DevTools: always use 16bit strings for inspector protocol.

third_party/WebKit/Source/platform/inspector_protocol/Parser.cpp

 // Copyright 2016 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "platform/inspector_protocol/Parser.h"
 
 #include "platform/inspector_protocol/String16.h"
 #include "platform/inspector_protocol/Values.h"
 
 namespace blink {
@@ skipping 15 matching lines @@
     NullToken,
     ListSeparator,
     ObjectPairSeparator,
     InvalidToken,
 };
 
 const char* const nullString = "null";
 const char* const trueString = "true";
 const char* const falseString = "false";
 
-template<typename CharType>
-bool parseConstToken(const CharType* start, const CharType* end, const CharType** tokenEnd, const char* token)
+bool parseConstToken(const UChar* start, const UChar* end, const UChar** tokenEnd, const char* token)
 {
     while (start < end && *token != '\0' && *start++ == *token++) { }
     if (*token != '\0')
         return false;
     *tokenEnd = start;
     return true;
 }
 
-template<typename CharType>
-bool readInt(const CharType* start, const CharType* end, const CharType** tokenEnd, bool canHaveLeadingZeros)
+bool readInt(const UChar* start, const UChar* end, const UChar** tokenEnd, bool canHaveLeadingZeros)
 {
     if (start == end)
         return false;
     bool haveLeadingZero = '0' == *start;
     int length = 0;
     while (start < end && '0' <= *start && *start <= '9') {
         ++start;
         ++length;
     }
     if (!length)
         return false;
     if (!canHaveLeadingZeros && length > 1 && haveLeadingZero)
         return false;
     *tokenEnd = start;
     return true;
 }
 
-template<typename CharType>
-bool parseNumberToken(const CharType* start, const CharType* end, const CharType** tokenEnd)
+bool parseNumberToken(const UChar* start, const UChar* end, const UChar** tokenEnd)
 {
     // We just grab the number here. We validate the size in DecodeNumber.
     // According to RFC4627, a valid number is: [minus] int [frac] [exp]
     if (start == end)
         return false;
-    CharType c = *start;
+    UChar c = *start;
     if ('-' == c)
         ++start;
 
     if (!readInt(start, end, &start, false))
         return false;
     if (start == end) {
         *tokenEnd = start;
         return true;
     }
 
@@ skipping 22 matching lines @@
                 return false;
         }
         if (!readInt(start, end, &start, true))
             return false;
     }
 
     *tokenEnd = start;
     return true;
 }
 
-template<typename CharType>
-bool readHexDigits(const CharType* start, const CharType* end, const CharType** tokenEnd, int digits)
+bool readHexDigits(const UChar* start, const UChar* end, const UChar** tokenEnd, int digits)
 {
     if (end - start < digits)
         return false;
     for (int i = 0; i < digits; ++i) {
-        CharType c = *start++;
+        UChar c = *start++;
         if (!(('0' <= c && c <= '9') || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')))
             return false;
     }
     *tokenEnd = start;
     return true;
 }
 
-template<typename CharType>
-bool parseStringToken(const CharType* start, const CharType* end, const CharType** tokenEnd)
+bool parseStringToken(const UChar* start, const UChar* end, const UChar** tokenEnd)
 {
     while (start < end) {
-        CharType c = *start++;
+        UChar c = *start++;
         if ('\\' == c) {
             c = *start++;
             // Make sure the escaped char is valid.
             switch (c) {
             case 'x':
                 if (!readHexDigits(start, end, &start, 2))
                     return false;
                 break;
             case 'u':
                 if (!readHexDigits(start, end, &start, 4))
@@ skipping 13 matching lines @@
                 return false;
             }
         } else if ('"' == c) {
             *tokenEnd = start;
             return true;
         }
     }
     return false;
 }
 
-template<typename CharType>
-bool skipComment(const CharType* start, const CharType* end, const CharType** commentEnd)
+bool skipComment(const UChar* start, const UChar* end, const UChar** commentEnd)
 {
     if (start == end)
         return false;
 
     if (*start != '/' || start + 1 >= end)
         return false;
     ++start;
 
     if (*start == '/') {
         // Single line comment, read to newline.
         for (++start; start < end; ++start) {
             if (*start == '\n' || *start == '\r') {
                 *commentEnd = start + 1;
                 return true;
             }
         }
         *commentEnd = end;
         // Comment reaches end-of-input, which is fine.
         return true;
     }
 
     if (*start == '*') {
-        CharType previous = '\0';
+        UChar previous = '\0';
         // Block comment, read until end marker.
         for (++start; start < end; previous = *start++) {
             if (previous == '*' && *start == '/') {
                 *commentEnd = start + 1;
                 return true;
             }
         }
         // Block comment must close before end-of-input.
         return false;
     }
 
     return false;
 }
 
-template<typename CharType>
-void skipWhitespaceAndComments(const CharType* start, const CharType* end, const CharType** whitespaceEnd)
+void skipWhitespaceAndComments(const UChar* start, const UChar* end, const UChar** whitespaceEnd)
 {
     while (start < end) {
         if (isSpaceOrNewline(*start)) {
             ++start;
         } else if (*start == '/') {
-            const CharType* commentEnd;
+            const UChar* commentEnd;
             if (!skipComment(start, end, &commentEnd))
                 break;
             start = commentEnd;
         } else {
             break;
         }
     }
     *whitespaceEnd = start;
 }
 
