[clang-format] Support globstar in .clang-format-ignore #121404

owenca · 2024-12-31T16:22:14Z

Closes #110160.
Closes #114969.

llvmbot · 2024-12-31T16:22:45Z

@llvm/pr-subscribers-clang

@llvm/pr-subscribers-clang-format

Author: Owen Pan (owenca)

Changes

Closes #114969.

Full diff: https://github.com/llvm/llvm-project/pull/121404.diff

4 Files Affected:

(modified) clang/docs/ClangFormat.rst (+4)
(modified) clang/docs/ReleaseNotes.rst (+1)
(modified) clang/lib/Format/MatchFilePath.cpp (+24-11)
(modified) clang/unittests/Format/MatchFilePathTest.cpp (+35)

diff --git a/clang/docs/ClangFormat.rst b/clang/docs/ClangFormat.rst
index c8f1d7f5a77581..21cf68236b8917 100644
--- a/clang/docs/ClangFormat.rst
+++ b/clang/docs/ClangFormat.rst
@@ -150,6 +150,10 @@ names. It has the following format:
 * Patterns follow the rules specified in `POSIX 2.13.1, 2.13.2, and Rule 1 of
   2.13.3 <https://pubs.opengroup.org/onlinepubs/9699919799/utilities/
   V3_chap02.html#tag_18_13>`_.
+* Bash globstar is supported, i.e. "two adjacent ``*``s used as a single pattern
+  will match all files and zero or more directories and subdirectories. If
+  followed by a ``/``, two adjacent ``*``s will match only directories and
+  subdirectories."
 * A pattern is negated if it starts with a bang (``!``).
 
 To match all files in a directory, use e.g. ``foo/bar/*``. To match all files in
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index b7da12bcf65818..2d59d74ddf35fc 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -1124,6 +1124,7 @@ clang-format
 - Adds ``RemoveEmptyLinesInUnwrappedLines`` option.
 - Adds ``KeepFormFeed`` option and set it to ``true`` for ``GNU`` style.
 - Adds ``AllowShortNamespacesOnASingleLine`` option.
+- Adds support for bash globstar in ``.clang-format-ignore``.
 
 libclang
 --------
diff --git a/clang/lib/Format/MatchFilePath.cpp b/clang/lib/Format/MatchFilePath.cpp
index 062b334dcdd8fd..aca3433dc7fa78 100644
--- a/clang/lib/Format/MatchFilePath.cpp
+++ b/clang/lib/Format/MatchFilePath.cpp
@@ -49,25 +49,38 @@ bool matchFilePath(StringRef Pattern, StringRef FilePath) {
         return false;
       break;
     case '*': {
-      while (++I < EOP && Pattern[I] == '*') { // Skip consecutive stars.
+      const bool MaybeGlobstar = I == 0 || Pattern[I - 1] == Separator;
+      int StarCount = 1;
+      for (; ++I < EOP && Pattern[I] == '*'; ++StarCount) {
+        // Skip consecutive stars.
       }
       const auto K = FilePath.find(Separator, J); // Index of next `Separator`.
       const bool NoMoreSeparatorsInFilePath = K == StringRef::npos;
+      bool Globstar = MaybeGlobstar && StarCount == 2;
       if (I == EOP) // `Pattern` ends with a star.
-        return NoMoreSeparatorsInFilePath;
-      // `Pattern` ends with a lone backslash.
-      if (Pattern[I] == '\\' && ++I == EOP)
-        return false;
+        return Globstar || NoMoreSeparatorsInFilePath;
+      if (Pattern[I] != Separator) {
+        Globstar = false;
+        // `Pattern` ends with a lone backslash.
+        if (Pattern[I] == '\\' && ++I == EOP)
+          return false;
+      }
       // The star is followed by a (possibly escaped) `Separator`.
       if (Pattern[I] == Separator) {
-        if (NoMoreSeparatorsInFilePath)
-          return false;
-        J = K; // Skip to next `Separator` in `FilePath`.
-        break;
+        if (!Globstar) {
+          if (NoMoreSeparatorsInFilePath)
+            return false;
+          J = K; // Skip to next `Separator` in `FilePath`.
+          break;
+        }
+        if (I + 1 < EOP &&
+            matchFilePath(Pattern.substr(I + 1), FilePath.substr(J))) {
+          return true;
+        }
       }
       // Recurse.
-      for (auto Pat = Pattern.substr(I); J < End && FilePath[J] != Separator;
-           ++J) {
+      for (auto Pat = Pattern.substr(I);
+           J < End && (Globstar || FilePath[J] != Separator); ++J) {
         if (matchFilePath(Pat, FilePath.substr(J)))
           return true;
       }
diff --git a/clang/unittests/Format/MatchFilePathTest.cpp b/clang/unittests/Format/MatchFilePathTest.cpp
index 28f665635718e5..9876a4583c226a 100644
--- a/clang/unittests/Format/MatchFilePathTest.cpp
+++ b/clang/unittests/Format/MatchFilePathTest.cpp
@@ -164,6 +164,41 @@ TEST_F(MatchFilePathTest, Path) {
   EXPECT_FALSE(match("foo\\", R"(foo*\)"));
 }
 
