[clang] [clang][Diagnostics] Highlight code snippets (PR #66514)

[Timm Baeder via cfe-commits]

Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>,
Timm =?utf-8?q?Bäder?= <tbaeder at redhat.com>
Message-ID:
In-Reply-To: <llvm/llvm-project/pull/66514/clang at github.com>


https://github.com/tbaederr updated https://github.com/llvm/llvm-project/pull/66514

>From 8e0d084529d159ee80dd3971f9a5bb112e24bb86 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Fri, 15 Sep 2023 15:51:39 +0200
Subject: [PATCH 01/21] [clang][Diagnostics] Highlight code snippets

Add some primitive syntax highlighting to our code snippet output.
---
 .../clang/Frontend/CodeSnippetHighlighter.h   |  46 +++++++
 clang/include/clang/Frontend/TextDiagnostic.h |   2 +
 clang/lib/Frontend/CMakeLists.txt             |   1 +
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 120 ++++++++++++++++++
 clang/lib/Frontend/TextDiagnostic.cpp         |  26 ++++
 5 files changed, 195 insertions(+)
 create mode 100644 clang/include/clang/Frontend/CodeSnippetHighlighter.h
 create mode 100644 clang/lib/Frontend/CodeSnippetHighlighter.cpp

diff --git a/clang/include/clang/Frontend/CodeSnippetHighlighter.h b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
new file mode 100644
index 000000000000000..776954b59e2e1a8
--- /dev/null
+++ b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
@@ -0,0 +1,46 @@
+//===--- CodeSnippetHighlighter.h - Code snippet highlighting ---*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_CLANG_FRONTEND_CODESNIPPETHIGHLIGHTER_H
+#define LLVM_CLANG_FRONTEND_CODESNIPPETHIGHLIGHTER_H
+
+#include "clang/Basic/LangOptions.h"
+#include "llvm/ADT/SmallSet.h"
+#include "llvm/Support/raw_ostream.h"
+#include <vector>
+
+namespace clang {
+
+struct StyleRange {
+  unsigned Start;
+  unsigned End;
+  const enum llvm::raw_ostream::Colors c;
+};
+
+class CodeSnippetHighlighter final {
+public:
+  CodeSnippetHighlighter() = default;
+
+  /// Produce StyleRanges for the given line.
+  /// The returned vector contains non-overlapping style ranges. They are sorted
+  /// from beginning of the line to the end.
+  std::vector<StyleRange> highlightLine(llvm::StringRef SourceLine,
+                                        const LangOptions &LangOpts);
+
+private:
+  bool Initialized = false;
+  /// Fills Keywords and Literals.
+  void ensureTokenData();
+
+  llvm::SmallSet<StringRef, 12> Keywords;
+  llvm::SmallSet<StringRef, 12> Literals;
+};
+
+} // namespace clang
+
+#endif
diff --git a/clang/include/clang/Frontend/TextDiagnostic.h b/clang/include/clang/Frontend/TextDiagnostic.h
index 7eb0ab0cdc9bca8..59fd4d4f9408d48 100644
--- a/clang/include/clang/Frontend/TextDiagnostic.h
+++ b/clang/include/clang/Frontend/TextDiagnostic.h
@@ -15,6 +15,7 @@
 #ifndef LLVM_CLANG_FRONTEND_TEXTDIAGNOSTIC_H
 #define LLVM_CLANG_FRONTEND_TEXTDIAGNOSTIC_H
 
+#include "clang/Frontend/CodeSnippetHighlighter.h"
 #include "clang/Frontend/DiagnosticRenderer.h"
 
 namespace clang {
@@ -33,6 +34,7 @@ namespace clang {
 /// printing coming out of libclang.
 class TextDiagnostic : public DiagnosticRenderer {
   raw_ostream &OS;
+  CodeSnippetHighlighter SnippetHighlighter;
 
 public:
   TextDiagnostic(raw_ostream &OS,
diff --git a/clang/lib/Frontend/CMakeLists.txt b/clang/lib/Frontend/CMakeLists.txt
index 1e5f0a859dfd568..f3547f771593093 100644
--- a/clang/lib/Frontend/CMakeLists.txt
+++ b/clang/lib/Frontend/CMakeLists.txt
@@ -42,6 +42,7 @@ add_clang_library(clangFrontend
   TextDiagnosticPrinter.cpp
   VerifyDiagnosticConsumer.cpp
   InterfaceStubFunctionsConsumer.cpp
+  CodeSnippetHighlighter.cpp
 
   DEPENDS
   ClangDriverOptions
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
new file mode 100644
index 000000000000000..829a533ad2692e5
--- /dev/null
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -0,0 +1,120 @@
+
+#include "clang/Frontend/CodeSnippetHighlighter.h"
+#include "clang/Basic/DiagnosticOptions.h"
+#include "clang/Basic/SourceManager.h"
+#include "clang/Lex/Lexer.h"
+#include "llvm/Support/raw_ostream.h"
+
+using namespace clang;
+
+void CodeSnippetHighlighter::ensureTokenData() {
+  if (Initialized)
+    return;
+
+  // List of keywords, literals and types we want to highlight.
+  // These are best-effort, as is everything we do wrt. highlighting.
+  Keywords.insert("_Static_assert");
+  Keywords.insert("auto");
+  Keywords.insert("concept");
+  Keywords.insert("const");
+  Keywords.insert("consteval");
+  Keywords.insert("constexpr");
+  Keywords.insert("delete");
+  Keywords.insert("do");
+  Keywords.insert("else");
+  Keywords.insert("final");
+  Keywords.insert("for");
+  Keywords.insert("if");
+  Keywords.insert("mutable");
+  Keywords.insert("namespace");
+  Keywords.insert("new");
+  Keywords.insert("private");
+  Keywords.insert("public");
+  Keywords.insert("requires");
+  Keywords.insert("return");
+  Keywords.insert("static");
+  Keywords.insert("static_assert");
+  Keywords.insert("using");
+  Keywords.insert("void");
+  Keywords.insert("volatile");
+  Keywords.insert("while");
+
+  // Builtin types we highlight
+  Keywords.insert("void");
+  Keywords.insert("char");
+  Keywords.insert("short");
+  Keywords.insert("int");
+  Keywords.insert("unsigned");
+  Keywords.insert("long");
+  Keywords.insert("float");
+  Keywords.insert("double");
+
+  Literals.insert("true");
+  Literals.insert("false");
+  Literals.insert("nullptr");
+
+  Initialized = true;
+}
+
+static SourceManager createTempSourceManager() {
+  FileSystemOptions FileOpts;
+  FileManager FileMgr(FileOpts);
+  llvm::IntrusiveRefCntPtr<DiagnosticIDs> DiagIDs(new DiagnosticIDs());
+  llvm::IntrusiveRefCntPtr<DiagnosticOptions> DiagOpts(new DiagnosticOptions());
+  DiagnosticsEngine diags(DiagIDs, DiagOpts);
+  return SourceManager(diags, FileMgr);
+}
+
+static Lexer createTempLexer(llvm::MemoryBufferRef B, SourceManager &FakeSM,
+                             const LangOptions &LangOpts) {
+  return Lexer(FakeSM.createFileID(B), B, FakeSM, LangOpts);
+}
+
+std::vector<StyleRange>
+CodeSnippetHighlighter::highlightLine(StringRef SourceLine,
+                                      const LangOptions &LangOpts) {
+  ensureTokenData();
+
+  constexpr raw_ostream::Colors CommentColor = raw_ostream::BLACK;
+  constexpr raw_ostream::Colors LiteralColor = raw_ostream::GREEN;
+  constexpr raw_ostream::Colors KeywordColor = raw_ostream::YELLOW;
+
+  const auto MemBuf = llvm::MemoryBuffer::getMemBuffer(SourceLine);
+  SourceManager FakeSM = createTempSourceManager();
+  Lexer L = createTempLexer(MemBuf->getMemBufferRef(), FakeSM, LangOpts);
+  L.SetKeepWhitespaceMode(true);
+
+  std::vector<StyleRange> Styles;
+  bool Stop = false;
+  while (!Stop) {
+    Token tok;
+    Stop = L.LexFromRawLexer(tok);
+    if (tok.is(tok::unknown))
+      continue;
+
+    bool Invalid;
+    unsigned Start =
+        FakeSM.getSpellingColumnNumber(tok.getLocation(), &Invalid) - 1;
+    if (Invalid)
+      continue;
+
+    if (tok.is(tok::raw_identifier)) {
+      // Almost everything we lex is an identifier, since we use a raw lexer.
+      // Some should be highlightes as literals, others as keywords.
+      if (Keywords.contains(tok.getRawIdentifier()))
+        Styles.push_back(
+            StyleRange{Start, Start + tok.getLength(), KeywordColor});
+      else if (Literals.contains(tok.getRawIdentifier()))
+        Styles.push_back(
+            StyleRange{Start, Start + tok.getLength(), LiteralColor});
+    } else if (tok::isLiteral(tok.getKind())) {
+      Styles.push_back(
+          StyleRange{Start, Start + tok.getLength(), LiteralColor});
+    } else if (tok.is(tok::comment)) {
+      Styles.push_back(
+          StyleRange{Start, Start + tok.getLength(), CommentColor});
+    }
+  }
+
+  return Styles;
+}
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index 779dead5d058d1a..13d7d1e048cf991 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -11,6 +11,7 @@
 #include "clang/Basic/DiagnosticOptions.h"
 #include "clang/Basic/FileManager.h"
 #include "clang/Basic/SourceManager.h"
+#include "clang/Frontend/CodeSnippetHighlighter.h"
 #include "clang/Lex/Lexer.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/StringExtras.h"
@@ -1278,6 +1279,9 @@ void TextDiagnostic::emitSnippetAndCaret(
 void TextDiagnostic::emitSnippet(StringRef SourceLine,
                                  unsigned MaxLineNoDisplayWidth,
                                  unsigned LineNo) {
+  std::vector<StyleRange> Styles =
+      SnippetHighlighter.highlightLine(SourceLine, LangOpts);
+
   // Emit line number.
   if (MaxLineNoDisplayWidth > 0) {
     unsigned LineNoDisplayWidth = getNumDisplayWidth(LineNo);
@@ -1287,11 +1291,33 @@ void TextDiagnostic::emitSnippet(StringRef SourceLine,
 
   // Print the source line one character at a time.
   bool PrintReversed = false;
+  bool HighlightingEnabled = DiagOpts->ShowColors;
   size_t I = 0;
   while (I < SourceLine.size()) {
     auto [Str, WasPrintable] =
         printableTextForNextCharacter(SourceLine, &I, DiagOpts->TabStop);
 
