mirror of
https://github.com/Mauler125/r5sdk.git
synced 2025-02-09 19:15:03 +01:00
2729 lines
97 KiB
C++
2729 lines
97 KiB
C++
// Protocol Buffers - Google's data interchange format
|
|
// Copyright 2008 Google Inc. All rights reserved.
|
|
// https://developers.google.com/protocol-buffers/
|
|
//
|
|
// Redistribution and use in source and binary forms, with or without
|
|
// modification, are permitted provided that the following conditions are
|
|
// met:
|
|
//
|
|
// * Redistributions of source code must retain the above copyright
|
|
// notice, this list of conditions and the following disclaimer.
|
|
// * Redistributions in binary form must reproduce the above
|
|
// copyright notice, this list of conditions and the following disclaimer
|
|
// in the documentation and/or other materials provided with the
|
|
// distribution.
|
|
// * Neither the name of Google Inc. nor the names of its
|
|
// contributors may be used to endorse or promote products derived from
|
|
// this software without specific prior written permission.
|
|
//
|
|
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
|
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
|
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
|
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
|
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
|
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
|
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
|
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
// Author: jschorr@google.com (Joseph Schorr)
|
|
// Based on original Protocol Buffers design by
|
|
// Sanjay Ghemawat, Jeff Dean, and others.
|
|
|
|
#include <thirdparty/protobuf/text_format.h>
|
|
|
|
#include <float.h>
|
|
#include <stdio.h>
|
|
|
|
#include <algorithm>
|
|
#include <atomic>
|
|
#include <climits>
|
|
#include <cmath>
|
|
#include <limits>
|
|
#include <vector>
|
|
|
|
#include <thirdparty/protobuf/stubs/stringprintf.h>
|
|
#include <thirdparty/protobuf/any.h>
|
|
#include <thirdparty/protobuf/descriptor.pb.h>
|
|
#include <thirdparty/protobuf/io/coded_stream.h>
|
|
#include <thirdparty/protobuf/io/tokenizer.h>
|
|
#include <thirdparty/protobuf/io/zero_copy_stream.h>
|
|
#include <thirdparty/protobuf/io/zero_copy_stream_impl.h>
|
|
#include <thirdparty/protobuf/descriptor.h>
|
|
#include <thirdparty/protobuf/dynamic_message.h>
|
|
#include <thirdparty/protobuf/map_field.h>
|
|
#include <thirdparty/protobuf/message.h>
|
|
#include <thirdparty/protobuf/repeated_field.h>
|
|
#include <thirdparty/protobuf/unknown_field_set.h>
|
|
#include <thirdparty/protobuf/wire_format_lite.h>
|
|
#include <thirdparty/protobuf/stubs/strutil.h>
|
|
#include <thirdparty/protobuf/io/strtod.h>
|
|
#include <thirdparty/protobuf/stubs/map_util.h>
|
|
#include <thirdparty/protobuf/stubs/stl_util.h>
|
|
|
|
// Must be included last.
|
|
#include <thirdparty/protobuf/port_def.inc>
|
|
|
|
#define DEBUG_STRING_SILENT_MARKER "\t "
|
|
|
|
namespace google {
|
|
namespace protobuf {
|
|
|
|
namespace {
|
|
|
|
inline bool IsHexNumber(const std::string& str) {
|
|
return (str.length() >= 2 && str[0] == '0' &&
|
|
(str[1] == 'x' || str[1] == 'X'));
|
|
}
|
|
|
|
inline bool IsOctNumber(const std::string& str) {
|
|
return (str.length() >= 2 && str[0] == '0' &&
|
|
(str[1] >= '0' && str[1] < '8'));
|
|
}
|
|
|
|
} // namespace
|
|
|
|
namespace internal {
|
|
// Controls insertion of DEBUG_STRING_SILENT_MARKER.
|
|
PROTOBUF_EXPORT std::atomic<bool> enable_debug_text_format_marker;
|
|
} // namespace internal
|
|
|
|
std::string Message::DebugString() const {
|
|
std::string debug_string;
|
|
|
|
TextFormat::Printer printer;
|
|
printer.SetExpandAny(true);
|
|
printer.SetInsertSilentMarker(internal::enable_debug_text_format_marker.load(
|
|
std::memory_order_relaxed));
|
|
|
|
printer.PrintToString(*this, &debug_string);
|
|
|
|
return debug_string;
|
|
}
|
|
|
|
std::string Message::ShortDebugString() const {
|
|
std::string debug_string;
|
|
|
|
TextFormat::Printer printer;
|
|
printer.SetSingleLineMode(true);
|
|
printer.SetExpandAny(true);
|
|
printer.SetInsertSilentMarker(internal::enable_debug_text_format_marker.load(
|
|
std::memory_order_relaxed));
|
|
|
|
printer.PrintToString(*this, &debug_string);
|
|
// Single line mode currently might have an extra space at the end.
|
|
if (!debug_string.empty() && debug_string[debug_string.size() - 1] == ' ') {
|
|
debug_string.resize(debug_string.size() - 1);
|
|
}
|
|
|
|
return debug_string;
|
|
}
|
|
|
|
std::string Message::Utf8DebugString() const {
|
|
std::string debug_string;
|
|
|
|
TextFormat::Printer printer;
|
|
printer.SetUseUtf8StringEscaping(true);
|
|
printer.SetExpandAny(true);
|
|
printer.SetInsertSilentMarker(internal::enable_debug_text_format_marker.load(
|
|
std::memory_order_relaxed));
|
|
|
|
printer.PrintToString(*this, &debug_string);
|
|
|
|
return debug_string;
|
|
}
|
|
|
|
void Message::PrintDebugString() const { printf("%s", DebugString().c_str()); }
|
|
|
|
|
|
// ===========================================================================
|
|
// Implementation of the parse information tree class.
|
|
void TextFormat::ParseInfoTree::RecordLocation(
|
|
const FieldDescriptor* field, TextFormat::ParseLocationRange range) {
|
|
locations_[field].push_back(range);
|
|
}
|
|
|
|
TextFormat::ParseInfoTree* TextFormat::ParseInfoTree::CreateNested(
|
|
const FieldDescriptor* field) {
|
|
// Owned by us in the map.
|
|
auto& vec = nested_[field];
|
|
vec.emplace_back(new TextFormat::ParseInfoTree());
|
|
return vec.back().get();
|
|
}
|
|
|
|
void CheckFieldIndex(const FieldDescriptor* field, int index) {
|
|
if (field == nullptr) {
|
|
return;
|
|
}
|
|
|
|
if (field->is_repeated() && index == -1) {
|
|
GOOGLE_LOG(DFATAL) << "Index must be in range of repeated field values. "
|
|
<< "Field: " << field->name();
|
|
} else if (!field->is_repeated() && index != -1) {
|
|
GOOGLE_LOG(DFATAL) << "Index must be -1 for singular fields."
|
|
<< "Field: " << field->name();
|
|
}
|
|
}
|
|
|
|
TextFormat::ParseLocationRange TextFormat::ParseInfoTree::GetLocationRange(
|
|
const FieldDescriptor* field, int index) const {
|
|
CheckFieldIndex(field, index);
|
|
if (index == -1) {
|
|
index = 0;
|
|
}
|
|
|
|
const std::vector<TextFormat::ParseLocationRange>* locations =
|
|
FindOrNull(locations_, field);
|
|
if (locations == nullptr ||
|
|
index >= static_cast<int64_t>(locations->size())) {
|
|
return TextFormat::ParseLocationRange();
|
|
}
|
|
|
|
return (*locations)[index];
|
|
}
|
|
|
|
TextFormat::ParseInfoTree* TextFormat::ParseInfoTree::GetTreeForNested(
|
|
const FieldDescriptor* field, int index) const {
|
|
CheckFieldIndex(field, index);
|
|
if (index == -1) {
|
|
index = 0;
|
|
}
|
|
|
|
auto it = nested_.find(field);
|
|
if (it == nested_.end() || index >= static_cast<int64_t>(it->second.size())) {
|
|
return nullptr;
|
|
}
|
|
|
|
return it->second[index].get();
|
|
}
|
|
|
|
namespace {
|
|
// These functions implement the behavior of the "default" TextFormat::Finder,
|
|
// they are defined as standalone to be called when finder_ is nullptr.
|
|
const FieldDescriptor* DefaultFinderFindExtension(Message* message,
|
|
const std::string& name) {
|
|
const Descriptor* descriptor = message->GetDescriptor();
|
|
return descriptor->file()->pool()->FindExtensionByPrintableName(descriptor,
|
|
name);
|
|
}
|
|
|
|
const FieldDescriptor* DefaultFinderFindExtensionByNumber(
|
|
const Descriptor* descriptor, int number) {
|
|
return descriptor->file()->pool()->FindExtensionByNumber(descriptor, number);
|
|
}
|
|
|
|
const Descriptor* DefaultFinderFindAnyType(const Message& message,
|
|
const std::string& prefix,
|
|
const std::string& name) {
|
|
if (prefix != internal::kTypeGoogleApisComPrefix &&
|
|
prefix != internal::kTypeGoogleProdComPrefix) {
|
|
return nullptr;
|
|
}
|
|
return message.GetDescriptor()->file()->pool()->FindMessageTypeByName(name);
|
|
}
|
|
} // namespace
|
|
|
|
// ===========================================================================
|
|
// Internal class for parsing an ASCII representation of a Protocol Message.
|
|
// This class makes use of the Protocol Message compiler's tokenizer found
|
|
// in //net/proto2/io/public/tokenizer.h. Note that class's Parse
|
|
// method is *not* thread-safe and should only be used in a single thread at
|
|
// a time.
|
|
|
|
// Makes code slightly more readable. The meaning of "DO(foo)" is
|
|
// "Execute foo and fail if it fails.", where failure is indicated by
|
|
// returning false. Borrowed from parser.cc (Thanks Kenton!).
|
|
#define DO(STATEMENT) \
|
|
if (STATEMENT) { \
|
|
} else { \
|
|
return false; \
|
|
}
|
|
|
|
class TextFormat::Parser::ParserImpl {
|
|
public:
|
|
// Determines if repeated values for non-repeated fields and
|
|
// oneofs are permitted, e.g., the string "foo: 1 foo: 2" for a
|
|
// required/optional field named "foo", or "baz: 1 qux: 2"
|
|
// where "baz" and "qux" are members of the same oneof.
|
|
enum SingularOverwritePolicy {
|
|
ALLOW_SINGULAR_OVERWRITES = 0, // the last value is retained
|
|
FORBID_SINGULAR_OVERWRITES = 1, // an error is issued
|
|
};
|
|
|
|
ParserImpl(const Descriptor* root_message_type,
|
|
io::ZeroCopyInputStream* input_stream,
|
|
io::ErrorCollector* error_collector,
|
|
const TextFormat::Finder* finder, ParseInfoTree* parse_info_tree,
|
|
SingularOverwritePolicy singular_overwrite_policy,
|
|
bool allow_case_insensitive_field, bool allow_unknown_field,
|
|
bool allow_unknown_extension, bool allow_unknown_enum,
|
|
bool allow_field_number, bool allow_relaxed_whitespace,
|
|
bool allow_partial, int recursion_limit)
|
|
: error_collector_(error_collector),
|
|
finder_(finder),
|
|
parse_info_tree_(parse_info_tree),
|
|
tokenizer_error_collector_(this),
|
|
tokenizer_(input_stream, &tokenizer_error_collector_),
|
|
root_message_type_(root_message_type),
|
|
singular_overwrite_policy_(singular_overwrite_policy),
|
|
allow_case_insensitive_field_(allow_case_insensitive_field),
|
|
allow_unknown_field_(allow_unknown_field),
|
|
allow_unknown_extension_(allow_unknown_extension),
|
|
allow_unknown_enum_(allow_unknown_enum),
|
|
allow_field_number_(allow_field_number),
|
|
allow_partial_(allow_partial),
|
|
initial_recursion_limit_(recursion_limit),
|
|
recursion_limit_(recursion_limit),
|
|
had_errors_(false) {
|
|
// For backwards-compatibility with proto1, we need to allow the 'f' suffix
|
|
// for floats.
|
|
tokenizer_.set_allow_f_after_float(true);
|
|
|
|
// '#' starts a comment.
|
|
tokenizer_.set_comment_style(io::Tokenizer::SH_COMMENT_STYLE);
|
|
|
|
if (allow_relaxed_whitespace) {
|
|
tokenizer_.set_require_space_after_number(false);
|
|
tokenizer_.set_allow_multiline_strings(true);
|
|
}
|
|
|
|
// Consume the starting token.
|
|
tokenizer_.Next();
|
|
}
|
|
~ParserImpl() {}
|
|
|
|
// Parses the ASCII representation specified in input and saves the
|
|
// information into the output pointer (a Message). Returns
|
|
// false if an error occurs (an error will also be logged to
|
|
// GOOGLE_LOG(ERROR)).
|
|
bool Parse(Message* output) {
|
|
// Consume fields until we cannot do so anymore.
|
|
while (true) {
|
|
if (LookingAtType(io::Tokenizer::TYPE_END)) {
|
|
// Ensures recursion limit properly unwinded, but only for success
|
|
// cases. This implicitly avoids the check when `Parse` returns false
|
|
// via `DO(...)`.
