1// Copyright (c) 2013 The Chromium Authors. All rights reserved.
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 "base/json/string_escape.h"
6
7#include "base/strings/string_util.h"
8#include "base/strings/utf_string_conversions.h"
9#include "testing/gtest/include/gtest/gtest.h"
10
11namespace base {
12
13TEST(JSONStringEscapeTest, EscapeUTF8) {
14  const struct {
15    const char* to_escape;
16    const char* escaped;
17  } cases[] = {
18    {"\b\001aZ\"\\wee", "\\b\\u0001aZ\\\"\\\\wee"},
19    {"a\b\f\n\r\t\v\1\\.\"z",
20        "a\\b\\f\\n\\r\\t\\u000B\\u0001\\\\.\\\"z"},
21    {"b\x0f\x7f\xf0\xff!",  // \xf0\xff is not a valid UTF-8 unit.
22        "b\\u000F\x7F\xEF\xBF\xBD\xEF\xBF\xBD!"},
23    {"c<>d", "c\\u003C>d"},
24  };
25
26  for (size_t i = 0; i < ARRAYSIZE_UNSAFE(cases); ++i) {
27    const char* in_ptr = cases[i].to_escape;
28    std::string in_str = in_ptr;
29
30    std::string out;
31    EscapeJSONString(in_ptr, false, &out);
32    EXPECT_EQ(std::string(cases[i].escaped), out);
33    EXPECT_TRUE(IsStringUTF8(out));
34
35    out.erase();
36    bool convert_ok = EscapeJSONString(in_str, false, &out);
37    EXPECT_EQ(std::string(cases[i].escaped), out);
38    EXPECT_TRUE(IsStringUTF8(out));
39
40    if (convert_ok) {
41      std::string fooout = GetQuotedJSONString(in_str);
42      EXPECT_EQ("\"" + std::string(cases[i].escaped) + "\"", fooout);
43      EXPECT_TRUE(IsStringUTF8(out));
44    }
45  }
46
47  std::string in = cases[0].to_escape;
48  std::string out;
49  EscapeJSONString(in, false, &out);
50  EXPECT_TRUE(IsStringUTF8(out));
51
52  // test quoting
53  std::string out_quoted;
54  EscapeJSONString(in, true, &out_quoted);
55  EXPECT_EQ(out.length() + 2, out_quoted.length());
56  EXPECT_EQ(out_quoted.find(out), 1U);
57  EXPECT_TRUE(IsStringUTF8(out_quoted));
58
59  // now try with a NULL in the string
60  std::string null_prepend = "test";
61  null_prepend.push_back(0);
62  in = null_prepend + in;
63  std::string expected = "test\\u0000";
64  expected += cases[0].escaped;
65  out.clear();
66  EscapeJSONString(in, false, &out);
67  EXPECT_EQ(expected, out);
68  EXPECT_TRUE(IsStringUTF8(out));
69}
70
71TEST(JSONStringEscapeTest, EscapeUTF16) {
72  const struct {
73    const wchar_t* to_escape;
74    const char* escaped;
75  } cases[] = {
76    {L"b\uffb1\u00ff", "b\xEF\xBE\xB1\xC3\xBF"},
77    {L"\b\001aZ\"\\wee", "\\b\\u0001aZ\\\"\\\\wee"},
78    {L"a\b\f\n\r\t\v\1\\.\"z",
79        "a\\b\\f\\n\\r\\t\\u000B\\u0001\\\\.\\\"z"},
80    {L"b\x0f\x7f\xf0\xff!", "b\\u000F\x7F\xC3\xB0\xC3\xBF!"},
81    {L"c<>d", "c\\u003C>d"},
82  };
83
84  for (size_t i = 0; i < ARRAYSIZE_UNSAFE(cases); ++i) {
85    string16 in = WideToUTF16(cases[i].to_escape);
86
87    std::string out;
88    EscapeJSONString(in, false, &out);
89    EXPECT_EQ(std::string(cases[i].escaped), out);
90    EXPECT_TRUE(IsStringUTF8(out));
91
92    out = GetQuotedJSONString(in);
