kbobyrev updated this revision to Diff 163680.
kbobyrev marked 17 inline comments as done.
kbobyrev added a comment.
Address a round of comments, refactor code.
https://reviews.llvm.org/D51481
Files:
clang-tools-extra/;
clang-tools-extra/clangd/CMakeLists.txt
clang-tools-extra/clangd/index/dex/DexIndex.cpp
clang-tools-extra/clangd/index/dex/DexIndex.h
clang-tools-extra/clangd/index/dex/Token.cpp
clang-tools-extra/clangd/index/dex/Token.h
clang-tools-extra/clangd/tool/ClangdMain.cpp
clang-tools-extra/unittests/clangd/DexIndexTests.cpp
clang-tools-extra/unittests/clangd/TestFS.cpp
clang-tools-extra/unittests/clangd/TestFS.h
Index: clang-tools-extra/unittests/clangd/TestFS.h
===================================================================
--- clang-tools-extra/unittests/clangd/TestFS.h
+++ clang-tools-extra/unittests/clangd/TestFS.h
@@ -59,7 +59,7 @@
};
// Returns an absolute (fake) test directory for this OS.
-const char *testRoot();
+std::string testRoot();
// Returns a suitable absolute path for this OS.
std::string testPath(PathRef File);
Index: clang-tools-extra/unittests/clangd/TestFS.cpp
===================================================================
--- clang-tools-extra/unittests/clangd/TestFS.cpp
+++ clang-tools-extra/unittests/clangd/TestFS.cpp
@@ -64,7 +64,7 @@
FileName, std::move(CommandLine), "")};
}
-const char *testRoot() {
+std::string testRoot() {
#ifdef _WIN32
return "C:\\clangd-test";
#else
Index: clang-tools-extra/unittests/clangd/DexIndexTests.cpp
===================================================================
--- clang-tools-extra/unittests/clangd/DexIndexTests.cpp
+++ clang-tools-extra/unittests/clangd/DexIndexTests.cpp
@@ -7,6 +7,8 @@
//
//===----------------------------------------------------------------------===//
+#include "FuzzyMatch.h"
+#include "TestFS.h"
#include "TestIndex.h"
#include "index/Index.h"
#include "index/Merge.h"
@@ -29,6 +31,8 @@
namespace dex {
namespace {
+std::vector<std::string> URISchemes = {"unittest"};
+
std::vector<DocID> consumeIDs(Iterator &It) {
auto IDAndScore = consume(It);
std::vector<DocID> IDs(IDAndScore.size());
@@ -325,14 +329,33 @@
}
testing::Matcher<std::vector<Token>>
-trigramsAre(std::initializer_list<std::string> Trigrams) {
+tokensAre(std::initializer_list<std::string> Strings, Token::Kind Kind) {
std::vector<Token> Tokens;
- for (const auto &Symbols : Trigrams) {
- Tokens.push_back(Token(Token::Kind::Trigram, Symbols));
+ for (const auto &TokenData : Strings) {
+ Tokens.push_back(Token(Kind, TokenData));
}
return testing::UnorderedElementsAreArray(Tokens);
}
+testing::Matcher<std::vector<Token>>
+trigramsAre(std::initializer_list<std::string> Trigrams) {
+ return tokensAre(Trigrams, Token::Kind::Trigram);
+}
+
+testing::Matcher<std::vector<Token>>
+pathsAre(std::initializer_list<std::string> Paths) {
+ return tokensAre(Paths, Token::Kind::PathURI);
+}
+
+testing::Matcher<std::vector<Token>>
+proximityPathsAre(std::initializer_list<std::string> ProximityPaths) {
+ std::vector<Token> Result;
+ for (const auto &Path : ProximityPaths) {