|
|
GOOGLE_DCHECK(had_errors_ || recursion_limit_ == initial_recursion_limit_)
|
|
<< "Recursion limit at end of parse should be "
|
|
<< initial_recursion_limit_ << ", but was " << recursion_limit_
|
|
<< ". Difference of " << initial_recursion_limit_ - recursion_limit_
|
|
<< " stack frames not accounted for stack unwind.";
|
|
|
|
return !had_errors_;
|
|
}
|
|
|
|
DO(ConsumeField(output));
|
|
}
|
|
}
|
|
|
|
bool ParseField(const FieldDescriptor* field, Message* output) {
|
|
bool suc;
|
|
if (field->cpp_type() == FieldDescriptor::CPPTYPE_MESSAGE) {
|
|
suc = ConsumeFieldMessage(output, output->GetReflection(), field);
|
|
} else {
|
|
suc = ConsumeFieldValue(output, output->GetReflection(), field);
|
|
}
|
|
return suc && LookingAtType(io::Tokenizer::TYPE_END);
|
|
}
|
|
|
|
void ReportError(int line, int col, const std::string& message) {
|
|
had_errors_ = true;
|
|
if (error_collector_ == nullptr) {
|
|
if (line >= 0) {
|
|
GOOGLE_LOG(ERROR) << "Error parsing text-format "
|
|
<< root_message_type_->full_name() << ": " << (line + 1)
|
|
<< ":" << (col + 1) << ": " << message;
|
|
} else {
|
|
GOOGLE_LOG(ERROR) << "Error parsing text-format "
|
|
<< root_message_type_->full_name() << ": " << message;
|
|
}
|
|
} else {
|
|
error_collector_->AddError(line, col, message);
|
|
}
|
|
}
|
|
|
|
void ReportWarning(int line, int col, const std::string& message) {
|
|
if (error_collector_ == nullptr) {
|
|
if (line >= 0) {
|
|
GOOGLE_LOG(WARNING) << "Warning parsing text-format "
|
|
<< root_message_type_->full_name() << ": " << (line + 1)
|
|
<< ":" << (col + 1) << ": " << message;
|
|
} else {
|
|
GOOGLE_LOG(WARNING) << "Warning parsing text-format "
|
|
<< root_message_type_->full_name() << ": " << message;
|
|
}
|
|
} else {
|
|
error_collector_->AddWarning(line, col, message);
|
|
}
|
|
}
|
|
|
|
private:
|
|
static constexpr int32_t kint32max = std::numeric_limits<int32_t>::max();
|
|
static constexpr uint32_t kuint32max = std::numeric_limits<uint32_t>::max();
|
|
static constexpr int64_t kint64min = std::numeric_limits<int64_t>::min();
|
|
static constexpr int64_t kint64max = std::numeric_limits<int64_t>::max();
|
|
static constexpr uint64_t kuint64max = std::numeric_limits<uint64_t>::max();
|
|
|
|
GOOGLE_DISALLOW_EVIL_CONSTRUCTORS(ParserImpl);
|
|
|
|
// Reports an error with the given message with information indicating
|
|
// the position (as derived from the current token).
|
|
void ReportError(const std::string& message) {
|
|
ReportError(tokenizer_.current().line, tokenizer_.current().column,
|
|
message);
|
|
}
|
|
|
|
// Reports a warning with the given message with information indicating
|
|
// the position (as derived from the current token).
|
|
void ReportWarning(const std::string& message) {
|
|
ReportWarning(tokenizer_.current().line, tokenizer_.current().column,
|
|
message);
|
|
}
|
|
|
|
// Consumes the specified message with the given starting delimiter.
|
|
// This method checks to see that the end delimiter at the conclusion of
|
|
// the consumption matches the starting delimiter passed in here.
|
|
bool ConsumeMessage(Message* message, const std::string delimiter) {
|
|
while (!LookingAt(">") && !LookingAt("}")) {
|
|
DO(ConsumeField(message));
|
|
}
|
|
|
|
// Confirm that we have a valid ending delimiter.
|
|
DO(Consume(delimiter));
|
|
return true;
|
|
}
|
|
|
|
// Consume either "<" or "{".
|
|
bool ConsumeMessageDelimiter(std::string* delimiter) {
|
|
if (TryConsume("<")) {
|
|
*delimiter = ">";
|
|
} else {
|
|
DO(Consume("{"));
|
|
*delimiter = "}";
|
|
}
|
|
return true;
|
|
}
|
|
|
|
|
|
// Consumes the current field (as returned by the tokenizer) on the
|
|
// passed in message.
|
|
bool ConsumeField(Message* message) {
|
|
const Reflection* reflection = message->GetReflection();
|
|
const Descriptor* descriptor = message->GetDescriptor();
|
|
|
|
std::string field_name;
|
|
bool reserved_field = false;
|
|
const FieldDescriptor* field = nullptr;
|
|
int start_line = tokenizer_.current().line;
|
|
int start_column = tokenizer_.current().column;
|
|
|
|
const FieldDescriptor* any_type_url_field;
|
|
const FieldDescriptor* any_value_field;
|
|
if (internal::GetAnyFieldDescriptors(*message, &any_type_url_field,
|
|
&any_value_field) &&
|
|
TryConsume("[")) {
|
|
std::string full_type_name, prefix;
|
|
DO(ConsumeAnyTypeUrl(&full_type_name, &prefix));
|
|
std::string prefix_and_full_type_name =
|
|
StrCat(prefix, full_type_name);
|
|
DO(ConsumeBeforeWhitespace("]"));
|
|
TryConsumeWhitespace(prefix_and_full_type_name, "Any");
|
|
// ':' is optional between message labels and values.
|
|
TryConsumeBeforeWhitespace(":");
|
|
TryConsumeWhitespace(prefix_and_full_type_name, "Any");
|
|
std::string serialized_value;
|
|
const Descriptor* value_descriptor =
|
|
finder_ ? finder_->FindAnyType(*message, prefix, full_type_name)
|
|
: DefaultFinderFindAnyType(*message, prefix, full_type_name);
|
|
if (value_descriptor == nullptr) {
|
|
ReportError("Could not find type \"" + prefix_and_full_type_name +
|
|
"\" stored in google.protobuf.Any.");
|
|
return false;
|
|
}
|
|
DO(ConsumeAnyValue(value_descriptor, &serialized_value));
|
|
if (singular_overwrite_policy_ == FORBID_SINGULAR_OVERWRITES) {
|
|
// Fail if any_type_url_field has already been specified.
|
|
if ((!any_type_url_field->is_repeated() &&
|
|
reflection->HasField(*message, any_type_url_field)) ||
|
|
(!any_value_field->is_repeated() &&
|
|
reflection->HasField(*message, any_value_field))) {
|
|
ReportError("Non-repeated Any specified multiple times.");
|
|
return false;
|
|
}
|
|
}
|
|
reflection->SetString(message, any_type_url_field,
|
|
prefix_and_full_type_name);
|
|
reflection->SetString(message, any_value_field, serialized_value);
|
|
return true;
|
|
}
|
|
if (TryConsume("[")) {
|
|
// Extension.
|
|
DO(ConsumeFullTypeName(&field_name));
|
|
DO(ConsumeBeforeWhitespace("]"));
|
|
TryConsumeWhitespace(message->GetTypeName(), "Extension");
|
|
|
|
field = finder_ ? finder_->FindExtension(message, field_name)
|
|
: DefaultFinderFindExtension(message, field_name);
|
|
|
|
if (field == nullptr) {
|
|
if (!allow_unknown_field_ && !allow_unknown_extension_) {
|
|
ReportError("Extension \"" + field_name +
|
|
"\" is not defined or "
|
|
"is not an extension of \"" +
|
|
descriptor->full_name() + "\".");
|
|
return false;
|
|
} else {
|
|
ReportWarning("Ignoring extension \"" + field_name +
|
|
"\" which is not defined or is not an extension of \"" +
|
|
descriptor->full_name() + "\".");
|
|
}
|
|
}
|
|
} else {
|
|
DO(ConsumeIdentifierBeforeWhitespace(&field_name));
|
|
TryConsumeWhitespace(message->GetTypeName(), "Normal");
|
|
|
|
int32_t field_number;
|
|
if (allow_field_number_ && safe_strto32(field_name, &field_number)) {
|
|
if (descriptor->IsExtensionNumber(field_number)) {
|
|
field = finder_
|
|
? finder_->FindExtensionByNumber(descriptor, field_number)
|
|
: DefaultFinderFindExtensionByNumber(descriptor,
|
|
field_number);
|
|
} else if (descriptor->IsReservedNumber(field_number)) {
|
|
reserved_field = true;
|
|
} else {
|
|
field = descriptor->FindFieldByNumber(field_number);
|
|
}
|
|
} else {
|
|
field = descriptor->FindFieldByName(field_name);
|
|
// Group names are expected to be capitalized as they appear in the
|
|
// .proto file, which actually matches their type names, not their
|
|
// field names.
|
|
if (field == nullptr) {
|
|
std::string lower_field_name = field_name;
|
|
LowerString(&lower_field_name);
|
|
field = descriptor->FindFieldByName(lower_field_name);
|
|
// If the case-insensitive match worked but the field is NOT a group,
|
|
if (field != nullptr &&
|
|
field->type() != FieldDescriptor::TYPE_GROUP) {
|
|
field = nullptr;
|
|
}
|
|
}
|
|
// Again, special-case group names as described above.
|
|
if (field != nullptr && field->type() == FieldDescriptor::TYPE_GROUP &&
|
|
field->message_type()->name() != field_name) {
|
|
field = nullptr;
|
|
}
|
|
|
|
if (field == nullptr && allow_case_insensitive_field_) {
|
|
std::string lower_field_name = field_name;
|
|
LowerString(&lower_field_name);
|
|
field = descriptor->FindFieldByLowercaseName(lower_field_name);
|
|
}
|
|
|
|
if (field == nullptr) {
|
|
reserved_field = descriptor->IsReservedName(field_name);
|
|
}
|
|
}
|
|
|
|
if (field == nullptr && !reserved_field) {
|
|
if (!allow_unknown_field_) {
|
|
ReportError("Message type \"" + descriptor->full_name() +
|
|
"\" has no field named \"" + field_name + "\".");
|
|
return false;
|
|
} else {
|
|
ReportWarning("Message type \"" + descriptor->full_name() +
|
|
"\" has no field named \"" + field_name + "\".");
|
|
}
|
|
}
|
|
}
|
|
|
|
// Skips unknown or reserved fields.
|
|
if (field == nullptr) {
|
|
GOOGLE_CHECK(allow_unknown_field_ || allow_unknown_extension_ || reserved_field);
|
|
|
|
// Try to guess the type of this field.
|
|
// If this field is not a message, there should be a ":" between the
|
|
// field name and the field value and also the field value should not
|
|
// start with "{" or "<" which indicates the beginning of a message body.
|
|
// If there is no ":" or there is a "{" or "<" after ":", this field has
|
|
// to be a message or the input is ill-formed.
|
|
if (TryConsumeBeforeWhitespace(":")) {
|
|
TryConsumeWhitespace(message->GetTypeName(), "Unknown/Reserved");
|
|
if (!LookingAt("{") && !LookingAt("<")) {
|
|
return SkipFieldValue();
|
|
}
|
|
}
|
|
return SkipFieldMessage();
|
|
}
|
|
|
|
if (singular_overwrite_policy_ == FORBID_SINGULAR_OVERWRITES) {
|
|
// Fail if the field is not repeated and it has already been specified.
|
|
if (!field->is_repeated() && reflection->HasField(*message, field)) {
|
|
ReportError("Non-repeated field \"" + field_name +
|
|
"\" is specified multiple times.");
|
|
return false;
|
|
}
|
|
// Fail if the field is a member of a oneof and another member has already
|
|
// been specified.
|
|
const OneofDescriptor* oneof = field->containing_oneof();
|
|
if (oneof != nullptr && reflection->HasOneof(*message, oneof)) {
|
|
const FieldDescriptor* other_field =
|
|
reflection->GetOneofFieldDescriptor(*message, oneof);
|
|
ReportError("Field \"" + field_name +
|
|
"\" is specified along with "
|
|
"field \"" +
|
|
other_field->name() +
|
|
"\", another member "
|
|
"of oneof \"" +
|
|
oneof->name() + "\".");
|
|
return false;
|
|
}
|
|
}
|
|
|
|
// Perform special handling for embedded message types.
|
|
if (field->cpp_type() == FieldDescriptor::CPPTYPE_MESSAGE) {
|
|
// ':' is optional here.
|
|
bool consumed_semicolon = TryConsumeBeforeWhitespace(":");
|
|
TryConsumeWhitespace(message->GetTypeName(), "Normal");
|
|
if (consumed_semicolon && field->options().weak() &&
|
|
LookingAtType(io::Tokenizer::TYPE_STRING)) {
|
|
// we are getting a bytes string for a weak field.
|
|
std::string tmp;
|
|
DO(ConsumeString(&tmp));
|
|
MessageFactory* factory =
|
|
finder_ ? finder_->FindExtensionFactory(field) : nullptr;
|
|
reflection->MutableMessage(message, field, factory)
|
|
->ParseFromString(tmp);
|
|
goto label_skip_parsing;
|
|
}
|
|
} else {
|
|
// ':' is required here.
|
|
DO(ConsumeBeforeWhitespace(":"));
|
|
TryConsumeWhitespace(message->GetTypeName(), "Normal");
|
|
}
|
|
|
|
if (field->is_repeated() && TryConsume("[")) {
|
|
// Short repeated format, e.g. "foo: [1, 2, 3]".
|
|
if (!TryConsume("]")) {
|
|
// "foo: []" is treated as empty.
|
|
while (true) {
|
|
if (field->cpp_type() == FieldDescriptor::CPPTYPE_MESSAGE) {
|
|
// Perform special handling for embedded message types.
|
|
DO(ConsumeFieldMessage(message, reflection, field));
|
|
} else {
|
|
DO(ConsumeFieldValue(message, reflection, field));
|
|
}
|
|
if (TryConsume("]")) {
|
|
break;
|
|
}
|
|
DO(Consume(","));
|
|
}
|
|
}
|
|
} else if (field->cpp_type() == FieldDescriptor::CPPTYPE_MESSAGE) {
|
|
DO(ConsumeFieldMessage(message, reflection, field));
|
|
} else {
|
|
DO(ConsumeFieldValue(message, reflection, field));
|
|
}
|
|
label_skip_parsing:
|
|
// For historical reasons, fields may optionally be separated by commas or
|
|
// semicolons.
