[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/PiotrZSL closed https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/dmpolukhin approved this pull request. https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
PiotrZSL wrote: No need for release notes as this realpath stuff were added in this release anyway. So this is just a fix for degradation. https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/5chmidti approved this pull request. This halves the time clang-tidy takes to run this check on `IdentifierNamingCheck.cpp` for me :) Maybe add a release note? https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/PiotrZSL edited https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/PiotrZSL updated
https://github.com/llvm/llvm-project/pull/92659
>From a2ae273d083f451f44a41bd928c42a5bbdd946ae Mon Sep 17 00:00:00 2001
From: Piotr Zegar
Date: Tue, 21 May 2024 17:36:08 +
Subject: [PATCH] [clang-tidy] Optimize readability-identifier-naming
- Reduce disk IO usage by adding cache to an realpath
introduced by #81985
---
.../clang-tidy/readability/IdentifierNamingCheck.cpp | 12 ++--
.../clang-tidy/readability/IdentifierNamingCheck.h | 2 ++
2 files changed, 12 insertions(+), 2 deletions(-)
diff --git a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
index c3208392df156..828f13805a698 100644
--- a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
+++ b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
@@ -1414,13 +1414,21 @@ IdentifierNamingCheck::getDiagInfo(const NamingCheckId
,
}};
}
+StringRef IdentifierNamingCheck::getRealFileName(StringRef FileName) const {
+ auto Iter = RealFileNameCache.try_emplace(FileName);
+ SmallString<256U> = Iter.first->getValue();
+ if (!Iter.second)
+return RealFileName;
+ llvm::sys::fs::real_path(FileName, RealFileName);
+ return RealFileName;
+}
+
const IdentifierNamingCheck::FileStyle &
IdentifierNamingCheck::getStyleForFile(StringRef FileName) const {
if (!GetConfigPerFile)
return *MainFileStyle;
- SmallString<128> RealFileName;
- llvm::sys::fs::real_path(FileName, RealFileName);
+ StringRef RealFileName = getRealFileName(FileName);
StringRef Parent = llvm::sys::path::parent_path(RealFileName);
auto Iter = NamingStylesCache.find(Parent);
if (Iter != NamingStylesCache.end())
diff --git a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.h
b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.h
index 27c8e4bc768c4..646ec0eac8dd1 100644
--- a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.h
+++ b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.h
@@ -205,6 +205,7 @@ class IdentifierNamingCheck final : public
RenamerClangTidyCheck {
const NamingCheckFailure ) const override;
const FileStyle (StringRef FileName) const;
+ StringRef getRealFileName(StringRef FileName) const;
/// Find the style kind of a field in an anonymous record.
StyleKind findStyleKindForAnonField(
@@ -222,6 +223,7 @@ class IdentifierNamingCheck final : public
RenamerClangTidyCheck {
/// Stores the style options as a vector, indexed by the specified \ref
/// StyleKind, for a given directory.
mutable llvm::StringMap NamingStylesCache;
+ mutable llvm::StringMap> RealFileNameCache;
FileStyle *MainFileStyle;
ClangTidyContext *Context;
const bool GetConfigPerFile;
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/PiotrZSL edited https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/PiotrZSL edited https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/PiotrZSL edited https://github.com/llvm/llvm-project/pull/92659 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
llvmbot wrote:
@llvm/pr-subscribers-clang-tools-extra
Author: Piotr Zegar (PiotrZSL)
Changes
- Reduce disk IO usage by adding cache to an realpath introduced by #81985
- Refactor HungarianNotation::getDeclTypeName to remove some string copies and
use more safe string API.
---
Full diff: https://github.com/llvm/llvm-project/pull/92659.diff
2 Files Affected:
- (modified) clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
(+35-33)
- (modified) clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.h
(+2)
``diff
diff --git a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
index c3208392df156..b2ab6e283c079 100644
--- a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
+++ b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
@@ -305,27 +305,20 @@ std::string
IdentifierNamingCheck::HungarianNotation::getDeclTypeName(
auto = VD->getASTContext().getSourceManager();
const char *Begin = SM.getCharacterData(VD->getBeginLoc());
const char *End = SM.getCharacterData(VD->getEndLoc());
- intptr_t StrLen = End - Begin;
+ if (End < Begin)
+return {};
// FIXME: Sometimes the value that returns from ValDecl->getEndLoc()
// is wrong(out of location of Decl). This causes `StrLen` will be assigned
// an unexpected large value. Current workaround to find the terminated
// character instead of the `getEndLoc()` function.
