Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(219)

Side by Side Diff: base/json/string_escape.cc

Issue 106793004: Revert of Stop doing unnecessary UTF-8 to UTF-16 conversions in JSONWriter. (Closed) Base URL: svn://svn.chromium.org/chrome/trunk/src
Patch Set: Created 7 years ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch | Annotate | Revision Log
« no previous file with comments | « base/json/string_escape.h ('k') | base/json/string_escape_unittest.cc » ('j') | no next file with comments »
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
OLDNEW
1 // Copyright (c) 2006-2008 The Chromium Authors. All rights reserved. 1 // Copyright (c) 2006-2008 The Chromium Authors. All rights reserved.
2 // Use of this source code is governed by a BSD-style license that can be 2 // Use of this source code is governed by a BSD-style license that can be
3 // found in the LICENSE file. 3 // found in the LICENSE file.
4 4
5 #include "base/json/string_escape.h" 5 #include "base/json/string_escape.h"
6 6
7 #include <string> 7 #include <string>
8 8
9 #include "base/strings/string_util.h" 9 #include "base/strings/string_util.h"
10 #include "base/strings/stringprintf.h" 10 #include "base/strings/stringprintf.h"
11 #include "base/strings/utf_string_conversion_utils.h"
12 #include "base/strings/utf_string_conversions.h"
13 #include "base/third_party/icu/icu_utf.h"
14 11
15 namespace base { 12 namespace base {
16 13
17 namespace { 14 namespace {
18 15
19 // Format string for printing a \uXXXX escape sequence. 16 // Try to escape |c| as a "SingleEscapeCharacter" (\n, etc). If successful,
20 const char kU16EscapeFormat[] = "\\u%04X"; 17 // returns true and appends the escape sequence to |dst|. This isn't required
21 18 // by the spec, but it's more readable by humans than the \uXXXX alternatives.
22 // The code point to output for an invalid input code unit. 19 template<typename CHAR>
23 const uint32 kReplacementCodePoint = 0xFFFD; 20 static bool JsonSingleEscapeChar(const CHAR c, std::string* dst) {
24
25 // Used below in EscapeSpecialCodePoint().
26 COMPILE_ASSERT('<' == 0x3C, less_than_sign_is_0x3c);
27
28 // Try to escape the |code_point| if it is a known special character. If
29 // successful, returns true and appends the escape sequence to |dest|. This
30 // isn't required by the spec, but it's more readable by humans.
31 bool EscapeSpecialCodePoint(uint32 code_point, std::string* dest) {
32 // WARNING: if you add a new case here, you need to update the reader as well. 21 // WARNING: if you add a new case here, you need to update the reader as well.
33 // Note: \v is in the reader, but not here since the JSON spec doesn't 22 // Note: \v is in the reader, but not here since the JSON spec doesn't
34 // allow it. 23 // allow it.
35 switch (code_point) { 24 switch (c) {
36 case '\b': 25 case '\b':
37 dest->append("\\b"); 26 dst->append("\\b");
38 break; 27 break;
39 case '\f': 28 case '\f':
40 dest->append("\\f"); 29 dst->append("\\f");
41 break; 30 break;
42 case '\n': 31 case '\n':
43 dest->append("\\n"); 32 dst->append("\\n");
44 break; 33 break;
45 case '\r': 34 case '\r':
46 dest->append("\\r"); 35 dst->append("\\r");
47 break; 36 break;
48 case '\t': 37 case '\t':
49 dest->append("\\t"); 38 dst->append("\\t");
50 break; 39 break;
51 case '\\': 40 case '\\':
52 dest->append("\\\\"); 41 dst->append("\\\\");
53 break; 42 break;
54 case '"': 43 case '"':
55 dest->append("\\\""); 44 dst->append("\\\"");
56 break;
57 // Escape < to prevent script execution; escaping > is not necessary and
58 // not doing so save a few bytes.