+TEST_F(MatchFilePathTest, Globstar) {
+  EXPECT_TRUE(match("/", "**"));
+  EXPECT_TRUE(match("foo", "**"));
+  EXPECT_TRUE(match("/foo", "**"));
+  EXPECT_TRUE(match("foo/", "**"));
+  EXPECT_TRUE(match("foo/bar", "**"));
+
+  EXPECT_TRUE(match("/", "**/"));
+  EXPECT_TRUE(match("foo/", "**/"));
+  EXPECT_TRUE(match("/foo/", "**/"));
+  EXPECT_TRUE(match("foo/bar/", "**/"));
+
+  EXPECT_TRUE(match("/", "/**"));
+  EXPECT_TRUE(match("/foo", "/**"));
+  EXPECT_TRUE(match("/foo/", "/**"));
+  EXPECT_TRUE(match("/foo/bar", "/**"));
+
+  EXPECT_TRUE(match("foo", "**/foo"));
+  EXPECT_TRUE(match("/foo", "**/foo"));
+  EXPECT_TRUE(match("foo/bar", "**/bar"));
+  EXPECT_TRUE(match("/foo/bar", "**/foo/bar"));
+  EXPECT_TRUE(match("foo/bar/baz", "**/bar/baz"));
+
+  EXPECT_TRUE(match("abc/foo", "abc/**"));
+  EXPECT_TRUE(match("abc/foo/", "abc/**"));
+  EXPECT_TRUE(match("abc/foo/bar", "abc/**"));
+
+  EXPECT_TRUE(match("a/b", "a/**/b"));
+  EXPECT_TRUE(match("a/x/b", "a/**/b"));
+  EXPECT_TRUE(match("a/x/y/b", "a/**/b"));
+
+  EXPECT_FALSE(match("a/x/b", "a**/b"));
+  EXPECT_FALSE(match("a/x/y/b", "a/**b"));
+}
+
 } // namespace
 } // namespace format
 } // namespace clang

llvmbot · 2024-12-31T16:22:46Z

@llvm/pr-subscribers-clang

Author: Owen Pan (owenca)

Changes

Closes #114969.

Full diff: https://github.com/llvm/llvm-project/pull/121404.diff

4 Files Affected:

(modified) clang/docs/ClangFormat.rst (+4)
(modified) clang/docs/ReleaseNotes.rst (+1)
(modified) clang/lib/Format/MatchFilePath.cpp (+24-11)
(modified) clang/unittests/Format/MatchFilePathTest.cpp (+35)

diff --git a/clang/docs/ClangFormat.rst b/clang/docs/ClangFormat.rst
index c8f1d7f5a77581..21cf68236b8917 100644
--- a/clang/docs/ClangFormat.rst
+++ b/clang/docs/ClangFormat.rst
@@ -150,6 +150,10 @@ names. It has the following format:
 * Patterns follow the rules specified in `POSIX 2.13.1, 2.13.2, and Rule 1 of
   2.13.3 <https://pubs.opengroup.org/onlinepubs/9699919799/utilities/
   V3_chap02.html#tag_18_13>`_.
+* Bash globstar is supported, i.e. "two adjacent ``*``s used as a single pattern
+  will match all files and zero or more directories and subdirectories. If
+  followed by a ``/``, two adjacent ``*``s will match only directories and
+  subdirectories."
 * A pattern is negated if it starts with a bang (``!``).
 
