annotator_impl.cc | Explore in Territory

// Copyright 2023 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

#ifdef UNSAFE_BUFFERS_BUILD
// TODO(crbug.com/40285824): Remove this and convert code to safer constructs.
#pragma allow_unsafe_buffers
#endif

#include "components/browsing_topics/annotator_impl.h"

#include <vector>

#include "base/barrier_closure.h"
#include "base/containers/contains.h"
#include "base/dcheck_is_on.h"
#include "base/files/file_util.h"
#include "base/ranges/algorithm.h"
#include "base/strings/string_number_conversions.h"
#include "base/strings/string_util.h"
#include "base/task/sequenced_task_runner.h"
#include "components/optimization_guide/core/optimization_guide_model_provider.h"
#include "components/optimization_guide/proto/models.pb.h"
#include "components/optimization_guide/proto/page_topics_model_metadata.pb.h"
#include "components/optimization_guide/proto/page_topics_override_list.pb.h"
#include "third_party/abseil-cpp/absl/strings/ascii.h"
#include "third_party/blink/public/common/features.h"
#include "third_party/zlib/google/compression_utils.h"

namespace browsing_topics {

namespace {

// The ID of the NONE category in the taxonomy. This node always exists.
// Semantically, the none category is attached to data for which we can say
// with certainty that no single label in the taxonomy is appropriate.
const int32_t kNoneCategoryId = …;

// The |kMeaninglessPrefixV2MinVersion| needed to support meaningless prefix v2.
// This should be compared with the version provided the model metadata.
const int32_t kMeaninglessPrefixV2MinVersion = …;

const base::FilePath::CharType kOverrideListBasePath[] = …);

// The result of an override list file load attempt. These values are logged to
// UMA histograms, do not change or reorder values. Make sure to update
// |BrowsingTopicsOverrideListFileLoadResult| in
// //tools/metrics/histograms/enums.xml.
enum class OverrideListFileLoadResult { … };

void RecordOverrideListFileLoadResult(OverrideListFileLoadResult result) { … }

std::optional<std::unordered_map<std::string, std::vector<int32_t>>>
LoadOverrideListFromFile(const base::FilePath& path) { … }

// Returns the length of the leading meaningless prefix of a host name as
// defined for the Topics Model.
//
// The full list of meaningless prefixes are:
//   ^(www[0-9]*|web|ftp|wap|home)$
//   ^(m|mobile|amp|w)$
int MeaninglessPrefixLength(const std::string& host) { … }

bool IsModelTaxonomyVersionSupported(int model_taxonomy_version) { … }

}  // namespace

AnnotatorImpl::AnnotatorImpl(
    optimization_guide::OptimizationGuideModelProvider* model_provider,
    scoped_refptr<base::SequencedTaskRunner> background_task_runner,
    const std::optional<optimization_guide::proto::Any>& model_metadata)
    : … { … }
AnnotatorImpl::~AnnotatorImpl() = default;

void AnnotatorImpl::NotifyWhenModelAvailable(base::OnceClosure callback) { … }

std::optional<optimization_guide::ModelInfo>
AnnotatorImpl::GetBrowsingTopicsModelInfo() const { … }

void AnnotatorImpl::BatchAnnotate(BatchAnnotationCallback callback,
                                  const std::vector<std::string>& inputs) { … }

void AnnotatorImpl::OnOverrideListLoadAttemptDone(
    BatchAnnotationCallback callback,
    const std::vector<std::string>& inputs,
    std::optional<std::unordered_map<std::string, std::vector<int32_t>>>
        override_list) { … }

void AnnotatorImpl::StartBatchAnnotate(BatchAnnotationCallback callback,
                                       const std::vector<std::string>& inputs) { … }

void AnnotatorImpl::OnBatchComplete(
    BatchAnnotationCallback callback,
    std::unique_ptr<std::vector<Annotation>> annotations_ptr) { … }

std::string AnnotatorImpl::PreprocessHost(const std::string& host) const { … }

void AnnotatorImpl::AnnotateSingleInput(
    base::OnceClosure single_input_done_signal,
    Annotation* annotation) { … }

void AnnotatorImpl::PostprocessCategoriesToBatchAnnotationResult(
    base::OnceClosure single_input_done_signal,
    Annotation* annotation,
    const std::optional<std::vector<tflite::task::core::Category>>& output) { … }

std::optional<std::vector<int32_t>>
AnnotatorImpl::ExtractCategoriesFromModelOutput(
    const std::vector<tflite::task::core::Category>& model_output) const { … }

void AnnotatorImpl::UnloadModel() { … }

void AnnotatorImpl::OnModelUpdated(
    optimization_guide::proto::OptimizationTarget optimization_target,
    base::optional_ref<const optimization_guide::ModelInfo> model_info) { … }

}  // namespace browsing_topics
chromium/components/browsing_topics/annotator_impl.cc
