[clang] b9924c7 - [clang] Make -dump-tokens option align tokens (#164894)
via cfe-commits
cfe-commits at lists.llvm.org
Fri Apr 3 05:33:41 PDT 2026
Author: alexpaniman
Date: 2026-04-03T08:33:36-04:00
New Revision: b9924c76da9126cdf1e4345cbb6b061c65c920db
URL: https://github.com/llvm/llvm-project/commit/b9924c76da9126cdf1e4345cbb6b061c65c920db
DIFF: https://github.com/llvm/llvm-project/commit/b9924c76da9126cdf1e4345cbb6b061c65c920db.diff
LOG: [clang] Make -dump-tokens option align tokens (#164894)
When using `-Xclang -dump-tokens`, the lexer dump output is currently
difficult to read because the data are misaligned. The existing
implementation simply separates the token name, spelling, flags, and
location using `'\t'`, which results in inconsistent spacing.
For example, the current output looks like this on provided in this
patch example **(BEFORE THIS PR)**:
<img width="2936" height="632" alt="image"
src="https://github.com/user-attachments/assets/ad893958-6d57-4a76-8838-7fc56e37e6a7"
/>
# Changes
This small PR improves the readability of the token dump by:
+ Adding padding after the token name and after the spelling (the
padding amount was chosen empirically to produce good average
alignment).
+ Swapping the order of location and flags (since flags can take up a
lot of space and disrupt alignment).
The result is a more readable output **(AFTER THIS PR)**:
<img width="1470" height="315" alt="image"
src="https://github.com/user-attachments/assets/c24f24e5-a431-42cc-b5b6-232bac5c635e"
/>
Added:
clang/test/Preprocessor/dump-tokens.cpp
Modified:
clang/lib/Lex/Preprocessor.cpp
Removed:
################################################################################
diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index 2cfefac1052a4..c430da67c1469 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -61,6 +61,7 @@
#include "llvm/ADT/StringRef.h"
#include "llvm/Support/Capacity.h"
#include "llvm/Support/ErrorHandling.h"
+#include "llvm/Support/FormatVariadic.h"
#include "llvm/Support/MemoryBuffer.h"
#include "llvm/Support/MemoryBufferRef.h"
#include "llvm/Support/SaveAndRestore.h"
@@ -240,14 +241,59 @@ void Preprocessor::FinalizeForModelFile() {
}
void Preprocessor::DumpToken(const Token &Tok, bool DumpFlags) const {
- llvm::errs() << tok::getTokenName(Tok.getKind());
+ std::string TokenStr;
+ llvm::raw_string_ostream OS(TokenStr);
+
+ // The alignment of 16 is chosen to comfortably fit most identifiers.
+ OS << llvm::formatv("{0,-16} ", tok::getTokenName(Tok.getKind()));
+
+ // Annotation tokens are just markers that don't have a spelling -- they
+ // indicate where something expanded.
+ if (!Tok.isAnnotation()) {
+ OS << "'";
+ // Escape string to prevent token spelling from spanning multiple lines.
+ OS.write_escaped(getSpelling(Tok));
+ OS << "'";
+ }
- if (!Tok.isAnnotation())
- llvm::errs() << " '" << getSpelling(Tok) << "'";
+ // The alignment of 48 (32 characters for the spelling + the 16 for
+ // the identifier name) fits most variable names, keywords and annotations.
+ llvm::errs() << llvm::formatv("{0,-48} ", OS.str());
if (!DumpFlags) return;
- llvm::errs() << "\t";
+ auto Loc = Tok.getLocation();
+ llvm::errs() << "Loc=<";
+ DumpLocation(Loc);
+ llvm::errs() << ">";
+
+ // If the token points directly to a file location (i.e. not a macro
+ // expansion), then add additional padding so that trailing markers
+ // align, provided the line/column numbers are reasonably sized.
+ //
+ // Otherwise, if it's a macro expansion, don't bother with alignment,
+ // as the line will include multiple locations and be very long.
+ //
+ // NOTE: To keep this stateless, it doesn't account for filename
+ // length, so when a header starts markers will be temporarily misaligned.
+ if (Loc.isFileID()) {
+ PresumedLoc PLoc = SourceMgr.getPresumedLoc(Loc);
+
+ if (!PLoc.isInvalid()) {
+ int LineWidth = llvm::utostr(PLoc.getLine()).size();
+ int ColumnWidth = llvm::utostr(PLoc.getColumn()).size();
+
+ // Reserve space for lines up to 9999 and columns up to 99,
+ // which is 4 + 2 = 6 characters in total.