+    // Just stop highlighting anything for this line if we found a non-printable
+    // character.
+    if (!WasPrintable)
+      HighlightingEnabled = false;
+
+    // FIXME: I hope we can do this in some nicer way.
+    if (HighlightingEnabled) {
+      std::optional<enum raw_ostream::Colors> H;
+      for (auto &P : Styles) {
+        if (P.Start < I && P.End >= I) {
+          H = P.c;
+          break;
+        }
+      }
+
+      if (H) {
+        OS.changeColor(*H, false);
+      } else
+        OS.resetColor();
+    }
+
     // Toggle inverted colors on or off for this character.
     if (DiagOpts->ShowColors) {
       if (WasPrintable == PrintReversed) {

>From 39f5e706e5cb2479d46210fc74eb4e5f8224feaf Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Wed, 20 Sep 2023 15:28:10 +0200
Subject: [PATCH 02/21] Get identifier table from Preprocessor

---
 .../clang/Frontend/CodeSnippetHighlighter.h   |  11 +-
 clang/include/clang/Frontend/TextDiagnostic.h |   7 +-
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 101 +++++-------------
 clang/lib/Frontend/TextDiagnostic.cpp         |   8 +-
 clang/lib/Frontend/TextDiagnosticPrinter.cpp  |   2 +-
 5 files changed, 39 insertions(+), 90 deletions(-)

diff --git a/clang/include/clang/Frontend/CodeSnippetHighlighter.h b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
index 776954b59e2e1a8..ec03375221f9ffc 100644
--- a/clang/include/clang/Frontend/CodeSnippetHighlighter.h
+++ b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
@@ -22,6 +22,8 @@ struct StyleRange {
   const enum llvm::raw_ostream::Colors c;
 };
 
+class Preprocessor;
+
 class CodeSnippetHighlighter final {
 public:
   CodeSnippetHighlighter() = default;
@@ -30,15 +32,8 @@ class CodeSnippetHighlighter final {
   /// The returned vector contains non-overlapping style ranges. They are sorted
   /// from beginning of the line to the end.
   std::vector<StyleRange> highlightLine(llvm::StringRef SourceLine,
+                                        const Preprocessor *PP,
                                         const LangOptions &LangOpts);
-
-private:
-  bool Initialized = false;
-  /// Fills Keywords and Literals.
-  void ensureTokenData();
-
-  llvm::SmallSet<StringRef, 12> Keywords;
-  llvm::SmallSet<StringRef, 12> Literals;
 };
 
 } // namespace clang
diff --git a/clang/include/clang/Frontend/TextDiagnostic.h b/clang/include/clang/Frontend/TextDiagnostic.h
index 59fd4d4f9408d48..8cdb9b141a8a4af 100644
--- a/clang/include/clang/Frontend/TextDiagnostic.h
+++ b/clang/include/clang/Frontend/TextDiagnostic.h
@@ -19,7 +19,6 @@
 #include "clang/Frontend/DiagnosticRenderer.h"
 
 namespace clang {
-
 /// Class to encapsulate the logic for formatting and printing a textual
 /// diagnostic message.
 ///
@@ -34,12 +33,12 @@ namespace clang {
 /// printing coming out of libclang.
 class TextDiagnostic : public DiagnosticRenderer {
   raw_ostream &OS;
+  const Preprocessor *PP;
   CodeSnippetHighlighter SnippetHighlighter;
 
 public:
-  TextDiagnostic(raw_ostream &OS,
-                 const LangOptions &LangOpts,
-                 DiagnosticOptions *DiagOpts);
+  TextDiagnostic(raw_ostream &OS, const LangOptions &LangOpts,
+                 const Preprocessor *PP, DiagnosticOptions *DiagOpts);
 
   ~TextDiagnostic() override;
 
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 829a533ad2692e5..63b3707fbb7ef83 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -3,59 +3,12 @@
 #include "clang/Basic/DiagnosticOptions.h"
 #include "clang/Basic/SourceManager.h"
 #include "clang/Lex/Lexer.h"
+#include "clang/Lex/Preprocessor.h"
+#include "clang/Lex/PreprocessorOptions.h"
 #include "llvm/Support/raw_ostream.h"
 
 using namespace clang;
 
-void CodeSnippetHighlighter::ensureTokenData() {
-  if (Initialized)
-    return;
-
-  // List of keywords, literals and types we want to highlight.
-  // These are best-effort, as is everything we do wrt. highlighting.
-  Keywords.insert("_Static_assert");
-  Keywords.insert("auto");
-  Keywords.insert("concept");
-  Keywords.insert("const");
-  Keywords.insert("consteval");
-  Keywords.insert("constexpr");
-  Keywords.insert("delete");
-  Keywords.insert("do");
-  Keywords.insert("else");
-  Keywords.insert("final");
-  Keywords.insert("for");
-  Keywords.insert("if");
-  Keywords.insert("mutable");
-  Keywords.insert("namespace");
-  Keywords.insert("new");
-  Keywords.insert("private");
-  Keywords.insert("public");
-  Keywords.insert("requires");
-  Keywords.insert("return");
-  Keywords.insert("static");
-  Keywords.insert("static_assert");
-  Keywords.insert("using");
-  Keywords.insert("void");
-  Keywords.insert("volatile");
-  Keywords.insert("while");
-
-  // Builtin types we highlight
-  Keywords.insert("void");
-  Keywords.insert("char");
-  Keywords.insert("short");
-  Keywords.insert("int");
-  Keywords.insert("unsigned");
-  Keywords.insert("long");
-  Keywords.insert("float");
-  Keywords.insert("double");
-
-  Literals.insert("true");
-  Literals.insert("false");
-  Literals.insert("nullptr");
-
-  Initialized = true;
-}
-
 static SourceManager createTempSourceManager() {
   FileSystemOptions FileOpts;
   FileManager FileMgr(FileOpts);
@@ -70,49 +23,51 @@ static Lexer createTempLexer(llvm::MemoryBufferRef B, SourceManager &FakeSM,
   return Lexer(FakeSM.createFileID(B), B, FakeSM, LangOpts);
 }
 
-std::vector<StyleRange>
-CodeSnippetHighlighter::highlightLine(StringRef SourceLine,
-                                      const LangOptions &LangOpts) {
-  ensureTokenData();
-
+std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
+    StringRef SourceLine, const Preprocessor *PP, const LangOptions &LangOpts) {
   constexpr raw_ostream::Colors CommentColor = raw_ostream::BLACK;
   constexpr raw_ostream::Colors LiteralColor = raw_ostream::GREEN;
   constexpr raw_ostream::Colors KeywordColor = raw_ostream::YELLOW;
 
-  const auto MemBuf = llvm::MemoryBuffer::getMemBuffer(SourceLine);
   SourceManager FakeSM = createTempSourceManager();
+  const auto MemBuf = llvm::MemoryBuffer::getMemBuffer(SourceLine);
   Lexer L = createTempLexer(MemBuf->getMemBufferRef(), FakeSM, LangOpts);
   L.SetKeepWhitespaceMode(true);
 
   std::vector<StyleRange> Styles;
   bool Stop = false;
   while (!Stop) {
-    Token tok;
-    Stop = L.LexFromRawLexer(tok);
-    if (tok.is(tok::unknown))
+    Token T;
+    Stop = L.LexFromRawLexer(T);
+    if (T.is(tok::unknown))
       continue;
 
     bool Invalid;
     unsigned Start =
-        FakeSM.getSpellingColumnNumber(tok.getLocation(), &Invalid) - 1;
+        FakeSM.getSpellingColumnNumber(T.getLocation(), &Invalid) - 1;
     if (Invalid)
       continue;
 
-    if (tok.is(tok::raw_identifier)) {
-      // Almost everything we lex is an identifier, since we use a raw lexer.
-      // Some should be highlightes as literals, others as keywords.
-      if (Keywords.contains(tok.getRawIdentifier()))
-        Styles.push_back(
-            StyleRange{Start, Start + tok.getLength(), KeywordColor});
-      else if (Literals.contains(tok.getRawIdentifier()))
+    if (T.is(tok::raw_identifier)) {
+      StringRef RawIdent = T.getRawIdentifier();
+      // Special case true/false/nullptr literals, since they will otherwise be
+      // treated as keywords.
+      if (RawIdent == "true" || RawIdent == "false" || RawIdent == "nullptr") {
         Styles.push_back(
-            StyleRange{Start, Start + tok.getLength(), LiteralColor});
-    } else if (tok::isLiteral(tok.getKind())) {
-      Styles.push_back(
-          StyleRange{Start, Start + tok.getLength(), LiteralColor});
-    } else if (tok.is(tok::comment)) {
-      Styles.push_back(
-          StyleRange{Start, Start + tok.getLength(), CommentColor});
+            StyleRange{Start, Start + T.getLength(), LiteralColor});
+      } else {
+        const IdentifierInfo *II = PP->getIdentifierInfo(RawIdent);
+        assert(II);
+
+        if (II->isKeyword(LangOpts)) {
+          Styles.push_back(
+              StyleRange{Start, Start + T.getLength(), KeywordColor});
+        }
+      }
+    } else if (tok::isLiteral(T.getKind())) {
+      Styles.push_back(StyleRange{Start, Start + T.getLength(), LiteralColor});
+    } else if (T.is(tok::comment)) {
+      Styles.push_back(StyleRange{Start, Start + T.getLength(), CommentColor});
     }
   }
 
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index 13d7d1e048cf991..e840cdd952d09f1 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -645,10 +645,10 @@ static bool printWordWrapped(raw_ostream &OS, StringRef Str, unsigned Columns,
   return Wrapped;
 }
 
-TextDiagnostic::TextDiagnostic(raw_ostream &OS,
-                               const LangOptions &LangOpts,
+TextDiagnostic::TextDiagnostic(raw_ostream &OS, const LangOptions &LangOpts,
+                               const Preprocessor *PP,
                                DiagnosticOptions *DiagOpts)
-  : DiagnosticRenderer(LangOpts, DiagOpts), OS(OS) {}
+    : DiagnosticRenderer(LangOpts, DiagOpts), OS(OS), PP(PP) {}
 
 TextDiagnostic::~TextDiagnostic() {}
 
@@ -1280,7 +1280,7 @@ void TextDiagnostic::emitSnippet(StringRef SourceLine,
                                  unsigned MaxLineNoDisplayWidth,
                                  unsigned LineNo) {
   std::vector<StyleRange> Styles =
-      SnippetHighlighter.highlightLine(SourceLine, LangOpts);
+      SnippetHighlighter.highlightLine(SourceLine, PP, LangOpts);
 
