1 // Copyright 2018 The Chromium Authors
2 // Use of this source code is governed by a BSD-style license that can be
3 // found in the LICENSE file.
4
5 #include <stddef.h>
6 #include <stdint.h>
7
8 #include <fuzzer/FuzzedDataProvider.h>
9
10 #include <vector>
11
12 #include "third_party/ced/src/compact_enc_det/compact_enc_det.h"
13
14 namespace {
15 constexpr size_t kMaxInputSize = 64 * 1024;
16 }
17
LLVMFuzzerTestOneInput(const uint8_t * data,size_t size)18 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
19 // Early out if there isn't enough data to extract options and pass data to
20 // the library.
21 if (size < 3 * sizeof(int32_t) + 1)
22 return 0;
23
24 // Limit the input size to avoid timing out on clusterfuzz.
25 if (size > kMaxInputSize)
26 return 0;
27
28 FuzzedDataProvider data_provider(data, size);
29
30 CompactEncDet::TextCorpusType corpus =
31 static_cast<CompactEncDet::TextCorpusType>(
32 data_provider.ConsumeIntegralInRange<int32_t>(
33 0, CompactEncDet::NUM_CORPA));
34 Encoding encoding_hint = static_cast<Encoding>(
35 data_provider.ConsumeIntegralInRange<int32_t>(0, NUM_ENCODINGS));
36 Language langauge_hint = static_cast<Language>(
37 data_provider.ConsumeIntegralInRange<int32_t>(0, NUM_LANGUAGES));
38 bool ignore_7bit_mail_encodings = data_provider.ConsumeBool();
39
40 std::vector<char> text = data_provider.ConsumeRemainingBytes<char>();
41 int bytes_consumed = 0;
42 bool is_reliable = false;
43 CompactEncDet::DetectEncoding(
44 text.data(), text.size(), nullptr /* url_hint */,
45 nullptr /* http_charset_hint */, nullptr /* meta_charset_hint */,
46 encoding_hint, langauge_hint, corpus, ignore_7bit_mail_encodings,
47 &bytes_consumed, &is_reliable);
48
49 return 0;
50 }
51