llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-clang
<details>
<summary>Changes</summary>
Add some primitive syntax highlighting to our code snippet output.
Before:
After:
_Obviously_ this is kinda WIP and more of a hack in general, but IMO it
increases readability of the source snippets (which people look at highlighted
all the time anyway...) and LLDB does something similar, so let's see.
--
Full diff: https://github.com/llvm/llvm-project/pull/66514.diff
5 Files Affected:
- (added) clang/include/clang/Frontend/CodeSnippetHighlighter.h (+46)
- (modified) clang/include/clang/Frontend/TextDiagnostic.h (+3-1)
- (modified) clang/lib/Frontend/CMakeLists.txt (+1)
- (added) clang/lib/Frontend/CodeSnippetHighlighter.cpp (+120)
- (modified) clang/lib/Frontend/TextDiagnostic.cpp (+29-2)
<pre>
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 &quot;clang/Basic/LangOptions.h&quot;
+#include &quot;llvm/ADT/SmallSet.h&quot;
+#include &quot;llvm/Support/raw_ostream.h&quot;
+#include &lt;vector&gt;
+
+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&lt;StyleRange&gt; highlightLine(llvm::StringRef
SourceLine,
+ const LangOptions &amp;LangOpts);
+
+private:
+ bool Initialized = false;
+ /// Fills Keywords and Literals.
+ void ensureTokenData();
+
+ llvm::SmallSet&lt;StringRef, 12&gt; Keywords;
+ llvm::SmallSet&lt;StringRef, 12&gt; Literals;
+};
+
+} // namespace clang
+
+#endif
diff --git a/clang/include/clang/Frontend/TextDiagnostic.h
b/clang/include/clang/Frontend/TextDiagnostic.h
index 7eb0ab0cdc9bca8..39e09fe553dd4b9 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 &quot;clang/Frontend/CodeSnippetHighlighter.h&quot;
#include &quot;clang/Frontend/DiagnosticRenderer.h&quot;
namespace clang {
@@ -33,6 +34,7 @@ namespace clang {
/// printing coming out of libclang.
class TextDiagnostic : public DiagnosticRenderer {
raw_ostream &amp;OS;
+ CodeSnippetHighlighter SnippetHighlighter;
public:
TextDiagnostic(raw_ostream &amp;OS,
@@ -104,7 +106,7 @@ class TextDiagnostic : public DiagnosticRenderer {
ArrayRef&lt;FixItHint&gt; Hints);
void emitSnippet(StringRef SourceLine, unsigned MaxLineNoDisplayWidth,
- unsigned LineNo);
+ unsigned LineNo, bool A, const SourceManager &amp;SM);
void emitParseableFixits(ArrayRef&lt;FixItHint&gt; Hints, const
SourceManager &amp;SM);
};
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 &quot;clang/Frontend/CodeSnippetHighlighter.h&quot;
+#include &quot;clang/Basic/DiagnosticOptions.h&quot;
+#include &quot;clang/Basic/SourceManager.h&quot;
+#include &quot;clang/Lex/Lexer.h&quot;
+#include &quot;llvm/Support/raw_ostream.h&quot;
+
+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(&quot;_Static_assert&quot;);
+ Keywords.insert(&quot;auto&quot;);
+ Keywords.insert(&quot;concept&quot;);
+ Keywords.insert(&quot;const&quot;);
+ Keywords.insert(&quot;consteval&quot;);
+ Keywords.insert(&quot;constexpr&quot;);
+ Keywords.insert(&quot;delete&quot;);
+ Keywords.insert(&quot;do&quot;);
+ Keywords.insert(&quot;else&quot;);
+ Keywords.insert(&quot;final&quot;);
+ Keywords.insert(&quot;for&quot;);
+ Keywords.insert(&quot;if&quot;);
+ Keywords.insert(&quot;mutable&quot;);
+ Keywords.insert(&quot;namespace&quot;);
+ Keywords.insert(&quot;new&quot;);
+ Keywords.insert(&quot;private&quot;);
+ Keywords.insert(&quot;public&quot;);
+ Keywords.insert(&quot;requires&quot;);
+ Keywords.insert(&quot;return&quot;);
+ Keywords.insert(&quot;static&quot;);
+ Keywords.insert(&quot;static_assert&quot;);
+ Keywords.insert(&quot;using&quot;);
