Java import sorting in clang-format

[SamMaier]

Summary

This is for https://bugs.chromium.org/p/chromium/issues/detail?id=768983 - however it will be useful for anyone using clang-format for Java, not just Chromium.

[mgrang]

lib/Format/Format.cpp

1856

Please use llvm::sort instead of std::sort. See https://llvm.org/docs/CodingStandards.html#beware-of-non-deterministic-sorting-order-of-equal-elements.

[SamMaier]

Changed std::sort to llvm::sort

[krasimir]

lib/Format/Format.cpp

848

It would be helpful to link to a style guide reference for this.

1837–1838

It would be helpful to spell out the sort order and add tests with respect to . and *. ASCII ordering has * < . < A < a which seems reasonable.

1922

Please add a test for this.

1932

Hm, this would also pick up the static in import a.*; // static, right? IMO not a big problem, just add a note about it.

unittests/Format/SortImportsTestJava.cpp

130

I wonder what happens if there are comments between import statements and comment lines after import statements. Consider adding some tests for that.

[krasimir]

Another example: we have to be careful with something like this as we don't want to break correct code:

import x;
import a.loooooong.
  c;
import y;

unittests/Format/SortImportsTestJava.cpp

130

Specifically, as discussed, we wanna keep comments about imports together with them when we reorder stuff around.

[krasimir]

lib/Format/Format.cpp

1823

nit: LLVM Style uses CamelCase for variables, please use I.

1834

nit: Sorts and deduplicates

1836

nit: Import declarations...

unittests/Format/SortImportsTestJava.cpp

99

Add a test where there are existing blocks of import statements separated by empty lines. I'd expect that a newly added import gets added to the right already existing block, right?

136

interested does the newline get taken into account? e.g. what happens with

sort("import org.a;\n"
     "import org.b;");

[SamMaier]

Addressing comments

lib/Format/Format.cpp

1932

It does not, since Static is the first group of the regex. Test has been added for this

unittests/Format/SortImportsTestJava.cpp

136

Added test, it works.

[krasimir]

This gooks great! Thanks for the contribution!
If you don't have commit access to Clang I can commit this for you.

include/clang/Format/Format.h

   /// \endcode
   bool IndentWrappedFunctionNames;
 
+  /// A vector of prefixes ordered by the desired groups for Java imports.
+  ///
+  /// Each group is seperated by a newline. Static imports will also follow the
+  /// same grouping convention above all non-static imports. One group's prefix
+  /// can be a subset of another - the longest prefix is always matched. Within
+  /// a group, the imports are ordered lexicographically.
+  ///
+  /// In the .clang-format configuration file, this can be configured like:
+  /// \code{.yaml}
+  ///   JavaImportGroups: ['com.example', 'com', 'org']
+  /// \endcode
+  /// Which will result in imports being formatted as so:
+  /// \code{.java}
+  ///    import static com.example.function1;
+  ///
+  ///    import static com.test.function2;
+  ///
+  ///    import static org.example.function3;
+  ///
+  ///    import com.example.ClassA;
+  ///    import com.example.Test;
+  ///    import com.example.a.ClassB;
+  ///
+  ///    import com.test.ClassC;
+  ///
+  ///    import org.example.ClassD;
+  /// \endcode
+  std::vector<std::string> JavaImportGroups;
+
   /// Quotation styles for JavaScript strings. Does not affect template
   /// strings.
   enum JavaScriptQuoteStyle {

            IndentPPDirectives == R.IndentPPDirectives &&
            IndentWidth == R.IndentWidth && Language == R.Language &&
            IndentWrappedFunctionNames == R.IndentWrappedFunctionNames &&
+           JavaImportGroups == R.JavaImportGroups &&
            JavaScriptQuotes == R.JavaScriptQuotes &&
            JavaScriptWrapImports == R.JavaScriptWrapImports &&
            KeepEmptyLinesAtTheStartOfBlocks ==

lib/Format/Format.cpp

     IO.mapOptional("IndentWidth", Style.IndentWidth);
     IO.mapOptional("IndentWrappedFunctionNames",
                    Style.IndentWrappedFunctionNames);
+    IO.mapOptional("JavaImportGroups", Style.JavaImportGroups);
     IO.mapOptional("JavaScriptQuotes", Style.JavaScriptQuotes);
     IO.mapOptional("JavaScriptWrapImports", Style.JavaScriptWrapImports);
     IO.mapOptional("KeepEmptyLinesAtTheStartOfBlocks",

     ChromiumStyle.BreakAfterJavaFieldAnnotations = true;
     ChromiumStyle.ContinuationIndentWidth = 8;
     ChromiumStyle.IndentWidth = 4;
+    ChromiumStyle.JavaImportGroups = {
+        "android",
+        "com",
+        "dalvik",
+        "junit",
+        "org",
+        "com.google.android.apps.chrome",
+        "org.chromium",
+        "java",
+        "javax",
+    };
+    ChromiumStyle.SortIncludes = true;
   } else if (Language == FormatStyle::LK_JavaScript) {
     ChromiumStyle.AllowShortIfStatementsOnASingleLine = false;
     ChromiumStyle.AllowShortLoopsOnASingleLine = false;

   int Category;
 };
 
