v8/third_party/inspector_protocol/lib/Parser_cpp.template
lukasza 01a93925c1 Adapt StringBuilder's append and toString methods via StringUtil helper.
This is needed to insulate generated code from blink::protocol namespace
from naming changes that we plan to do in the Great Blink Rename (which
in particular will change wtf::StringBuilder::toString into ToString,
and similarily will rename reserveCapacity and append methods).

This CL also includes roll of inspector_protocol which starts to
generate code that uses the new methods of StringUtil adapter:
rolling third_party/inspector to 1a131872167f0f7653629326891aa3ec94417f27.

BUG=683447

Review-Url: https://codereview.chromium.org/2660503002
Cr-Commit-Position: refs/heads/master@{#42750}
2017-01-27 21:11:15 +00:00

554 lines
14 KiB
Plaintext

// 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.
{% for namespace in config.protocol.namespace %}
namespace {{namespace}} {
{% endfor %}
namespace {
const int stackLimit = 1000;
enum Token {
ObjectBegin,
ObjectEnd,
ArrayBegin,
ArrayEnd,
StringLiteral,
Number,
BoolTrue,
BoolFalse,
NullToken,
ListSeparator,
ObjectPairSeparator,
InvalidToken,
};
const char* const nullString = "null";
const char* const trueString = "true";
const char* const falseString = "false";
bool isASCII(uint16_t c)
{
return !(c & ~0x7F);
}
bool isSpaceOrNewLine(uint16_t c)
{
return isASCII(c) && c <= ' ' && (c == ' ' || (c <= 0xD && c >= 0x9));
}
double charactersToDouble(const uint16_t* characters, size_t length, bool* ok)
{
std::vector<char> buffer;
buffer.reserve(length + 1);
for (size_t i = 0; i < length; ++i) {
if (!isASCII(characters[i])) {
*ok = false;
return 0;
}
buffer.push_back(static_cast<char>(characters[i]));
}
buffer.push_back('\0');
char* endptr;
double result = std::strtod(buffer.data(), &endptr);
*ok = !(*endptr);
return result;
}
double charactersToDouble(const uint8_t* characters, size_t length, bool* ok)
{
std::string buffer(reinterpret_cast<const char*>(characters), length);
char* endptr;
double result = std::strtod(buffer.data(), &endptr);
*ok = !(*endptr);
return result;
}
template<typename Char>
bool parseConstToken(const Char* start, const Char* end, const Char** tokenEnd, const char* token)
{
while (start < end && *token != '\0' && *start++ == *token++) { }
if (*token != '\0')
return false;
*tokenEnd = start;
return true;
}
template<typename Char>
bool readInt(const Char* start, const Char* end, const Char** 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 Char>
bool parseNumberToken(const Char* start, const Char* end, const Char** 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;
Char c = *start;
if ('-' == c)
++start;
if (!readInt(start, end, &start, false))
return false;
if (start == end) {
*tokenEnd = start;
return true;
}
// Optional fraction part
c = *start;
if ('.' == c) {
++start;
if (!readInt(start, end, &start, true))
return false;
if (start == end) {
*tokenEnd = start;
return true;
}
c = *start;
}
// Optional exponent part
if ('e' == c || 'E' == c) {
++start;
if (start == end)
return false;
c = *start;
if ('-' == c || '+' == c) {
++start;
if (start == end)
return false;
}
if (!readInt(start, end, &start, true))
return false;
}
*tokenEnd = start;
return true;
}
template<typename Char>
bool readHexDigits(const Char* start, const Char* end, const Char** tokenEnd, int digits)
{
if (end - start < digits)
return false;
for (int i = 0; i < digits; ++i) {
Char c = *start++;
if (!(('0' <= c && c <= '9') || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')))
return false;
}
*tokenEnd = start;
return true;
}
template<typename Char>
bool parseStringToken(const Char* start, const Char* end, const Char** tokenEnd)
{
while (start < end) {
Char c = *start++;
if ('\\' == c) {
if (start == end)
return false;
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))
return false;
break;
case '\\':
case '/':
case 'b':
case 'f':
case 'n':
case 'r':
case 't':
case 'v':
case '"':
break;
default:
return false;
}
} else if ('"' == c) {
*tokenEnd = start;
return true;
}
}
return false;
}
template<typename Char>
bool skipComment(const Char* start, const Char* end, const Char** 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 == '*') {
Char 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 Char>
void skipWhitespaceAndComments(const Char* start, const Char* end, const Char** whitespaceEnd)
{
while (start < end) {
if (isSpaceOrNewLine(*start)) {
++start;
} else if (*start == '/') {
const Char* commentEnd;
if (!skipComment(start, end, &commentEnd))
break;
start = commentEnd;
} else {
break;
}
}
*whitespaceEnd = start;
}
template<typename Char>
Token parseToken(const Char* start, const Char* end, const Char** tokenStart, const Char** tokenEnd)
{
skipWhitespaceAndComments(start, end, tokenStart);
start = *tokenStart;
if (start == end)
return InvalidToken;
switch (*start) {
case 'n':
if (parseConstToken(start, end, tokenEnd, nullString))
return NullToken;
break;
case 't':
if (parseConstToken(start, end, tokenEnd, trueString))
return BoolTrue;
break;
case 'f':
if (parseConstToken(start, end, tokenEnd, falseString))
return BoolFalse;
break;
case '[':
*tokenEnd = start + 1;
return ArrayBegin;
case ']':
*tokenEnd = start + 1;
return ArrayEnd;