+ const int ReservedSpace = 6;
+
+ int LeftSpace = ReservedSpace - LineWidth - ColumnWidth;
+ int Padding = std::max<int>(0, LeftSpace);
+
+ llvm::errs().indent(Padding);
+ }
+ }
+
if (Tok.isAtStartOfLine())
llvm::errs() << " [StartOfLine]";
if (Tok.hasLeadingSpace())
@@ -256,13 +302,8 @@ void Preprocessor::DumpToken(const Token &Tok, bool DumpFlags) const {
llvm::errs() << " [ExpandDisabled]";
if (Tok.needsCleaning()) {
const char *Start = SourceMgr.getCharacterData(Tok.getLocation());
- llvm::errs() << " [UnClean='" << StringRef(Start, Tok.getLength())
- << "']";
+ llvm::errs() << " [UnClean='" << StringRef(Start, Tok.getLength()) << "']";
}
-
- llvm::errs() << "\tLoc=<";
- DumpLocation(Tok.getLocation());
- llvm::errs() << ">";
}
void Preprocessor::DumpLocation(SourceLocation Loc) const {
diff --git a/clang/test/Preprocessor/dump-tokens.cpp b/clang/test/Preprocessor/dump-tokens.cpp
new file mode 100644
index 0000000000000..315548530d4df
--- /dev/null
+++ b/clang/test/Preprocessor/dump-tokens.cpp
@@ -0,0 +1,37 @@
+// RUN: %clang_cc1 -dump-tokens %s 2>&1 | FileCheck %s --strict-whitespace
+
+// To make location reporting in the test more robust, provide line number and file name explicitly.
+#line 2 "dump-tokens.cpp"
+
+// Different kinds of identifiers with
diff erent spelling lengths.
+-> // CHECK: arrow '->' Loc=<{{.*}}:4:1> [StartOfLine]
+5 // CHECK-NEXT: numeric_constant '5' Loc=<{{.*}}:5:1> [StartOfLine]
+id // CHECK-NEXT: identifier 'id' Loc=<{{.*}}:6:1> [StartOfLine]
+& // CHECK-NEXT: amp '&' Loc=<{{.*}}:7:1> [StartOfLine]
+) // CHECK-NEXT: r_paren ')' Loc=<{{.*}}:8:1> [StartOfLine]
+unsigned // CHECK-NEXT: unsigned 'unsigned' Loc=<{{.*}}:9:1> [StartOfLine]
+~ // CHECK-NEXT: tilde '~' Loc=<{{.*}}:10:1> [StartOfLine]
+long_variable_name_very_long // CHECK-NEXT: identifier 'long_variable_name_very_long' Loc=<{{.*}}:11:1> [StartOfLine]
+union // CHECK-NEXT: union 'union' Loc=<{{.*}}:12:1> [StartOfLine]
+42 // CHECK-NEXT: numeric_constant '42' Loc=<{{.*}}:13:1> [StartOfLine]
+j // CHECK-NEXT: identifier 'j' Loc=<{{.*}}:14:1> [StartOfLine]
+&= // CHECK-NEXT: ampequal '&=' Loc=<{{.*}}:15:1> [StartOfLine]
+15 // CHECK-NEXT: numeric_constant '15' Loc=<{{.*}}:16:1> [StartOfLine]
+
+// Different locations in line and trailing markers.
+ at
diff erent locations= in line // CHECK-NEXT: identifier 'at' Loc=<{{.*}}:19:2> [StartOfLine] [LeadingSpace]
+ // CHECK-NEXT: identifier '
diff erent' Loc=<{{.*}}:19:5> [LeadingSpace]
+ // CHECK-NEXT: identifier 'locations' Loc=<{{.*}}:19:15> [LeadingSpace]
+ // CHECK-NEXT: equal '=' Loc=<{{.*}}:19:24>
+ // CHECK-NEXT: identifier 'in' Loc=<{{.*}}:19:26> [LeadingSpace]
+ // CHECK-NEXT: identifier 'line' Loc=<{{.*}}:19:29> [LeadingSpace]
+
+// Tokens that require escaping & annotations.
+#pragma clang __debug parser_crash // CHECK-NEXT: annot_pragma_parser_crash Loc=<{{.*}}:27:23>
+ // CHECK-NEXT: eod '\n' Loc=<{{.*}}:27:119> [LeadingSpace]
+#pragma clang __debug captured // CHECK-NEXT: annot_pragma_captured Loc=<{{.*}}:29:120>
+#pragma clang __debug dump X // CHECK-NEXT: annot_pragma_dump Loc=<{{.*}}:30:23>
+ // CHECK-NEXT: identifier 'X' Loc=<{{.*}}:30:28> [LeadingSpace]
+ // CHECK-NEXT: eod '\n' Loc=<{{.*}}:30:119> [LeadingSpace]
+ // CHECK-NEXT: eof '' Loc=<{{.*}}:34:1>
+
More information about the cfe-commits
mailing list