+struct JavaImportDirective {
+  StringRef Identifier;
+  StringRef Text;
+  unsigned Offset;
+  bool IsStatic;
+};
+
 } // end anonymous namespace
 
 // Determines whether 'Ranges' intersects with ('Start', 'End').

 
 namespace {
 
-const char IncludeRegexPattern[] =
+const char CppIncludeRegexPattern[] =
     R"(^[\t\ ]*#[\t\ ]*(import|include)[^"<]*(["<][^">]*[">]))";
 
 } // anonymous namespace

                                       unsigned *Cursor) {
   unsigned Prev = 0;
   unsigned SearchFrom = 0;
-  llvm::Regex IncludeRegex(IncludeRegexPattern);
+  llvm::Regex IncludeRegex(CppIncludeRegexPattern);
   SmallVector<StringRef, 4> Matches;
   SmallVector<IncludeDirective, 16> IncludesInBlock;
 

   return Replaces;
 }
 
+// Returns group number to use as a first order sort on imports. Gives UINT_MAX
+// if the import does not match any given groups.
+static unsigned findJavaImportGroup(const FormatStyle &Style,
+                                    StringRef ImportIdentifier) {
+  unsigned LongestMatchIndex = UINT_MAX;
+  unsigned LongestMatchLength = 0;
+  for (unsigned i = 0; i < Style.JavaImportGroups.size(); i++) {
+    std::string GroupPrefix = Style.JavaImportGroups[i];
+    if (ImportIdentifier.startswith(GroupPrefix) &&
+        GroupPrefix.length() > LongestMatchLength) {
+      LongestMatchIndex = i;
+      LongestMatchLength = GroupPrefix.length();
+    }
+  }
+  return LongestMatchIndex;
+}
+
+// Sorts and deduplicate a block of includes given by 'Imports' based on
+// JavaImportGroups, then adding the necessary replacement to 'Replaces'.
+// import declarations with the same text will be deduplicated. Between each
+// import group, a newline is inserted, and within each import group, a
+// case-sensitive sort is performed.
+static void sortJavaImports(const FormatStyle &Style,
+                            const SmallVectorImpl<JavaImportDirective> &Imports,
+                            ArrayRef<tooling::Range> Ranges, StringRef FileName,
+                            tooling::Replacements &Replaces) {
+  unsigned ImportsBeginOffset = Imports.front().Offset;
+  unsigned ImportsEndOffset =
+      Imports.back().Offset + Imports.back().Text.size();
+  unsigned ImportsBlockSize = ImportsEndOffset - ImportsBeginOffset;
+  if (!affectsRange(Ranges, ImportsBeginOffset, ImportsEndOffset))
+    return;
+  SmallVector<unsigned, 16> Indices;
+  SmallVector<unsigned, 16> JavaImportGroups;
+  for (unsigned i = 0, e = Imports.size(); i != e; ++i) {
+    Indices.push_back(i);
+    JavaImportGroups.push_back(
+        findJavaImportGroup(Style, Imports[i].Identifier));
+  }
+  llvm::sort(Indices.begin(), Indices.end(), [&](unsigned LHSI, unsigned RHSI) {
+        // Negating IsStatic to push static imports above non-static imports.
+        return std::make_tuple(!Imports[LHSI].IsStatic, JavaImportGroups[LHSI],
+                               Imports[LHSI].Identifier) <
+               std::make_tuple(!Imports[RHSI].IsStatic, JavaImportGroups[RHSI],
+                               Imports[RHSI].Identifier);
+      });
+
+  // Deduplicate imports.
+  Indices.erase(std::unique(Indices.begin(), Indices.end(),
+                            [&](unsigned LHSI, unsigned RHSI) {
+                              return Imports[LHSI].Text == Imports[RHSI].Text;
+                            }),
+                Indices.end());
+
+  bool CurrentIsStatic = Imports[Indices.front()].IsStatic;
+  unsigned CurrentImportGroup = JavaImportGroups[Indices.front()];
+
+  std::string result;
+  for (unsigned Index : Indices) {
+    if (!result.empty()) {
+      result += "\n";
+      if (CurrentIsStatic != Imports[Index].IsStatic ||
+          CurrentImportGroup != JavaImportGroups[Index])
+        result += "\n";
+    }
+    result += Imports[Index].Text;
+    CurrentIsStatic = Imports[Index].IsStatic;
+    CurrentImportGroup = JavaImportGroups[Index];
+  }
+
+  auto Err = Replaces.add(tooling::Replacement(FileName, Imports.front().Offset,
+                                               ImportsBlockSize, result));
+  // FIXME: better error handling. For now, just skip the replacement for the
+  // release version.
+  if (Err) {
+    llvm::errs() << llvm::toString(std::move(Err)) << "\n";
+    assert(false);
+  }
+}
+
+namespace {
+
+const char JavaImportRegexPattern[] =
+    "^[\t ]*import[\t ]*(static[\t ]*)?([^\t ]*)[\t ]*;";
+
+} // anonymous namespace
+
+tooling::Replacements sortJavaImports(const FormatStyle &Style, StringRef Code,
+                                      ArrayRef<tooling::Range> Ranges,
+                                      StringRef FileName,
+                                      tooling::Replacements &Replaces) {
+  unsigned Prev = 0;
+  unsigned SearchFrom = 0;
+  llvm::Regex ImportRegex(JavaImportRegexPattern);
+  SmallVector<StringRef, 4> Matches;
+  SmallVector<JavaImportDirective, 16> ImportsInBlock;
+
+  bool FormattingOff = false;
+
+  for (;;) {
+    auto Pos = Code.find('\n', SearchFrom);
+    StringRef Line =
+        Code.substr(Prev, (Pos != StringRef::npos ? Pos : Code.size()) - Prev);
+
+    StringRef Trimmed = Line.trim();
+    if (Trimmed == "// clang-format off")
+      FormattingOff = true;
+    else if (Trimmed == "// clang-format on")
+      FormattingOff = false;
+
+    if (!FormattingOff) {
+      if (ImportRegex.match(Line, &Matches)) {
+        StringRef Static = Matches[1];
+        StringRef Identifier = Matches[2];
+        bool IsStatic = false;
+        if (Static.contains("static")) {
+          IsStatic = true;
+        }
+        ImportsInBlock.push_back({Identifier, Line, Prev, IsStatic});
+      }
+      Prev = Pos + 1;
+    }
+    if (Pos == StringRef::npos || Pos + 1 == Code.size())
+      break;
+    SearchFrom = Pos + 1;
+  }
+  if (!ImportsInBlock.empty())
+    sortJavaImports(Style, ImportsInBlock, Ranges, FileName, Replaces);
+  return Replaces;
+}
+
 bool isMpegTS(StringRef Code) {
   // MPEG transport streams use the ".ts" file extension. clang-format should
   // not attempt to format those. MPEG TS' frame format starts with 0x47 every