+ Result.emplace_back(Token::Kind::PathURI, Path);
+ }
+ return testing::UnorderedElementsAreArray(Result);
+}
+
TEST(DexIndexTrigrams, IdentifierTrigrams) {
EXPECT_THAT(generateIdentifierTrigrams("X86"),
trigramsAre({"x86", "x$$", "x8$"}));
@@ -407,8 +430,27 @@
"hij", "ijk", "jkl", "klm"}));
}
+TEST(DexSearchTokens, SymbolPath) {
+ EXPECT_THAT(generateProximityPathURIs(
+ "unittest:///clang-tools-extra/clangd/index/Token.h"),
+ pathsAre({"unittest:///clang-tools-extra/clangd/index/Token.h",
+ "unittest:///clang-tools-extra/clangd/index",
+ "unittest:///clang-tools-extra/clangd",
+ "unittest:///clang-tools-extra", "unittest:///"}));
+}
+
+TEST(DexSearchTokens, QueryProximityDistances) {
+ EXPECT_THAT(generateQueryProximityPathURIs(testRoot() + "/a/b/c/d/e/f/g.h",
+ URISchemes),
+ proximityPathsAre({{"unittest:///a/b/c/d/e/f/g.h"},
+ {"unittest:///a/b/c/d/e/f"},
+ {"unittest:///a/b/c/d/e"},
+ {"unittest:///a/b/c/d"},
+ {"unittest:///a/b/c"}}));
+}
+
TEST(DexIndex, Lookup) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"ns::abc", "ns::xyz"}));
EXPECT_THAT(lookup(I, SymbolID("ns::abc")), UnorderedElementsAre("ns::abc"));
EXPECT_THAT(lookup(I, {SymbolID("ns::abc"), SymbolID("ns::xyz")}),
@@ -419,7 +461,7 @@
}
TEST(DexIndex, FuzzyFind) {
- DexIndex Index;
+ DexIndex Index(URISchemes);
Index.build(generateSymbols({"ns::ABC", "ns::BCD", "::ABC", "ns::nested::ABC",
"other::ABC", "other::A"}));
FuzzyFindRequest Req;
@@ -442,7 +484,7 @@
}
TEST(DexIndexTest, FuzzyMatchQ) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(
generateSymbols({"LaughingOutLoud", "LionPopulation", "LittleOldLady"}));
FuzzyFindRequest Req;
@@ -453,7 +495,7 @@
}
TEST(DexIndexTest, DexIndexSymbolsRecycled) {
- DexIndex I;
+ DexIndex I(URISchemes);
std::weak_ptr<SlabAndPointers> Symbols;
I.build(generateNumSymbols(0, 10, &Symbols));
FuzzyFindRequest Req;
@@ -482,14 +524,14 @@
FuzzyFindRequest Req;
Req.Query = "7";
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(std::move(Symbols));
auto Matches = match(I, Req);
EXPECT_EQ(Matches.size(), 4u);
}
TEST(DexIndexTest, DexIndexLimitedNumMatches) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateNumSymbols(0, 100));
FuzzyFindRequest Req;
Req.Query = "5";
@@ -501,7 +543,7 @@
}
TEST(DexIndexTest, FuzzyMatch) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(
generateSymbols({"LaughingOutLoud", "LionPopulation", "LittleOldLady"}));
FuzzyFindRequest Req;
@@ -512,60 +554,60 @@
}
TEST(DexIndexTest, MatchQualifiedNamesWithoutSpecificScope) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"a::y1", "b::y2", "y3"}));
FuzzyFindRequest Req;
Req.Query = "y";
EXPECT_THAT(match(I, Req), UnorderedElementsAre("a::y1", "b::y2", "y3"));
}
TEST(DexIndexTest, MatchQualifiedNamesWithGlobalScope) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"a::y1", "b::y2", "y3"}));
FuzzyFindRequest Req;
Req.Query = "y";
Req.Scopes = {""};
EXPECT_THAT(match(I, Req), UnorderedElementsAre("y3"));
}
TEST(DexIndexTest, MatchQualifiedNamesWithOneScope) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"a::y1", "a::y2", "a::x", "b::y2", "y3"}));