   // Emit line number.
   if (MaxLineNoDisplayWidth > 0) {
diff --git a/clang/lib/Frontend/TextDiagnosticPrinter.cpp b/clang/lib/Frontend/TextDiagnosticPrinter.cpp
index 0ff5376098ffe8d..3bc3935078baada 100644
--- a/clang/lib/Frontend/TextDiagnosticPrinter.cpp
+++ b/clang/lib/Frontend/TextDiagnosticPrinter.cpp
@@ -36,7 +36,7 @@ TextDiagnosticPrinter::~TextDiagnosticPrinter() {
 void TextDiagnosticPrinter::BeginSourceFile(const LangOptions &LO,
                                             const Preprocessor *PP) {
   // Build the TextDiagnostic utility.
-  TextDiag.reset(new TextDiagnostic(OS, LO, &*DiagOpts));
+  TextDiag.reset(new TextDiagnostic(OS, LO, PP, &*DiagOpts));
 }
 
 void TextDiagnosticPrinter::EndSourceFile() {

>From 6ba5a8c9c62ccbccd8acb482bf1604effaa00248 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Wed, 20 Sep 2023 17:24:42 +0200
Subject: [PATCH 03/21] Move the PP parameter to the end of the TextDiagnostic
 ctor

---
 clang/include/clang/Frontend/TextDiagnostic.h | 2 +-
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 2 ++
 clang/lib/Frontend/TextDiagnostic.cpp         | 4 ++--
 clang/lib/Frontend/TextDiagnosticPrinter.cpp  | 2 +-
 4 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/clang/include/clang/Frontend/TextDiagnostic.h b/clang/include/clang/Frontend/TextDiagnostic.h
index 8cdb9b141a8a4af..43c39ff96a2d1ce 100644
--- a/clang/include/clang/Frontend/TextDiagnostic.h
+++ b/clang/include/clang/Frontend/TextDiagnostic.h
@@ -38,7 +38,7 @@ class TextDiagnostic : public DiagnosticRenderer {
 
 public:
   TextDiagnostic(raw_ostream &OS, const LangOptions &LangOpts,
-                 const Preprocessor *PP, DiagnosticOptions *DiagOpts);
+                 DiagnosticOptions *DiagOpts, const Preprocessor *PP = nullptr);
 
   ~TextDiagnostic() override;
 
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 63b3707fbb7ef83..32bd61f3746023c 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -25,6 +25,8 @@ static Lexer createTempLexer(llvm::MemoryBufferRef B, SourceManager &FakeSM,
 
 std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
     StringRef SourceLine, const Preprocessor *PP, const LangOptions &LangOpts) {
+  if (!PP)
+    return {};
   constexpr raw_ostream::Colors CommentColor = raw_ostream::BLACK;
   constexpr raw_ostream::Colors LiteralColor = raw_ostream::GREEN;
   constexpr raw_ostream::Colors KeywordColor = raw_ostream::YELLOW;
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index e840cdd952d09f1..c9207e9dfbf9217 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -646,8 +646,8 @@ static bool printWordWrapped(raw_ostream &OS, StringRef Str, unsigned Columns,
 }
 
 TextDiagnostic::TextDiagnostic(raw_ostream &OS, const LangOptions &LangOpts,
-                               const Preprocessor *PP,
-                               DiagnosticOptions *DiagOpts)
+                               DiagnosticOptions *DiagOpts,
+                               const Preprocessor *PP)
     : DiagnosticRenderer(LangOpts, DiagOpts), OS(OS), PP(PP) {}
 
 TextDiagnostic::~TextDiagnostic() {}
diff --git a/clang/lib/Frontend/TextDiagnosticPrinter.cpp b/clang/lib/Frontend/TextDiagnosticPrinter.cpp
index 3bc3935078baada..b2fb762537573ef 100644
--- a/clang/lib/Frontend/TextDiagnosticPrinter.cpp
+++ b/clang/lib/Frontend/TextDiagnosticPrinter.cpp
@@ -36,7 +36,7 @@ TextDiagnosticPrinter::~TextDiagnosticPrinter() {
 void TextDiagnosticPrinter::BeginSourceFile(const LangOptions &LO,
                                             const Preprocessor *PP) {
   // Build the TextDiagnostic utility.
-  TextDiag.reset(new TextDiagnostic(OS, LO, PP, &*DiagOpts));
+  TextDiag.reset(new TextDiagnostic(OS, LO, &*DiagOpts, PP));
 }
 
 void TextDiagnosticPrinter::EndSourceFile() {

>From 6fffd4c5cac11cc72514bd81635f851f97bf5909 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Thu, 21 Sep 2023 06:38:24 +0200
Subject: [PATCH 04/21] Tune colors

---
 clang/include/clang/Frontend/CodeSnippetHighlighter.h | 2 +-
 clang/lib/Frontend/CodeSnippetHighlighter.cpp         | 6 +++---
 clang/lib/Frontend/TextDiagnostic.cpp                 | 7 ++++---
 3 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/clang/include/clang/Frontend/CodeSnippetHighlighter.h b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
index ec03375221f9ffc..c2a0184085d5da4 100644
--- a/clang/include/clang/Frontend/CodeSnippetHighlighter.h
+++ b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
@@ -19,7 +19,7 @@ namespace clang {
 struct StyleRange {
   unsigned Start;
   unsigned End;
-  const enum llvm::raw_ostream::Colors c;
+  const enum llvm::raw_ostream::Colors color;
 };
 
 class Preprocessor;
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 32bd61f3746023c..dba7f5d2848505a 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -27,9 +27,9 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
     StringRef SourceLine, const Preprocessor *PP, const LangOptions &LangOpts) {
   if (!PP)
     return {};
-  constexpr raw_ostream::Colors CommentColor = raw_ostream::BLACK;
-  constexpr raw_ostream::Colors LiteralColor = raw_ostream::GREEN;
-  constexpr raw_ostream::Colors KeywordColor = raw_ostream::YELLOW;
+  constexpr raw_ostream::Colors CommentColor = raw_ostream::GREEN;
+  constexpr raw_ostream::Colors LiteralColor = raw_ostream::CYAN;
+  constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
 
   SourceManager FakeSM = createTempSourceManager();
   const auto MemBuf = llvm::MemoryBuffer::getMemBuffer(SourceLine);
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index c9207e9dfbf9217..35a92a8044f2e52 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -19,6 +19,7 @@
 #include "llvm/Support/ErrorHandling.h"
 #include "llvm/Support/Locale.h"
 #include "llvm/Support/Path.h"
+#include "llvm/Support/Process.h"
 #include "llvm/Support/raw_ostream.h"
 #include <algorithm>
 #include <optional>
@@ -1307,14 +1308,14 @@ void TextDiagnostic::emitSnippet(StringRef SourceLine,
       std::optional<enum raw_ostream::Colors> H;
       for (auto &P : Styles) {
         if (P.Start < I && P.End >= I) {
-          H = P.c;
+          H = P.color;
           break;
         }
       }
 
-      if (H) {
+      if (H)
         OS.changeColor(*H, false);
-      } else
+      else
         OS.resetColor();
     }
 

>From 2ef42d04a15fb769f4b8a8f3748e93425b46f115 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Thu, 21 Sep 2023 11:01:43 +0200
Subject: [PATCH 05/21] Lex the entire file

---
 .../clang/Frontend/CodeSnippetHighlighter.h   |   7 +-
 clang/include/clang/Frontend/TextDiagnostic.h |   3 +-
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 138 ++++++++++++------
 clang/lib/Frontend/TextDiagnostic.cpp         |  14 +-
 4 files changed, 111 insertions(+), 51 deletions(-)

diff --git a/clang/include/clang/Frontend/CodeSnippetHighlighter.h b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
index c2a0184085d5da4..51c14880fb95485 100644
--- a/clang/include/clang/Frontend/CodeSnippetHighlighter.h
+++ b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
@@ -23,6 +23,8 @@ struct StyleRange {
 };
 
 class Preprocessor;
+class FileID;
+class SourceManager;
 
 class CodeSnippetHighlighter final {
 public:
@@ -31,9 +33,10 @@ class CodeSnippetHighlighter final {
   /// Produce StyleRanges for the given line.
   /// The returned vector contains non-overlapping style ranges. They are sorted
   /// from beginning of the line to the end.
-  std::vector<StyleRange> highlightLine(llvm::StringRef SourceLine,
+  std::vector<StyleRange> highlightLine(unsigned LineNumber,
                                         const Preprocessor *PP,
-                                        const LangOptions &LangOpts);
+                                        const LangOptions &LangOpts, FileID FID,
+                                        const SourceManager &SM);
 };
 
 } // namespace clang
diff --git a/clang/include/clang/Frontend/TextDiagnostic.h b/clang/include/clang/Frontend/TextDiagnostic.h
index 43c39ff96a2d1ce..102b33aedd5ef98 100644
--- a/clang/include/clang/Frontend/TextDiagnostic.h
+++ b/clang/include/clang/Frontend/TextDiagnostic.h
@@ -105,7 +105,8 @@ class TextDiagnostic : public DiagnosticRenderer {
                            ArrayRef<FixItHint> Hints);
 
   void emitSnippet(StringRef SourceLine, unsigned MaxLineNoDisplayWidth,
-                   unsigned LineNo);
+                   FileID FID, const SourceManager &SM, unsigned LineNo,
+                   unsigned DisplayLineNo);
 
   void emitParseableFixits(ArrayRef<FixItHint> Hints, const SourceManager &SM);
 };
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index dba7f5d2848505a..d319e690a355b2b 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -9,34 +9,47 @@
 
 using namespace clang;
 
-static SourceManager createTempSourceManager() {
-  FileSystemOptions FileOpts;
-  FileManager FileMgr(FileOpts);
-  llvm::IntrusiveRefCntPtr<DiagnosticIDs> DiagIDs(new DiagnosticIDs());
-  llvm::IntrusiveRefCntPtr<DiagnosticOptions> DiagOpts(new DiagnosticOptions());
-  DiagnosticsEngine diags(DiagIDs, DiagOpts);
-  return SourceManager(diags, FileMgr);
-}
-
-static Lexer createTempLexer(llvm::MemoryBufferRef B, SourceManager &FakeSM,
-                             const LangOptions &LangOpts) {
-  return Lexer(FakeSM.createFileID(B), B, FakeSM, LangOpts);
-}
+static constexpr raw_ostream::Colors CommentColor = raw_ostream::GREEN;
+static constexpr raw_ostream::Colors LiteralColor = raw_ostream::CYAN;
+static constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
 
 std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
-    StringRef SourceLine, const Preprocessor *PP, const LangOptions &LangOpts) {
+    unsigned LineNumber, const Preprocessor *PP, const LangOptions &LangOpts,
+    FileID FID, const SourceManager &SM) {
   if (!PP)
     return {};
-  constexpr raw_ostream::Colors CommentColor = raw_ostream::GREEN;
-  constexpr raw_ostream::Colors LiteralColor = raw_ostream::CYAN;
-  constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
 
