You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

324 lines
12 KiB

/*
* Copyright (C) 2018 The Android Open Source Project
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "annotator/grammar/grammar-annotator.h"
#include "annotator/feature-processor.h"
#include "annotator/grammar/utils.h"
#include "annotator/types.h"
#include "utils/base/arena.h"
#include "utils/base/logging.h"
#include "utils/normalization.h"
#include "utils/optional.h"
#include "utils/utf8/unicodetext.h"
namespace libtextclassifier3 {
namespace {
// Retrieves all capturing nodes from a parse tree.
std::unordered_map<uint16, const grammar::ParseTree*> GetCapturingNodes(
const grammar::ParseTree* parse_tree) {
std::unordered_map<uint16, const grammar::ParseTree*> capturing_nodes;
for (const grammar::MappingNode* mapping_node :
grammar::SelectAllOfType<grammar::MappingNode>(
parse_tree, grammar::ParseTree::Type::kMapping)) {
capturing_nodes[mapping_node->id] = mapping_node;
}
return capturing_nodes;
}
// Computes the selection boundaries from a parse tree.
CodepointSpan MatchSelectionBoundaries(
const grammar::ParseTree* parse_tree,
const GrammarModel_::RuleClassificationResult* classification) {
if (classification->capturing_group() == nullptr) {
// Use full match as selection span.
return parse_tree->codepoint_span;
}
// Set information from capturing matches.
CodepointSpan span{kInvalidIndex, kInvalidIndex};
std::unordered_map<uint16, const grammar::ParseTree*> capturing_nodes =
GetCapturingNodes(parse_tree);
// Compute span boundaries.
for (int i = 0; i < classification->capturing_group()->size(); i++) {
auto it = capturing_nodes.find(i);
if (it == capturing_nodes.end()) {
// Capturing group is not active, skip.
continue;
}
const CapturingGroup* group = classification->capturing_group()->Get(i);
if (group->extend_selection()) {
if (span.first == kInvalidIndex) {
span = it->second->codepoint_span;
} else {
span.first = std::min(span.first, it->second->codepoint_span.first);
span.second = std::max(span.second, it->second->codepoint_span.second);
}
}
}
return span;
}
} // namespace
GrammarAnnotator::GrammarAnnotator(
const UniLib* unilib, const GrammarModel* model,
const MutableFlatbufferBuilder* entity_data_builder)
: unilib_(*unilib),
model_(model),
tokenizer_(BuildTokenizer(unilib, model->tokenizer_options())),
entity_data_builder_(entity_data_builder),
analyzer_(unilib, model->rules(), &tokenizer_) {}
// Filters out results that do not overlap with a reference span.
std::vector<grammar::Derivation> GrammarAnnotator::OverlappingDerivations(
const CodepointSpan& selection,
const std::vector<grammar::Derivation>& derivations,
const bool only_exact_overlap) const {
std::vector<grammar::Derivation> result;
for (const grammar::Derivation& derivation : derivations) {
// Discard matches that do not match the selection.
// Simple check.
if (!SpansOverlap(selection, derivation.parse_tree->codepoint_span)) {
continue;
}
// Compute exact selection boundaries (without assertions and
// non-capturing parts).
const CodepointSpan span = MatchSelectionBoundaries(
derivation.parse_tree,
model_->rule_classification_result()->Get(derivation.rule_id));
if (!SpansOverlap(selection, span) ||
(only_exact_overlap && span != selection)) {
continue;
}
result.push_back(derivation);
}
return result;
}
bool GrammarAnnotator::InstantiateAnnotatedSpanFromDerivation(
const grammar::TextContext& input_context,
const grammar::ParseTree* parse_tree,
const GrammarModel_::RuleClassificationResult* interpretation,
AnnotatedSpan* result) const {
result->span = MatchSelectionBoundaries(parse_tree, interpretation);
ClassificationResult classification;
if (!InstantiateClassificationFromDerivation(
input_context, parse_tree, interpretation, &classification)) {
return false;
}
result->classification.push_back(classification);
return true;
}
// Instantiates a classification result from a rule match.
bool GrammarAnnotator::InstantiateClassificationFromDerivation(
const grammar::TextContext& input_context,
const grammar::ParseTree* parse_tree,
const GrammarModel_::RuleClassificationResult* interpretation,
ClassificationResult* classification) const {
classification->collection = interpretation->collection_name()->str();
classification->score = interpretation->target_classification_score();
classification->priority_score = interpretation->priority_score();
// Assemble entity data.
if (entity_data_builder_ == nullptr) {
return true;
}
std::unique_ptr<MutableFlatbuffer> entity_data =
entity_data_builder_->NewRoot();
if (interpretation->serialized_entity_data() != nullptr) {
entity_data->MergeFromSerializedFlatbuffer(
StringPiece(interpretation->serialized_entity_data()->data(),
interpretation->serialized_entity_data()->size()));
}
if (interpretation->entity_data() != nullptr) {
entity_data->MergeFrom(reinterpret_cast<const flatbuffers::Table*>(
interpretation->entity_data()));
}
// Populate entity data from the capturing matches.
if (interpretation->capturing_group() != nullptr) {
// Gather active capturing matches.
std::unordered_map<uint16, const grammar::ParseTree*> capturing_nodes =
GetCapturingNodes(parse_tree);
for (int i = 0; i < interpretation->capturing_group()->size(); i++) {
auto it = capturing_nodes.find(i);
if (it == capturing_nodes.end()) {
// Capturing group is not active, skip.