|
|
TryConsume(";") || TryConsume(",");
|
|
|
|
if (field->options().deprecated()) {
|
|
ReportWarning("text format contains deprecated field \"" + field_name +
|
|
"\"");
|
|
}
|
|
|
|
// If a parse info tree exists, add the location for the parsed
|
|
// field.
|
|
if (parse_info_tree_ != nullptr) {
|
|
int end_line = tokenizer_.previous().line;
|
|
int end_column = tokenizer_.previous().end_column;
|
|
|
|
RecordLocation(parse_info_tree_, field,
|
|
ParseLocationRange(ParseLocation(start_line, start_column),
|
|
ParseLocation(end_line, end_column)));
|
|
}
|
|
|
|
return true;
|
|
}
|
|
|
|
// Skips the next field including the field's name and value.
|
|
bool SkipField() {
|
|
if (TryConsume("[")) {
|
|
// Extension name or type URL.
|
|
DO(ConsumeTypeUrlOrFullTypeName());
|
|
DO(ConsumeBeforeWhitespace("]"));
|
|
} else {
|
|
std::string field_name;
|
|
DO(ConsumeIdentifierBeforeWhitespace(&field_name));
|
|
}
|
|
TryConsumeWhitespace("Unknown/Reserved", "n/a");
|
|
|
|
// Try to guess the type of this field.
|
|
// If this field is not a message, there should be a ":" between the
|
|
// field name and the field value and also the field value should not
|
|
// start with "{" or "<" which indicates the beginning of a message body.
|
|
// If there is no ":" or there is a "{" or "<" after ":", this field has
|
|
// to be a message or the input is ill-formed.
|
|
if (TryConsumeBeforeWhitespace(":")) {
|
|
TryConsumeWhitespace("Unknown/Reserved", "n/a");
|
|
if (!LookingAt("{") && !LookingAt("<")) {
|
|
DO(SkipFieldValue());
|
|
} else {
|
|
DO(SkipFieldMessage());
|
|
}
|
|
} else {
|
|
DO(SkipFieldMessage());
|
|
}
|
|
// For historical reasons, fields may optionally be separated by commas or
|
|
// semicolons.
|
|
TryConsume(";") || TryConsume(",");
|
|
return true;
|
|
}
|
|
|
|
bool ConsumeFieldMessage(Message* message, const Reflection* reflection,
|
|
const FieldDescriptor* field) {
|
|
if (--recursion_limit_ < 0) {
|
|
ReportError(
|
|
StrCat("Message is too deep, the parser exceeded the "
|
|
"configured recursion limit of ",
|
|
initial_recursion_limit_, "."));
|
|
return false;
|
|
}
|
|
// If the parse information tree is not nullptr, create a nested one
|
|
// for the nested message.
|
|
ParseInfoTree* parent = parse_info_tree_;
|
|
if (parent != nullptr) {
|
|
parse_info_tree_ = CreateNested(parent, field);
|
|
}
|
|
|
|
std::string delimiter;
|
|
DO(ConsumeMessageDelimiter(&delimiter));
|
|
MessageFactory* factory =
|
|
finder_ ? finder_->FindExtensionFactory(field) : nullptr;
|
|
if (field->is_repeated()) {
|
|
DO(ConsumeMessage(reflection->AddMessage(message, field, factory),
|
|
delimiter));
|
|
} else {
|
|
DO(ConsumeMessage(reflection->MutableMessage(message, field, factory),
|
|
delimiter));
|
|
}
|
|
|
|
++recursion_limit_;
|
|
|
|
// Reset the parse information tree.
|
|
parse_info_tree_ = parent;
|
|
return true;
|
|
}
|
|
|
|
// Skips the whole body of a message including the beginning delimiter and
|
|
// the ending delimiter.
|
|
bool SkipFieldMessage() {
|
|
if (--recursion_limit_ < 0) {
|
|
ReportError(
|
|
StrCat("Message is too deep, the parser exceeded the "
|
|
"configured recursion limit of ",
|
|
initial_recursion_limit_, "."));
|
|
return false;
|
|
}
|
|
|
|
std::string delimiter;
|
|
DO(ConsumeMessageDelimiter(&delimiter));
|
|
while (!LookingAt(">") && !LookingAt("}")) {
|
|
DO(SkipField());
|
|
}
|
|
DO(Consume(delimiter));
|
|
|
|
++recursion_limit_;
|
|
return true;
|
|
}
|
|
|
|
bool ConsumeFieldValue(Message* message, const Reflection* reflection,
|
|
const FieldDescriptor* field) {
|
|
// Define an easy to use macro for setting fields. This macro checks
|
|
// to see if the field is repeated (in which case we need to use the Add
|
|
// methods or not (in which case we need to use the Set methods).
|
|
#define SET_FIELD(CPPTYPE, VALUE) \
|
|
if (field->is_repeated()) { \
|
|
reflection->Add##CPPTYPE(message, field, VALUE); \
|
|
} else { \
|
|
reflection->Set##CPPTYPE(message, field, VALUE); \
|
|
}
|
|
|
|
switch (field->cpp_type()) {
|
|
case FieldDescriptor::CPPTYPE_INT32: {
|
|
int64_t value;
|
|
DO(ConsumeSignedInteger(&value, kint32max));
|
|
SET_FIELD(Int32, static_cast<int32_t>(value));
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_UINT32: {
|
|
uint64_t value;
|
|
DO(ConsumeUnsignedInteger(&value, kuint32max));
|
|
SET_FIELD(UInt32, static_cast<uint32_t>(value));
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_INT64: {
|
|
int64_t value;
|
|
DO(ConsumeSignedInteger(&value, kint64max));
|
|
SET_FIELD(Int64, value);
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_UINT64: {
|
|
uint64_t value;
|
|
DO(ConsumeUnsignedInteger(&value, kuint64max));
|
|
SET_FIELD(UInt64, value);
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_FLOAT: {
|
|
double value;
|
|
DO(ConsumeDouble(&value));
|
|
SET_FIELD(Float, io::SafeDoubleToFloat(value));
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_DOUBLE: {
|
|
double value;
|
|
DO(ConsumeDouble(&value));
|
|
SET_FIELD(Double, value);
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_STRING: {
|
|
std::string value;
|
|
DO(ConsumeString(&value));
|
|
SET_FIELD(String, value);
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_BOOL: {
|
|
if (LookingAtType(io::Tokenizer::TYPE_INTEGER)) {
|
|
uint64_t value;
|
|
DO(ConsumeUnsignedInteger(&value, 1));
|
|
SET_FIELD(Bool, value);
|
|
} else {
|
|
std::string value;
|
|
DO(ConsumeIdentifier(&value));
|
|
if (value == "true" || value == "True" || value == "t") {
|
|
SET_FIELD(Bool, true);
|
|
} else if (value == "false" || value == "False" || value == "f") {
|
|
SET_FIELD(Bool, false);
|
|
} else {
|
|
ReportError("Invalid value for boolean field \"" + field->name() +
|
|
"\". Value: \"" + value + "\".");
|
|
return false;
|
|
}
|
|
}
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_ENUM: {
|
|
std::string value;
|
|
int64_t int_value = kint64max;
|
|
const EnumDescriptor* enum_type = field->enum_type();
|
|
const EnumValueDescriptor* enum_value = nullptr;
|
|
|
|
if (LookingAtType(io::Tokenizer::TYPE_IDENTIFIER)) {
|
|
DO(ConsumeIdentifier(&value));
|
|
// Find the enumeration value.
|
|
enum_value = enum_type->FindValueByName(value);
|
|
|
|
} else if (LookingAt("-") ||
|
|
LookingAtType(io::Tokenizer::TYPE_INTEGER)) {
|
|
DO(ConsumeSignedInteger(&int_value, kint32max));
|
|
value = StrCat(int_value); // for error reporting
|
|
enum_value = enum_type->FindValueByNumber(int_value);
|
|
} else {
|
|
ReportError("Expected integer or identifier, got: " +
|
|
tokenizer_.current().text);
|
|
return false;
|
|
}
|
|
|
|
if (enum_value == nullptr) {
|
|
if (int_value != kint64max &&
|
|
reflection->SupportsUnknownEnumValues()) {
|
|
SET_FIELD(EnumValue, int_value);
|
|
return true;
|
|
} else if (!allow_unknown_enum_) {
|
|
ReportError("Unknown enumeration value of \"" + value +
|
|
"\" for "
|
|
"field \"" +
|
|
field->name() + "\".");
|
|
return false;
|
|
} else {
|
|
ReportWarning("Unknown enumeration value of \"" + value +
|
|
"\" for "
|
|
"field \"" +
|
|
field->name() + "\".");
|
|
return true;
|
|
}
|
|
}
|
|
|
|
SET_FIELD(Enum, enum_value);
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_MESSAGE: {
|
|
// We should never get here. Put here instead of a default
|
|
// so that if new types are added, we get a nice compiler warning.
|
|
GOOGLE_LOG(FATAL) << "Reached an unintended state: CPPTYPE_MESSAGE";
|
|
break;
|
|
}
|
|
}
|
|
#undef SET_FIELD
|
|
return true;
|
|
}
|
|
|
|
bool SkipFieldValue() {
|
|
if (--recursion_limit_ < 0) {
|
|
ReportError(
|
|
StrCat("Message is too deep, the parser exceeded the "
|
|
"configured recursion limit of ",
|
|
initial_recursion_limit_, "."));
|
|
return false;
|
|
}
|
|
|
|
if (LookingAtType(io::Tokenizer::TYPE_STRING)) {
|
|
while (LookingAtType(io::Tokenizer::TYPE_STRING)) {
|
|
tokenizer_.Next();
|
|
}
|
|
++recursion_limit_;
|
|
return true;
|
|
}
|
|
if (TryConsume("[")) {
|
|
while (true) {
|
|
if (!LookingAt("{") && !LookingAt("<")) {
|
|
DO(SkipFieldValue());
|
|
} else {
|
|
DO(SkipFieldMessage());
|
|
}
|
|
if (TryConsume("]")) {
|
|
break;
|
|
}
|
|
DO(Consume(","));
|
|
}
|
|
++recursion_limit_;
|
|
return true;
|
|
}
|
|
// Possible field values other than string:
|
|
// 12345 => TYPE_INTEGER
|
|
// -12345 => TYPE_SYMBOL + TYPE_INTEGER
|
|
// 1.2345 => TYPE_FLOAT
|
|
// -1.2345 => TYPE_SYMBOL + TYPE_FLOAT
|
|
// inf => TYPE_IDENTIFIER
|
|
// -inf => TYPE_SYMBOL + TYPE_IDENTIFIER
|
|
// TYPE_INTEGER => TYPE_IDENTIFIER
|
|
// Divides them into two group, one with TYPE_SYMBOL
|
|
// and the other without:
|
|
// Group one:
|
|
// 12345 => TYPE_INTEGER
|
|
// 1.2345 => TYPE_FLOAT
|
|
// inf => TYPE_IDENTIFIER
|
|
// TYPE_INTEGER => TYPE_IDENTIFIER
|
|
// Group two:
|
|
// -12345 => TYPE_SYMBOL + TYPE_INTEGER
|
|
// -1.2345 => TYPE_SYMBOL + TYPE_FLOAT
|
|
// -inf => TYPE_SYMBOL + TYPE_IDENTIFIER
|
|
// As we can see, the field value consists of an optional '-' and one of
|
|
// TYPE_INTEGER, TYPE_FLOAT and TYPE_IDENTIFIER.
|
|
bool has_minus = TryConsume("-");
|
|
if (!LookingAtType(io::Tokenizer::TYPE_INTEGER) &&
|
|
!LookingAtType(io::Tokenizer::TYPE_FLOAT) &&
|
|
!LookingAtType(io::Tokenizer::TYPE_IDENTIFIER)) {
|
|
std::string text = tokenizer_.current().text;
|
|
ReportError("Cannot skip field value, unexpected token: " + text);
|
|
++recursion_limit_;
|
|
return false;
|
|
}
|
|
// Combination of '-' and TYPE_IDENTIFIER may result in an invalid field
|
|
// value while other combinations all generate valid values.
|
|
// We check if the value of this combination is valid here.
|
|
// TYPE_IDENTIFIER after a '-' should be one of the float values listed
|
|
// below:
|
|
// inf, inff, infinity, nan
|
|
if (has_minus && LookingAtType(io::Tokenizer::TYPE_IDENTIFIER)) {
|
|
std::string text = tokenizer_.current().text;
|
|
LowerString(&text);
|
|
if (text != "inf" &&
|
|
text != "infinity" && text != "nan") {
|
|
ReportError("Invalid float number: " + text);
|
|
++recursion_limit_;
|
|
return false;
|
|
}
|
|
}
|
|
tokenizer_.Next();
|
|
++recursion_limit_;
|
|
return true;
|
|
}
|
|
|
|
// Returns true if the current token's text is equal to that specified.
|
|
bool LookingAt(const std::string& text) {
|
|
return tokenizer_.current().text == text;
|
|
}
|
|
|
|
// Returns true if the current token's type is equal to that specified.
|
|
bool LookingAtType(io::Tokenizer::TokenType token_type) {
|
|
return tokenizer_.current().type == token_type;
|
|
}
|
|
|
|
// Consumes an identifier and saves its value in the identifier parameter.
|
|
// Returns false if the token is not of type IDENTFIER.
|
|
bool ConsumeIdentifier(std::string* identifier) {
|
|
if (LookingAtType(io::Tokenizer::TYPE_IDENTIFIER)) {
|
|
*identifier = tokenizer_.current().text;
|
|
tokenizer_.Next();
|
|
return true;
|
|
}
|
|
|
|
// If allow_field_numer_ or allow_unknown_field_ is true, we should able
|
|
// to parse integer identifiers.