-  SourceManager FakeSM = createTempSourceManager();
-  const auto MemBuf = llvm::MemoryBuffer::getMemBuffer(SourceLine);
-  Lexer L = createTempLexer(MemBuf->getMemBufferRef(), FakeSM, LangOpts);
+  // Classify the given token and append it to the given vector.
+  auto appendStyle = [PP, &LangOpts](std::vector<StyleRange> &Vec,
+                                     const Token &T, unsigned Start,
+                                     unsigned Length) -> void {
+    if (T.is(tok::raw_identifier)) {
+      StringRef RawIdent = T.getRawIdentifier();
+      // Special case true/false/nullptr literals, since they will otherwise be
+      // treated as keywords.
+      if (RawIdent == "true" || RawIdent == "false" || RawIdent == "nullptr") {
+        Vec.push_back(StyleRange{Start, Start + Length, LiteralColor});
+      } else {
+        const IdentifierInfo *II = PP->getIdentifierInfo(RawIdent);
+        assert(II);
+
+        if (II->isKeyword(LangOpts)) {
+          Vec.push_back(StyleRange{Start, Start + Length, KeywordColor});
+        }
+      }
+    } else if (tok::isLiteral(T.getKind())) {
+      Vec.push_back(StyleRange{Start, Start + Length, LiteralColor});
+    } else if (T.is(tok::comment)) {
+      Vec.push_back(StyleRange{Start, Start + Length, CommentColor});
+    }
+  };
+
+  auto Buff = SM.getBufferOrNone(FID);
+  assert(Buff);
+  Lexer L = Lexer(FID, *Buff, SM, LangOpts);
   L.SetKeepWhitespaceMode(true);
+  std::vector<std::vector<StyleRange>> Lines;
 
-  std::vector<StyleRange> Styles;
   bool Stop = false;
   while (!Stop) {
     Token T;
@@ -45,33 +58,74 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
       continue;
 
     bool Invalid;
-    unsigned Start =
-        FakeSM.getSpellingColumnNumber(T.getLocation(), &Invalid) - 1;
+    unsigned StartCol =
+        SM.getSpellingColumnNumber(T.getLocation(), &Invalid) - 1;
+    if (Invalid)
+      continue;
+    unsigned StartLine =
+        SM.getSpellingLineNumber(T.getLocation(), &Invalid) - 1;
     if (Invalid)
       continue;
 
-    if (T.is(tok::raw_identifier)) {
-      StringRef RawIdent = T.getRawIdentifier();
-      // Special case true/false/nullptr literals, since they will otherwise be
-      // treated as keywords.
-      if (RawIdent == "true" || RawIdent == "false" || RawIdent == "nullptr") {
-        Styles.push_back(
-            StyleRange{Start, Start + T.getLength(), LiteralColor});
-      } else {
-        const IdentifierInfo *II = PP->getIdentifierInfo(RawIdent);
-        assert(II);
+    while (Lines.size() <= StartLine)
+      Lines.push_back({});
 
-        if (II->isKeyword(LangOpts)) {
-          Styles.push_back(
-              StyleRange{Start, Start + T.getLength(), KeywordColor});
-        }
+    unsigned EndLine = SM.getSpellingLineNumber(T.getEndLoc(), &Invalid) - 1;
+    if (Invalid)
+      continue;
+
+    // Simple tokens.
+    if (StartLine == EndLine) {
+      appendStyle(Lines[StartLine], T, StartCol, T.getLength());
+      continue;
+    }
+    unsigned NumLines = EndLine - StartLine;
+
+    // For tokens that span multiple lines (think multiline comments), we
+    // divide them into multiple StyleRanges.
+    unsigned EndCol = SM.getSpellingColumnNumber(T.getEndLoc(), &Invalid) - 1;
+    if (Invalid)
+      continue;
+
+    std::string Spelling = Lexer::getSpelling(T, SM, LangOpts);
+
+    unsigned L = 0;
+    unsigned LineLength = 0;
+    for (unsigned I = 0; I <= Spelling.size(); ++I) {
+      // This line is done.
+      if (Spelling[I] == '\n' || Spelling[I] == '\r' || I == Spelling.size()) {
+        while (Lines.size() <= StartLine + L)
+          Lines.push_back({});
+
+        if (L == 0) // First line
+          appendStyle(Lines[StartLine + L], T, StartCol, LineLength);
+        else if (L == NumLines) // Last line
+          appendStyle(Lines[StartLine + L], T, 0, EndCol);
+        else
+          appendStyle(Lines[StartLine + L], T, 0, LineLength);
+        ++L;
+        LineLength = 0;
+        continue;
       }
-    } else if (tok::isLiteral(T.getKind())) {
-      Styles.push_back(StyleRange{Start, Start + T.getLength(), LiteralColor});
-    } else if (T.is(tok::comment)) {
-      Styles.push_back(StyleRange{Start, Start + T.getLength(), CommentColor});
+      ++LineLength;
+    }
+  }
+
+#if 0
+  llvm::errs() << "--\nLine Style info: \n";
+  int I = 0;
+  for (std::vector<StyleRange> &Line : Lines) {
+    llvm::errs() << I << ": ";
+    for (const auto &R : Line) {
+      llvm::errs() << "{" << R.Start << ", " << R.End << "}, ";
     }
+    llvm::errs() << "\n";
+
+    ++I;
   }
+#endif
 
-  return Styles;
+  while (Lines.size() <= LineNumber)
+    Lines.push_back({});
+  return Lines[LineNumber];
 }
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index 35a92a8044f2e52..5aea7b8f4210749 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -1249,7 +1249,8 @@ void TextDiagnostic::emitSnippetAndCaret(
     }
 
     // Emit what we have computed.
-    emitSnippet(SourceLine, MaxLineNoDisplayWidth, DisplayLineNo);
+    emitSnippet(SourceLine, MaxLineNoDisplayWidth, FID, SM, LineNo,
+                DisplayLineNo);
 
     if (!CaretLine.empty()) {
       indentForLineNumbers();
@@ -1278,16 +1279,17 @@ void TextDiagnostic::emitSnippetAndCaret(
 }
 
 void TextDiagnostic::emitSnippet(StringRef SourceLine,
-                                 unsigned MaxLineNoDisplayWidth,
-                                 unsigned LineNo) {
+                                 unsigned MaxLineNoDisplayWidth, FileID FID,
+                                 const SourceManager &SM, unsigned LineNo,
+                                 unsigned DisplayLineNo) {
   std::vector<StyleRange> Styles =
-      SnippetHighlighter.highlightLine(SourceLine, PP, LangOpts);
+      SnippetHighlighter.highlightLine(LineNo - 1, PP, LangOpts, FID, SM);
 
   // Emit line number.
   if (MaxLineNoDisplayWidth > 0) {
-    unsigned LineNoDisplayWidth = getNumDisplayWidth(LineNo);
+    unsigned LineNoDisplayWidth = getNumDisplayWidth(DisplayLineNo);
     OS.indent(MaxLineNoDisplayWidth - LineNoDisplayWidth + 1)
-        << LineNo << " | ";
+        << DisplayLineNo << " | ";
   }
 
   // Print the source line one character at a time.

>From 6b63886afb16e780b85937c329d02da620f14309 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Fri, 22 Sep 2023 06:48:55 +0200
Subject: [PATCH 06/21] Try to fix PCH test

---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index d319e690a355b2b..8905fbfb29b8927 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -19,6 +19,10 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
   if (!PP)
     return {};
 
+  // Might cause emission of another diagnostic.
+  if (PP->getIdentifierTable().getExternalIdentifierLookup())
+    return {};
+
   // Classify the given token and append it to the given vector.
   auto appendStyle = [PP, &LangOpts](std::vector<StyleRange> &Vec,
                                      const Token &T, unsigned Start,

>From 27a94727c9283a370d596612fb26347ac9b30df9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Tue, 26 Sep 2023 08:11:58 +0200
Subject: [PATCH 07/21] Measurements

---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 24 +++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 8905fbfb29b8927..17614a962ee6a7f 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -6,6 +6,7 @@
 #include "clang/Lex/Preprocessor.h"
 #include "clang/Lex/PreprocessorOptions.h"
 #include "llvm/Support/raw_ostream.h"
+#include <chrono>
 
 using namespace clang;
 
@@ -16,6 +17,9 @@ static constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
 std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
     unsigned LineNumber, const Preprocessor *PP, const LangOptions &LangOpts,
     FileID FID, const SourceManager &SM) {
+  std::chrono::steady_clock::time_point begin =
+      std::chrono::steady_clock::now();
+
   if (!PP)
     return {};
 
@@ -23,6 +27,7 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
   if (PP->getIdentifierTable().getExternalIdentifierLookup())
     return {};
 
+  size_t NTokens = 0;
   // Classify the given token and append it to the given vector.
   auto appendStyle = [PP, &LangOpts](std::vector<StyleRange> &Vec,
                                      const Token &T, unsigned Start,
@@ -56,6 +61,7 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
 
   bool Stop = false;
   while (!Stop) {
+    ++NTokens;
     Token T;
     Stop = L.LexFromRawLexer(T);
     if (T.is(tok::unknown))
@@ -131,5 +137,23 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
 
   while (Lines.size() <= LineNumber)
     Lines.push_back({});
+
+  std::chrono::steady_clock::time_point end = std::chrono::steady_clock::now();
+  llvm::errs() << "Lexed " << Lines.size() << " lines and " << NTokens
+               << " Tokens\n";
+  llvm::errs() << "That took "
+               << std::chrono::duration_cast<std::chrono::microseconds>(end -
+                                                                        begin)
+                      .count()
+               << " microseconds\n";
+  llvm::errs() << "That took "
+               << std::chrono::duration_cast<std::chrono::milliseconds>(end -
+                                                                        begin)
+                      .count()
+               << " milliseconds\n";
+  llvm::errs()
+      << "That took "
+      << std::chrono::duration_cast<std::chrono::seconds>(end - begin).count()
+      << " seconds\n";
   return Lines[LineNumber];
 }

>From 334ba2d626fbce02ff08adf2c9269edf455b3e90 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Wed, 27 Sep 2023 08:05:09 +0200
Subject: [PATCH 08/21] Slightly improve performance by bailing out earlier