case ',':
*tokenEnd = start + 1;
return ListSeparator;
case '{':
*tokenEnd = start + 1;
return ObjectBegin;
case '}':
*tokenEnd = start + 1;
return ObjectEnd;
case ':':
*tokenEnd = start + 1;
return ObjectPairSeparator;
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
case '-':
if (parseNumberToken(start, end, tokenEnd))
return Number;
break;
case '"':
if (parseStringToken(start + 1, end, tokenEnd))
return StringLiteral;
break;
}
return InvalidToken;
}
template<typename Char>
int hexToInt(Char 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;
DCHECK(false);
return 0;
}
template<typename Char>
bool decodeString(const Char* start, const Char* end, StringBuilder* output)
{
while (start < end) {
uint16_t c = *start++;
if ('\\' != c) {
StringUtil::builderAppend(*output, c);
continue;
}
if (start == end)
return false;
c = *start++;
if (c == 'x') {
// \x is not supported.
return false;
}
switch (c) {
case '"':
case '/':
case '\\':
break;
case 'b':
c = '\b';
break;
case 'f':
c = '\f';
break;
case 'n':
c = '\n';
break;
case 'r':
c = '\r';
break;
case 't':
c = '\t';
break;
case 'v':
c = '\v';
break;
case 'u':
c = (hexToInt(*start) << 12) +
(hexToInt(*(start + 1)) << 8) +
(hexToInt(*(start + 2)) << 4) +
hexToInt(*(start + 3));
start += 4;
break;
default:
return false;
}
StringUtil::builderAppend(*output, c);
}
return true;
}
template<typename Char>
bool decodeString(const Char* start, const Char* end, String* output)
{
if (start == end) {
*output = "";
return true;
}
if (start > end)
return false;
StringBuilder buffer;
StringUtil::builderReserve(buffer, end - start);
if (!decodeString(start, end, &buffer))
return false;
*output = StringUtil::builderToString(buffer);
return true;
}
template<typename Char>
std::unique_ptr<Value> buildValue(const Char* start, const Char* end, const Char** valueTokenEnd, int depth)
{
if (depth > stackLimit)
return nullptr;
std::unique_ptr<Value> result;
const Char* tokenStart;
const Char* 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;
case BoolFalse:
result = FundamentalValue::create(false);
break;
case Number: {
bool ok;
double value = charactersToDouble(tokenStart, tokenEnd - tokenStart, &ok);
if (!ok)
return nullptr;
int number = static_cast<int>(value);
if (number == value)
result = FundamentalValue::create(number);
else
result = FundamentalValue::create(value);
break;
}
case StringLiteral: {
String value;
bool ok = decodeString(tokenStart + 1, tokenEnd - 1, &value);
if (!ok)
return nullptr;
result = StringValue::create(value);
break;
}
case ArrayBegin: {
std::unique_ptr<ListValue> array = ListValue::create();
start = tokenEnd;
token = parseToken(start, end, &tokenStart, &tokenEnd);
while (token != ArrayEnd) {
std::unique_ptr<Value> arrayNode = buildValue(start, end, &tokenEnd, depth + 1);
if (!arrayNode)
return nullptr;
array->pushValue(std::move(arrayNode));
// After a list value, we expect a comma or the end of the list.
start = tokenEnd;
token = parseToken(start, end, &tokenStart, &tokenEnd);
if (token == ListSeparator) {
start = tokenEnd;
token = parseToken(start, end, &tokenStart, &tokenEnd);
if (token == ArrayEnd)
return nullptr;
} else if (token != ArrayEnd) {
// Unexpected value after list value. Bail out.
return nullptr;
}
}
if (token != ArrayEnd)
return nullptr;
result = std::move(array);
break;
}
case ObjectBegin: {
std::unique_ptr<DictionaryValue> object = DictionaryValue::create();
start = tokenEnd;
token = parseToken(start, end, &tokenStart, &tokenEnd);
while (token != ObjectEnd) {
if (token != StringLiteral)
return nullptr;
String key;
if (!decodeString(tokenStart + 1, tokenEnd - 1, &key))
return nullptr;
start = tokenEnd;
token = parseToken(start, end, &tokenStart, &tokenEnd);
if (token != ObjectPairSeparator)
return nullptr;
start = tokenEnd;
std::unique_ptr<Value> value = buildValue(start, end, &tokenEnd, depth + 1);
if (!value)
return nullptr;
object->setValue(key, std::move(value));
start = tokenEnd;
// After a key/value pair, we expect a comma or the end of the
// object.
token = parseToken(start, end, &tokenStart, &tokenEnd);
if (token == ListSeparator) {
start = tokenEnd;
token = parseToken(start, end, &tokenStart, &tokenEnd);
if (token == ObjectEnd)
return nullptr;
} else if (token != ObjectEnd) {
// Unexpected value after last object value. Bail out.
return nullptr;
}
}
if (token != ObjectEnd)
return nullptr;
result = std::move(object);
break;
}
default:
// We got a token that's not a value.
return nullptr;
}
skipWhitespaceAndComments(tokenEnd, end, valueTokenEnd);
return result;
}
template<typename Char>
std::unique_ptr<Value> parseJSONInternal(const Char* start, unsigned length)
{
const Char* end = start + length;
const Char *tokenEnd;
std::unique_ptr<Value> value = buildValue(start, end, &tokenEnd, 0);
if (!value || tokenEnd != end)
return nullptr;
return value;
}
} // anonymous namespace
std::unique_ptr<Value> parseJSONCharacters(const uint16_t* characters, unsigned length)
{
return parseJSONInternal<uint16_t>(characters, length);
}
std::unique_ptr<Value> parseJSONCharacters(const uint8_t* characters, unsigned length)
{
return parseJSONInternal<uint8_t>(characters, length);
}
{% for namespace in config.protocol.namespace %}
} // namespace {{namespace}}
{% endfor %}