|
|
if ((allow_field_number_ || allow_unknown_field_ ||
|
|
allow_unknown_extension_) &&
|
|
LookingAtType(io::Tokenizer::TYPE_INTEGER)) {
|
|
*identifier = tokenizer_.current().text;
|
|
tokenizer_.Next();
|
|
return true;
|
|
}
|
|
|
|
ReportError("Expected identifier, got: " + tokenizer_.current().text);
|
|
return false;
|
|
}
|
|
|
|
// Similar to `ConsumeIdentifier`, but any following whitespace token may
|
|
// be reported.
|
|
bool ConsumeIdentifierBeforeWhitespace(std::string* identifier) {
|
|
tokenizer_.set_report_whitespace(true);
|
|
bool result = ConsumeIdentifier(identifier);
|
|
tokenizer_.set_report_whitespace(false);
|
|
return result;
|
|
}
|
|
|
|
// Consume a string of form "<id1>.<id2>....<idN>".
|
|
bool ConsumeFullTypeName(std::string* name) {
|
|
DO(ConsumeIdentifier(name));
|
|
while (TryConsume(".")) {
|
|
std::string part;
|
|
DO(ConsumeIdentifier(&part));
|
|
*name += ".";
|
|
*name += part;
|
|
}
|
|
return true;
|
|
}
|
|
|
|
bool ConsumeTypeUrlOrFullTypeName() {
|
|
std::string discarded;
|
|
DO(ConsumeIdentifier(&discarded));
|
|
while (TryConsume(".") || TryConsume("/")) {
|
|
DO(ConsumeIdentifier(&discarded));
|
|
}
|
|
return true;
|
|
}
|
|
|
|
// Consumes a string and saves its value in the text parameter.
|
|
// Returns false if the token is not of type STRING.
|
|
bool ConsumeString(std::string* text) {
|
|
if (!LookingAtType(io::Tokenizer::TYPE_STRING)) {
|
|
ReportError("Expected string, got: " + tokenizer_.current().text);
|
|
return false;
|
|
}
|
|
|
|
text->clear();
|
|
while (LookingAtType(io::Tokenizer::TYPE_STRING)) {
|
|
io::Tokenizer::ParseStringAppend(tokenizer_.current().text, text);
|
|
|
|
tokenizer_.Next();
|
|
}
|
|
|
|
return true;
|
|
}
|
|
|
|
// Consumes a uint64_t and saves its value in the value parameter.
|
|
// Returns false if the token is not of type INTEGER.
|
|
bool ConsumeUnsignedInteger(uint64_t* value, uint64_t max_value) {
|
|
if (!LookingAtType(io::Tokenizer::TYPE_INTEGER)) {
|
|
ReportError("Expected integer, got: " + tokenizer_.current().text);
|
|
return false;
|
|
}
|
|
|
|
if (!io::Tokenizer::ParseInteger(tokenizer_.current().text, max_value,
|
|
value)) {
|
|
ReportError("Integer out of range (" + tokenizer_.current().text + ")");
|
|
return false;
|
|
}
|
|
|
|
tokenizer_.Next();
|
|
return true;
|
|
}
|
|
|
|
// Consumes an int64_t and saves its value in the value parameter.
|
|
// Note that since the tokenizer does not support negative numbers,
|
|
// we actually may consume an additional token (for the minus sign) in this
|
|
// method. Returns false if the token is not an integer
|
|
// (signed or otherwise).
|
|
bool ConsumeSignedInteger(int64_t* value, uint64_t max_value) {
|
|
bool negative = false;
|
|
|
|
if (TryConsume("-")) {
|
|
negative = true;
|
|
// Two's complement always allows one more negative integer than
|
|
// positive.
|
|
++max_value;
|
|
}
|
|
|
|
uint64_t unsigned_value;
|
|
|
|
DO(ConsumeUnsignedInteger(&unsigned_value, max_value));
|
|
|
|
if (negative) {
|
|
if ((static_cast<uint64_t>(kint64max) + 1) == unsigned_value) {
|
|
*value = kint64min;
|
|
} else {
|
|
*value = -static_cast<int64_t>(unsigned_value);
|
|
}
|
|
} else {
|
|
*value = static_cast<int64_t>(unsigned_value);
|
|
}
|
|
|
|
return true;
|
|
}
|
|
|
|
// Consumes a double and saves its value in the value parameter.
|
|
// Accepts decimal numbers only, rejects hex or oct numbers.
|
|
bool ConsumeUnsignedDecimalAsDouble(double* value, uint64_t max_value) {
|
|
if (!LookingAtType(io::Tokenizer::TYPE_INTEGER)) {
|
|
ReportError("Expected integer, got: " + tokenizer_.current().text);
|
|
return false;
|
|
}
|
|
|
|
const std::string& text = tokenizer_.current().text;
|
|
if (IsHexNumber(text) || IsOctNumber(text)) {
|
|
ReportError("Expect a decimal number, got: " + text);
|
|
return false;
|
|
}
|
|
|
|
uint64_t uint64_value;
|
|
if (io::Tokenizer::ParseInteger(text, max_value, &uint64_value)) {
|
|
*value = static_cast<double>(uint64_value);
|
|
} else {
|
|
// Uint64 overflow, attempt to parse as a double instead.
|
|
*value = io::Tokenizer::ParseFloat(text);
|
|
}
|
|
|
|
tokenizer_.Next();
|
|
return true;
|
|
}
|
|
|
|
// Consumes a double and saves its value in the value parameter.
|
|
// Note that since the tokenizer does not support negative numbers,
|
|
// we actually may consume an additional token (for the minus sign) in this
|
|
// method. Returns false if the token is not a double
|
|
// (signed or otherwise).
|
|
bool ConsumeDouble(double* value) {
|
|
bool negative = false;
|
|
|
|
if (TryConsume("-")) {
|
|
negative = true;
|
|
}
|
|
|
|
// A double can actually be an integer, according to the tokenizer.
|
|
// Therefore, we must check both cases here.
|
|
if (LookingAtType(io::Tokenizer::TYPE_INTEGER)) {
|
|
// We have found an integer value for the double.
|
|
DO(ConsumeUnsignedDecimalAsDouble(value, kuint64max));
|
|
} else if (LookingAtType(io::Tokenizer::TYPE_FLOAT)) {
|
|
// We have found a float value for the double.
|
|
*value = io::Tokenizer::ParseFloat(tokenizer_.current().text);
|
|
|
|
// Mark the current token as consumed.
|
|
tokenizer_.Next();
|
|
} else if (LookingAtType(io::Tokenizer::TYPE_IDENTIFIER)) {
|
|
std::string text = tokenizer_.current().text;
|
|
LowerString(&text);
|
|
if (text == "inf" ||
|
|
text == "infinity") {
|
|
*value = std::numeric_limits<double>::infinity();
|
|
tokenizer_.Next();
|
|
} else if (text == "nan") {
|
|
*value = std::numeric_limits<double>::quiet_NaN();
|
|
tokenizer_.Next();
|
|
} else {
|
|
ReportError("Expected double, got: " + text);
|
|
return false;
|
|
}
|
|
} else {
|
|
ReportError("Expected double, got: " + tokenizer_.current().text);
|
|
return false;
|
|
}
|
|
|
|
if (negative) {
|
|
*value = -*value;
|
|
}
|
|
|
|
return true;
|
|
}
|
|
|
|
// Consumes Any::type_url value, of form "type.googleapis.com/full.type.Name"
|
|
// or "type.googleprod.com/full.type.Name"
|
|
bool ConsumeAnyTypeUrl(std::string* full_type_name, std::string* prefix) {
|
|
// TODO(saito) Extend Consume() to consume multiple tokens at once, so that
|
|
// this code can be written as just DO(Consume(kGoogleApisTypePrefix)).
|
|
DO(ConsumeIdentifier(prefix));
|
|
while (TryConsume(".")) {
|
|
std::string url;
|
|
DO(ConsumeIdentifier(&url));
|
|
*prefix += "." + url;
|
|
}
|
|
DO(Consume("/"));
|
|
*prefix += "/";
|
|
DO(ConsumeFullTypeName(full_type_name));
|
|
|
|
return true;
|
|
}
|
|
|
|
// A helper function for reconstructing Any::value. Consumes a text of
|
|
// full_type_name, then serializes it into serialized_value.
|
|
bool ConsumeAnyValue(const Descriptor* value_descriptor,
|
|
std::string* serialized_value) {
|
|
DynamicMessageFactory factory;
|
|
const Message* value_prototype = factory.GetPrototype(value_descriptor);
|
|
if (value_prototype == nullptr) {
|
|
return false;
|
|
}
|
|
std::unique_ptr<Message> value(value_prototype->New());
|
|
std::string sub_delimiter;
|
|
DO(ConsumeMessageDelimiter(&sub_delimiter));
|
|
DO(ConsumeMessage(value.get(), sub_delimiter));
|
|
|
|
if (allow_partial_) {
|
|
value->AppendPartialToString(serialized_value);
|
|
} else {
|
|
if (!value->IsInitialized()) {
|
|
ReportError(
|
|
"Value of type \"" + value_descriptor->full_name() +
|
|
"\" stored in google.protobuf.Any has missing required fields");
|
|
return false;
|
|
}
|
|
value->AppendToString(serialized_value);
|
|
}
|
|
return true;
|
|
}
|
|
|
|
// Consumes a token and confirms that it matches that specified in the
|
|
// value parameter. Returns false if the token found does not match that
|
|
// which was specified.
|
|
bool Consume(const std::string& value) {
|
|
const std::string& current_value = tokenizer_.current().text;
|
|
|
|
if (current_value != value) {
|
|
ReportError("Expected \"" + value + "\", found \"" + current_value +
|
|
"\".");
|
|
return false;
|
|
}
|
|
|
|
tokenizer_.Next();
|
|
|
|
return true;
|
|
}
|
|
|
|
// Similar to `Consume`, but the following token may be tokenized as
|
|
// TYPE_WHITESPACE.
|
|
bool ConsumeBeforeWhitespace(const std::string& value) {
|
|
// Report whitespace after this token, but only once.
|
|
tokenizer_.set_report_whitespace(true);
|
|
bool result = Consume(value);
|
|
tokenizer_.set_report_whitespace(false);
|
|
return result;
|
|
}
|
|
|
|
// Attempts to consume the supplied value. Returns false if a the
|
|
// token found does not match the value specified.
|
|
bool TryConsume(const std::string& value) {
|
|
if (tokenizer_.current().text == value) {
|
|
tokenizer_.Next();
|
|
return true;
|
|
} else {
|
|
return false;
|
|
}
|
|
}
|
|
|
|
// Similar to `TryConsume`, but the following token may be tokenized as
|
|
// TYPE_WHITESPACE.
|
|
bool TryConsumeBeforeWhitespace(const std::string& value) {
|
|
// Report whitespace after this token, but only once.
|
|
tokenizer_.set_report_whitespace(true);
|
|
bool result = TryConsume(value);
|
|
tokenizer_.set_report_whitespace(false);
|
|
return result;
|
|
}
|
|
|
|
bool TryConsumeWhitespace(const std::string& message_type,
|
|
const char* field_type) {
|
|
if (LookingAtType(io::Tokenizer::TYPE_WHITESPACE)) {
|
|
tokenizer_.Next();
|
|
return true;
|
|
}
|
|
|
|
return false;
|
|
}
|
|
|
|
// An internal instance of the Tokenizer's error collector, used to
|
|
// collect any base-level parse errors and feed them to the ParserImpl.
|
|
class ParserErrorCollector : public io::ErrorCollector {
|
|
public:
|
|
explicit ParserErrorCollector(TextFormat::Parser::ParserImpl* parser)
|
|
: parser_(parser) {}
|
|
|
|
~ParserErrorCollector() override {}
|
|
|
|
void AddError(int line, int column, const std::string& message) override {
|
|
parser_->ReportError(line, column, message);
|
|
}
|
|
|
|
void AddWarning(int line, int column, const std::string& message) override {
|
|
parser_->ReportWarning(line, column, message);
|
|
}
|
|
|
|
private:
|
|
GOOGLE_DISALLOW_EVIL_CONSTRUCTORS(ParserErrorCollector);
|
|
TextFormat::Parser::ParserImpl* parser_;
|
|
};
|
|
|
|
io::ErrorCollector* error_collector_;
|
|
const TextFormat::Finder* finder_;
|
|
ParseInfoTree* parse_info_tree_;
|
|
ParserErrorCollector tokenizer_error_collector_;
|
|
io::Tokenizer tokenizer_;
|
|
const Descriptor* root_message_type_;
|
|
SingularOverwritePolicy singular_overwrite_policy_;
|
|
const bool allow_case_insensitive_field_;
|
|
const bool allow_unknown_field_;
|
|
const bool allow_unknown_extension_;
|
|
const bool allow_unknown_enum_;
|
|
const bool allow_field_number_;
|
|
const bool allow_partial_;
|
|
const int initial_recursion_limit_;
|
|
int recursion_limit_;
|
|
bool had_errors_;
|
|
};
|
|
|
|
// ===========================================================================
|
|
// Internal class for writing text to the io::ZeroCopyOutputStream. Adapted
|
|
// from the Printer found in //net/proto2/io/public/printer.h
|
|
class TextFormat::Printer::TextGenerator
|
|
: public TextFormat::BaseTextGenerator {
|
|
public:
|
|
explicit TextGenerator(io::ZeroCopyOutputStream* output,
|
|
int initial_indent_level)
|
|
: output_(output),
|
|
buffer_(nullptr),
|
|
buffer_size_(0),
|
|
at_start_of_line_(true),
|
|
failed_(false),
|
|
insert_silent_marker_(false),
|
|
indent_level_(initial_indent_level),
|
|
initial_indent_level_(initial_indent_level) {}
|
|
|
|
explicit TextGenerator(io::ZeroCopyOutputStream* output,
|
|
bool insert_silent_marker, int initial_indent_level)
|
|
: output_(output),
|
|
buffer_(nullptr),
|
|
buffer_size_(0),
|
|
at_start_of_line_(true),
|
|
failed_(false),
|
|
insert_silent_marker_(insert_silent_marker),
|
|
indent_level_(initial_indent_level),
|
|
initial_indent_level_(initial_indent_level) {}
|
|
|
|
~TextGenerator() {
|
|
// Only BackUp() if we're sure we've successfully called Next() at least
|
|
// once.