FuzzyFindRequest Req;
Req.Query = "y";
Req.Scopes = {"a::"};
EXPECT_THAT(match(I, Req), UnorderedElementsAre("a::y1", "a::y2"));
}
TEST(DexIndexTest, MatchQualifiedNamesWithMultipleScopes) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"a::y1", "a::y2", "a::x", "b::y3", "y3"}));
FuzzyFindRequest Req;
Req.Query = "y";
Req.Scopes = {"a::", "b::"};
EXPECT_THAT(match(I, Req), UnorderedElementsAre("a::y1", "a::y2", "b::y3"));
}
TEST(DexIndexTest, NoMatchNestedScopes) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"a::y1", "a::b::y2"}));
FuzzyFindRequest Req;
Req.Query = "y";
Req.Scopes = {"a::"};
EXPECT_THAT(match(I, Req), UnorderedElementsAre("a::y1"));
}
TEST(DexIndexTest, IgnoreCases) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"ns::ABC", "ns::abc"}));
FuzzyFindRequest Req;
Req.Query = "AB";
Req.Scopes = {"ns::"};
EXPECT_THAT(match(I, Req), UnorderedElementsAre("ns::ABC", "ns::abc"));
}
TEST(DexIndexTest, Lookup) {
- DexIndex I;
+ DexIndex I(URISchemes);
I.build(generateSymbols({"ns::abc", "ns::xyz"}));
EXPECT_THAT(lookup(I, SymbolID("ns::abc")), UnorderedElementsAre("ns::abc"));
EXPECT_THAT(lookup(I, {SymbolID("ns::abc"), SymbolID("ns::xyz")}),
@@ -575,6 +617,58 @@
EXPECT_THAT(lookup(I, SymbolID("ns::nonono")), UnorderedElementsAre());
}
+TEST(DexIndexTest, ProximityPathsBoosting) {
+ DexIndex I(URISchemes);
+
+ // All strings should be stored in local variables to ensure their lifetime is
+ // the same as Slab's.
+ auto RootURI = URI::create(testRoot() + "/file.h", "unittest");
+ assert(static_cast<bool>(RootURI));
+ auto CloseFileURI = URI::create(testRoot() + "/a/b/c/d/e/file.h", "unittest");
+ assert(static_cast<bool>(CloseFileURI));
+
+ // Populate Symbol Slab
+ std::vector<const Symbol *> Slab;
+ // Symbol in the root will perfectly match the request, but its canonical
+ // declaration is in the root directory.
+ Symbol RootSymbol;
+ std::string RootSymbolName = "abc";
+ RootSymbol.Name = RootSymbolName;
+ std::string RootFilename = RootURI->toString();
+ RootSymbol.CanonicalDeclaration.FileURI = RootFilename;
+ Slab.push_back(&RootSymbol);
+
+ // Another symbol has much lower Fuzzy Matching Score, but it is very close to
+ // the directory the request would come from.
+ Symbol CloseSymbol;
+ std::string CloseSymbolName = "AbsolutelyBizzareCoincidenceMatch";
+ CloseSymbol.Name = CloseSymbolName;
+ std::string CloseFileName = CloseFileURI->toString();
+ CloseSymbol.CanonicalDeclaration.FileURI = CloseFileName;
+
+ FuzzyFindRequest Req;
+ Req.Query = "abc";
+ Req.MaxCandidateCount = 1;
+ // FuzzyFind request comes from the file which is far from the root.
+ Req.ProximityPaths.push_back(testRoot() + "/a/b/c/d/e/f/file.h");
+
+ FuzzyMatcher Matcher(Req.Query);
+ EXPECT_TRUE(Matcher.match(RootSymbolName) > Matcher.match(CloseSymbolName));
+
+ // Insert duplicates so that only the boosted symbols end up in the scoring
+ // stage.
+ // FIXME(kbobyrev): 10 is the number of items passed to the last stage which
+ // involves fuzzy scoring. Would having access to that information be better
+ // here?