 To match all files in a directory, use e.g. ``foo/bar/*``. To match all files in
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index b7da12bcf65818..2d59d74ddf35fc 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -1124,6 +1124,7 @@ clang-format
 - Adds ``RemoveEmptyLinesInUnwrappedLines`` option.
 - Adds ``KeepFormFeed`` option and set it to ``true`` for ``GNU`` style.
 - Adds ``AllowShortNamespacesOnASingleLine`` option.
+- Adds support for bash globstar in ``.clang-format-ignore``.
 
 libclang
 --------
diff --git a/clang/lib/Format/MatchFilePath.cpp b/clang/lib/Format/MatchFilePath.cpp
index 062b334dcdd8fd..aca3433dc7fa78 100644
--- a/clang/lib/Format/MatchFilePath.cpp
+++ b/clang/lib/Format/MatchFilePath.cpp
@@ -49,25 +49,38 @@ bool matchFilePath(StringRef Pattern, StringRef FilePath) {
         return false;
       break;
     case '*': {
-      while (++I < EOP && Pattern[I] == '*') { // Skip consecutive stars.
+      const bool MaybeGlobstar = I == 0 || Pattern[I - 1] == Separator;
+      int StarCount = 1;
+      for (; ++I < EOP && Pattern[I] == '*'; ++StarCount) {
+        // Skip consecutive stars.
       }
       const auto K = FilePath.find(Separator, J); // Index of next `Separator`.
       const bool NoMoreSeparatorsInFilePath = K == StringRef::npos;
+      bool Globstar = MaybeGlobstar && StarCount == 2;
       if (I == EOP) // `Pattern` ends with a star.
-        return NoMoreSeparatorsInFilePath;
-      // `Pattern` ends with a lone backslash.
-      if (Pattern[I] == '\\' && ++I == EOP)
-        return false;
+        return Globstar || NoMoreSeparatorsInFilePath;
+      if (Pattern[I] != Separator) {
+        Globstar = false;
+        // `Pattern` ends with a lone backslash.
+        if (Pattern[I] == '\\' && ++I == EOP)
+          return false;
+      }
       // The star is followed by a (possibly escaped) `Separator`.
       if (Pattern[I] == Separator) {
-        if (NoMoreSeparatorsInFilePath)
-          return false;
-        J = K; // Skip to next `Separator` in `FilePath`.
-        break;
+        if (!Globstar) {
+          if (NoMoreSeparatorsInFilePath)
+            return false;
+          J = K; // Skip to next `Separator` in `FilePath`.
+          break;
+        }
+        if (I + 1 < EOP &&
+            matchFilePath(Pattern.substr(I + 1), FilePath.substr(J))) {
+          return true;
+        }
       }
       // Recurse.
-      for (auto Pat = Pattern.substr(I); J < End && FilePath[J] != Separator;
-           ++J) {
+      for (auto Pat = Pattern.substr(I);
+           J < End && (Globstar || FilePath[J] != Separator); ++J) {
         if (matchFilePath(Pat, FilePath.substr(J)))
           return true;
       }
diff --git a/clang/unittests/Format/MatchFilePathTest.cpp b/clang/unittests/Format/MatchFilePathTest.cpp
index 28f665635718e5..9876a4583c226a 100644
--- a/clang/unittests/Format/MatchFilePathTest.cpp
+++ b/clang/unittests/Format/MatchFilePathTest.cpp
@@ -164,6 +164,41 @@ TEST_F(MatchFilePathTest, Path) {
   EXPECT_FALSE(match("foo\\", R"(foo*\)"));
 }
 