|
|
if (!failed_ && buffer_size_ > 0) {
|
|
output_->BackUp(buffer_size_);
|
|
}
|
|
}
|
|
|
|
// Indent text by two spaces. After calling Indent(), two spaces will be
|
|
// inserted at the beginning of each line of text. Indent() may be called
|
|
// multiple times to produce deeper indents.
|
|
void Indent() override { ++indent_level_; }
|
|
|
|
// Reduces the current indent level by two spaces, or crashes if the indent
|
|
// level is zero.
|
|
void Outdent() override {
|
|
if (indent_level_ == 0 || indent_level_ < initial_indent_level_) {
|
|
GOOGLE_LOG(DFATAL) << " Outdent() without matching Indent().";
|
|
return;
|
|
}
|
|
|
|
--indent_level_;
|
|
}
|
|
|
|
size_t GetCurrentIndentationSize() const override {
|
|
return 2 * indent_level_;
|
|
}
|
|
|
|
// Print text to the output stream.
|
|
void Print(const char* text, size_t size) override {
|
|
if (indent_level_ > 0) {
|
|
size_t pos = 0; // The number of bytes we've written so far.
|
|
for (size_t i = 0; i < size; i++) {
|
|
if (text[i] == '\n') {
|
|
// Saw newline. If there is more text, we may need to insert an
|
|
// indent here. So, write what we have so far, including the '\n'.
|
|
Write(text + pos, i - pos + 1);
|
|
pos = i + 1;
|
|
|
|
// Setting this true will cause the next Write() to insert an indent
|
|
// first.
|
|
at_start_of_line_ = true;
|
|
}
|
|
}
|
|
// Write the rest.
|
|
Write(text + pos, size - pos);
|
|
} else {
|
|
Write(text, size);
|
|
if (size > 0 && text[size - 1] == '\n') {
|
|
at_start_of_line_ = true;
|
|
}
|
|
}
|
|
}
|
|
|
|
// True if any write to the underlying stream failed. (We don't just
|
|
// crash in this case because this is an I/O failure, not a programming
|
|
// error.)
|
|
bool failed() const { return failed_; }
|
|
|
|
void PrintMaybeWithMarker(StringPiece text) {
|
|
Print(text.data(), text.size());
|
|
if (ConsumeInsertSilentMarker()) {
|
|
PrintLiteral(DEBUG_STRING_SILENT_MARKER);
|
|
}
|
|
}
|
|
|
|
void PrintMaybeWithMarker(StringPiece text_head,
|
|
StringPiece text_tail) {
|
|
Print(text_head.data(), text_head.size());
|
|
if (ConsumeInsertSilentMarker()) {
|
|
PrintLiteral(DEBUG_STRING_SILENT_MARKER);
|
|
}
|
|
Print(text_tail.data(), text_tail.size());
|
|
}
|
|
|
|
private:
|
|
GOOGLE_DISALLOW_EVIL_CONSTRUCTORS(TextGenerator);
|
|
|
|
void Write(const char* data, size_t size) {
|
|
if (failed_) return;
|
|
if (size == 0) return;
|
|
|
|
if (at_start_of_line_) {
|
|
// Insert an indent.
|
|
at_start_of_line_ = false;
|
|
WriteIndent();
|
|
if (failed_) return;
|
|
}
|
|
|
|
while (static_cast<int64_t>(size) > buffer_size_) {
|
|
// Data exceeds space in the buffer. Copy what we can and request a
|
|
// new buffer.
|
|
if (buffer_size_ > 0) {
|
|
memcpy(buffer_, data, buffer_size_);
|
|
data += buffer_size_;
|
|
size -= buffer_size_;
|
|
}
|
|
void* void_buffer = nullptr;
|
|
failed_ = !output_->Next(&void_buffer, &buffer_size_);
|
|
if (failed_) return;
|
|
buffer_ = reinterpret_cast<char*>(void_buffer);
|
|
}
|
|
|
|
// Buffer is big enough to receive the data; copy it.
|
|
memcpy(buffer_, data, size);
|
|
buffer_ += size;
|
|
buffer_size_ -= size;
|
|
}
|
|
|
|
void WriteIndent() {
|
|
if (indent_level_ == 0) {
|
|
return;
|
|
}
|
|
GOOGLE_DCHECK(!failed_);
|
|
int size = GetCurrentIndentationSize();
|
|
|
|
while (size > buffer_size_) {
|
|
// Data exceeds space in the buffer. Write what we can and request a new
|
|
// buffer.
|
|
if (buffer_size_ > 0) {
|
|
memset(buffer_, ' ', buffer_size_);
|
|
}
|
|
size -= buffer_size_;
|
|
void* void_buffer;
|
|
failed_ = !output_->Next(&void_buffer, &buffer_size_);
|
|
if (failed_) return;
|
|
buffer_ = reinterpret_cast<char*>(void_buffer);
|
|
}
|
|
|
|
// Buffer is big enough to receive the data; copy it.
|
|
memset(buffer_, ' ', size);
|
|
buffer_ += size;
|
|
buffer_size_ -= size;
|
|
}
|
|
|
|
// Return the current value of insert_silent_marker_. If it is true, set it
|
|
// to false as we assume that a silent marker is inserted after a call to this
|
|
// function.
|
|
bool ConsumeInsertSilentMarker() {
|
|
if (insert_silent_marker_) {
|
|
insert_silent_marker_ = false;
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
|
|
io::ZeroCopyOutputStream* const output_;
|
|
char* buffer_;
|
|
int buffer_size_;
|
|
bool at_start_of_line_;
|
|
bool failed_;
|
|
// This flag is false when inserting silent marker is disabled or a silent
|
|
// marker has been inserted.
|
|
bool insert_silent_marker_;
|
|
|
|
int indent_level_;
|
|
int initial_indent_level_;
|
|
};
|
|
|
|
// ===========================================================================
|
|
// An internal field value printer that may insert a silent marker in
|
|
// DebugStrings.
|
|
class TextFormat::Printer::DebugStringFieldValuePrinter
|
|
: public TextFormat::FastFieldValuePrinter {
|
|
public:
|
|
void PrintMessageStart(const Message& /*message*/, int /*field_index*/,
|
|
int /*field_count*/, bool single_line_mode,
|
|
BaseTextGenerator* generator) const override {
|
|
// This is safe as only TextGenerator is used with
|
|
// DebugStringFieldValuePrinter.
|
|
TextGenerator* text_generator = static_cast<TextGenerator*>(generator);
|
|
if (single_line_mode) {
|
|
text_generator->PrintMaybeWithMarker(" ", "{ ");
|
|
} else {
|
|
text_generator->PrintMaybeWithMarker(" ", "{\n");
|
|
}
|
|
}
|
|
};
|
|
|
|
// ===========================================================================
|
|
// An internal field value printer that escape UTF8 strings.
|
|
class TextFormat::Printer::FastFieldValuePrinterUtf8Escaping
|
|
: public TextFormat::Printer::DebugStringFieldValuePrinter {
|
|
public:
|
|
void PrintString(const std::string& val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintLiteral("\"");
|
|
generator->PrintString(strings::Utf8SafeCEscape(val));
|
|
generator->PrintLiteral("\"");
|
|
}
|
|
void PrintBytes(const std::string& val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
return FastFieldValuePrinter::PrintString(val, generator);
|
|
}
|
|
};
|
|
|
|
// ===========================================================================
|
|
// Implementation of the default Finder for extensions.
|
|
TextFormat::Finder::~Finder() {}
|
|
|
|
const FieldDescriptor* TextFormat::Finder::FindExtension(
|
|
Message* message, const std::string& name) const {
|
|
return DefaultFinderFindExtension(message, name);
|
|
}
|
|
|
|
const FieldDescriptor* TextFormat::Finder::FindExtensionByNumber(
|
|
const Descriptor* descriptor, int number) const {
|
|
return DefaultFinderFindExtensionByNumber(descriptor, number);
|
|
}
|
|
|
|
const Descriptor* TextFormat::Finder::FindAnyType(
|
|
const Message& message, const std::string& prefix,
|
|
const std::string& name) const {
|
|
return DefaultFinderFindAnyType(message, prefix, name);
|
|
}
|
|
|
|
MessageFactory* TextFormat::Finder::FindExtensionFactory(
|
|
const FieldDescriptor* /*field*/) const {
|
|
return nullptr;
|
|
}
|
|
|
|
// ===========================================================================
|
|
|
|
TextFormat::Parser::Parser()
|
|
: error_collector_(nullptr),
|
|
finder_(nullptr),
|
|
parse_info_tree_(nullptr),
|
|
allow_partial_(false),
|
|
allow_case_insensitive_field_(false),
|
|
allow_unknown_field_(false),
|
|
allow_unknown_extension_(false),
|
|
allow_unknown_enum_(false),
|
|
allow_field_number_(false),
|
|
allow_relaxed_whitespace_(false),
|
|
allow_singular_overwrites_(false),
|
|
recursion_limit_(std::numeric_limits<int>::max()) {}
|
|
|
|
TextFormat::Parser::~Parser() {}
|
|
|
|
namespace {
|
|
|
|
bool CheckParseInputSize(StringPiece input,
|
|
io::ErrorCollector* error_collector) {
|
|
if (input.size() > INT_MAX) {
|
|
error_collector->AddError(
|
|
-1, 0,
|
|
StrCat(
|
|
"Input size too large: ", static_cast<int64_t>(input.size()),
|
|
" bytes", " > ", INT_MAX, " bytes."));
|
|
return false;
|
|
}
|
|
return true;
|
|
}
|
|
|
|
} // namespace
|
|
|
|
bool TextFormat::Parser::Parse(io::ZeroCopyInputStream* input,
|
|
Message* output) {
|
|
output->Clear();
|
|
|
|
ParserImpl::SingularOverwritePolicy overwrites_policy =
|
|
allow_singular_overwrites_ ? ParserImpl::ALLOW_SINGULAR_OVERWRITES
|
|
: ParserImpl::FORBID_SINGULAR_OVERWRITES;
|
|
|
|
ParserImpl parser(output->GetDescriptor(), input, error_collector_, finder_,
|
|
parse_info_tree_, overwrites_policy,
|
|
allow_case_insensitive_field_, allow_unknown_field_,
|
|
allow_unknown_extension_, allow_unknown_enum_,
|
|
allow_field_number_, allow_relaxed_whitespace_,
|
|
allow_partial_, recursion_limit_);
|
|
return MergeUsingImpl(input, output, &parser);
|
|
}
|
|
|
|
bool TextFormat::Parser::ParseFromString(ConstStringParam input,
|
|
Message* output) {
|
|
DO(CheckParseInputSize(input, error_collector_));
|
|
io::ArrayInputStream input_stream(input.data(), input.size());
|
|
return Parse(&input_stream, output);
|
|
}
|
|
|
|
bool TextFormat::Parser::Merge(io::ZeroCopyInputStream* input,
|
|
Message* output) {
|
|
ParserImpl parser(output->GetDescriptor(), input, error_collector_, finder_,
|
|
parse_info_tree_, ParserImpl::ALLOW_SINGULAR_OVERWRITES,
|
|
allow_case_insensitive_field_, allow_unknown_field_,
|
|
allow_unknown_extension_, allow_unknown_enum_,
|
|
allow_field_number_, allow_relaxed_whitespace_,
|
|
allow_partial_, recursion_limit_);
|
|
return MergeUsingImpl(input, output, &parser);
|
|
}
|
|
|
|
bool TextFormat::Parser::MergeFromString(ConstStringParam input,
|
|
Message* output) {
|
|
DO(CheckParseInputSize(input, error_collector_));
|
|
io::ArrayInputStream input_stream(input.data(), input.size());
|
|
return Merge(&input_stream, output);
|
|
}
|
|
|
|
|
|
bool TextFormat::Parser::MergeUsingImpl(io::ZeroCopyInputStream* /* input */,
|
|
Message* output,
|
|
ParserImpl* parser_impl) {
|
|
if (!parser_impl->Parse(output)) return false;
|
|
if (!allow_partial_ && !output->IsInitialized()) {
|
|
std::vector<std::string> missing_fields;
|
|
output->FindInitializationErrors(&missing_fields);
|
|
parser_impl->ReportError(-1, 0,
|
|
"Message missing required fields: " +
|
|
Join(missing_fields, ", "));
|
|
return false;
|
|
}
|
|
return true;
|
|
}
|
|
|
|
bool TextFormat::Parser::ParseFieldValueFromString(const std::string& input,
|
|
const FieldDescriptor* field,
|
|
Message* output) {
|
|
io::ArrayInputStream input_stream(input.data(), input.size());
|
|
ParserImpl parser(
|
|
output->GetDescriptor(), &input_stream, error_collector_, finder_,
|
|
parse_info_tree_, ParserImpl::ALLOW_SINGULAR_OVERWRITES,
|
|
allow_case_insensitive_field_, allow_unknown_field_,
|
|
allow_unknown_extension_, allow_unknown_enum_, allow_field_number_,
|
|
allow_relaxed_whitespace_, allow_partial_, recursion_limit_);
|
|
return parser.ParseField(field, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::Parse(io::ZeroCopyInputStream* input,
|
|
Message* output) {
|
|
return Parser().Parse(input, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::Merge(io::ZeroCopyInputStream* input,
|
|
Message* output) {
|
|
return Parser().Merge(input, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::ParseFromString(ConstStringParam input,
|
|
Message* output) {
|
|
return Parser().ParseFromString(input, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::MergeFromString(ConstStringParam input,
|
|
Message* output) {
|
|
return Parser().MergeFromString(input, output);
|
|
}
|
|
|
|
|
|
#undef DO
|
|
|
|
// ===========================================================================
|
|
|
|
TextFormat::BaseTextGenerator::~BaseTextGenerator() {}
|
|
|
|
namespace {
|
|
|
|
// A BaseTextGenerator that writes to a string.