---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 17614a962ee6a7f..71e5c30e56d146a 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -41,14 +41,13 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
       } else {
         const IdentifierInfo *II = PP->getIdentifierInfo(RawIdent);
         assert(II);
-
-        if (II->isKeyword(LangOpts)) {
+        if (II->isKeyword(LangOpts))
           Vec.push_back(StyleRange{Start, Start + Length, KeywordColor});
-        }
       }
     } else if (tok::isLiteral(T.getKind())) {
       Vec.push_back(StyleRange{Start, Start + Length, LiteralColor});
-    } else if (T.is(tok::comment)) {
+    } else {
+      assert(T.is(tok::comment));
       Vec.push_back(StyleRange{Start, Start + Length, CommentColor});
     }
   };
@@ -67,6 +66,11 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
     if (T.is(tok::unknown))
       continue;
 
+    // We are only interested in identifiers, literals and comments.
+    if (!T.is(tok::raw_identifier) && !T.is(tok::comment) &&
+        !tok::isLiteral(T.getKind()))
+      continue;
+
     bool Invalid;
     unsigned StartCol =
         SM.getSpellingColumnNumber(T.getLocation(), &Invalid) - 1;
@@ -138,6 +142,7 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
   while (Lines.size() <= LineNumber)
     Lines.push_back({});
 
+#if 0
   std::chrono::steady_clock::time_point end = std::chrono::steady_clock::now();
   llvm::errs() << "Lexed " << Lines.size() << " lines and " << NTokens
                << " Tokens\n";
@@ -155,5 +160,6 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
       << "That took "
       << std::chrono::duration_cast<std::chrono::seconds>(end - begin).count()
       << " seconds\n";
+#endif
   return Lines[LineNumber];
 }

>From 867b13df70d1e6b29156ec6eecf4a78f08c7e8a5 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Wed, 27 Sep 2023 10:45:36 +0200
Subject: [PATCH 09/21] Only care about tokens that touch our LineNumber.

---
 .../clang/Frontend/CodeSnippetHighlighter.h   |  2 +-
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 19 +++++++++++++------
 clang/lib/Frontend/TextDiagnostic.cpp         |  2 +-
 3 files changed, 15 insertions(+), 8 deletions(-)

diff --git a/clang/include/clang/Frontend/CodeSnippetHighlighter.h b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
index 51c14880fb95485..a65bd3991d4eff2 100644
--- a/clang/include/clang/Frontend/CodeSnippetHighlighter.h
+++ b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
@@ -19,7 +19,7 @@ namespace clang {
 struct StyleRange {
   unsigned Start;
   unsigned End;
-  const enum llvm::raw_ostream::Colors color;
+  const enum llvm::raw_ostream::Colors Color;
 };
 
 class Preprocessor;
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 71e5c30e56d146a..7663155c6c83923 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -71,23 +71,30 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
         !tok::isLiteral(T.getKind()))
       continue;
 
-    bool Invalid;
-    unsigned StartCol =
-        SM.getSpellingColumnNumber(T.getLocation(), &Invalid) - 1;
+    bool Invalid = false;
+    unsigned EndLine = SM.getSpellingLineNumber(T.getEndLoc(), &Invalid) - 1;
     if (Invalid)
       continue;
+
+    if (EndLine < LineNumber)
+      continue;
     unsigned StartLine =
         SM.getSpellingLineNumber(T.getLocation(), &Invalid) - 1;
     if (Invalid)
       continue;
+    if (StartLine > LineNumber)
+      break;
 
-    while (Lines.size() <= StartLine)
-      Lines.push_back({});
+    // Must have an intersection at this point
+    assert(StartLine <= LineNumber && EndLine >= LineNumber);
 
-    unsigned EndLine = SM.getSpellingLineNumber(T.getEndLoc(), &Invalid) - 1;
+    unsigned StartCol =
+        SM.getSpellingColumnNumber(T.getLocation(), &Invalid) - 1;
     if (Invalid)
       continue;
 
+    while (Lines.size() <= StartLine)
+      Lines.push_back({});
     // Simple tokens.
     if (StartLine == EndLine) {
       appendStyle(Lines[StartLine], T, StartCol, T.getLength());
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index 5aea7b8f4210749..f2793d23522f1a5 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -1310,7 +1310,7 @@ void TextDiagnostic::emitSnippet(StringRef SourceLine,
       std::optional<enum raw_ostream::Colors> H;
       for (auto &P : Styles) {
         if (P.Start < I && P.End >= I) {
-          H = P.color;
+          H = P.Color;
           break;
         }
       }

>From b6f546891fbe3993265b74897ad3d2666ef53180 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Fri, 6 Oct 2023 15:28:25 +0200
Subject: [PATCH 10/21] Add checkpoints to Preprocessor

---
 .../clang/Frontend/CodeSnippetHighlighter.h   | 13 ++--
 clang/include/clang/Frontend/TextDiagnostic.h |  2 +-
 clang/include/clang/Lex/Preprocessor.h        |  5 ++
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 64 +++++++++++--------
 clang/lib/Frontend/TextDiagnostic.cpp         | 10 +--
 clang/lib/Lex/Preprocessor.cpp                | 32 ++++++++++
 6 files changed, 86 insertions(+), 40 deletions(-)

diff --git a/clang/include/clang/Frontend/CodeSnippetHighlighter.h b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
index a65bd3991d4eff2..451a182b3e35317 100644
--- a/clang/include/clang/Frontend/CodeSnippetHighlighter.h
+++ b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
@@ -10,16 +10,15 @@
 #define LLVM_CLANG_FRONTEND_CODESNIPPETHIGHLIGHTER_H
 
 #include "clang/Basic/LangOptions.h"
-#include "llvm/ADT/SmallSet.h"
+#include "llvm/ADT/SmallVector.h"
 #include "llvm/Support/raw_ostream.h"
-#include <vector>
 
 namespace clang {
 
 struct StyleRange {
   unsigned Start;
   unsigned End;
-  const enum llvm::raw_ostream::Colors Color;
+  enum llvm::raw_ostream::Colors Color;
 };
 
 class Preprocessor;
@@ -33,10 +32,10 @@ class CodeSnippetHighlighter final {
   /// Produce StyleRanges for the given line.
   /// The returned vector contains non-overlapping style ranges. They are sorted
   /// from beginning of the line to the end.
-  std::vector<StyleRange> highlightLine(unsigned LineNumber,
-                                        const Preprocessor *PP,
-                                        const LangOptions &LangOpts, FileID FID,
-                                        const SourceManager &SM);
+  llvm::SmallVector<StyleRange>
+  highlightLine(unsigned LineNumber, const Preprocessor *PP,
+                const LangOptions &LangOpts, FileID FID,
+                const SourceManager &SM, const char *LineStart);
 };
 
 } // namespace clang
diff --git a/clang/include/clang/Frontend/TextDiagnostic.h b/clang/include/clang/Frontend/TextDiagnostic.h
index 102b33aedd5ef98..ecd5bb4a4f568dc 100644
--- a/clang/include/clang/Frontend/TextDiagnostic.h
+++ b/clang/include/clang/Frontend/TextDiagnostic.h
@@ -106,7 +106,7 @@ class TextDiagnostic : public DiagnosticRenderer {
 
   void emitSnippet(StringRef SourceLine, unsigned MaxLineNoDisplayWidth,
                    FileID FID, const SourceManager &SM, unsigned LineNo,
-                   unsigned DisplayLineNo);
+                   unsigned DisplayLineNo, const char *LineStart);
 
   void emitParseableFixits(ArrayRef<FixItHint> Hints, const SourceManager &SM);
 };
diff --git a/clang/include/clang/Lex/Preprocessor.h b/clang/include/clang/Lex/Preprocessor.h
index 18d88407ae12c90..c8bd2814a85d6f3 100644
--- a/clang/include/clang/Lex/Preprocessor.h
+++ b/clang/include/clang/Lex/Preprocessor.h
@@ -128,6 +128,7 @@ enum MacroUse {
 class Preprocessor {
   friend class VAOptDefinitionContext;
   friend class VariadicMacroScopeGuard;
+  friend class CodeSnippetHighlighter;
 
   llvm::unique_function<void(const clang::Token &)> OnToken;
   std::shared_ptr<PreprocessorOptions> PPOpts;
@@ -141,6 +142,10 @@ class Preprocessor {
   HeaderSearch      &HeaderInfo;
   ModuleLoader      &TheModuleLoader;
 
+  llvm::SmallVector<const char *> CheckPoints;
+  void saveCheckPoint(const char *P);
+  const char *getSaveFor(const char *S) const;
+
   /// External source of macros.
   ExternalPreprocessorSource *ExternalSource;
 
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 7663155c6c83923..28b66d4c05b8c8b 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -14,9 +14,9 @@ static constexpr raw_ostream::Colors CommentColor = raw_ostream::GREEN;
 static constexpr raw_ostream::Colors LiteralColor = raw_ostream::CYAN;
 static constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
 
-std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
+llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
     unsigned LineNumber, const Preprocessor *PP, const LangOptions &LangOpts,
-    FileID FID, const SourceManager &SM) {
+    FileID FID, const SourceManager &SM, const char *LineStart) {
   std::chrono::steady_clock::time_point begin =
       std::chrono::steady_clock::now();
 
@@ -29,7 +29,7 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
 
   size_t NTokens = 0;
   // Classify the given token and append it to the given vector.
-  auto appendStyle = [PP, &LangOpts](std::vector<StyleRange> &Vec,
+  auto appendStyle = [PP, &LangOpts](llvm::SmallVector<StyleRange> &Vec,
                                      const Token &T, unsigned Start,
                                      unsigned Length) -> void {
     if (T.is(tok::raw_identifier)) {
@@ -52,12 +52,23 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
     }
   };
 
+  // Figure out where to start lexing from.
   auto Buff = SM.getBufferOrNone(FID);
   assert(Buff);
   Lexer L = Lexer(FID, *Buff, SM, LangOpts);
   L.SetKeepWhitespaceMode(true);
-  std::vector<std::vector<StyleRange>> Lines;
 
+  // Seek to the last save point before the start of the line.
+  if (const char *Save = PP->getSaveFor(LineStart);
+      Buff->getBufferStart() <= Save && Save < Buff->getBufferEnd()) {
+    size_t Offset = Save - Buff->getBufferStart();
+    assert(Save >= Buff->getBufferStart());
+    assert(Save <= Buff->getBufferEnd());
+
+    L.seek(Offset, /*IsAtStartOfLine=*/true);
+  }
+
+  llvm::SmallVector<StyleRange> LineRanges;
   bool Stop = false;
   while (!Stop) {
     ++NTokens;
@@ -93,14 +104,13 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
     if (Invalid)
       continue;
 