+ for (size_t Index = 0; Index < 10; ++Index)
+ Slab.push_back(&CloseSymbol);
+
+ I.build(std::make_shared<std::vector<const Symbol *>>(Slab));
+
+ EXPECT_THAT(match(I, Req),
+ UnorderedElementsAre("AbsolutelyBizzareCoincidenceMatch"));
+}
+
} // namespace
} // namespace dex
} // namespace clangd
Index: clang-tools-extra/clangd/tool/ClangdMain.cpp
===================================================================
--- clang-tools-extra/clangd/tool/ClangdMain.cpp
+++ clang-tools-extra/clangd/tool/ClangdMain.cpp
@@ -42,7 +42,9 @@
// Build an in-memory static index for global symbols from a YAML-format file.
// The size of global symbols should be relatively small, so that all symbols
// can be managed in memory.
-std::unique_ptr<SymbolIndex> buildStaticIndex(llvm::StringRef YamlSymbolFile) {
+std::unique_ptr<SymbolIndex>
+buildStaticIndex(llvm::StringRef YamlSymbolFile,
+ std::vector<std::string> &URISchemes) {
auto Buffer = llvm::MemoryBuffer::getFile(YamlSymbolFile);
if (!Buffer) {
llvm::errs() << "Can't open " << YamlSymbolFile << "\n";
@@ -53,9 +55,10 @@
for (auto Sym : Slab)
SymsBuilder.insert(Sym);
- return UseDex ? dex::DexIndex::build(std::move(SymsBuilder).build())
- : MemIndex::build(std::move(SymsBuilder).build(),
- SymbolOccurrenceSlab::createEmpty());
+ return UseDex
+ ? dex::DexIndex::build(std::move(SymsBuilder).build(), URISchemes)
+ : MemIndex::build(std::move(SymsBuilder).build(),
+ SymbolOccurrenceSlab::createEmpty());
}
} // namespace
@@ -299,7 +302,7 @@
Opts.BuildDynamicSymbolIndex = EnableIndex;
std::unique_ptr<SymbolIndex> StaticIdx;
if (EnableIndex && !YamlSymbolFile.empty()) {
- StaticIdx = buildStaticIndex(YamlSymbolFile);
+ StaticIdx = buildStaticIndex(YamlSymbolFile, Opts.URISchemes);
Opts.StaticIndex = StaticIdx.get();
}
Opts.AsyncThreadsCount = WorkerThreadsCount;
Index: clang-tools-extra/clangd/index/dex/Token.h
===================================================================
--- clang-tools-extra/clangd/index/dex/Token.h
+++ clang-tools-extra/clangd/index/dex/Token.h
@@ -41,22 +41,32 @@
/// Kind specifies Token type which defines semantics for the internal
/// representation. Each Kind has different representation stored in Data
/// field.
+ // FIXME(kbobyrev): Storing Data hash would be more efficient than storing raw
+ // strings. For example, PathURI store URIs of each directory and its parents,
+ // which induces a lot of overhead because these paths tend to be long and
+ // each parent directory is a prefix.
enum class Kind {
/// Represents trigram used for fuzzy search of unqualified symbol names.
///
/// Data contains 3 bytes with trigram contents.
Trigram,
/// Scope primitives, e.g. "symbol belongs to namespace foo::bar".
///
- /// Data stroes full scope name , e.g. "foo::bar::baz::" or "" (for global
+ /// Data stroes full scope name, e.g. "foo::bar::baz::" or "" (for global
/// scope).
Scope,
+ /// Path URI to symbol declaration.
+ ///
+ /// Data stores path URI of file and each parent directory of symbol
+ /// declaration location.
+ ///
+ /// Example: "file:///path/to/clang-tools-extra/clangd/index/SymbolIndex.h"
+ /// and all its parents.
+ PathURI,
/// Internal Token type for invalid/special tokens, e.g. empty tokens for
/// llvm::DenseMap.
Sentinel,
/// FIXME(kbobyrev): Add other Token Kinds
- /// * Path with full or relative path to the directory in which symbol is
- /// defined
/// * Type with qualified type name or its USR
};
@@ -81,6 +91,16 @@
}
};
+/// Returns Search Token for a number of parent directories of given Path.
+/// Should be used within the index build process.