- const char *EOL = strchr(Begin, '\n');
- if (!EOL)
-EOL = Begin + strlen(Begin);
-
- const char *PosList[] = {strchr(Begin, '='), strchr(Begin, ';'),
- strchr(Begin, ','), strchr(Begin, ')'), EOL};
- for (const auto : PosList) {
-if (Pos > Begin)
- EOL = std::min(EOL, Pos);
- }
+ llvm::StringRef NameRef(Begin, End - Begin);
+ auto Pos = NameRef.find_first_of("\n\r=;,)");
+ if (Pos != llvm::StringRef::npos)
+NameRef = NameRef.substr(0, Pos);
- StrLen = EOL - Begin;
- std::string TypeName;
- if (StrLen > 0) {
-std::string Type(Begin, StrLen);
+ if (!NameRef.empty()) {
+std::string Type(NameRef.begin(), NameRef.end());
static constexpr StringRef Keywords[] = {
// Constexpr specifiers
@@ -339,66 +332,67 @@ std::string
IdentifierNamingCheck::HungarianNotation::getDeclTypeName(
// Remove keywords
for (StringRef Kw : Keywords) {
- for (size_t Pos = 0;
- (Pos = Type.find(Kw.data(), Pos)) != std::string::npos;) {
+ for (size_t Pos = 0; (Pos = Type.find(Kw, Pos)) != std::string::npos;) {
Type.replace(Pos, Kw.size(), "");
}
}
-TypeName = Type.erase(0, Type.find_first_not_of(' '));
// Remove template parameters
const size_t Pos = Type.find('<');
if (Pos != std::string::npos) {
- TypeName = Type.erase(Pos, Type.size() - Pos);
+ Type.erase(Pos);
}
// Replace spaces with single space.
for (size_t Pos = 0; (Pos = Type.find(" ", Pos)) != std::string::npos;
- Pos += strlen(" ")) {
- Type.replace(Pos, strlen(" "), " ");
+ Pos += 2U) {
+ Type.replace(Pos, 2U, " ");
}
// Replace " &" with "&".
for (size_t Pos = 0; (Pos = Type.find(" &", Pos)) != std::string::npos;
- Pos += strlen("&")) {
- Type.replace(Pos, strlen(" &"), "&");
+ Pos += 2U) {
+ Type.replace(Pos, 2U, "&");
}
// Replace " *" with "* ".
for (size_t Pos = 0; (Pos = Type.find(" *", Pos)) != std::string::npos;
- Pos += strlen("*")) {
- Type.replace(Pos, strlen(" *"), "* ");
+ Pos += 1U) {
+ Type.replace(Pos, 2U, "* ");
}
+Type.erase(0, Type.find_first_not_of(' '));
+
// Remove redundant tailing.
static constexpr StringRef TailsOfMultiWordType[] = {
" int", " char", " double", " long", " short"};
bool RedundantRemoved = false;
for (auto Kw : TailsOfMultiWordType) {
- size_t Pos = Type.rfind(Kw.data());
+ size_t Pos = Type.rfind(Kw);
if (Pos != std::string::npos) {
const size_t PtrCount = getAsteriskCount(Type, ND);
-Type = Type.substr(0, Pos + Kw.size() + PtrCount);
+Type.erase(Pos + Kw.size() + PtrCount);
RedundantRemoved = true;
break;
}
}
-TypeName = Type.erase(0, Type.find_first_not_of(' '));
+Type.erase(0, Type.find_first_not_of(' '));
if (!RedundantRemoved) {
std::size_t FoundSpace = Type.find(' ');
if (FoundSpace != std::string::npos)
Type = Type.substr(0, FoundSpace);
}
-TypeName = Type.erase(0, Type.find_first_not_of(' '));
+Type.erase(0, Type.find_first_not_of(' '));
QualType QT = VD->getType();
if (!QT.isNull() && QT->isArrayType())
- TypeName.append("[]");
+ Type.append("[]");
+return Type;
}
- return TypeName;
+ return {};
}
[clang-tools-extra] [clang-tidy] Optimize realpath in readability-identifier-naming (PR #92659)
https://github.com/PiotrZSL created
https://github.com/llvm/llvm-project/pull/92659
- Reduce disk IO usage by adding cache to an realpath introduced by #81985
- Refactor HungarianNotation::getDeclTypeName to remove some string copies and
use more safe string API.
>From 45bce19a4852ab95c882948ea85ba8d87a4eeef2 Mon Sep 17 00:00:00 2001
From: Piotr Zegar
Date: Sat, 18 May 2024 16:56:31 +
Subject: [PATCH] [clang-tidy] Optimize readability-identifier-naming
- Reduce disk IO usage by adding cache to an realpath
introduced by #81985
- Refactor HungarianNotation::getDeclTypeName to remove some
string copies and use more safe string API.