|
|
class StringBaseTextGenerator : public TextFormat::BaseTextGenerator {
|
|
public:
|
|
void Print(const char* text, size_t size) override {
|
|
output_.append(text, size);
|
|
}
|
|
|
|
// Some compilers do not support ref-qualifiers even in C++11 mode.
|
|
// Disable the optimization for now and revisit it later.
|
|
#if 0 // LANG_CXX11
|
|
std::string Consume() && { return std::move(output_); }
|
|
#else // !LANG_CXX11
|
|
const std::string& Get() { return output_; }
|
|
#endif // LANG_CXX11
|
|
|
|
private:
|
|
std::string output_;
|
|
};
|
|
|
|
} // namespace
|
|
|
|
// The default implementation for FieldValuePrinter. We just delegate the
|
|
// implementation to the default FastFieldValuePrinter to avoid duplicating the
|
|
// logic.
|
|
TextFormat::FieldValuePrinter::FieldValuePrinter() {}
|
|
TextFormat::FieldValuePrinter::~FieldValuePrinter() {}
|
|
|
|
#if 0 // LANG_CXX11
|
|
#define FORWARD_IMPL(fn, ...) \
|
|
StringBaseTextGenerator generator; \
|
|
delegate_.fn(__VA_ARGS__, &generator); \
|
|
return std::move(generator).Consume()
|
|
#else // !LANG_CXX11
|
|
#define FORWARD_IMPL(fn, ...) \
|
|
StringBaseTextGenerator generator; \
|
|
delegate_.fn(__VA_ARGS__, &generator); \
|
|
return generator.Get()
|
|
#endif // LANG_CXX11
|
|
|
|
std::string TextFormat::FieldValuePrinter::PrintBool(bool val) const {
|
|
FORWARD_IMPL(PrintBool, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintInt32(int32_t val) const {
|
|
FORWARD_IMPL(PrintInt32, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintUInt32(uint32_t val) const {
|
|
FORWARD_IMPL(PrintUInt32, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintInt64(int64_t val) const {
|
|
FORWARD_IMPL(PrintInt64, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintUInt64(uint64_t val) const {
|
|
FORWARD_IMPL(PrintUInt64, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintFloat(float val) const {
|
|
FORWARD_IMPL(PrintFloat, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintDouble(double val) const {
|
|
FORWARD_IMPL(PrintDouble, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintString(
|
|
const std::string& val) const {
|
|
FORWARD_IMPL(PrintString, val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintBytes(
|
|
const std::string& val) const {
|
|
return PrintString(val);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintEnum(
|
|
int32_t val, const std::string& name) const {
|
|
FORWARD_IMPL(PrintEnum, val, name);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintFieldName(
|
|
const Message& message, const Reflection* reflection,
|
|
const FieldDescriptor* field) const {
|
|
FORWARD_IMPL(PrintFieldName, message, reflection, field);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintMessageStart(
|
|
const Message& message, int field_index, int field_count,
|
|
bool single_line_mode) const {
|
|
FORWARD_IMPL(PrintMessageStart, message, field_index, field_count,
|
|
single_line_mode);
|
|
}
|
|
std::string TextFormat::FieldValuePrinter::PrintMessageEnd(
|
|
const Message& message, int field_index, int field_count,
|
|
bool single_line_mode) const {
|
|
FORWARD_IMPL(PrintMessageEnd, message, field_index, field_count,
|
|
single_line_mode);
|
|
}
|
|
#undef FORWARD_IMPL
|
|
|
|
TextFormat::FastFieldValuePrinter::FastFieldValuePrinter() {}
|
|
TextFormat::FastFieldValuePrinter::~FastFieldValuePrinter() {}
|
|
void TextFormat::FastFieldValuePrinter::PrintBool(
|
|
bool val, BaseTextGenerator* generator) const {
|
|
if (val) {
|
|
generator->PrintLiteral("true");
|
|
} else {
|
|
generator->PrintLiteral("false");
|
|
}
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintInt32(
|
|
int32_t val, BaseTextGenerator* generator) const {
|
|
generator->PrintString(StrCat(val));
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintUInt32(
|
|
uint32_t val, BaseTextGenerator* generator) const {
|
|
generator->PrintString(StrCat(val));
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintInt64(
|
|
int64_t val, BaseTextGenerator* generator) const {
|
|
generator->PrintString(StrCat(val));
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintUInt64(
|
|
uint64_t val, BaseTextGenerator* generator) const {
|
|
generator->PrintString(StrCat(val));
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintFloat(
|
|
float val, BaseTextGenerator* generator) const {
|
|
generator->PrintString(!std::isnan(val) ? SimpleFtoa(val) : "nan");
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintDouble(
|
|
double val, BaseTextGenerator* generator) const {
|
|
generator->PrintString(!std::isnan(val) ? SimpleDtoa(val) : "nan");
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintEnum(
|
|
int32_t /*val*/, const std::string& name,
|
|
BaseTextGenerator* generator) const {
|
|
generator->PrintString(name);
|
|
}
|
|
|
|
void TextFormat::FastFieldValuePrinter::PrintString(
|
|
const std::string& val, BaseTextGenerator* generator) const {
|
|
generator->PrintLiteral("\"");
|
|
generator->PrintString(CEscape(val));
|
|
generator->PrintLiteral("\"");
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintBytes(
|
|
const std::string& val, BaseTextGenerator* generator) const {
|
|
PrintString(val, generator);
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintFieldName(
|
|
const Message& message, int /*field_index*/, int /*field_count*/,
|
|
const Reflection* reflection, const FieldDescriptor* field,
|
|
BaseTextGenerator* generator) const {
|
|
PrintFieldName(message, reflection, field, generator);
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintFieldName(
|
|
const Message& /*message*/, const Reflection* /*reflection*/,
|
|
const FieldDescriptor* field, BaseTextGenerator* generator) const {
|
|
if (field->is_extension()) {
|
|
generator->PrintLiteral("[");
|
|
generator->PrintString(field->PrintableNameForExtension());
|
|
generator->PrintLiteral("]");
|
|
} else if (field->type() == FieldDescriptor::TYPE_GROUP) {
|
|
// Groups must be serialized with their original capitalization.
|
|
generator->PrintString(field->message_type()->name());
|
|
} else {
|
|
generator->PrintString(field->name());
|
|
}
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintMessageStart(
|
|
const Message& /*message*/, int /*field_index*/, int /*field_count*/,
|
|
bool single_line_mode, BaseTextGenerator* generator) const {
|
|
if (single_line_mode) {
|
|
generator->PrintLiteral(" { ");
|
|
} else {
|
|
generator->PrintLiteral(" {\n");
|
|
}
|
|
}
|
|
bool TextFormat::FastFieldValuePrinter::PrintMessageContent(
|
|
const Message& /*message*/, int /*field_index*/, int /*field_count*/,
|
|
bool /*single_line_mode*/, BaseTextGenerator* /*generator*/) const {
|
|
return false; // Use the default printing function.
|
|
}
|
|
void TextFormat::FastFieldValuePrinter::PrintMessageEnd(
|
|
const Message& /*message*/, int /*field_index*/, int /*field_count*/,
|
|
bool single_line_mode, BaseTextGenerator* generator) const {
|
|
if (single_line_mode) {
|
|
generator->PrintLiteral("} ");
|
|
} else {
|
|
generator->PrintLiteral("}\n");
|
|
}
|
|
}
|
|
|
|
namespace {
|
|
|
|
// A legacy compatibility wrapper. Takes ownership of the delegate.
|
|
class FieldValuePrinterWrapper : public TextFormat::FastFieldValuePrinter {
|
|
public:
|
|
explicit FieldValuePrinterWrapper(
|
|
const TextFormat::FieldValuePrinter* delegate)
|
|
: delegate_(delegate) {}
|
|
|
|
void SetDelegate(const TextFormat::FieldValuePrinter* delegate) {
|
|
delegate_.reset(delegate);
|
|
}
|
|
|
|
void PrintBool(bool val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintBool(val));
|
|
}
|
|
void PrintInt32(int32_t val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintInt32(val));
|
|
}
|
|
void PrintUInt32(uint32_t val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintUInt32(val));
|
|
}
|
|
void PrintInt64(int64_t val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintInt64(val));
|
|
}
|
|
void PrintUInt64(uint64_t val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintUInt64(val));
|
|
}
|
|
void PrintFloat(float val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintFloat(val));
|
|
}
|
|
void PrintDouble(double val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintDouble(val));
|
|
}
|
|
void PrintString(const std::string& val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintString(val));
|
|
}
|
|
void PrintBytes(const std::string& val,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintBytes(val));
|
|
}
|
|
void PrintEnum(int32_t val, const std::string& name,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintEnum(val, name));
|
|
}
|
|
void PrintFieldName(const Message& message, int /*field_index*/,
|
|
int /*field_count*/, const Reflection* reflection,
|
|
const FieldDescriptor* field,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(
|
|
delegate_->PrintFieldName(message, reflection, field));
|
|
}
|
|
void PrintFieldName(const Message& message, const Reflection* reflection,
|
|
const FieldDescriptor* field,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(
|
|
delegate_->PrintFieldName(message, reflection, field));
|
|
}
|
|
void PrintMessageStart(
|
|
const Message& message, int field_index, int field_count,
|
|
bool single_line_mode,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintMessageStart(
|
|
message, field_index, field_count, single_line_mode));
|
|
}
|
|
void PrintMessageEnd(
|
|
const Message& message, int field_index, int field_count,
|
|
bool single_line_mode,
|
|
TextFormat::BaseTextGenerator* generator) const override {
|
|
generator->PrintString(delegate_->PrintMessageEnd(
|
|
message, field_index, field_count, single_line_mode));
|
|
}
|
|
|
|
private:
|
|
std::unique_ptr<const TextFormat::FieldValuePrinter> delegate_;
|
|
};
|
|
|
|
} // namespace
|
|
|
|
const char* const TextFormat::Printer::kDoNotParse =
|
|
"DO NOT PARSE: fields may be stripped and missing.\n";
|
|
|
|
TextFormat::Printer::Printer()
|
|
: initial_indent_level_(0),
|
|
single_line_mode_(false),
|
|
use_field_number_(false),
|
|
use_short_repeated_primitives_(false),
|
|
insert_silent_marker_(false),
|
|
hide_unknown_fields_(false),
|
|
print_message_fields_in_index_order_(false),
|
|
expand_any_(false),
|
|
truncate_string_field_longer_than_(0LL),
|
|
finder_(nullptr) {
|
|
SetUseUtf8StringEscaping(false);
|
|
}
|
|
|
|
void TextFormat::Printer::SetUseUtf8StringEscaping(bool as_utf8) {
|
|
SetDefaultFieldValuePrinter(as_utf8 ? new FastFieldValuePrinterUtf8Escaping()
|
|
: new DebugStringFieldValuePrinter());
|
|
}
|
|
|
|
void TextFormat::Printer::SetDefaultFieldValuePrinter(
|
|
const FieldValuePrinter* printer) {
|
|
default_field_value_printer_.reset(new FieldValuePrinterWrapper(printer));
|
|
}
|
|
|
|
void TextFormat::Printer::SetDefaultFieldValuePrinter(
|
|
const FastFieldValuePrinter* printer) {
|
|
default_field_value_printer_.reset(printer);
|
|
}
|
|
|
|
bool TextFormat::Printer::RegisterFieldValuePrinter(
|
|
const FieldDescriptor* field, const FieldValuePrinter* printer) {
|
|
if (field == nullptr || printer == nullptr) {
|
|
return false;
|
|
}
|
|
std::unique_ptr<FieldValuePrinterWrapper> wrapper(
|
|
new FieldValuePrinterWrapper(nullptr));
|
|
auto pair = custom_printers_.insert(std::make_pair(field, nullptr));
|
|
if (pair.second) {
|
|
wrapper->SetDelegate(printer);
|
|
pair.first->second = std::move(wrapper);
|
|
return true;
|
|
} else {
|
|
return false;
|
|
}
|
|
}
|
|
|
|
bool TextFormat::Printer::RegisterFieldValuePrinter(
|
|
const FieldDescriptor* field, const FastFieldValuePrinter* printer) {
|
|
if (field == nullptr || printer == nullptr) {
|
|
return false;
|
|
}
|
|
auto pair = custom_printers_.insert(std::make_pair(field, nullptr));
|
|
if (pair.second) {
|
|
pair.first->second.reset(printer);
|
|
return true;
|
|
} else {
|
|
return false;
|
|
}
|
|
}
|
|
|
|
bool TextFormat::Printer::RegisterMessagePrinter(
|
|
const Descriptor* descriptor, const MessagePrinter* printer) {
|
|
if (descriptor == nullptr || printer == nullptr) {
|
|
return false;
|
|
}
|
|
auto pair =
|
|
custom_message_printers_.insert(std::make_pair(descriptor, nullptr));
|
|
if (pair.second) {
|
|
pair.first->second.reset(printer);
|
|
return true;
|
|
} else {
|
|
return false;
|
|
}
|
|
}
|
|
|
|
bool TextFormat::Printer::PrintToString(const Message& message,
|
|
std::string* output) const {
|
|
GOOGLE_DCHECK(output) << "output specified is nullptr";
|
|
|
|
output->clear();
|
|
io::StringOutputStream output_stream(output);
|
|
|
|
return Print(message, &output_stream);
|
|
}
|
|
|
|
bool TextFormat::Printer::PrintUnknownFieldsToString(
|
|
const UnknownFieldSet& unknown_fields, std::string* output) const {
|
|
GOOGLE_DCHECK(output) << "output specified is nullptr";
|
|
|
|
output->clear();
|
|
io::StringOutputStream output_stream(output);
|
|
return PrintUnknownFields(unknown_fields, &output_stream);
|
|
}
|
|
|
|
bool TextFormat::Printer::Print(const Message& message,
|
|
io::ZeroCopyOutputStream* output) const {
|
|
TextGenerator generator(output, insert_silent_marker_, initial_indent_level_);
|
|
|
|
Print(message, &generator);
|
|
|
|
// Output false if the generator failed internally.