+TEST_F(MatchFilePathTest, Globstar) {
+  EXPECT_TRUE(match("/", "**"));
+  EXPECT_TRUE(match("foo", "**"));
+  EXPECT_TRUE(match("/foo", "**"));
+  EXPECT_TRUE(match("foo/", "**"));
+  EXPECT_TRUE(match("foo/bar", "**"));
+
+  EXPECT_TRUE(match("/", "**/"));
+  EXPECT_TRUE(match("foo/", "**/"));
+  EXPECT_TRUE(match("/foo/", "**/"));
+  EXPECT_TRUE(match("foo/bar/", "**/"));
+
+  EXPECT_TRUE(match("/", "/**"));
+  EXPECT_TRUE(match("/foo", "/**"));
+  EXPECT_TRUE(match("/foo/", "/**"));
+  EXPECT_TRUE(match("/foo/bar", "/**"));
+
+  EXPECT_TRUE(match("foo", "**/foo"));
+  EXPECT_TRUE(match("/foo", "**/foo"));
+  EXPECT_TRUE(match("foo/bar", "**/bar"));
+  EXPECT_TRUE(match("/foo/bar", "**/foo/bar"));
+  EXPECT_TRUE(match("foo/bar/baz", "**/bar/baz"));
+
+  EXPECT_TRUE(match("abc/foo", "abc/**"));
+  EXPECT_TRUE(match("abc/foo/", "abc/**"));
+  EXPECT_TRUE(match("abc/foo/bar", "abc/**"));
+
+  EXPECT_TRUE(match("a/b", "a/**/b"));
+  EXPECT_TRUE(match("a/x/b", "a/**/b"));
+  EXPECT_TRUE(match("a/x/y/b", "a/**/b"));
+
+  EXPECT_FALSE(match("a/x/b", "a**/b"));
+  EXPECT_FALSE(match("a/x/y/b", "a/**b"));
+}
+
 } // namespace
 } // namespace format
 } // namespace clang

Closes llvm#114969.

clang/unittests/Format/MatchFilePathTest.cpp

llvmbot added clang Clang issues not falling into any other category clang-format labels Dec 31, 2024

owenca force-pushed the globstar branch from 8b78021 to a1060e9 Compare December 31, 2024 16:32

owenca removed the clang Clang issues not falling into any other category label Dec 31, 2024

owenca mentioned this pull request Dec 31, 2024

[clang-format] clang-format-ignore: Add support for double asterisk patterns #110560

Closed

owenca force-pushed the globstar branch from a1060e9 to 13c13f0 Compare December 31, 2024 16:49

llvmbot added the clang Clang issues not falling into any other category label Dec 31, 2024

[clang-format] Support globstar in .clang-format-ignore

36efaa2

Closes llvm#114969.

owenca force-pushed the globstar branch from 13c13f0 to 36efaa2 Compare December 31, 2024 16:57

owenca removed the clang Clang issues not falling into any other category label Dec 31, 2024

leijurv reviewed Jan 1, 2025

View reviewed changes

clang/unittests/Format/MatchFilePathTest.cpp Outdated Show resolved Hide resolved

NFC: get rid of MaybeGlobstar.

1ffe78c

llvmbot added the clang Clang issues not falling into any other category label Jan 1, 2025

owenca requested review from rymiel, mydeveloperday and HazardyKnusperkeks January 1, 2025 03:58

owenca commented Jan 1, 2025

View reviewed changes

clang/unittests/Format/MatchFilePathTest.cpp Outdated Show resolved Hide resolved

owenca added 2 commits January 1, 2025 02:05

NFC: make a test case less confusing.

d1ca628

NFC: simplify logic

7294a56

owenca removed the clang Clang issues not falling into any other category label Jan 1, 2025

mydeveloperday approved these changes Jan 1, 2025

View reviewed changes

owenca merged commit cd23949 into llvm:main Jan 1, 2025
9 checks passed

owenca deleted the globstar branch January 1, 2025 23:38

llvm deleted a comment from llvm-ci Jan 1, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[clang-format] Support globstar in .clang-format-ignore #121404

[clang-format] Support globstar in .clang-format-ignore #121404

Uh oh!

owenca commented Dec 31, 2024 •

edited

Loading

Uh oh!

llvmbot commented Dec 31, 2024 •

edited

Loading

Uh oh!

llvmbot commented Dec 31, 2024

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

[clang-format] Support globstar in .clang-format-ignore #121404

[clang-format] Support globstar in .clang-format-ignore #121404

Uh oh!

Conversation

owenca commented Dec 31, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

llvmbot commented Dec 31, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

llvmbot commented Dec 31, 2024

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

owenca commented Dec 31, 2024 •

edited

Loading

llvmbot commented Dec 31, 2024 •

edited

Loading