Skip to content

Commit

Permalink
ICU-22549 Add Normalizer2 Fuzzer
Browse files Browse the repository at this point in the history
ICU-22549 Remove unnecessary include files
  • Loading branch information
FrankYFTang committed Dec 15, 2023
1 parent e76094c commit 7775942
Show file tree
Hide file tree
Showing 2 changed files with 80 additions and 0 deletions.
1 change: 1 addition & 0 deletions icu4c/source/test/fuzzer/Makefile.in
Original file line number Diff line number Diff line change
Expand Up @@ -41,6 +41,7 @@ FUZZER_TARGETS = \
date_time_pattern_generator_fuzzer \
list_format_fuzzer locale_fuzzer \
locale_morph_fuzzer \
normalizer2_fuzzer \
number_format_fuzzer \
number_formatter_fuzzer \
relative_date_time_formatter_fuzzer \
Expand Down
79 changes: 79 additions & 0 deletions icu4c/source/test/fuzzer/normalizer2_fuzzer.cpp
Original file line number Diff line number Diff line change
@@ -0,0 +1,79 @@
// © 2023 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html

// Fuzzer for ICU Calendar.

#include <cstring>

#include "unicode/normalizer2.h"

extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
uint16_t rnd;
UChar32 char1, char2;
if (size < sizeof(rnd) + sizeof(char1) + sizeof(char2)) return 0;
icu::StringPiece fuzzData(reinterpret_cast<const char *>(data), size);

std::memcpy(&rnd, fuzzData.data(), sizeof(rnd));
fuzzData.remove_prefix(sizeof(rnd));
std::memcpy(&char1, fuzzData.data(), sizeof(char1));
fuzzData.remove_prefix(sizeof(char1));
std::memcpy(&char2, fuzzData.data(), sizeof(char2));
fuzzData.remove_prefix(sizeof(char2));

size_t len = fuzzData.size() / sizeof(char16_t);
icu::UnicodeString text(false, reinterpret_cast<const char16_t*>(fuzzData.data()), len);

UErrorCode status = U_ZERO_ERROR;
const icu::Normalizer2* norm = nullptr;
switch (rnd % 6) {
case 0:
norm = icu::Normalizer2::getNFCInstance(status);
break;
case 1:
norm = icu::Normalizer2::getNFDInstance(status);
break;
case 2:
norm = icu::Normalizer2::getNFKCInstance(status);
break;
case 3:
norm = icu::Normalizer2::getNFKDInstance(status);
break;
case 4:
norm = icu::Normalizer2::getNFKCCasefoldInstance(status);
break;
case 5:
norm = icu::Normalizer2::getNFKCSimpleCasefoldInstance(status);
break;
}
if (U_SUCCESS(status)) {
norm->normalize(text, status);
status = U_ZERO_ERROR;

icu::UnicodeString out;

norm->normalize(text, out, status);
status = U_ZERO_ERROR;

norm->normalizeSecondAndAppend(out, text, status);
status = U_ZERO_ERROR;

norm->append(out, text, status);
status = U_ZERO_ERROR;

norm->getDecomposition(char1, out);
norm->getRawDecomposition(char1, out);
norm->composePair(char1, char2);
norm->getCombiningClass(char1);
norm->isNormalized(text, status);
status = U_ZERO_ERROR;

norm->quickCheck(text, status);
status = U_ZERO_ERROR;

norm->hasBoundaryBefore(char1);
norm->hasBoundaryAfter(char1);
norm->isInert(char1);
}

return EXIT_SUCCESS;
}

0 comments on commit 7775942

Please sign in to comment.