+std::vector<Token> generateProximityPathURIs(llvm::StringRef Path);
+
+/// Returns Search Token and its distance from the absolute path for the first
+/// URI scheme allows valid conversions from AbsolutePath.
+std::vector<Token>
+generateQueryProximityPathURIs(llvm::StringRef AbsolutePath,
+ llvm::ArrayRef<std::string> URISchemes);
+
} // namespace dex
} // namespace clangd
} // namespace clang
Index: clang-tools-extra/clangd/index/dex/Token.cpp
===================================================================
--- /dev/null
+++ clang-tools-extra/clangd/index/dex/Token.cpp
@@ -0,0 +1,70 @@
+//===--- Token.cpp - Symbol Search primitive --------------------*- C++ -*-===//
+// The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#include "Token.h"
+#include "URI.h"
+#include "llvm/Support/Path.h"
+#include <cassert>
+
+namespace clang {
+namespace clangd {
+namespace dex {
+
+std::vector<Token> generateProximityPathURIs(llvm::StringRef URIPath) {
+ std::vector<Token> Result;
+ // Don't generate any tokens for empty paths.
+ if (URIPath.empty())
+ return Result;
+ auto ParsedURI = URI::parse(URIPath);
+ assert(static_cast<bool>(ParsedURI) &&
+ "Non-empty argument of generateProximityPathURIs() should be a valid "
+ "URI.");
+ const auto Scheme = ParsedURI->scheme();
+ const auto Authority = ParsedURI->authority();
+ StringRef Path = ParsedURI->body();
+ // FIXME(kbobyrev): Currently, this is a heuristic which defines the maximum
+ // size of resulting vector. Some projects might want to have higher limit if
+ // the file hierarchy is deeper. For the generic case, it would be useful to
+ // calculate Limit in the index build stage by calculating the maximum depth
+ // of the project source tree at runtime.
+ size_t Limit = 5;
+ while (!Path.empty() && Limit--) {
+ // FIXME(kbobyrev): Parsing and encoding path to URIs is not necessary and
+ // could be optimized.
+ URI TokenURI(Scheme, Authority, Path);
+ Result.emplace_back(Token::Kind::PathURI, TokenURI.toString());
+ Path = llvm::sys::path::parent_path(Path, llvm::sys::path::Style::posix);
+ }
+ return Result;
+}
+
+std::vector<Token>
+generateQueryProximityPathURIs(llvm::StringRef AbsolutePath,
+ llvm::ArrayRef<std::string> URISchemes) {
+ std::vector<Token> Result;
+ for (const auto Scheme : URISchemes) {
+ auto URI = URI::create(AbsolutePath, Scheme);
+ // For some paths, conversion to different URI schemes is impossible. These
+ // should be just skipped.
+ if (!static_cast<bool>(URI)) {
+ // Ignore the error.
+ llvm::consumeError(URI.takeError());
+ continue;
+ }
+ for (const auto ProximityPath : generateProximityPathURIs(URI->toString()))
+ Result.emplace_back(ProximityPath);
+ // As soon as one valid URI scheme for given absolute path is found, the
+ // token generation process should be stopped.
+ break;
+ }
+ return Result;
+}
+
+} // namespace dex
+} // namespace clangd
+} // namespace clang
Index: clang-tools-extra/clangd/index/dex/DexIndex.h
===================================================================
--- clang-tools-extra/clangd/index/dex/DexIndex.h
+++ clang-tools-extra/clangd/index/dex/DexIndex.h
@@ -39,12 +39,21 @@
// index on disk and then load it if available.
class DexIndex : public SymbolIndex {
public:
+ /// URISchemes are used to convert proximity paths from FuzzyFindRequest to
+ /// the desired URIs. URISchemes should contain a list of schemes sorted by
+ /// the priority, because each proximity path from FuzzyFindRequest will be
+ /// converted to the first scheme from URISchemes list which is valid for that
+ /// path.
+ explicit DexIndex(llvm::ArrayRef<std::string> URISchemes)
+ : URISchemes(URISchemes) {}
+
/// \brief (Re-)Build index for `Symbols`. All symbol pointers must remain
/// accessible as long as `Symbols` is kept alive.