---
.../readability/IdentifierNamingCheck.cpp | 68 ++-
.../readability/IdentifierNamingCheck.h | 2 +
2 files changed, 37 insertions(+), 33 deletions(-)
diff --git a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
index c3208392df156..b2ab6e283c079 100644
--- a/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
+++ b/clang-tools-extra/clang-tidy/readability/IdentifierNamingCheck.cpp
@@ -305,27 +305,20 @@ std::string
IdentifierNamingCheck::HungarianNotation::getDeclTypeName(
auto = VD->getASTContext().getSourceManager();
const char *Begin = SM.getCharacterData(VD->getBeginLoc());
const char *End = SM.getCharacterData(VD->getEndLoc());
- intptr_t StrLen = End - Begin;
+ if (End < Begin)
+return {};
// FIXME: Sometimes the value that returns from ValDecl->getEndLoc()
// is wrong(out of location of Decl). This causes `StrLen` will be assigned
// an unexpected large value. Current workaround to find the terminated
// character instead of the `getEndLoc()` function.
- const char *EOL = strchr(Begin, '\n');
- if (!EOL)
-EOL = Begin + strlen(Begin);
-
- const char *PosList[] = {strchr(Begin, '='), strchr(Begin, ';'),
- strchr(Begin, ','), strchr(Begin, ')'), EOL};
- for (const auto : PosList) {
-if (Pos > Begin)
- EOL = std::min(EOL, Pos);
- }
+ llvm::StringRef NameRef(Begin, End - Begin);
+ auto Pos = NameRef.find_first_of("\n\r=;,)");
+ if (Pos != llvm::StringRef::npos)
+NameRef = NameRef.substr(0, Pos);
- StrLen = EOL - Begin;
- std::string TypeName;
- if (StrLen > 0) {
-std::string Type(Begin, StrLen);
+ if (!NameRef.empty()) {
+std::string Type(NameRef.begin(), NameRef.end());
static constexpr StringRef Keywords[] = {
// Constexpr specifiers
@@ -339,66 +332,67 @@ std::string
IdentifierNamingCheck::HungarianNotation::getDeclTypeName(
// Remove keywords
for (StringRef Kw : Keywords) {
- for (size_t Pos = 0;
- (Pos = Type.find(Kw.data(), Pos)) != std::string::npos;) {
+ for (size_t Pos = 0; (Pos = Type.find(Kw, Pos)) != std::string::npos;) {
Type.replace(Pos, Kw.size(), "");
}
}
-TypeName = Type.erase(0, Type.find_first_not_of(' '));
// Remove template parameters
const size_t Pos = Type.find('<');
if (Pos != std::string::npos) {
- TypeName = Type.erase(Pos, Type.size() - Pos);
+ Type.erase(Pos);
}
// Replace spaces with single space.
for (size_t Pos = 0; (Pos = Type.find(" ", Pos)) != std::string::npos;
- Pos += strlen(" ")) {
- Type.replace(Pos, strlen(" "), " ");
+ Pos += 2U) {
+ Type.replace(Pos, 2U, " ");
}
// Replace " &" with "&".
for (size_t Pos = 0; (Pos = Type.find(" &", Pos)) != std::string::npos;
- Pos += strlen("&")) {
- Type.replace(Pos, strlen(" &"), "&");
+ Pos += 2U) {
+ Type.replace(Pos, 2U, "&");
}
// Replace " *" with "* ".
for (size_t Pos = 0; (Pos = Type.find(" *", Pos)) != std::string::npos;
- Pos += strlen("*")) {
- Type.replace(Pos, strlen(" *"), "* ");
+ Pos += 1U) {
+ Type.replace(Pos, 2U, "* ");
}
+Type.erase(0, Type.find_first_not_of(' '));
+
// Remove redundant tailing.
static constexpr StringRef TailsOfMultiWordType[] = {
" int", " char", " double", " long", " short"};
bool RedundantRemoved = false;
for (auto Kw : TailsOfMultiWordType) {
- size_t Pos = Type.rfind(Kw.data());
+ size_t Pos = Type.rfind(Kw);
if (Pos != std::string::npos) {
const size_t PtrCount = getAsteriskCount(Type, ND);
-Type = Type.substr(0, Pos + Kw.size() + PtrCount);
+Type.erase(Pos + Kw.size() + PtrCount);
RedundantRemoved = true;
break;
}
}
-TypeName = Type.erase(0, Type.find_first_not_of(' '));
+Type.erase(0, Type.find_first_not_of(' '));
if (!RedundantRemoved) {
std::size_t FoundSpace = Type.find(' ');
if (FoundSpace != std::string::npos)
Type = Type.substr(0, FoundSpace);
}
-TypeName = Type.erase(0, Type.find_first_not_of(' '));
+Type.erase(0,