-    while (Lines.size() <= StartLine)
-      Lines.push_back({});
     // Simple tokens.
     if (StartLine == EndLine) {
-      appendStyle(Lines[StartLine], T, StartCol, T.getLength());
+      appendStyle(LineRanges, T, StartCol, T.getLength());
       continue;
     }
     unsigned NumLines = EndLine - StartLine;
+    assert(NumLines >= 1);
 
     // For tokens that span multiple lines (think multiline comments), we
     // divide them into multiple StyleRanges.
@@ -115,15 +125,17 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
     for (unsigned I = 0; I <= Spelling.size(); ++I) {
       // This line is done.
       if (Spelling[I] == '\n' || Spelling[I] == '\r' || I == Spelling.size()) {
-        while (Lines.size() <= StartLine + L)
-          Lines.push_back({});
-
-        if (L == 0) // First line
-          appendStyle(Lines[StartLine + L], T, StartCol, LineLength);
-        else if (L == NumLines) // Last line
-          appendStyle(Lines[StartLine + L], T, 0, EndCol);
-        else
-          appendStyle(Lines[StartLine + L], T, 0, LineLength);
+        if (StartLine + L == LineNumber) {
+          if (L == 0) // First line
+            appendStyle(LineRanges, T, StartCol, LineLength);
+          else if (L == NumLines) // Last line
+            appendStyle(LineRanges, T, 0, EndCol);
+          else
+            appendStyle(LineRanges, T, 0, LineLength);
+
+          // We only do one line, so we're done.
+          break;
+        }
         ++L;
         LineLength = 0;
         continue;
@@ -134,25 +146,21 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
 
 #if 0
   llvm::errs() << "--\nLine Style info: \n";
-  int I = 0;
-  for (std::vector<StyleRange> &Line : Lines) {
-    llvm::errs() << I << ": ";
-    for (const auto &R : Line) {
+  //int I = 0;
+  //for (std::vector<StyleRange> &Line : Lines) {
+    //llvm::errs() << I << ": ";
+    for (const auto &R : LineRanges) {
       llvm::errs() << "{" << R.Start << ", " << R.End << "}, ";
     }
     llvm::errs() << "\n";
 
-    ++I;
-  }
+    //++I;
+  //}
 #endif
 
-  while (Lines.size() <= LineNumber)
-    Lines.push_back({});
-
 #if 0
   std::chrono::steady_clock::time_point end = std::chrono::steady_clock::now();
-  llvm::errs() << "Lexed " << Lines.size() << " lines and " << NTokens
-               << " Tokens\n";
+  llvm::errs() << "Lexed " << NTokens << " Tokens\n";
   llvm::errs() << "That took "
                << std::chrono::duration_cast<std::chrono::microseconds>(end -
                                                                         begin)
@@ -168,5 +176,5 @@ std::vector<StyleRange> CodeSnippetHighlighter::highlightLine(
       << std::chrono::duration_cast<std::chrono::seconds>(end - begin).count()
       << " seconds\n";
 #endif
-  return Lines[LineNumber];
+  return LineRanges;
 }
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index f2793d23522f1a5..cbc0cfacec20f0e 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -13,6 +13,7 @@
 #include "clang/Basic/SourceManager.h"
 #include "clang/Frontend/CodeSnippetHighlighter.h"
 #include "clang/Lex/Lexer.h"
+#include "clang/Lex/Preprocessor.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/StringExtras.h"
 #include "llvm/Support/ConvertUTF.h"
@@ -1250,7 +1251,7 @@ void TextDiagnostic::emitSnippetAndCaret(
 
     // Emit what we have computed.
     emitSnippet(SourceLine, MaxLineNoDisplayWidth, FID, SM, LineNo,
-                DisplayLineNo);
+                DisplayLineNo, LineStart);
 
     if (!CaretLine.empty()) {
       indentForLineNumbers();
@@ -1281,9 +1282,10 @@ void TextDiagnostic::emitSnippetAndCaret(
 void TextDiagnostic::emitSnippet(StringRef SourceLine,
                                  unsigned MaxLineNoDisplayWidth, FileID FID,
                                  const SourceManager &SM, unsigned LineNo,
-                                 unsigned DisplayLineNo) {
-  std::vector<StyleRange> Styles =
-      SnippetHighlighter.highlightLine(LineNo - 1, PP, LangOpts, FID, SM);
+                                 unsigned DisplayLineNo,
+                                 const char *LineStart) {
+  llvm::SmallVector<StyleRange> Styles = SnippetHighlighter.highlightLine(
+      LineNo - 1, PP, LangOpts, FID, SM, LineStart);
 
   // Emit line number.
   if (MaxLineNoDisplayWidth > 0) {
diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index ede4c51487ffbe7..e40ba5f33272bec 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -546,6 +546,7 @@ void Preprocessor::EnterMainSourceFile() {
   // information) and predefined macros aren't guaranteed to be set properly.
   assert(NumEnteredSourceFiles == 0 && "Cannot reenter the main file!");
   FileID MainFileID = SourceMgr.getMainFileID();
+  // llvm::errs() << "##### Main source file: " << (int)MainFileID << "\n";
 
   // If MainFileID is loaded it means we loaded an AST file, no need to enter
   // a main file.
@@ -878,15 +879,46 @@ bool Preprocessor::HandleIdentifier(Token &Identifier) {
   return true;
 }
 
+void Preprocessor::saveCheckPoint(const char *P) {
+  static constexpr ptrdiff_t Limit = 1000;
+  if (CheckPoints.empty()) {
+    CheckPoints.push_back(P);
+    return;
+  }
+
+  const char *Cur = CheckPoints.back();
+  if (Cur == P)
+    return;
+  if ((P - Cur) > Limit)
+    CheckPoints.push_back(P);
+}
+
+const char *Preprocessor::getSaveFor(const char *S) const {
+  const char *C = S;
+  // FIXME: Use std::lower_bound or something smart. Aaron knows what I'm
+  // talking about.
+  for (ssize_t I = CheckPoints.size() - 1; I >= 0; --I) {
+    C = CheckPoints[I];
+    if (CheckPoints[I] <= S)
+      break;
+  }
+  return C;
+}
+
 void Preprocessor::Lex(Token &Result) {
   ++LexLevel;
 
   // We loop here until a lex function returns a token; this avoids recursion.
   bool ReturnedToken;
+  // const char *Save = nullptr;
   do {
     switch (CurLexerKind) {
     case CLK_Lexer:
       ReturnedToken = CurLexer->Lex(Result);
+      if (ReturnedToken && CurLexer &&
+          CurLexer->getFileID() == SourceMgr.getMainFileID()) {
+        saveCheckPoint(CurLexer->BufferPtr);
+      }
       break;
     case CLK_TokenLexer:
       ReturnedToken = CurTokenLexer->Lex(Result);

>From e554e03d43757a11b58d6f9ac3830aa695446b37 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Sat, 7 Oct 2023 12:35:21 +0200
Subject: [PATCH 11/21] Add missing license header

---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 28b66d4c05b8c8b..042745fc639ec39 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -1,3 +1,10 @@
+//===-- CodeSnippetHighlighter.cpp - Code snippet highlighting --*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
 
 #include "clang/Frontend/CodeSnippetHighlighter.h"
 #include "clang/Basic/DiagnosticOptions.h"

>From ff0bb283ff9951fed3d8905d5ced21c20d98d9a1 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Tue, 10 Oct 2023 14:04:43 +0200
Subject: [PATCH 12/21] Fewer checkpoints

---
 clang/lib/Lex/Preprocessor.cpp | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index e40ba5f33272bec..74a57851e1a795c 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -880,7 +880,7 @@ bool Preprocessor::HandleIdentifier(Token &Identifier) {
 }
 
 void Preprocessor::saveCheckPoint(const char *P) {
-  static constexpr ptrdiff_t Limit = 1000;
+  static constexpr ptrdiff_t Limit = 1024 * 8;
   if (CheckPoints.empty()) {
     CheckPoints.push_back(P);
     return;
@@ -915,10 +915,8 @@ void Preprocessor::Lex(Token &Result) {
     switch (CurLexerKind) {
     case CLK_Lexer:
       ReturnedToken = CurLexer->Lex(Result);
-      if (ReturnedToken && CurLexer &&
-          CurLexer->getFileID() == SourceMgr.getMainFileID()) {
+      if (ReturnedToken && CurLexer->getFileID() == SourceMgr.getMainFileID())
         saveCheckPoint(CurLexer->BufferPtr);
-      }
       break;
     case CLK_TokenLexer:
       ReturnedToken = CurTokenLexer->Lex(Result);

>From 3fde8e9e2a4af494b95e41a3a0c5f485c6ef8261 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Fri, 13 Oct 2023 11:11:06 +0200
Subject: [PATCH 13/21] Cleanup

---
 clang/lib/Lex/Preprocessor.cpp | 16 ++++------------
 1 file changed, 4 insertions(+), 12 deletions(-)

diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index 74a57851e1a795c..f77a38d3e3113a0 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -546,7 +546,6 @@ void Preprocessor::EnterMainSourceFile() {
   // information) and predefined macros aren't guaranteed to be set properly.
   assert(NumEnteredSourceFiles == 0 && "Cannot reenter the main file!");
   FileID MainFileID = SourceMgr.getMainFileID();
-  // llvm::errs() << "##### Main source file: " << (int)MainFileID << "\n";
 
   // If MainFileID is loaded it means we loaded an AST file, no need to enter
   // a main file.
@@ -894,15 +893,8 @@ void Preprocessor::saveCheckPoint(const char *P) {
 }
 
 const char *Preprocessor::getSaveFor(const char *S) const {
-  const char *C = S;
-  // FIXME: Use std::lower_bound or something smart. Aaron knows what I'm
-  // talking about.
-  for (ssize_t I = CheckPoints.size() - 1; I >= 0; --I) {
-    C = CheckPoints[I];
-    if (CheckPoints[I] <= S)
-      break;
-  }
-  return C;
+  auto It = llvm::lower_bound(CheckPoints, S, std::less<const char *>());
+  return *It;
 }
 
 void Preprocessor::Lex(Token &Result) {
@@ -910,12 +902,12 @@ void Preprocessor::Lex(Token &Result) {
 
   // We loop here until a lex function returns a token; this avoids recursion.
   bool ReturnedToken;
-  // const char *Save = nullptr;
   do {
     switch (CurLexerKind) {
     case CLK_Lexer:
       ReturnedToken = CurLexer->Lex(Result);
-      if (ReturnedToken && CurLexer->getFileID() == SourceMgr.getMainFileID())
+      if (ReturnedToken && CurLexer &&
+          CurLexer->getFileID() == SourceMgr.getMainFileID())
         saveCheckPoint(CurLexer->BufferPtr);
       break;
     case CLK_TokenLexer:

>From 927fd55d1b03c77fca8a0543108ab71f6ad41a45 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Sun, 15 Oct 2023 17:00:36 +0200
Subject: [PATCH 14/21] Address some review comments

---
 clang/include/clang/Frontend/CodeSnippetHighlighter.h | 2 ++
 clang/lib/Frontend/CodeSnippetHighlighter.cpp         | 9 +++++----
 clang/lib/Frontend/TextDiagnostic.cpp                 | 1 -
 3 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/clang/include/clang/Frontend/CodeSnippetHighlighter.h b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
index 451a182b3e35317..cb3c96f69293795 100644
--- a/clang/include/clang/Frontend/CodeSnippetHighlighter.h
+++ b/clang/include/clang/Frontend/CodeSnippetHighlighter.h
@@ -19,6 +19,8 @@ struct StyleRange {
   unsigned Start;
   unsigned End;
   enum llvm::raw_ostream::Colors Color;
+  StyleRange(unsigned S, unsigned E, enum llvm::raw_ostream::Colors C)
+      : Start(S), End(E), Color(C){};
 };
 
 class Preprocessor;
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 042745fc639ec39..30c4c791cb4f824 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -7,6 +7,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "clang/Frontend/CodeSnippetHighlighter.h"
+#include "clang/Basic/CharInfo.h"
 #include "clang/Basic/DiagnosticOptions.h"
 #include "clang/Basic/SourceManager.h"
 #include "clang/Lex/Lexer.h"
@@ -49,13 +50,13 @@ llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
         const IdentifierInfo *II = PP->getIdentifierInfo(RawIdent);
         assert(II);
         if (II->isKeyword(LangOpts))
-          Vec.push_back(StyleRange{Start, Start + Length, KeywordColor});
+          Vec.emplace_back(Start, Start + Length, KeywordColor);
       }
     } else if (tok::isLiteral(T.getKind())) {
-      Vec.push_back(StyleRange{Start, Start + Length, LiteralColor});
+      Vec.emplace_back(Start, Start + Length, LiteralColor);
     } else {
       assert(T.is(tok::comment));
-      Vec.push_back(StyleRange{Start, Start + Length, CommentColor});
+      Vec.emplace_back(Start, Start + Length, CommentColor);
     }
   };
 
@@ -131,7 +132,7 @@ llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
     unsigned LineLength = 0;
     for (unsigned I = 0; I <= Spelling.size(); ++I) {
       // This line is done.
-      if (Spelling[I] == '\n' || Spelling[I] == '\r' || I == Spelling.size()) {
+      if (isVerticalWhitespace(Spelling[I]) || I == Spelling.size()) {
         if (StartLine + L == LineNumber) {
           if (L == 0) // First line
             appendStyle(LineRanges, T, StartCol, LineLength);
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index cbc0cfacec20f0e..033d21656b12724 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -20,7 +20,6 @@
 #include "llvm/Support/ErrorHandling.h"
 #include "llvm/Support/Locale.h"
 #include "llvm/Support/Path.h"
-#include "llvm/Support/Process.h"
 #include "llvm/Support/raw_ostream.h"
 #include <algorithm>
 #include <optional>

>From 36f2aff876a1b182b5fd736d3a04bb6707f2f81a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Mon, 16 Oct 2023 07:21:41 +0200
Subject: [PATCH 15/21] Fix highlighting and add another assertion

---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp |  1 +
 clang/lib/Lex/Preprocessor.cpp                | 14 ++++++++++++--
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 30c4c791cb4f824..7a3fdc1e0d16ea8 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -72,6 +72,7 @@ llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
     size_t Offset = Save - Buff->getBufferStart();
     assert(Save >= Buff->getBufferStart());
     assert(Save <= Buff->getBufferEnd());
+    assert(Save <= LineStart);
 
     L.seek(Offset, /*IsAtStartOfLine=*/true);
   }
diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index f77a38d3e3113a0..8bcd03dc5209ad5 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -892,9 +892,19 @@ void Preprocessor::saveCheckPoint(const char *P) {
     CheckPoints.push_back(P);
 }
 
+/// We want to always return a value lower than \p S.
+/// If there is no such checkpoint, return nullptr.
 const char *Preprocessor::getSaveFor(const char *S) const {
-  auto It = llvm::lower_bound(CheckPoints, S, std::less<const char *>());
-  return *It;
+  const char *Result = nullptr;
+  for (ssize_t I = CheckPoints.size() - 1; I >= 0; --I) {
+    const char *C = CheckPoints[I];
+    if (C <= S) {
+      Result = C;
+      break;
+    }
+  }
+
+  return Result;
 }
 
 void Preprocessor::Lex(Token &Result) {

>From e6f5b4035d47d3cba9ff6c3af064a05cb478c6da Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Mon, 16 Oct 2023 07:51:10 +0200
Subject: [PATCH 16/21] Change colors one last time

To match those used in LLDB
---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 7a3fdc1e0d16ea8..316a151c9943c90 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -18,9 +18,9 @@
 
 using namespace clang;
 
-static constexpr raw_ostream::Colors CommentColor = raw_ostream::GREEN;
-static constexpr raw_ostream::Colors LiteralColor = raw_ostream::CYAN;
-static constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
+static constexpr raw_ostream::Colors CommentColor = raw_ostream::MAGENTA;
+static constexpr raw_ostream::Colors LiteralColor = raw_ostream::RED;
+static constexpr raw_ostream::Colors KeywordColor = raw_ostream::GREEN;
 
 llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
     unsigned LineNumber, const Preprocessor *PP, const LangOptions &LangOpts,
@@ -45,7 +45,7 @@ llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
       // Special case true/false/nullptr literals, since they will otherwise be
       // treated as keywords.
       if (RawIdent == "true" || RawIdent == "false" || RawIdent == "nullptr") {
-        Vec.push_back(StyleRange{Start, Start + Length, LiteralColor});
+        Vec.emplace_back(Start, Start + Length, LiteralColor);
       } else {
         const IdentifierInfo *II = PP->getIdentifierInfo(RawIdent);
         assert(II);

>From a5cd2b4621690fbbbc4e855bb5001e22c3d5816a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Tue, 17 Oct 2023 05:35:29 +0200
Subject: [PATCH 17/21] Add isAtStartOfLine() check

---
 clang/lib/Lex/Preprocessor.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index 8bcd03dc5209ad5..f49cb815f65bdc8 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -916,7 +916,7 @@ void Preprocessor::Lex(Token &Result) {
     switch (CurLexerKind) {
     case CLK_Lexer:
       ReturnedToken = CurLexer->Lex(Result);
-      if (ReturnedToken && CurLexer &&
+      if (ReturnedToken && Result.isAtStartOfLine() && CurLexer &&
           CurLexer->getFileID() == SourceMgr.getMainFileID())
         saveCheckPoint(CurLexer->BufferPtr);
       break;

>From bc577170cbfc3fa7b9b0b253526e9b2461d4d469 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Tue, 17 Oct 2023 06:13:12 +0200
Subject: [PATCH 18/21] Address review comments

---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp |  2 +-
 clang/lib/Frontend/TextDiagnostic.cpp         | 15 +++++----------
 clang/lib/Lex/Preprocessor.cpp                | 18 +++++++-----------
 3 files changed, 13 insertions(+), 22 deletions(-)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 316a151c9943c90..715c113d519438d 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -20,7 +20,7 @@ using namespace clang;
 
 static constexpr raw_ostream::Colors CommentColor = raw_ostream::MAGENTA;
 static constexpr raw_ostream::Colors LiteralColor = raw_ostream::RED;
-static constexpr raw_ostream::Colors KeywordColor = raw_ostream::GREEN;
+static constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
 
 llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
     unsigned LineNumber, const Preprocessor *PP, const LangOptions &LangOpts,
diff --git a/clang/lib/Frontend/TextDiagnostic.cpp b/clang/lib/Frontend/TextDiagnostic.cpp
index 033d21656b12724..1378b3a0812e598 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -1306,18 +1306,13 @@ void TextDiagnostic::emitSnippet(StringRef SourceLine,
     if (!WasPrintable)
       HighlightingEnabled = false;
 
-    // FIXME: I hope we can do this in some nicer way.
     if (HighlightingEnabled) {
-      std::optional<enum raw_ostream::Colors> H;
-      for (auto &P : Styles) {
-        if (P.Start < I && P.End >= I) {
-          H = P.Color;
-          break;
-        }
-      }
+      const auto *CharStyle = llvm::find_if(Styles, [I](const StyleRange &R) {
+        return (R.Start < I && R.End >= I);
+      });
 
-      if (H)
-        OS.changeColor(*H, false);
+      if (CharStyle != Styles.end())
+        OS.changeColor(CharStyle->Color, false);
       else
         OS.resetColor();
     }
diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index f49cb815f65bdc8..5c9eaa5d9d559db 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -163,6 +163,8 @@ Preprocessor::Preprocessor(std::shared_ptr<PreprocessorOptions> PPOpts,
     PreambleConditionalStack.startRecording();
 
   MaxTokens = LangOpts.MaxTokens;
+
+  CheckPoints.push_back(nullptr);
 }
 
 Preprocessor::~Preprocessor() {
@@ -878,17 +880,11 @@ bool Preprocessor::HandleIdentifier(Token &Identifier) {
   return true;
 }
 
+static constexpr ptrdiff_t CheckPointLimit = 1024 * 8;
 void Preprocessor::saveCheckPoint(const char *P) {
-  static constexpr ptrdiff_t Limit = 1024 * 8;
-  if (CheckPoints.empty()) {
-    CheckPoints.push_back(P);
-    return;
-  }
-
-  const char *Cur = CheckPoints.back();
-  if (Cur == P)
-    return;
-  if ((P - Cur) > Limit)
+  assert(!CheckPoints.empty());
+  assert(CheckPoints.back() != P);
+  if ((P - CheckPoints.back()) > CheckPointLimit)
     CheckPoints.push_back(P);
 }
 
@@ -916,7 +912,7 @@ void Preprocessor::Lex(Token &Result) {
     switch (CurLexerKind) {
     case CLK_Lexer:
       ReturnedToken = CurLexer->Lex(Result);
-      if (ReturnedToken && Result.isAtStartOfLine() && CurLexer &&
+      if (ReturnedToken && CurLexer && Result.isAtStartOfLine() &&
           CurLexer->getFileID() == SourceMgr.getMainFileID())
         saveCheckPoint(CurLexer->BufferPtr);
       break;

