
Log::write(SSH::LOG, [$s="\xf0\x28\x8c\xbc"]);
Log::write(SSH::LOG, [$s="\xf0\x80\x8c\xbc"]);
Log::write(SSH::LOG, [$s="\xf2\x28\x8c\xbc"]);
Log::write(SSH::LOG, [$s="\xf4\x90\x8c\xbc"]);
# Invalid 4 Octet Sequence (in 3rd Octet)
Log::write(SSH::LOG, [$s="\xf0\x90\x28\xbc"]);
# Invalid 4 Octet Sequence (in 4th Octet)
Log::write(SSH::LOG, [$s="\xf0\x28\x8c\x28"]);
# Invalid 4 Octet Sequence (too short)
Log::write(SSH::LOG, [$s="\xf4\x80\x8c"]);
Log::write(SSH::LOG, [$s="\xf0"]);
}