- void build(std::shared_ptr<std::vector<const Symbol *>> Syms);
+ void build(std::shared_ptr<std::vector<const Symbol *>> Symbols);
/// \brief Build index from a symbol slab.
- static std::unique_ptr<SymbolIndex> build(SymbolSlab Slab);
+ static std::unique_ptr<SymbolIndex>
+ build(SymbolSlab Slab, llvm::ArrayRef<std::string> URISchemes);
bool
fuzzyFind(const FuzzyFindRequest &Req,
@@ -60,23 +69,24 @@
size_t estimateMemoryUsage() const override;
private:
-
mutable std::mutex Mutex;
std::shared_ptr<std::vector<const Symbol *>> Symbols /*GUARDED_BY(Mutex)*/;
llvm::DenseMap<SymbolID, const Symbol *> LookupTable /*GUARDED_BY(Mutex)*/;
- llvm::DenseMap<const Symbol *, float> SymbolQuality /*GUARDED_BY(Mutex)*/;
+ std::vector<float> SymbolQuality /*GUARDED_BY(Mutex)*/;
// Inverted index is a mapping from the search token to the posting list,
// which contains all items which can be characterized by such search token.
// For example, if the search token is scope "std::", the corresponding
// posting list would contain all indices of symbols defined in namespace std.
// Inverted index is used to retrieve posting lists which are processed during
// the fuzzyFind process.
llvm::DenseMap<Token, PostingList> InvertedIndex /*GUARDED_BY(Mutex)*/;
+
+ const std::vector<std::string> URISchemes;
};
} // namespace dex
} // namespace clangd
} // namespace clang
-#endif
+#endif // LLVM_CLANG_TOOLS_EXTRA_CLANGD_INDEX_DEX_DEXINDEX_H
Index: clang-tools-extra/clangd/index/dex/DexIndex.cpp
===================================================================
--- clang-tools-extra/clangd/index/dex/DexIndex.cpp
+++ clang-tools-extra/clangd/index/dex/DexIndex.cpp
@@ -8,6 +8,7 @@
//===----------------------------------------------------------------------===//
#include "DexIndex.h"
+#include "../../FileDistance.h"
#include "../../FuzzyMatch.h"
#include "../../Logger.h"
#include <algorithm>
@@ -26,29 +27,31 @@
// Returns the tokens which are given symbols's characteristics. For example,
// trigrams and scopes.
// FIXME(kbobyrev): Support more token types:
-// * Path proximity
// * Types
std::vector<Token> generateSearchTokens(const Symbol &Sym) {
std::vector<Token> Result = generateIdentifierTrigrams(Sym.Name);
Result.emplace_back(Token::Kind::Scope, Sym.Scope);
+ for (const auto &PathToken :
+ generateProximityPathURIs(Sym.CanonicalDeclaration.FileURI))
+ Result.push_back(PathToken);
return Result;
}
} // namespace
void DexIndex::build(std::shared_ptr<std::vector<const Symbol *>> Syms) {
llvm::DenseMap<SymbolID, const Symbol *> TempLookupTable;
- llvm::DenseMap<const Symbol *, float> TempSymbolQuality;
+ llvm::DenseMap<const Symbol *, float> SymbolQualityLookup;
for (const Symbol *Sym : *Syms) {
TempLookupTable[Sym->ID] = Sym;
- TempSymbolQuality[Sym] = quality(*Sym);
+ SymbolQualityLookup[Sym] = quality(*Sym);
}
// Symbols are sorted by symbol qualities so that items in the posting lists
// are stored in the descending order of symbol quality.
std::sort(begin(*Syms), end(*Syms),
[&](const Symbol *LHS, const Symbol *RHS) {
- return TempSymbolQuality[LHS] > TempSymbolQuality[RHS];
+ return SymbolQualityLookup[LHS] > SymbolQualityLookup[RHS];
});
llvm::DenseMap<Token, PostingList> TempInvertedIndex;
// Populate TempInvertedIndex with posting lists for index symbols.