>From 2a90e905d381f1e79a312c35e7d128c0d3acdde3 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Tue, 17 Oct 2023 07:44:11 +0200
Subject: [PATCH 19/21] Rename lexer API

---
 clang/include/clang/Lex/Preprocessor.h        | 9 ++++++---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 2 +-
 clang/lib/Lex/Preprocessor.cpp                | 4 ++--
 3 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/clang/include/clang/Lex/Preprocessor.h b/clang/include/clang/Lex/Preprocessor.h
index c8bd2814a85d6f3..593fee6803f9aa4 100644
--- a/clang/include/clang/Lex/Preprocessor.h
+++ b/clang/include/clang/Lex/Preprocessor.h
@@ -141,10 +141,7 @@ class Preprocessor {
   std::unique_ptr<ScratchBuffer> ScratchBuf;
   HeaderSearch      &HeaderInfo;
   ModuleLoader      &TheModuleLoader;
-
   llvm::SmallVector<const char *> CheckPoints;
-  void saveCheckPoint(const char *P);
-  const char *getSaveFor(const char *S) const;
 
   /// External source of macros.
   ExternalPreprocessorSource *ExternalSource;
@@ -1328,6 +1325,11 @@ class Preprocessor {
     OnToken = std::move(F);
   }
 
+  /// Returns a pointer into the main file's buffer that's guaranteed to be
+  /// after a fully lexed token. This can be used to partially lex a file
+  /// without starting in the middle of a token.
+  const char *getCompleteTokenCheckpoint(const char *P) const;
+
   void setPreprocessToken(bool Preprocess) { PreprocessToken = Preprocess; }
 
   bool isMacroDefined(StringRef Id) {
@@ -2268,6 +2270,7 @@ class Preprocessor {
 
   const char *getCurLexerEndPos();
   void diagnoseMissingHeaderInUmbrellaDir(const Module &Mod);
+  void saveCheckPoint(const char *P);
 
 public:
   void PoisonSEHIdentifiers(bool Poison = true); // Borland
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 715c113d519438d..1292469f80a0734 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -67,7 +67,7 @@ llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
   L.SetKeepWhitespaceMode(true);
 
   // Seek to the last save point before the start of the line.
-  if (const char *Save = PP->getSaveFor(LineStart);
+  if (const char *Save = PP->getCompleteTokenCheckpoint(LineStart);
       Buff->getBufferStart() <= Save && Save < Buff->getBufferEnd()) {
     size_t Offset = Save - Buff->getBufferStart();
     assert(Save >= Buff->getBufferStart());
diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index 5c9eaa5d9d559db..af42353df832fa8 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -890,11 +890,11 @@ void Preprocessor::saveCheckPoint(const char *P) {
 
 /// We want to always return a value lower than \p S.
 /// If there is no such checkpoint, return nullptr.
-const char *Preprocessor::getSaveFor(const char *S) const {
+const char *Preprocessor::getCompleteTokenCheckpoint(const char *P) const {
   const char *Result = nullptr;
   for (ssize_t I = CheckPoints.size() - 1; I >= 0; --I) {
     const char *C = CheckPoints[I];
-    if (C <= S) {
+    if (C <= P) {
       Result = C;
       break;
     }

>From 87efff0935e9d3d7741883ca68875d669b78784a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Fri, 20 Oct 2023 09:43:46 +0200
Subject: [PATCH 20/21] Just don't highlight in files >1MB

---
 clang/include/clang/Lex/Preprocessor.h        |  7 ----
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 34 ++++++++-----------
 clang/lib/Lex/Preprocessor.cpp                | 28 ---------------
 3 files changed, 15 insertions(+), 54 deletions(-)

diff --git a/clang/include/clang/Lex/Preprocessor.h b/clang/include/clang/Lex/Preprocessor.h
index 593fee6803f9aa4..57ae43ff14ab6c7 100644
--- a/clang/include/clang/Lex/Preprocessor.h
+++ b/clang/include/clang/Lex/Preprocessor.h
@@ -141,7 +141,6 @@ class Preprocessor {
   std::unique_ptr<ScratchBuffer> ScratchBuf;
   HeaderSearch      &HeaderInfo;
   ModuleLoader      &TheModuleLoader;
-  llvm::SmallVector<const char *> CheckPoints;
 
   /// External source of macros.
   ExternalPreprocessorSource *ExternalSource;
@@ -1325,11 +1324,6 @@ class Preprocessor {
     OnToken = std::move(F);
   }
 
-  /// Returns a pointer into the main file's buffer that's guaranteed to be
-  /// after a fully lexed token. This can be used to partially lex a file
-  /// without starting in the middle of a token.
-  const char *getCompleteTokenCheckpoint(const char *P) const;
-
   void setPreprocessToken(bool Preprocess) { PreprocessToken = Preprocess; }
 
   bool isMacroDefined(StringRef Id) {
@@ -2270,7 +2264,6 @@ class Preprocessor {
 
   const char *getCurLexerEndPos();
   void diagnoseMissingHeaderInUmbrellaDir(const Module &Mod);
-  void saveCheckPoint(const char *P);
 
 public:
   void PoisonSEHIdentifiers(bool Poison = true); // Borland
diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index 1292469f80a0734..a1ca68227323469 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -18,9 +18,15 @@
 
 using namespace clang;
 
-static constexpr raw_ostream::Colors CommentColor = raw_ostream::MAGENTA;
-static constexpr raw_ostream::Colors LiteralColor = raw_ostream::RED;
+// Magenta is taken for 'warning'. Red is already 'error' and 'cya'
+// is already taken for 'note'. Green is already used to underline
+// source ranges. White and black are bad because of the usual
+// terminal backgrounds. Which leaves us only with TWO options.
+static constexpr raw_ostream::Colors CommentColor = raw_ostream::YELLOW;
+static constexpr raw_ostream::Colors LiteralColor = raw_ostream::GREEN;
 static constexpr raw_ostream::Colors KeywordColor = raw_ostream::BLUE;
+/// Maximum size of file we still highlight.
+static constexpr size_t MaxBufferSize = 1024 * 1024; // 1MB.
 
 llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
     unsigned LineNumber, const Preprocessor *PP, const LangOptions &LangOpts,
@@ -35,6 +41,13 @@ llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
   if (PP->getIdentifierTable().getExternalIdentifierLookup())
     return {};
 
+  auto Buff = SM.getBufferOrNone(FID);
+  if (!Buff || Buff->getBufferSize() > MaxBufferSize)
+    return {};
+
+  Lexer L = Lexer(FID, *Buff, SM, LangOpts);
+  L.SetKeepWhitespaceMode(true);
+
   size_t NTokens = 0;
   // Classify the given token and append it to the given vector.
   auto appendStyle = [PP, &LangOpts](llvm::SmallVector<StyleRange> &Vec,
@@ -60,23 +73,6 @@ llvm::SmallVector<StyleRange> CodeSnippetHighlighter::highlightLine(
     }
   };
 
-  // Figure out where to start lexing from.
-  auto Buff = SM.getBufferOrNone(FID);
-  assert(Buff);
-  Lexer L = Lexer(FID, *Buff, SM, LangOpts);
-  L.SetKeepWhitespaceMode(true);
-
-  // Seek to the last save point before the start of the line.
-  if (const char *Save = PP->getCompleteTokenCheckpoint(LineStart);
-      Buff->getBufferStart() <= Save && Save < Buff->getBufferEnd()) {
-    size_t Offset = Save - Buff->getBufferStart();
-    assert(Save >= Buff->getBufferStart());
-    assert(Save <= Buff->getBufferEnd());
-    assert(Save <= LineStart);
-
-    L.seek(Offset, /*IsAtStartOfLine=*/true);
-  }
-
   llvm::SmallVector<StyleRange> LineRanges;
   bool Stop = false;
   while (!Stop) {
diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index af42353df832fa8..ede4c51487ffbe7 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -163,8 +163,6 @@ Preprocessor::Preprocessor(std::shared_ptr<PreprocessorOptions> PPOpts,
     PreambleConditionalStack.startRecording();
 
   MaxTokens = LangOpts.MaxTokens;
-
-  CheckPoints.push_back(nullptr);
 }
 
 Preprocessor::~Preprocessor() {
@@ -880,29 +878,6 @@ bool Preprocessor::HandleIdentifier(Token &Identifier) {
   return true;
 }
 
-static constexpr ptrdiff_t CheckPointLimit = 1024 * 8;
-void Preprocessor::saveCheckPoint(const char *P) {
-  assert(!CheckPoints.empty());
-  assert(CheckPoints.back() != P);
-  if ((P - CheckPoints.back()) > CheckPointLimit)
-    CheckPoints.push_back(P);
-}
-
-/// We want to always return a value lower than \p S.
-/// If there is no such checkpoint, return nullptr.
-const char *Preprocessor::getCompleteTokenCheckpoint(const char *P) const {
-  const char *Result = nullptr;
-  for (ssize_t I = CheckPoints.size() - 1; I >= 0; --I) {
-    const char *C = CheckPoints[I];
-    if (C <= P) {
-      Result = C;
-      break;
-    }
-  }
-
-  return Result;
-}
-
 void Preprocessor::Lex(Token &Result) {
   ++LexLevel;
 
@@ -912,9 +887,6 @@ void Preprocessor::Lex(Token &Result) {
     switch (CurLexerKind) {
     case CLK_Lexer:
       ReturnedToken = CurLexer->Lex(Result);
-      if (ReturnedToken && CurLexer && Result.isAtStartOfLine() &&
-          CurLexer->getFileID() == SourceMgr.getMainFileID())
-        saveCheckPoint(CurLexer->BufferPtr);
       break;
     case CLK_TokenLexer:
       ReturnedToken = CurTokenLexer->Lex(Result);

>From 0989eb91d9ac5d36d116776cf1c7fe815fa40f82 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= <tbaeder at redhat.com>
Date: Sun, 29 Oct 2023 09:22:47 +0100
Subject: [PATCH 21/21] Fix a typo

---
 clang/lib/Frontend/CodeSnippetHighlighter.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/clang/lib/Frontend/CodeSnippetHighlighter.cpp b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
index a1ca68227323469..90ab5a4927efb39 100644
--- a/clang/lib/Frontend/CodeSnippetHighlighter.cpp
+++ b/clang/lib/Frontend/CodeSnippetHighlighter.cpp
@@ -18,7 +18,7 @@
 
 using namespace clang;
 
-// Magenta is taken for 'warning'. Red is already 'error' and 'cya'
+// Magenta is taken for 'warning'. Red is already 'error' and 'cyan'
 // is already taken for 'note'. Green is already used to underline
 // source ranges. White and black are bad because of the usual
 // terminal backgrounds. Which leaves us only with TWO options.