93    EXPECT_EQ("\"" + std::string(cases[i].escaped) + "\"", out);
94    EXPECT_TRUE(IsStringUTF8(out));
95  }
96
97  string16 in = WideToUTF16(cases[0].to_escape);
98  std::string out;
99  EscapeJSONString(in, false, &out);
100  EXPECT_TRUE(IsStringUTF8(out));
101
102  // test quoting
103  std::string out_quoted;
104  EscapeJSONString(in, true, &out_quoted);
105  EXPECT_EQ(out.length() + 2, out_quoted.length());
106  EXPECT_EQ(out_quoted.find(out), 1U);
107  EXPECT_TRUE(IsStringUTF8(out));
108
109  // now try with a NULL in the string
110  string16 null_prepend = WideToUTF16(L"test");
111  null_prepend.push_back(0);
112  in = null_prepend + in;
113  std::string expected = "test\\u0000";
114  expected += cases[0].escaped;
115  out.clear();
116  EscapeJSONString(in, false, &out);
117  EXPECT_EQ(expected, out);
118  EXPECT_TRUE(IsStringUTF8(out));
119}
120
121TEST(JSONStringEscapeTest, EscapeUTF16OutsideBMP) {
122  {
123    // {a, U+10300, !}, SMP.
124    string16 test;
125    test.push_back('a');
126    test.push_back(0xD800);
127    test.push_back(0xDF00);
128    test.push_back('!');
129    std::string actual;
130    EXPECT_TRUE(EscapeJSONString(test, false, &actual));
131    EXPECT_EQ("a\xF0\x90\x8C\x80!", actual);
132  }
133  {
134    // {U+20021, U+2002B}, SIP.
135    string16 test;
136    test.push_back(0xD840);
137    test.push_back(0xDC21);
138    test.push_back(0xD840);
139    test.push_back(0xDC2B);
140    std::string actual;
141    EXPECT_TRUE(EscapeJSONString(test, false, &actual));
142    EXPECT_EQ("\xF0\xA0\x80\xA1\xF0\xA0\x80\xAB", actual);
143  }
144  {
145    // {?, U+D800, @}, lone surrogate.
146    string16 test;
147    test.push_back('?');
148    test.push_back(0xD800);
149    test.push_back('@');
150    std::string actual;
151    EXPECT_FALSE(EscapeJSONString(test, false, &actual));
152    EXPECT_EQ("?\xEF\xBF\xBD@", actual);
153  }
154}
155
156TEST(JSONStringEscapeTest, EscapeBytes) {
157  const struct {
158    const char* to_escape;
159    const char* escaped;
160  } cases[] = {
161    {"b\x0f\x7f\xf0\xff!", "b\\u000F\\u007F\\u00F0\\u00FF!"},
162    {"\xe5\xc4\x4f\x05\xb6\xfd\0", "\\u00E5\\u00C4O\\u0005\\u00B6\\u00FD"},
163  };
164
165  for (size_t i = 0; i < ARRAYSIZE_UNSAFE(cases); ++i) {
166    std::string in = std::string(cases[i].to_escape);
167    EXPECT_FALSE(IsStringUTF8(in));
168
169    EXPECT_EQ(std::string(cases[i].escaped),
170        EscapeBytesAsInvalidJSONString(in, false));
171    EXPECT_EQ("\"" + std::string(cases[i].escaped) + "\"",
172        EscapeBytesAsInvalidJSONString(in, true));
173  }
174
175  const char kEmbedNull[] = { '\xab', '\x39', '\0', '\x9f', '\xab' };
176  std::string in(kEmbedNull, ARRAYSIZE_UNSAFE(kEmbedNull));
177  EXPECT_FALSE(IsStringUTF8(in));
178  EXPECT_EQ(std::string("\\u00AB9\\u0000\\u009F\\u00AB"),
179            EscapeBytesAsInvalidJSONString(in, false));
180}
181
182}  // namespace base
183