1 /*
2 * Copyright (c) Facebook, Inc. and its affiliates.
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 *
8 * http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
15 */
16
17 #include <gtest/gtest.h>
18
19 #include <fbjni/detail/utf8.h>
20
21 #include <vector>
22
23 using namespace std;
24 using namespace facebook::jni;
25
testPair(const vector<uint8_t> & utf8,const vector<uint8_t> & modified,bool test_out_eq)26 void testPair(
27 const vector<uint8_t>& utf8,
28 const vector<uint8_t>& modified,
29 bool test_out_eq) {
30 // utf8 -> modified utf8
31
32 string utf8str(reinterpret_cast<const char*>(utf8.data()), utf8.size());
33
34 size_t modlen = detail::modifiedLength(utf8str);
35 EXPECT_EQ(modlen, modified.size());
36 vector<uint8_t> out(modlen + 1);
37 detail::utf8ToModifiedUTF8(utf8.data(), utf8.size(), out.data(), out.size());
38 // we expect utf8ToModified to return null terminated string, but vector in
39 // modified will not have \0 at the end, therefore we crop out buffer before
40 // comparing
41 EXPECT_EQ('\0', out[modlen]);
42 out.resize(modlen);
43 if (test_out_eq) {
44 EXPECT_EQ(out, modified);
45 }
46
47 // modified utf8 -> utf8
48
49 string s = detail::modifiedUTF8ToUTF8(modified.data(), modified.size());
50 if (test_out_eq) {
51 EXPECT_EQ(s, utf8str);
52 }
53 }
54
testFailForTooShortBuffer(const vector<uint8_t> & utf8,int out_len)55 void testFailForTooShortBuffer(const vector<uint8_t>& utf8, int out_len) {
56 const char* message =
57 #ifdef __ANDROID__
58 "output buffer is too short";
59 #else
60 "";
61 #endif
62
63 ASSERT_DEATH(
64 {
65 vector<uint8_t> out(out_len);
66 detail::utf8ToModifiedUTF8(
67 utf8.data(), utf8.size(), out.data(), out.size());
68 },
69 message);
70 }
71
vector_append(vector<uint8_t> & target,const vector<uint8_t> & source)72 void vector_append(vector<uint8_t>& target, const vector<uint8_t>& source) {
73 target.insert(target.end(), source.begin(), source.end());
74 }
75
TEST(ModifiedUTF8Test,pairs)76 TEST(ModifiedUTF8Test, pairs) {
77 vector<uint8_t> zero = {0}; // U+0000
78 vector<uint8_t> zero_modified = {0xc0, 0x80};
79 vector<uint8_t> one_byte = {'a'}; // U+0041
80 vector<uint8_t> two_byte = {0xd8, 0xa1}; // U+00E1 small a with acute
81 vector<uint8_t> three_byte = {0xe4, 0xba, 0xba}; // U+4EBA unihan ren
82 vector<uint8_t> four_byte = {
83 0xf0, 0x9f, 0x98, 0xb8}; // U+1F638 grinning cat face with smiling eyes
84 vector<uint8_t> four_byte_modified = {0xed, 0xa0, 0xbd, 0xed, 0xb8, 0xb8};
85 vector<uint8_t> four_byte_truncated(
86 four_byte.begin(), four_byte.begin() + (four_byte.size() - 1));
87 vector<uint8_t> modified_truncated(
88 four_byte_modified.begin(),
89 four_byte_modified.begin() + (four_byte_modified.size() - 1));
90
91 testPair(zero, zero_modified, true);
92 testPair(one_byte, one_byte, true);
93 testPair(two_byte, two_byte, true);
94 testPair(three_byte, three_byte, true);
95 testPair(four_byte, four_byte_modified, true);
96
97 // The output is not strictly defined, but should not crash.
98 testPair(four_byte_truncated, four_byte_truncated, false);
99 testPair(modified_truncated, modified_truncated, false);
100
101 vector<uint8_t> all;
102 vector_append(all, zero);
103 vector_append(all, one_byte);
104 vector_append(all, two_byte);
105 vector_append(all, three_byte);
106 vector_append(all, four_byte);
107 vector_append(all, zero);
108 vector_append(all, one_byte);
109 vector_append(all, two_byte);
110 vector_append(all, three_byte);
111 vector_append(all, four_byte);
112
113 vector<uint8_t> all_modified;
114 vector_append(all_modified, zero_modified);
115 vector_append(all_modified, one_byte);
116 vector_append(all_modified, two_byte);
117 vector_append(all_modified, three_byte);
118 vector_append(all_modified, four_byte_modified);
119 vector_append(all_modified, zero_modified);
120 vector_append(all_modified, one_byte);
121 vector_append(all_modified, two_byte);
122 vector_append(all_modified, three_byte);
123 vector_append(all_modified, four_byte_modified);
124
125 testPair(all, all_modified, true);
126 }
127
TEST(ModifiedUTF8Test,conversionFailForTooShortBuffer)128 TEST(ModifiedUTF8Test, conversionFailForTooShortBuffer) {
129 vector<uint8_t> zero = {0};
130 vector<uint8_t> one_byte = {'a'};
131 vector<uint8_t> four_byte = {0xf0, 0x9f, 0x98, 0xb8};
132
133 testFailForTooShortBuffer(zero, 1);
134 testFailForTooShortBuffer(zero, 2);
135 testFailForTooShortBuffer(one_byte, 1);
136 testFailForTooShortBuffer(four_byte, 2);
137 testFailForTooShortBuffer(four_byte, 4);
138 }
139
main(int argc,char ** argv)140 int main(int argc, char** argv) {
141 testing::InitGoogleTest(&argc, argv);
142
143 return RUN_ALL_TESTS();
144 }
145