Mirror/zeek
1
0
Fork 0
mirror of https://github.com/zeek/zeek.git synced 2025-10-09 10:08:20 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

Expand unit test for json_escape_utf8 to include all of the strings from the ascii-json-utf8 btest

Browse source
This commit is contained in:
Tim Wojtulewicz 2020-01-08 12:06:52 -07:00
parent 46e7308422
commit ee0619f999
1 changed files with 54 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

54
src/util.cc
View file

@ -2190,6 +2190,60 @@ TEST_CASE("util json_escape_utf8")
CHECK(json_escape_utf8("string") == "string");
CHECK(json_escape_utf8("string\n") == "string\n");
CHECK(json_escape_utf8("string\x82") == "string\\x82");
CHECK(json_escape_utf8("\x07\xd4\xb7o") == "\\x07Էo");
// These strings are duplicated from the scripts.base.frameworks.logging.ascii-json-utf8 btest
// Valid ASCII and valid ASCII control characters
CHECK(json_escape_utf8("a") == "a");
CHECK(json_escape_utf8("\b\f\n\r\t\x00\x15") == "\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
CHECK(json_escape_utf8("\xc3\xb1") == "\xc3\xb1");
// Invalid 2 Octet Sequence
CHECK(json_escape_utf8("\xc3\x28") == "\\xc3(");
CHECK(json_escape_utf8("\xc0\x81") == "\\xc0\\x81");
CHECK(json_escape_utf8("\xc1\x81") == "\\xc1\\x81");
CHECK(json_escape_utf8("\xc2\xcf") == "\\xc2\\xcf");
// Invalid Sequence Identifier
CHECK(json_escape_utf8("\xa0\xa1") == "\\xa0\\xa1");
// Valid 3 Octet Sequence
CHECK(json_escape_utf8("\xe2\x82\xa1") == "\xe2\x82\xa1");
CHECK(json_escape_utf8("\xe0\xa3\xa1") == "\xe0\xa3\xa1");
// Invalid 3 Octet Sequence (in 2nd Octet)
CHECK(json_escape_utf8("\xe0\x80\xa1") == "\\xe0\\x80\\xa1");
CHECK(json_escape_utf8("\xe2\x28\xa1") == "\\xe2(\\xa1");
CHECK(json_escape_utf8("\xed\xa0\xa1") == "\\xed\\xa0\\xa1");
// Invalid 3 Octet Sequence (in 3rd Octet)
CHECK(json_escape_utf8("\xe2\x82\x28") == "\\xe2\\x82(");
// Valid 4 Octet Sequence
CHECK(json_escape_utf8("\xf0\x90\x8c\xbc") == "\xf0\x90\x8c\xbc");
CHECK(json_escape_utf8("\xf1\x80\x8c\xbc") == "\xf1\x80\x8c\xbc");
CHECK(json_escape_utf8("\xf4\x80\x8c\xbc") == "\xf4\x80\x8c\xbc");
// Invalid 4 Octet Sequence (in 2nd Octet)
CHECK(json_escape_utf8("\xf0\x80\x8c\xbc") == "\\xf0\\x80\\x8c\\xbc");
CHECK(json_escape_utf8("\xf2\x28\x8c\xbc") == "\\xf2(\\x8c\\xbc");
CHECK(json_escape_utf8("\xf4\x90\x8c\xbc") == "\\xf4\\x90\\x8c\\xbc");
// Invalid 4 Octet Sequence (in 3rd Octet)
CHECK(json_escape_utf8("\xf0\x90\x28\xbc") == "\\xf0\\x90(\\xbc");
// Invalid 4 Octet Sequence (in 4th Octet)
CHECK(json_escape_utf8("\xf0\x28\x8c\x28") == "\\xf0(\\x8c(");
// Invalid 4 Octet Sequence (too short)
CHECK(json_escape_utf8("\xf4\x80\x8c") == "\\xf4\\x80\\x8c");
CHECK(json_escape_utf8("\xf0") == "\\xf0");
}
string json_escape_utf8(const string& val)