-template<typename CharType>
-Token parseToken(const CharType* start, const CharType* end, const CharType** tokenStart, const CharType** tokenEnd)
+Token parseToken(const UChar* start, const UChar* end, const UChar** tokenStart, const UChar** tokenEnd)
 {
     skipWhitespaceAndComments(start, end, tokenStart);
     start = *tokenStart;
 
     if (start == end)
         return InvalidToken;
 
     switch (*start) {
     case 'n':
         if (parseConstToken(start, end, tokenEnd, nullString))
@@ skipping 40 matching lines @@
             return Number;
         break;
     case '"':
         if (parseStringToken(start + 1, end, tokenEnd))
             return StringLiteral;
         break;
     }
     return InvalidToken;
 }
 
-template<typename CharType>
-inline int hexToInt(CharType c)
+inline int hexToInt(UChar c)
 {
     if ('0' <= c && c <= '9')
         return c - '0';
     if ('A' <= c && c <= 'F')
         return c - 'A' + 10;
     if ('a' <= c && c <= 'f')
         return c - 'a' + 10;
     ASSERT_NOT_REACHED();
     return 0;
 }
 
-template<typename CharType>
-bool decodeUTF8(const CharType* start, const CharType* end, const CharType** utf8charEnd, String16Builder* output)
-{
-    UChar utf16[4] = {0};
-    char utf8[6] = {0};
-    size_t utf8count = 0;
-
-    while (start < end) {
-        if (start + 1 >= end || *start != '\\' || *(start + 1) != 'x')
-            return false;
-        start += 2;
-
-        // Accumulate one more utf8 character and try converting to utf16.
-        if (start + 1 >= end)
-            return false;
-        utf8[utf8count++] = (hexToInt(*start) << 4) + hexToInt(*(start + 1));
-        start += 2;
-
-        const char* utf8start = utf8;
-        UChar* utf16start = utf16;
-        String16::ConversionResult conversionResult = String16::convertUTF8ToUTF16(&utf8start, utf8start + utf8count, &utf16start, utf16start + 4, nullptr, true);
-
-        if (conversionResult == String16::sourceIllegal)
-            return false;
-
-        if (conversionResult == String16::conversionOK) {
-            // Not all utf8 characters were consumed - failed parsing.
-            if (utf8start != utf8 + utf8count)
-                return false;
-
-            size_t utf16length = utf16start - utf16;
-            output->append(utf16, utf16length);
-            *utf8charEnd = start;
-            return true;
-        }
-
-        // Keep accumulating utf8 characters up to buffer length (6 should be enough).
-        if (utf8count >= 6)
-            return false;
-    }
-
-    return false;
-}
-
-template<typename CharType>
-bool decodeString(const CharType* start, const CharType* end, String16Builder* output)
+bool decodeString(const UChar* start, const UChar* end, String16Builder* output)
 {
     while (start < end) {
         UChar c = *start++;
         if ('\\' != c) {
             output->append(c);
             continue;
         }
         c = *start++;
 
         if (c == 'x') {
-            // Rewind "\x".
-            if (!decodeUTF8(start - 2, end, &start, output))
-                return false;
-            continue;
+            // \x is not supported.
+            return false;
         }
 
         switch (c) {
         case '"':
         case '/':
         case '\\':
             break;
         case 'b':
             c = '\b';
             break;
@@ skipping 20 matching lines @@
             start += 4;
             break;
         default:
             return false;
         }
         output->append(c);
     }
     return true;
 }
 
-template<typename CharType>
-bool decodeString(const CharType* start, const CharType* end, String16* output)
+bool decodeString(const UChar* start, const UChar* end, String16* output)
 {
     if (start == end) {
         *output = "";
         return true;
     }
     if (start > end)
         return false;
     String16Builder buffer;
     buffer.reserveCapacity(end - start);
     if (!decodeString(start, end, &buffer))
         return false;
     *output = buffer.toString();
-    // Validate constructed utf16 string.
-    if (!output->validateUTF8())
-        return false;
     return true;
 }
 
-template<typename CharType>
-PassOwnPtr<Value> buildValue(const CharType* start, const CharType* end, const CharType** valueTokenEnd, int depth)
+PassOwnPtr<Value> buildValue(const UChar* start, const UChar* end, const UChar** valueTokenEnd, int depth)
 {
     if (depth > stackLimit)
         return nullptr;
 
     OwnPtr<Value> result;
-    const CharType* tokenStart;
-    const CharType* tokenEnd;
+    const UChar* tokenStart;
+    const UChar* tokenEnd;
     Token token = parseToken(start, end, &tokenStart, &tokenEnd);
     switch (token) {
     case InvalidToken:
         return nullptr;
     case NullToken:
         result = Value::null();
         break;
     case BoolTrue:
         result = FundamentalValue::create(true);
         break;
@@ skipping 88 matching lines @@
 
     default:
         // We got a token that's not a value.
         return nullptr;
     }
 
     skipWhitespaceAndComments(tokenEnd, end, valueTokenEnd);
     return result.release();
 }
 
-template<typename CharType>
-PassOwnPtr<Value> parseJSONInternal(const CharType* start, unsigned length)
+PassOwnPtr<Value> parseJSONInternal(const UChar* start, unsigned length)
 {
-    const CharType* end = start + length;
-    const CharType *tokenEnd;
+    const UChar* end = start + length;
+    const UChar *tokenEnd;
     OwnPtr<Value> value = buildValue(start, end, &tokenEnd, 0);
     if (!value || tokenEnd != end)
         return nullptr;
     return value.release();
 }
 
 } // anonymous namespace
 
 PassOwnPtr<Value> parseJSON(const String16& json)
 {
     if (json.isEmpty())
         return nullptr;
-    if (json.is8Bit())
-        return parseJSONInternal(json.characters8(), json.length());
     return parseJSONInternal(json.characters16(), json.length());
 }
 
 } // namespace protocol
 } // namespace blink