document_metadata_extractor.cc | Explore in Territory

// Copyright 2017 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

#include "third_party/blink/renderer/modules/document_metadata/document_metadata_extractor.h"

#include <algorithm>
#include <memory>
#include <utility>

#include "components/schema_org/common/metadata.mojom-blink.h"
#include "third_party/blink/public/mojom/document_metadata/document_metadata.mojom-blink.h"
#include "third_party/blink/renderer/core/dom/document.h"
#include "third_party/blink/renderer/core/dom/element_traversal.h"
#include "third_party/blink/renderer/core/frame/local_frame.h"
#include "third_party/blink/renderer/core/html/html_element.h"
#include "third_party/blink/renderer/core/html_names.h"
#include "third_party/blink/renderer/platform/instrumentation/tracing/trace_event.h"
#include "third_party/blink/renderer/platform/json/json_parser.h"
#include "third_party/blink/renderer/platform/json/json_values.h"
#include "third_party/blink/renderer/platform/wtf/text/atomic_string.h"
#include "third_party/blink/renderer/platform/wtf/text/string_builder.h"
#include "third_party/blink/renderer/platform/wtf/vector.h"

namespace blink {

namespace {

WebPage;
WebPagePtr;
Entity;
EntityPtr;
Property;
PropertyPtr;
Values;
ValuesPtr;

// App Indexing enforces a max nesting depth of 5. Our top level message
// corresponds to the WebPage, so this only leaves 4 more levels. We will parse
// entites up to this depth, and ignore any further nesting. If an object at the
// max nesting depth has a property corresponding to an entity, that property
// will be dropped. Note that we will still parse json-ld blocks deeper than
// this, but it won't be passed to App Indexing.
constexpr int kMaxDepth = …;
// Some strings are very long, and we don't currently use those, so limit string
// length to something reasonable to avoid undue pressure on Icing. Note that
// App Indexing supports strings up to length 20k.
constexpr wtf_size_t kMaxStringLength = …;
// Enforced by App Indexing, so stop processing early if possible.
constexpr wtf_size_t kMaxNumFields = …;
// Enforced by App Indexing, so stop processing early if possible.
constexpr wtf_size_t kMaxRepeatedSize = …;

constexpr char kJSONLDKeyType[] = …;
constexpr char kJSONLDKeyGraph[] = …;
bool IsSupportedType(AtomicString type) { … }

void ExtractEntity(const JSONObject&, int recursion_level, Entity&);

bool ParseRepeatedValue(const JSONArray& arr,
                        int recursion_level,
                        ValuesPtr& values) { … }

void ExtractEntity(const JSONObject& val, int recursion_level, Entity& entity) { … }

void ExtractTopLevelEntity(const JSONObject& val, Vector<EntityPtr>& entities) { … }

void ExtractEntitiesFromArray(const JSONArray& arr,
                              Vector<EntityPtr>& entities) { … }

void ExtractEntityFromTopLevelObject(const JSONObject& val,
                                     Vector<EntityPtr>& entities) { … }

// These values are persisted to logs. Entries should not be renumbered and
// numeric values should never be reused.
enum ExtractionStatus { … };

ExtractionStatus ExtractMetadata(const Element& root,
                                 Vector<EntityPtr>& entities) { … }

}  // namespace

WebPagePtr DocumentMetadataExtractor::Extract(const Document& document) { … }

}  // namespace blink
chromium/third_party/blink/renderer/modules/document_metadata/document_metadata_extractor.cc