@@ -58,6 +61,10 @@
TempInvertedIndex[Token].push_back(SymbolRank);
}
+ std::vector<float> TempSymbolQuality(Syms->size());
+ for (size_t I = 0; I < Syms->size(); ++I)
+ TempSymbolQuality[I] = SymbolQualityLookup[(*Syms)[I]];
+
{
std::lock_guard<std::mutex> Lock(Mutex);
@@ -72,8 +79,9 @@
estimateMemoryUsage());
}
-std::unique_ptr<SymbolIndex> DexIndex::build(SymbolSlab Slab) {
- auto Idx = llvm::make_unique<DexIndex>();
+std::unique_ptr<SymbolIndex>
+DexIndex::build(SymbolSlab Slab, llvm::ArrayRef<std::string> URISchemes) {
+ auto Idx = llvm::make_unique<DexIndex>(URISchemes);
Idx->build(getSymbolsFromSlab(std::move(Slab)));
return std::move(Idx);
}
@@ -117,6 +125,33 @@
if (!ScopeIterators.empty())
TopLevelChildren.push_back(createOr(move(ScopeIterators)));
+ // Add proximity paths boosting.
+ std::vector<std::unique_ptr<Iterator>> BoostingIterators;
+ // Try to build BOOST iterator for each Proximity Path provided by
+ // FuzzyFindRequest. Boosting factor should depend on the distance to the
+ // Proximity Path: the closer processed path is, the higher boosting factor.
+ for (const auto &ProximityPath : Req.ProximityPaths) {
+ const auto ProximityPathURIs =
+ generateQueryProximityPathURIs(ProximityPath, URISchemes);
+ for (size_t Distance = 0; Distance < ProximityPathURIs.size();
+ ++Distance) {
+ const auto It = InvertedIndex.find(ProximityPathURIs[Distance]);
+ if (It != InvertedIndex.end()) {
+ // FIXME(kbobyrev): Append LIMIT on top of every BOOST iterator.
+ // Each parent directory is boosted by PARENTS_TO_BOOST + 1 - Level.
+ float BoostFactor =
+ std::exp(Distance * 0.4f / FileDistanceOptions().UpCost);
+ BoostingIterators.push_back(
+ createBoost(create(It->second), BoostFactor));
+ }
+ }
+ }
+ // Boosting iterators do not actually filter symbols. In order to preserve
+ // the validity of resulting query, TRUE iterator should be added along
+ // BOOSTs.
+ BoostingIterators.push_back(createTrue(Symbols->size()));
+ TopLevelChildren.push_back(createOr(move(BoostingIterators)));
+
// Use TRUE iterator if both trigrams and scopes from the query are not
// present in the symbol index.
auto QueryIterator = TopLevelChildren.empty()
@@ -127,23 +162,35 @@
// FIXME(kbobyrev): Pre-scoring retrieval threshold should be adjusted as
// using 100x of the requested number might not be good in practice, e.g.
// when the requested number of items is small.
- const unsigned ItemsToRetrieve = 100 * Req.MaxCandidateCount;
+ const size_t ItemsToRetrieve = 100 * Req.MaxCandidateCount;
auto Root = createLimit(move(QueryIterator), ItemsToRetrieve);
- // FIXME(kbobyrev): Add boosting to the query and utilize retrieved
- // boosting scores.
+
std::vector<std::pair<DocID, float>> SymbolDocIDs = consume(*Root);
+ // Sort items using a combination of boosting score and quality.
+ // FIXME(kbobyrev): Consider merging this stage with the next one? This is
+ // a performance/quality trade-off and if the performance is not an issue
+ // there's plenty of quality to be bought by sacrificing part of
+ // performance.
+ std::sort(begin(SymbolDocIDs), end(SymbolDocIDs),
+ [&](const std::pair<DocID, float> &LHS,
+ const std::pair<DocID, float> &RHS) {
+ return SymbolQuality[LHS.first] * LHS.second >
+ SymbolQuality[RHS.first] * RHS.second;
+ });
+
// Retrieve top Req.MaxCandidateCount items.