continue;
}
const CapturingGroup* group = interpretation->capturing_group()->Get(i);
// Add static entity data.
if (group->serialized_entity_data() != nullptr) {
entity_data->MergeFromSerializedFlatbuffer(
StringPiece(interpretation->serialized_entity_data()->data(),
interpretation->serialized_entity_data()->size()));
}
// Set entity field from captured text.
if (group->entity_field_path() != nullptr) {
const grammar::ParseTree* capturing_match = it->second;
UnicodeText match_text =
input_context.Span(capturing_match->codepoint_span);
if (group->normalization_options() != nullptr) {
match_text = NormalizeText(unilib_, group->normalization_options(),
match_text);
}
if (!entity_data->ParseAndSet(group->entity_field_path(),
match_text.ToUTF8String())) {
TC3_LOG(ERROR) << "Could not set entity data from capturing match.";
return false;
}
}
}
}
if (entity_data && entity_data->HasExplicitlySetFields()) {
classification->serialized_entity_data = entity_data->Serialize();
}
return true;
}
bool GrammarAnnotator::Annotate(const std::vector<Locale>& locales,
const UnicodeText& text,
std::vector<AnnotatedSpan>* result) const {
grammar::TextContext input_context =
analyzer_.BuildTextContextForInput(text, locales);
UnsafeArena arena(/*block_size=*/16 << 10);
for (const grammar::Derivation& derivation : ValidDeduplicatedDerivations(
analyzer_.parser().Parse(input_context, &arena))) {
const GrammarModel_::RuleClassificationResult* interpretation =
model_->rule_classification_result()->Get(derivation.rule_id);
if ((interpretation->enabled_modes() & ModeFlag_ANNOTATION) == 0) {
continue;
}
result->emplace_back();
if (!InstantiateAnnotatedSpanFromDerivation(
input_context, derivation.parse_tree, interpretation,
&result->back())) {
return false;
}
}
return true;
}
bool GrammarAnnotator::SuggestSelection(const std::vector<Locale>& locales,
const UnicodeText& text,
const CodepointSpan& selection,
AnnotatedSpan* result) const {
if (!selection.IsValid() || selection.IsEmpty()) {
return false;
}
grammar::TextContext input_context =
analyzer_.BuildTextContextForInput(text, locales);
UnsafeArena arena(/*block_size=*/16 << 10);
const GrammarModel_::RuleClassificationResult* best_interpretation = nullptr;
const grammar::ParseTree* best_match = nullptr;
for (const grammar::Derivation& derivation :
ValidDeduplicatedDerivations(OverlappingDerivations(
selection, analyzer_.parser().Parse(input_context, &arena),
/*only_exact_overlap=*/false))) {
const GrammarModel_::RuleClassificationResult* interpretation =
model_->rule_classification_result()->Get(derivation.rule_id);
if ((interpretation->enabled_modes() & ModeFlag_SELECTION) == 0) {
continue;
}
if (best_interpretation == nullptr ||
interpretation->priority_score() >
best_interpretation->priority_score()) {
best_interpretation = interpretation;
best_match = derivation.parse_tree;
}
}
if (best_interpretation == nullptr) {
return false;
}
return InstantiateAnnotatedSpanFromDerivation(input_context, best_match,
best_interpretation, result);
}
bool GrammarAnnotator::ClassifyText(
const std::vector<Locale>& locales, const UnicodeText& text,
const CodepointSpan& selection,
ClassificationResult* classification_result) const {
if (!selection.IsValid() || selection.IsEmpty()) {
// Nothing to do.
return false;
}
grammar::TextContext input_context =
analyzer_.BuildTextContextForInput(text, locales);
if (const TokenSpan context_span = CodepointSpanToTokenSpan(
input_context.tokens, selection,
/*snap_boundaries_to_containing_tokens=*/true);
context_span.IsValid()) {
if (model_->context_left_num_tokens() != kInvalidIndex) {
input_context.context_span.first =
std::max(0, context_span.first - model_->context_left_num_tokens());
}
if (model_->context_right_num_tokens() != kInvalidIndex) {
input_context.context_span.second =
std::min(static_cast<int>(input_context.tokens.size()),
context_span.second + model_->context_right_num_tokens());
}
}
UnsafeArena arena(/*block_size=*/16 << 10);
const GrammarModel_::RuleClassificationResult* best_interpretation = nullptr;
const grammar::ParseTree* best_match = nullptr;
for (const grammar::Derivation& derivation :
ValidDeduplicatedDerivations(OverlappingDerivations(
selection, analyzer_.parser().Parse(input_context, &arena),
/*only_exact_overlap=*/true))) {
const GrammarModel_::RuleClassificationResult* interpretation =
model_->rule_classification_result()->Get(derivation.rule_id);
if ((interpretation->enabled_modes() & ModeFlag_CLASSIFICATION) == 0) {
continue;
}
if (best_interpretation == nullptr ||
interpretation->priority_score() >
best_interpretation->priority_score()) {
best_interpretation = interpretation;
best_match = derivation.parse_tree;
}
}
if (best_interpretation == nullptr) {
return false;
}
return InstantiateClassificationFromDerivation(
input_context, best_match, best_interpretation, classification_result);
}
} // namespace libtextclassifier3