59 case '<':
60 dest->append("\\u003C");
61 break; 45 break;
62 default: 46 default:
63 return false; 47 return false;
64 } 48 }
65 return true; 49 return true;
66 } 50 }
67 51
68 template <typename S> 52 template <class STR>
69 bool EscapeJSONStringImpl(const S& str, bool put_in_quotes, std::string* dest) { 53 void JsonDoubleQuoteT(const STR& str,
70 bool did_replacement = false; 54 bool put_in_quotes,
55 std::string* dst) {
56 if (put_in_quotes)
57 dst->push_back('"');
71 58
72 if (put_in_quotes) 59 for (typename STR::const_iterator it = str.begin(); it != str.end(); ++it) {
73 dest->push_back('"'); 60 typename ToUnsigned<typename STR::value_type>::Unsigned c = *it;
74 61 if (!JsonSingleEscapeChar(c, dst)) {
75 // Casting is necessary because ICU uses int32. Try and do so safely. 62 if (c < 32 || c > 126 || c == '<' || c == '>') {
76 CHECK_LE(str.length(), static_cast<size_t>(kint32max)); 63 // 1. Escaping <, > to prevent script execution.
77 const int32 length = static_cast<int32>(str.length()); 64 // 2. Technically, we could also pass through c > 126 as UTF8, but this
78 65 // is also optional. It would also be a pain to implement here.
79 for (int32 i = 0; i < length; ++i) { 66 unsigned int as_uint = static_cast<unsigned int>(c);
80 uint32 code_point; 67 base::StringAppendF(dst, "\\u%04X", as_uint);
81 if (!ReadUnicodeCharacter(str.data(), length, &i, &code_point)) { 68 } else {
82 code_point = kReplacementCodePoint; 69 unsigned char ascii = static_cast<unsigned char>(*it);
83 did_replacement = true; 70 dst->push_back(ascii);
71 }
84 } 72 }
85
86 if (EscapeSpecialCodePoint(code_point, dest))
87 continue;
88
89 // Escape non-printing characters.
90 if (code_point < 32)
91 base::StringAppendF(dest, kU16EscapeFormat, code_point);
92 else
93 WriteUnicodeCharacter(code_point, dest);
94 } 73 }
95 74
96 if (put_in_quotes) 75 if (put_in_quotes)
97 dest->push_back('"'); 76 dst->push_back('"');
98
99 return !did_replacement;
100 } 77 }
101 78
102 } // namespace 79 } // namespace
103 80
104 bool EscapeJSONString(const StringPiece& str, 81 void JsonDoubleQuote(const StringPiece& str,
105 bool put_in_quotes, 82 bool put_in_quotes,
106 std::string* dest) { 83 std::string* dst) {
107 return EscapeJSONStringImpl(str, put_in_quotes, dest); 84 JsonDoubleQuoteT(str, put_in_quotes, dst);
108 } 85 }
109 86
110 bool EscapeJSONString(const StringPiece16& str, 87 std::string GetDoubleQuotedJson(const StringPiece& str) {
111 bool put_in_quotes, 88 std::string dst;
112 std::string* dest) { 89 JsonDoubleQuote(str, true, &dst);
113 return EscapeJSONStringImpl(str, put_in_quotes, dest); 90 return dst;
114 } 91 }
115 92
116 std::string GetQuotedJSONString(const StringPiece& str) { 93 void JsonDoubleQuote(const StringPiece16& str,
117 std::string dest; 94 bool put_in_quotes,
118 bool ok = EscapeJSONStringImpl(str, true, &dest); 95 std::string* dst) {
119 DCHECK(ok); 96 JsonDoubleQuoteT(str, put_in_quotes, dst);
120 return dest;
121 } 97 }
122 98
123 std::string GetQuotedJSONString(const StringPiece16& str) { 99 std::string GetDoubleQuotedJson(const StringPiece16& str) {
124 std::string dest; 100 std::string dst;
125 bool ok = EscapeJSONStringImpl(str, true, &dest); 101 JsonDoubleQuote(str, true, &dst);
126 DCHECK(ok); 102 return dst;
127 return dest;
128 }
129
130 std::string EscapeBytesAsInvalidJSONString(const StringPiece& str,
131 bool put_in_quotes) {
132 std::string dest;
133
134 if (put_in_quotes)
135 dest.push_back('"');
136
137 for (StringPiece::const_iterator it = str.begin(); it != str.end(); ++it) {
138 ToUnsigned<StringPiece::value_type>::Unsigned c = *it;
139 if (EscapeSpecialCodePoint(c, &dest))
140 continue;
141
142 if (c < 32 || c > 126)
143 base::StringAppendF(&dest, kU16EscapeFormat, c);
144 else
145 dest.push_back(*it);
146 }
147
148 if (put_in_quotes)
149 dest.push_back('"');
150
151 return dest;
152 } 103 }
153 104
154 } // namespace base 105 } // namespace base
OLDNEW
« no previous file with comments | « base/json/string_escape.h ('k') | base/json/string_escape_unittest.cc » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698