+ const size_t ItemsToScore = 10 * Req.MaxCandidateCount;
std::priority_queue<std::pair<float, const Symbol *>> Top;
- for (const auto &P : SymbolDocIDs) {
- const DocID SymbolDocID = P.first;
+ for (size_t I = 0; I < std::min(SymbolDocIDs.size(), ItemsToScore); ++I) {
+ const DocID SymbolDocID = SymbolDocIDs[I].first;
const auto *Sym = (*Symbols)[SymbolDocID];
const llvm::Optional<float> Score = Filter.match(Sym->Name);
if (!Score)
continue;
// Multiply score by a negative factor so that Top stores items with the
// highest actual score.
- Top.emplace(-(*Score) * SymbolQuality.find(Sym)->second, Sym);
+ Top.emplace(-(*Score) * SymbolQuality[SymbolDocID], Sym);
if (Top.size() > Req.MaxCandidateCount) {
More = true;
Top.pop();
Index: clang-tools-extra/clangd/CMakeLists.txt
===================================================================
--- clang-tools-extra/clangd/CMakeLists.txt
+++ clang-tools-extra/clangd/CMakeLists.txt
@@ -46,6 +46,7 @@
index/dex/DexIndex.cpp
index/dex/Iterator.cpp
+ index/dex/Token.cpp
index/dex/Trigram.cpp
LINK_LIBS
Index: clang-tools-extra/;
===================================================================
--- /dev/null
+++ clang-tools-extra/;
@@ -0,0 +1,68 @@
+//===--- Token.cpp - Symbol Search primitive --------------------*- C++ -*-===//
+// The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#include "Token.h"
+#include "URI.h"
+#include "llvm/Support/Path.h"
+#include <cassert>
+
+namespace clang {
+namespace clangd {
+namespace dex {
+
+std::vector<Token> generateProximityPaths(llvm::StringRef URIPath,
+ size_t Limit) {
+ std::vector<Token> Result;
+ // Don't generate any tokens for empty paths.
+ if (URIPath.empty())
+ return Result;
+ auto ParsedURI = URI::parse(URIPath);
+ assert(
+ static_cast<bool>(ParsedURI) &&
+ "Non-empty argument of generateProximityPaths() should be a valid URI.");
+ const auto Scheme = ParsedURI.get().scheme();
+ const auto Authority = ParsedURI.get().authority();
+ StringRef Path = ParsedURI.get().body();
+ // Get parent directory of the file with symbol declaration.
+ Path = llvm::sys::path::parent_path(Path);
+ while (!Path.empty() && Limit--) {
+ URI TokenURI(Scheme, Authority, Path);
+ Result.emplace_back(Token::Kind::PathURI, TokenURI.toString());
+ Path = llvm::sys::path::parent_path(Path);
+ }
+ return Result;
+}
+
+std::vector<std::pair<Token, unsigned>>
+generateQueryProximityPaths(llvm::StringRef AbsolutePath,
+ llvm::ArrayRef<std::string> URISchemes,
+ size_t Count) {
+ std::vector<std::pair<Token, unsigned>> Result;
+ for (const auto Scheme : URISchemes) {
+ unsigned Distance = 0;
+ auto URI = URI::create(AbsolutePath, Scheme);
+ // For some paths, conversion to different URI schemes is impossible. These
+ // should be just skipped.
+ if (!static_cast<bool>(URI)) {
+ // Ignore the error.
+ handleAllErrors(URI.takeError(), [](const llvm::ErrorInfoBase &) {});
+ continue;
+ }
+ for (const auto ProximityPath :
+ generateProximityPaths(URI.get().toString(), Count))
+ Result.emplace_back(ProximityPath, Distance++);
+ // As soon as one valid URI scheme for given absolute path is found, the
+ // token generation process should be stopped.
+ break;
+ }
+ return Result;
+}
+
+} // namespace dex
+} // namespace clangd
+} // namespace clang
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