|
|
return !generator.failed();
|
|
}
|
|
|
|
// Maximum recursion depth for heuristically printing out length-delimited
|
|
// unknown fields as messages.
|
|
static constexpr int kUnknownFieldRecursionLimit = 10;
|
|
|
|
bool TextFormat::Printer::PrintUnknownFields(
|
|
const UnknownFieldSet& unknown_fields,
|
|
io::ZeroCopyOutputStream* output) const {
|
|
TextGenerator generator(output, initial_indent_level_);
|
|
|
|
PrintUnknownFields(unknown_fields, &generator, kUnknownFieldRecursionLimit);
|
|
|
|
// Output false if the generator failed internally.
|
|
return !generator.failed();
|
|
}
|
|
|
|
namespace {
|
|
// Comparison functor for sorting FieldDescriptors by field index.
|
|
// Normal fields have higher precedence than extensions.
|
|
struct FieldIndexSorter {
|
|
bool operator()(const FieldDescriptor* left,
|
|
const FieldDescriptor* right) const {
|
|
if (left->is_extension() && right->is_extension()) {
|
|
return left->number() < right->number();
|
|
} else if (left->is_extension()) {
|
|
return false;
|
|
} else if (right->is_extension()) {
|
|
return true;
|
|
} else {
|
|
return left->index() < right->index();
|
|
}
|
|
}
|
|
};
|
|
|
|
} // namespace
|
|
|
|
bool TextFormat::Printer::PrintAny(const Message& message,
|
|
TextGenerator* generator) const {
|
|
const FieldDescriptor* type_url_field;
|
|
const FieldDescriptor* value_field;
|
|
if (!internal::GetAnyFieldDescriptors(message, &type_url_field,
|
|
&value_field)) {
|
|
return false;
|
|
}
|
|
|
|
const Reflection* reflection = message.GetReflection();
|
|
|
|
// Extract the full type name from the type_url field.
|
|
const std::string& type_url = reflection->GetString(message, type_url_field);
|
|
std::string url_prefix;
|
|
std::string full_type_name;
|
|
if (!internal::ParseAnyTypeUrl(type_url, &url_prefix, &full_type_name)) {
|
|
return false;
|
|
}
|
|
|
|
// Print the "value" in text.
|
|
const Descriptor* value_descriptor =
|
|
finder_ ? finder_->FindAnyType(message, url_prefix, full_type_name)
|
|
: DefaultFinderFindAnyType(message, url_prefix, full_type_name);
|
|
if (value_descriptor == nullptr) {
|
|
GOOGLE_LOG(WARNING) << "Can't print proto content: proto type " << type_url
|
|
<< " not found";
|
|
return false;
|
|
}
|
|
DynamicMessageFactory factory;
|
|
std::unique_ptr<Message> value_message(
|
|
factory.GetPrototype(value_descriptor)->New());
|
|
std::string serialized_value = reflection->GetString(message, value_field);
|
|
if (!value_message->ParseFromString(serialized_value)) {
|
|
GOOGLE_LOG(WARNING) << type_url << ": failed to parse contents";
|
|
return false;
|
|
}
|
|
generator->PrintLiteral("[");
|
|
generator->PrintString(type_url);
|
|
generator->PrintLiteral("]");
|
|
const FastFieldValuePrinter* printer = GetFieldPrinter(value_field);
|
|
printer->PrintMessageStart(message, -1, 0, single_line_mode_, generator);
|
|
generator->Indent();
|
|
Print(*value_message, generator);
|
|
generator->Outdent();
|
|
printer->PrintMessageEnd(message, -1, 0, single_line_mode_, generator);
|
|
return true;
|
|
}
|
|
|
|
void TextFormat::Printer::Print(const Message& message,
|
|
TextGenerator* generator) const {
|
|
const Reflection* reflection = message.GetReflection();
|
|
if (!reflection) {
|
|
// This message does not provide any way to describe its structure.
|
|
// Parse it again in an UnknownFieldSet, and display this instead.
|
|
UnknownFieldSet unknown_fields;
|
|
{
|
|
std::string serialized = message.SerializeAsString();
|
|
io::ArrayInputStream input(serialized.data(), serialized.size());
|
|
unknown_fields.ParseFromZeroCopyStream(&input);
|
|
}
|
|
PrintUnknownFields(unknown_fields, generator, kUnknownFieldRecursionLimit);
|
|
return;
|
|
}
|
|
const Descriptor* descriptor = message.GetDescriptor();
|
|
auto itr = custom_message_printers_.find(descriptor);
|
|
if (itr != custom_message_printers_.end()) {
|
|
itr->second->Print(message, single_line_mode_, generator);
|
|
return;
|
|
}
|
|
if (descriptor->full_name() == internal::kAnyFullTypeName && expand_any_ &&
|
|
PrintAny(message, generator)) {
|
|
return;
|
|
}
|
|
std::vector<const FieldDescriptor*> fields;
|
|
if (descriptor->options().map_entry()) {
|
|
fields.push_back(descriptor->field(0));
|
|
fields.push_back(descriptor->field(1));
|
|
} else {
|
|
reflection->ListFieldsOmitStripped(message, &fields);
|
|
if (reflection->IsMessageStripped(message.GetDescriptor())) {
|
|
generator->Print(kDoNotParse, std::strlen(kDoNotParse));
|
|
}
|
|
}
|
|
|
|
if (print_message_fields_in_index_order_) {
|
|
std::sort(fields.begin(), fields.end(), FieldIndexSorter());
|
|
}
|
|
for (const FieldDescriptor* field : fields) {
|
|
PrintField(message, reflection, field, generator);
|
|
}
|
|
if (!hide_unknown_fields_) {
|
|
PrintUnknownFields(reflection->GetUnknownFields(message), generator,
|
|
kUnknownFieldRecursionLimit);
|
|
}
|
|
}
|
|
|
|
void TextFormat::Printer::PrintFieldValueToString(const Message& message,
|
|
const FieldDescriptor* field,
|
|
int index,
|
|
std::string* output) const {
|
|
GOOGLE_DCHECK(output) << "output specified is nullptr";
|
|
|
|
output->clear();
|
|
io::StringOutputStream output_stream(output);
|
|
TextGenerator generator(&output_stream, initial_indent_level_);
|
|
|
|
PrintFieldValue(message, message.GetReflection(), field, index, &generator);
|
|
}
|
|
|
|
class MapEntryMessageComparator {
|
|
public:
|
|
explicit MapEntryMessageComparator(const Descriptor* descriptor)
|
|
: field_(descriptor->field(0)) {}
|
|
|
|
bool operator()(const Message* a, const Message* b) {
|
|
const Reflection* reflection = a->GetReflection();
|
|
switch (field_->cpp_type()) {
|
|
case FieldDescriptor::CPPTYPE_BOOL: {
|
|
bool first = reflection->GetBool(*a, field_);
|
|
bool second = reflection->GetBool(*b, field_);
|
|
return first < second;
|
|
}
|
|
case FieldDescriptor::CPPTYPE_INT32: {
|
|
int32_t first = reflection->GetInt32(*a, field_);
|
|
int32_t second = reflection->GetInt32(*b, field_);
|
|
return first < second;
|
|
}
|
|
case FieldDescriptor::CPPTYPE_INT64: {
|
|
int64_t first = reflection->GetInt64(*a, field_);
|
|
int64_t second = reflection->GetInt64(*b, field_);
|
|
return first < second;
|
|
}
|
|
case FieldDescriptor::CPPTYPE_UINT32: {
|
|
uint32_t first = reflection->GetUInt32(*a, field_);
|
|
uint32_t second = reflection->GetUInt32(*b, field_);
|
|
return first < second;
|
|
}
|
|
case FieldDescriptor::CPPTYPE_UINT64: {
|
|
uint64_t first = reflection->GetUInt64(*a, field_);
|
|
uint64_t second = reflection->GetUInt64(*b, field_);
|
|
return first < second;
|
|
}
|
|
case FieldDescriptor::CPPTYPE_STRING: {
|
|
std::string first = reflection->GetString(*a, field_);
|
|
std::string second = reflection->GetString(*b, field_);
|
|
return first < second;
|
|
}
|
|
default:
|
|
GOOGLE_LOG(DFATAL) << "Invalid key for map field.";
|
|
return true;
|
|
}
|
|
}
|
|
|
|
private:
|
|
const FieldDescriptor* field_;
|
|
};
|
|
|
|
namespace internal {
|
|
class MapFieldPrinterHelper {
|
|
public:
|
|
// DynamicMapSorter::Sort cannot be used because it enfores syncing with
|
|
// repeated field.
|
|
static bool SortMap(const Message& message, const Reflection* reflection,
|
|
const FieldDescriptor* field,
|
|
std::vector<const Message*>* sorted_map_field);
|
|
static void CopyKey(const MapKey& key, Message* message,
|
|
const FieldDescriptor* field_desc);
|
|
static void CopyValue(const MapValueRef& value, Message* message,
|
|
const FieldDescriptor* field_desc);
|
|
};
|
|
|
|
// Returns true if elements contained in sorted_map_field need to be released.
|
|
bool MapFieldPrinterHelper::SortMap(
|
|
const Message& message, const Reflection* reflection,
|
|
const FieldDescriptor* field,
|
|
std::vector<const Message*>* sorted_map_field) {
|
|
bool need_release = false;
|
|
const MapFieldBase& base = *reflection->GetMapData(message, field);
|
|
|
|
if (base.IsRepeatedFieldValid()) {
|
|
const RepeatedPtrField<Message>& map_field =
|
|
reflection->GetRepeatedPtrFieldInternal<Message>(message, field);
|
|
for (int i = 0; i < map_field.size(); ++i) {
|
|
sorted_map_field->push_back(
|
|
const_cast<RepeatedPtrField<Message>*>(&map_field)->Mutable(i));
|
|
}
|
|
} else {
|
|
// TODO(teboring): For performance, instead of creating map entry message
|
|
// for each element, just store map keys and sort them.
|
|
const Descriptor* map_entry_desc = field->message_type();
|
|
const Message* prototype =
|
|
reflection->GetMessageFactory()->GetPrototype(map_entry_desc);
|
|
for (MapIterator iter =
|
|
reflection->MapBegin(const_cast<Message*>(&message), field);
|
|
iter != reflection->MapEnd(const_cast<Message*>(&message), field);
|
|
++iter) {
|
|
Message* map_entry_message = prototype->New();
|
|
CopyKey(iter.GetKey(), map_entry_message, map_entry_desc->field(0));
|
|
CopyValue(iter.GetValueRef(), map_entry_message,
|
|
map_entry_desc->field(1));
|
|
sorted_map_field->push_back(map_entry_message);
|
|
}
|
|
need_release = true;
|
|
}
|
|
|
|
MapEntryMessageComparator comparator(field->message_type());
|
|
std::stable_sort(sorted_map_field->begin(), sorted_map_field->end(),
|
|
comparator);
|
|
return need_release;
|
|
}
|
|
|
|
void MapFieldPrinterHelper::CopyKey(const MapKey& key, Message* message,
|
|
const FieldDescriptor* field_desc) {
|
|
const Reflection* reflection = message->GetReflection();
|
|
switch (field_desc->cpp_type()) {
|
|
case FieldDescriptor::CPPTYPE_DOUBLE:
|
|
case FieldDescriptor::CPPTYPE_FLOAT:
|
|
case FieldDescriptor::CPPTYPE_ENUM:
|
|
case FieldDescriptor::CPPTYPE_MESSAGE:
|
|
GOOGLE_LOG(ERROR) << "Not supported.";
|
|
break;
|
|
case FieldDescriptor::CPPTYPE_STRING:
|
|
reflection->SetString(message, field_desc, key.GetStringValue());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_INT64:
|
|
reflection->SetInt64(message, field_desc, key.GetInt64Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_INT32:
|
|
reflection->SetInt32(message, field_desc, key.GetInt32Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_UINT64:
|
|
reflection->SetUInt64(message, field_desc, key.GetUInt64Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_UINT32:
|
|
reflection->SetUInt32(message, field_desc, key.GetUInt32Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_BOOL:
|
|
reflection->SetBool(message, field_desc, key.GetBoolValue());
|
|
return;
|
|
}
|
|
}
|
|
|
|
void MapFieldPrinterHelper::CopyValue(const MapValueRef& value,
|
|
Message* message,
|
|
const FieldDescriptor* field_desc) {
|
|
const Reflection* reflection = message->GetReflection();
|
|
switch (field_desc->cpp_type()) {
|
|
case FieldDescriptor::CPPTYPE_DOUBLE:
|
|
reflection->SetDouble(message, field_desc, value.GetDoubleValue());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_FLOAT:
|
|
reflection->SetFloat(message, field_desc, value.GetFloatValue());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_ENUM:
|
|
reflection->SetEnumValue(message, field_desc, value.GetEnumValue());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_MESSAGE: {
|
|
Message* sub_message = value.GetMessageValue().New();
|
|
sub_message->CopyFrom(value.GetMessageValue());
|
|
reflection->SetAllocatedMessage(message, sub_message, field_desc);
|
|
return;
|
|
}
|
|
case FieldDescriptor::CPPTYPE_STRING:
|
|
reflection->SetString(message, field_desc, value.GetStringValue());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_INT64:
|
|
reflection->SetInt64(message, field_desc, value.GetInt64Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_INT32:
|
|
reflection->SetInt32(message, field_desc, value.GetInt32Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_UINT64:
|
|
reflection->SetUInt64(message, field_desc, value.GetUInt64Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_UINT32:
|
|
reflection->SetUInt32(message, field_desc, value.GetUInt32Value());
|
|
return;
|
|
case FieldDescriptor::CPPTYPE_BOOL:
|
|
reflection->SetBool(message, field_desc, value.GetBoolValue());
|
|
return;
|
|
}
|
|
}
|
|
} // namespace internal
|
|
|
|
void TextFormat::Printer::PrintField(const Message& message,
|
|
const Reflection* reflection,
|
|
const FieldDescriptor* field,
|
|
TextGenerator* generator) const {
|
|
if (use_short_repeated_primitives_ && field->is_repeated() &&
|
|
field->cpp_type() != FieldDescriptor::CPPTYPE_STRING &&
|
|
field->cpp_type() != FieldDescriptor::CPPTYPE_MESSAGE) {
|
|
PrintShortRepeatedField(message, reflection, field, generator);
|
|
return;
|
|
}
|
|
|
|
int count = 0;
|
|
|
|
if (field->is_repeated()) {
|
|
count = reflection->FieldSize(message, field);
|
|
} else if (reflection->HasField(message, field) ||
|
|
field->containing_type()->options().map_entry()) {
|
|
count = 1;
|
|
}
|
|
|
|
std::vector<const Message*> sorted_map_field;
|
|
bool need_release = false;
|
|
bool is_map = field->is_map();
|
|
if (is_map) {
|
|
need_release = internal::MapFieldPrinterHelper::SortMap(
|
|
message, reflection, field, &sorted_map_field);
|
|
}
|
|
|
|
for (int j = 0; j < count; ++j) {
|
|
const int field_index = field->is_repeated() ? j : -1;
|
|
|
|
PrintFieldName(message, field_index, count, reflection, field, generator);
|
|
|
|
if (field->cpp_type() == FieldDescriptor::CPPTYPE_MESSAGE) {
|
|
const FastFieldValuePrinter* printer = GetFieldPrinter(field);
|
|
const Message& sub_message =
|
|
field->is_repeated()
|
|
? (is_map ? *sorted_map_field[j]
|
|
: reflection->GetRepeatedMessage(message, field, j))
|
|
: reflection->GetMessage(message, field);
|
|
printer->PrintMessageStart(sub_message, field_index, count,
|
|
single_line_mode_, generator);
|
|
generator->Indent();
|
|
if (!printer->PrintMessageContent(sub_message, field_index, count,
|
|
single_line_mode_, generator)) {
|
|
Print(sub_message, generator);
|
|
}
|
|
generator->Outdent();
|
|
printer->PrintMessageEnd(sub_message, field_index, count,
|
|
single_line_mode_, generator);
|
|
} else {
|
|
generator->PrintMaybeWithMarker(": ");
|
|
// Write the field value.
