ICU-22549 Add RuleBasedBreakIterator fuzzer

unicode-org · Nov 29, 2023 · 5d3e84a · 5d3e84a
1 parent 757d2cd
commit 5d3e84a
Show file tree

Hide file tree

Showing 2 changed files with 27 additions and 1 deletion.
diff --git a/icu4c/source/test/fuzzer/Makefile.in b/icu4c/source/test/fuzzer/Makefile.in
@@ -33,7 +33,7 @@ CPPFLAGS += -I$(srcdir) -I$(top_srcdir)/common -I$(top_srcdir)/i18n -I$(top_srcd
 DEFS += -D'U_TOPSRCDIR="$(top_srcdir)/"' -D'U_TOPBUILDDIR="$(BUILDDIR)"'
 LIBS = $(LIBCTESTFW) $(LIBICUTOOLUTIL) $(LIBICUIO) $(LIBICUI18N) $(LIBICUUC) $(DEFAULT_LIBS) $(LIB_M)
 
-FUZZER_TARGETS = break_iterator_fuzzer calendar_fuzzer collator_compare_fuzzer collator_rulebased_fuzzer converter_fuzzer date_format_fuzzer list_format_fuzzer locale_fuzzer locale_morph_fuzzer number_format_fuzzer relative_date_time_formatter_fuzzer ucasemap_fuzzer uloc_canonicalize_fuzzer uloc_for_language_tag_fuzzer uloc_get_name_fuzzer uloc_is_right_to_left_fuzzer uloc_open_keywords_fuzzer unicode_string_codepage_create_fuzzer uregex_open_fuzzer
+FUZZER_TARGETS = break_iterator_fuzzer calendar_fuzzer collator_compare_fuzzer collator_rulebased_fuzzer converter_fuzzer date_format_fuzzer list_format_fuzzer locale_fuzzer locale_morph_fuzzer number_format_fuzzer relative_date_time_formatter_fuzzer rule_based_break_iterator_fuzzer ucasemap_fuzzer uloc_canonicalize_fuzzer uloc_for_language_tag_fuzzer uloc_get_name_fuzzer uloc_is_right_to_left_fuzzer uloc_open_keywords_fuzzer unicode_string_codepage_create_fuzzer uregex_open_fuzzer
 
 OBJECTS = $(FUZZER_TARGETS:%=%.o)
 OBJECTS += fuzzer_driver.o locale_util.o

diff --git a/icu4c/source/test/fuzzer/rule_based_break_iterator_fuzzer.cpp b/icu4c/source/test/fuzzer/rule_based_break_iterator_fuzzer.cpp
@@ -0,0 +1,26 @@
+// © 2023 and later: Unicode, Inc. and others.
+// License & terms of use: http:https://www.unicode.org/copyright.html
+
+#include <cstring>
+
+#include "fuzzer_utils.h"
+#include "unicode/localpointer.h"
+#include "unicode/locid.h"
+#include "unicode/rbbi.h"
+
+IcuEnvironment* env = new IcuEnvironment();
+
+extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
+ UErrorCode status = U_ZERO_ERROR;
+
+ size_t unistr_size = size/2;
+ std::unique_ptr<char16_t[]> fuzzbuff(new char16_t[unistr_size]);
+ std::memcpy(fuzzbuff.get(), data, unistr_size * 2);
+ icu::UnicodeString fuzzstr(false, fuzzbuff.get(), unistr_size);
+
+ UParseError parse;
+ icu::LocalPointer<icu::RuleBasedBreakIterator> brk(
+ new icu::RuleBasedBreakIterator(fuzzstr, parse, status));
+
+ return 0;
+}