Author: Utkarsh Saxena Date: 2021-03-02T10:05:37+01:00 New Revision: 438b5bb05a429d697674088d400e4800c1108658
URL: https://github.com/llvm/llvm-project/commit/438b5bb05a429d697674088d400e4800c1108658 DIFF: https://github.com/llvm/llvm-project/commit/438b5bb05a429d697674088d400e4800c1108658.diff LOG: [clangd] Use ML Code completion ranking as default. This makes code completion use a Decision Forest based ranking algorithm to rank completion candidates. [Esitmated 6% accuracy boost]. This was previously hidden behind the flag --ranking-model=decision_forest. This patch makes it the default ranking algorithm. Note: this is a generic model, not specialized for any particular project. clangd does not collect or upload data to train code completion. Also treat Keywords separately as they are not recorded by the training set generator. Differential Revision: https://reviews.llvm.org/D96353 Added: Modified: clang-tools-extra/clangd/CodeComplete.h clang-tools-extra/clangd/Quality.cpp clang-tools-extra/clangd/unittests/CodeCompleteTests.cpp Removed: ################################################################################ diff --git a/clang-tools-extra/clangd/CodeComplete.h b/clang-tools-extra/clangd/CodeComplete.h index debf71d4117c..40a528caa939 100644 --- a/clang-tools-extra/clangd/CodeComplete.h +++ b/clang-tools-extra/clangd/CodeComplete.h @@ -133,7 +133,7 @@ struct CodeCompleteOptions { enum CodeCompletionRankingModel { Heuristics, DecisionForest, - } RankingModel = Heuristics; + } RankingModel = DecisionForest; /// Callback used to score a CompletionCandidate if DecisionForest ranking /// model is enabled. diff --git a/clang-tools-extra/clangd/Quality.cpp b/clang-tools-extra/clangd/Quality.cpp index b49392bc7d04..99421009c71c 100644 --- a/clang-tools-extra/clangd/Quality.cpp +++ b/clang-tools-extra/clangd/Quality.cpp @@ -580,12 +580,16 @@ evaluateDecisionForest(const SymbolQualitySignals &Quality, // multiplciative boost (like NameMatch). This allows us to weigh the // prediciton score and NameMatch appropriately. Scores.ExcludingName = pow(Base, Evaluate(E)); - // NeedsFixIts is not part of the DecisionForest as generating training - // data that needs fixits is not-feasible. + // Following cases are not part of the generated training dataset: + // - Symbols with `NeedsFixIts`. + // - Forbidden symbols. + // - Keywords: Dataset contains only macros and decls. if (Relevance.NeedsFixIts) Scores.ExcludingName *= 0.5; if (Relevance.Forbidden) Scores.ExcludingName *= 0; + if (Quality.Category == SymbolQualitySignals::Keyword) + Scores.ExcludingName *= 4; // NameMatch should be a multiplier on total score to support rescoring. Scores.Total = Relevance.NameMatch * Scores.ExcludingName; diff --git a/clang-tools-extra/clangd/unittests/CodeCompleteTests.cpp b/clang-tools-extra/clangd/unittests/CodeCompleteTests.cpp index b7a40179aa98..0ff1e83b7613 100644 --- a/clang-tools-extra/clangd/unittests/CodeCompleteTests.cpp +++ b/clang-tools-extra/clangd/unittests/CodeCompleteTests.cpp @@ -647,13 +647,13 @@ TEST(CompletionTest, ScopedWithFilter) { } TEST(CompletionTest, ReferencesAffectRanking) { - auto Results = completions("int main() { abs^ }", {ns("absl"), func("absb")}); - EXPECT_THAT(Results.Completions, - HasSubsequence(Named("absb"), Named("absl"))); - Results = completions("int main() { abs^ }", - {withReferences(10000, ns("absl")), func("absb")}); - EXPECT_THAT(Results.Completions, - HasSubsequence(Named("absl"), Named("absb"))); + EXPECT_THAT(completions("int main() { abs^ }", {func("absA"), func("absB")}) + .Completions, + HasSubsequence(Named("absA"), Named("absB"))); + EXPECT_THAT(completions("int main() { abs^ }", + {func("absA"), withReferences(1000, func("absB"))}) + .Completions, + HasSubsequence(Named("absB"), Named("absA"))); } TEST(CompletionTest, ContextWords) { _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