|
|
PrintFieldValue(message, reflection, field, field_index, generator);
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral(" ");
|
|
} else {
|
|
generator->PrintLiteral("\n");
|
|
}
|
|
}
|
|
}
|
|
|
|
if (need_release) {
|
|
for (const Message* message_to_delete : sorted_map_field) {
|
|
delete message_to_delete;
|
|
}
|
|
}
|
|
}
|
|
|
|
void TextFormat::Printer::PrintShortRepeatedField(
|
|
const Message& message, const Reflection* reflection,
|
|
const FieldDescriptor* field, TextGenerator* generator) const {
|
|
// Print primitive repeated field in short form.
|
|
int size = reflection->FieldSize(message, field);
|
|
PrintFieldName(message, /*field_index=*/-1, /*field_count=*/size, reflection,
|
|
field, generator);
|
|
generator->PrintMaybeWithMarker(": ", "[");
|
|
for (int i = 0; i < size; i++) {
|
|
if (i > 0) generator->PrintLiteral(", ");
|
|
PrintFieldValue(message, reflection, field, i, generator);
|
|
}
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral("] ");
|
|
} else {
|
|
generator->PrintLiteral("]\n");
|
|
}
|
|
}
|
|
|
|
void TextFormat::Printer::PrintFieldName(const Message& message,
|
|
int field_index, int field_count,
|
|
const Reflection* reflection,
|
|
const FieldDescriptor* field,
|
|
TextGenerator* generator) const {
|
|
// if use_field_number_ is true, prints field number instead
|
|
// of field name.
|
|
if (use_field_number_) {
|
|
generator->PrintString(StrCat(field->number()));
|
|
return;
|
|
}
|
|
|
|
const FastFieldValuePrinter* printer = GetFieldPrinter(field);
|
|
printer->PrintFieldName(message, field_index, field_count, reflection, field,
|
|
generator);
|
|
}
|
|
|
|
void TextFormat::Printer::PrintFieldValue(const Message& message,
|
|
const Reflection* reflection,
|
|
const FieldDescriptor* field,
|
|
int index,
|
|
TextGenerator* generator) const {
|
|
GOOGLE_DCHECK(field->is_repeated() || (index == -1))
|
|
<< "Index must be -1 for non-repeated fields";
|
|
|
|
const FastFieldValuePrinter* printer = GetFieldPrinter(field);
|
|
|
|
switch (field->cpp_type()) {
|
|
#define OUTPUT_FIELD(CPPTYPE, METHOD) \
|
|
case FieldDescriptor::CPPTYPE_##CPPTYPE: \
|
|
printer->Print##METHOD( \
|
|
field->is_repeated() \
|
|
? reflection->GetRepeated##METHOD(message, field, index) \
|
|
: reflection->Get##METHOD(message, field), \
|
|
generator); \
|
|
break
|
|
|
|
OUTPUT_FIELD(INT32, Int32);
|
|
OUTPUT_FIELD(INT64, Int64);
|
|
OUTPUT_FIELD(UINT32, UInt32);
|
|
OUTPUT_FIELD(UINT64, UInt64);
|
|
OUTPUT_FIELD(FLOAT, Float);
|
|
OUTPUT_FIELD(DOUBLE, Double);
|
|
OUTPUT_FIELD(BOOL, Bool);
|
|
#undef OUTPUT_FIELD
|
|
|
|
case FieldDescriptor::CPPTYPE_STRING: {
|
|
std::string scratch;
|
|
const std::string& value =
|
|
field->is_repeated()
|
|
? reflection->GetRepeatedStringReference(message, field, index,
|
|
&scratch)
|
|
: reflection->GetStringReference(message, field, &scratch);
|
|
const std::string* value_to_print = &value;
|
|
std::string truncated_value;
|
|
if (truncate_string_field_longer_than_ > 0 &&
|
|
static_cast<size_t>(truncate_string_field_longer_than_) <
|
|
value.size()) {
|
|
truncated_value = value.substr(0, truncate_string_field_longer_than_) +
|
|
"...<truncated>...";
|
|
value_to_print = &truncated_value;
|
|
}
|
|
if (field->type() == FieldDescriptor::TYPE_STRING) {
|
|
printer->PrintString(*value_to_print, generator);
|
|
} else {
|
|
GOOGLE_DCHECK_EQ(field->type(), FieldDescriptor::TYPE_BYTES);
|
|
printer->PrintBytes(*value_to_print, generator);
|
|
}
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_ENUM: {
|
|
int enum_value =
|
|
field->is_repeated()
|
|
? reflection->GetRepeatedEnumValue(message, field, index)
|
|
: reflection->GetEnumValue(message, field);
|
|
const EnumValueDescriptor* enum_desc =
|
|
field->enum_type()->FindValueByNumber(enum_value);
|
|
if (enum_desc != nullptr) {
|
|
printer->PrintEnum(enum_value, enum_desc->name(), generator);
|
|
} else {
|
|
// Ordinarily, enum_desc should not be null, because proto2 has the
|
|
// invariant that set enum field values must be in-range, but with the
|
|
// new integer-based API for enums (or the RepeatedField<int> loophole),
|
|
// it is possible for the user to force an unknown integer value. So we
|
|
// simply use the integer value itself as the enum value name in this
|
|
// case.
|
|
printer->PrintEnum(enum_value, StrCat(enum_value), generator);
|
|
}
|
|
break;
|
|
}
|
|
|
|
case FieldDescriptor::CPPTYPE_MESSAGE:
|
|
Print(field->is_repeated()
|
|
? reflection->GetRepeatedMessage(message, field, index)
|
|
: reflection->GetMessage(message, field),
|
|
generator);
|
|
break;
|
|
}
|
|
}
|
|
|
|
/* static */ bool TextFormat::Print(const Message& message,
|
|
io::ZeroCopyOutputStream* output) {
|
|
return Printer().Print(message, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::PrintUnknownFields(
|
|
const UnknownFieldSet& unknown_fields, io::ZeroCopyOutputStream* output) {
|
|
return Printer().PrintUnknownFields(unknown_fields, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::PrintToString(const Message& message,
|
|
std::string* output) {
|
|
return Printer().PrintToString(message, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::PrintUnknownFieldsToString(
|
|
const UnknownFieldSet& unknown_fields, std::string* output) {
|
|
return Printer().PrintUnknownFieldsToString(unknown_fields, output);
|
|
}
|
|
|
|
/* static */ void TextFormat::PrintFieldValueToString(
|
|
const Message& message, const FieldDescriptor* field, int index,
|
|
std::string* output) {
|
|
return Printer().PrintFieldValueToString(message, field, index, output);
|
|
}
|
|
|
|
/* static */ bool TextFormat::ParseFieldValueFromString(
|
|
const std::string& input, const FieldDescriptor* field, Message* message) {
|
|
return Parser().ParseFieldValueFromString(input, field, message);
|
|
}
|
|
|
|
void TextFormat::Printer::PrintUnknownFields(
|
|
const UnknownFieldSet& unknown_fields, TextGenerator* generator,
|
|
int recursion_budget) const {
|
|
for (int i = 0; i < unknown_fields.field_count(); i++) {
|
|
const UnknownField& field = unknown_fields.field(i);
|
|
std::string field_number = StrCat(field.number());
|
|
|
|
switch (field.type()) {
|
|
case UnknownField::TYPE_VARINT:
|
|
generator->PrintString(field_number);
|
|
generator->PrintMaybeWithMarker(": ");
|
|
generator->PrintString(StrCat(field.varint()));
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral(" ");
|
|
} else {
|
|
generator->PrintLiteral("\n");
|
|
}
|
|
break;
|
|
case UnknownField::TYPE_FIXED32: {
|
|
generator->PrintString(field_number);
|
|
generator->PrintMaybeWithMarker(": ", "0x");
|
|
generator->PrintString(
|
|
StrCat(strings::Hex(field.fixed32(), strings::ZERO_PAD_8)));
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral(" ");
|
|
} else {
|
|
generator->PrintLiteral("\n");
|
|
}
|
|
break;
|
|
}
|
|
case UnknownField::TYPE_FIXED64: {
|
|
generator->PrintString(field_number);
|
|
generator->PrintMaybeWithMarker(": ", "0x");
|
|
generator->PrintString(
|
|
StrCat(strings::Hex(field.fixed64(), strings::ZERO_PAD_16)));
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral(" ");
|
|
} else {
|
|
generator->PrintLiteral("\n");
|
|
}
|
|
break;
|
|
}
|
|
case UnknownField::TYPE_LENGTH_DELIMITED: {
|
|
generator->PrintString(field_number);
|
|
const std::string& value = field.length_delimited();
|
|
// We create a CodedInputStream so that we can adhere to our recursion
|
|
// budget when we attempt to parse the data. UnknownFieldSet parsing is
|
|
// recursive because of groups.
|
|
io::CodedInputStream input_stream(
|
|
reinterpret_cast<const uint8_t*>(value.data()), value.size());
|
|
input_stream.SetRecursionLimit(recursion_budget);
|
|
UnknownFieldSet embedded_unknown_fields;
|
|
if (!value.empty() && recursion_budget > 0 &&
|
|
embedded_unknown_fields.ParseFromCodedStream(&input_stream)) {
|
|
// This field is parseable as a Message.
|
|
// So it is probably an embedded message.
|
|
if (single_line_mode_) {
|
|
generator->PrintMaybeWithMarker(" ", "{ ");
|
|
} else {
|
|
generator->PrintMaybeWithMarker(" ", "{\n");
|
|
generator->Indent();
|
|
}
|
|
PrintUnknownFields(embedded_unknown_fields, generator,
|
|
recursion_budget - 1);
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral("} ");
|
|
} else {
|
|
generator->Outdent();
|
|
generator->PrintLiteral("}\n");
|
|
}
|
|
} else {
|
|
// This field is not parseable as a Message (or we ran out of
|
|
// recursion budget). So it is probably just a plain string.
|
|
generator->PrintMaybeWithMarker(": ", "\"");
|
|
generator->PrintString(CEscape(value));
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral("\" ");
|
|
} else {
|
|
generator->PrintLiteral("\"\n");
|
|
}
|
|
}
|
|
break;
|
|
}
|
|
case UnknownField::TYPE_GROUP:
|
|
generator->PrintString(field_number);
|
|
if (single_line_mode_) {
|
|
generator->PrintMaybeWithMarker(" ", "{ ");
|
|
} else {
|
|
generator->PrintMaybeWithMarker(" ", "{\n");
|
|
generator->Indent();
|
|
}
|
|
// For groups, we recurse without checking the budget. This is OK,
|
|
// because if the groups were too deeply nested then we would have
|
|
// already rejected the message when we originally parsed it.
|
|
PrintUnknownFields(field.group(), generator, recursion_budget - 1);
|
|
if (single_line_mode_) {
|
|
generator->PrintLiteral("} ");
|
|
} else {
|
|
generator->Outdent();
|
|
generator->PrintLiteral("}\n");
|
|
}
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
|
|
} // namespace protobuf
|
|
} // namespace google
|
|
|
|
#include <thirdparty/protobuf/port_undef.inc>
|