+ Keywords.insert(&quot;void&quot;);
+ Keywords.insert(&quot;volatile&quot;);
+ Keywords.insert(&quot;while&quot;);
+
+ // Builtin types we highlight
+ Keywords.insert(&quot;void&quot;);
+ Keywords.insert(&quot;char&quot;);
+ Keywords.insert(&quot;short&quot;);
+ Keywords.insert(&quot;int&quot;);
+ Keywords.insert(&quot;unsigned&quot;);
+ Keywords.insert(&quot;long&quot;);
+ Keywords.insert(&quot;float&quot;);
+ Keywords.insert(&quot;double&quot;);
+
+ Literals.insert(&quot;true&quot;);
+ Literals.insert(&quot;false&quot;);
+ Literals.insert(&quot;nullptr&quot;);
+
+ Initialized = true;
+}
+
+static SourceManager createTempSourceManager() {
+ FileSystemOptions FileOpts;
+ FileManager FileMgr(FileOpts);
+ llvm::IntrusiveRefCntPtr&lt;DiagnosticIDs&gt; DiagIDs(new
DiagnosticIDs());
+ llvm::IntrusiveRefCntPtr&lt;DiagnosticOptions&gt; DiagOpts(new
DiagnosticOptions());
+ DiagnosticsEngine diags(DiagIDs, DiagOpts);
+ return SourceManager(diags, FileMgr);
+}
+
+static Lexer createTempLexer(llvm::MemoryBufferRef B, SourceManager
&amp;FakeSM,
+ const LangOptions &amp;LangOpts) {
+ return Lexer(FakeSM.createFileID(B), B, FakeSM, LangOpts);
+}
+
+std::vector&lt;StyleRange&gt;
+CodeSnippetHighlighter::highlightLine(StringRef SourceLine,
+ const LangOptions &amp;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-&gt;getMemBufferRef(), FakeSM,
LangOpts);
+ L.SetKeepWhitespaceMode(true);
+
+ std::vector&lt;StyleRange&gt; 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(), &amp;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 eaa6e8d29a1dece..a7a2405d9bbae6d 100644
--- a/clang/lib/Frontend/TextDiagnostic.cpp
+++ b/clang/lib/Frontend/TextDiagnostic.cpp
@@ -11,6 +11,7 @@
#include &quot;clang/Basic/DiagnosticOptions.h&quot;
#include &quot;clang/Basic/FileManager.h&quot;
#include &quot;clang/Basic/SourceManager.h&quot;
+#include &quot;clang/Frontend/CodeSnippetHighlighter.h&quot;
#include &quot;clang/Lex/Lexer.h&quot;
#include &quot;llvm/ADT/SmallString.h&quot;
#include &quot;llvm/ADT/StringExtras.h&quot;
@@ -1248,7 +1249,7 @@ void TextDiagnostic::emitSnippetAndCaret(
}
// Emit what we have computed.
- emitSnippet(SourceLine, MaxLineNoDisplayWidth, DisplayLineNo);
+ emitSnippet(SourceLine, MaxLineNoDisplayWidth, DisplayLineNo, false, SM);
if (!CaretLine.empty()) {
indentForLineNumbers();
@@ -1278,7 +1279,11 @@ void TextDiagnostic::emitSnippetAndCaret(
void TextDiagnostic::emitSnippet(StringRef SourceLine,
unsigned MaxLineNoDisplayWidth,
- unsigned LineNo) {
+ unsigned LineNo, bool IsSlash,
+ const SourceManager &amp;SM) {
+ std::vector&lt;StyleRange&gt; Styles =
+ SnippetHighlighter.highlightLine(SourceLine, LangOpts);
+
// Emit line number.
if (MaxLineNoDisplayWidth &gt; 0) {
unsigned LineNoDisplayWidth = getNumDisplayWidth(LineNo);
@@ -1288,11 +1293,33 @@ void TextDiagnostic::emitSnippet(StringRef SourceLine,
// Print the source line one character at a time.
bool PrintReversed = false;
+ bool HighlightingEnabled = DiagOpts-&gt;ShowColors;
size_t I = 0;
while (I &lt; SourceLine.size()) {
auto [Str, WasPrintable] =
printableTextForNextCharacter(SourceLine, &amp;I,
DiagOpts-&gt;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&lt;enum raw_ostream::Colors&gt; H;
+ for (auto &amp;P : Styles) {
+ if (P.Start &lt; I &amp;&amp; P.End &gt;= 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-&gt;ShowColors) {
if (WasPrintable == PrintReversed) {
</pre>
</details>
https://github.com/llvm/llvm-project/pull/66514
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