     return Replaces;
   if (Style.Language == FormatStyle::LanguageKind::LK_JavaScript)
     return sortJavaScriptImports(Style, Code, Ranges, FileName);
+  if (Style.Language == FormatStyle::LanguageKind::LK_Java)
+    return sortJavaImports(Style, Code, Ranges, FileName, Replaces);
   sortCppIncludes(Style, Code, Ranges, FileName, Replaces, Cursor);
   return Replaces;
 }

 
 inline bool isHeaderInsertion(const tooling::Replacement &Replace) {
   return Replace.getOffset() == UINT_MAX && Replace.getLength() == 0 &&
-         llvm::Regex(IncludeRegexPattern).match(Replace.getReplacementText());
+         llvm::Regex(CppIncludeRegexPattern)
+             .match(Replace.getReplacementText());
 }
 
 inline bool isHeaderDeletion(const tooling::Replacement &Replace) {

     }
   }
 
-  llvm::Regex IncludeRegex = llvm::Regex(IncludeRegexPattern);
+  llvm::Regex IncludeRegex = llvm::Regex(CppIncludeRegexPattern);
   llvm::SmallVector<StringRef, 4> Matches;
   for (const auto &R : HeaderInsertions) {
     auto IncludeDirective = R.getReplacementText();

unittests/Format/CMakeLists.txt

   FormatTestTextProto.cpp
   NamespaceEndCommentsFixerTest.cpp
   SortImportsTestJS.cpp
+  SortImportsTestJava.cpp
   SortIncludesTest.cpp
   UsingDeclarationsSorterTest.cpp
   )

unittests/Format/SortImportsTestJava.cpp

+//===- unittest/Format/SortImportsTestJava.cpp - Import sort unit tests ---===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#include "clang/Format/Format.h"
+#include "gtest/gtest.h"
+
+#define DEBUG_TYPE "format-test"
+
+namespace clang {
+namespace format {
+namespace {
+
+class SortImportsTestJava : public ::testing::Test {
+protected:
+  std::vector<tooling::Range> GetCodeRange(StringRef Code) {
+    return std::vector<tooling::Range>(1, tooling::Range(0, Code.size()));
+  }
+
+  std::string sort(StringRef Code, std::vector<tooling::Range> Ranges) {
+    auto Replaces = sortIncludes(FmtStyle, Code, Ranges, "input.java");
+    Ranges = tooling::calculateRangesAfterReplacements(Replaces, Ranges);
+    auto Sorted = applyAllReplacements(Code, Replaces);
+    EXPECT_TRUE(static_cast<bool>(Sorted));
+    auto Result = applyAllReplacements(
+        *Sorted, reformat(FmtStyle, *Sorted, Ranges, "input.java"));
+    EXPECT_TRUE(static_cast<bool>(Result));
+    return *Result;
+  }
+
+  std::string sort(StringRef Code) { return sort(Code, GetCodeRange(Code)); }
+
+  FormatStyle FmtStyle;
+
+public:
+  SortImportsTestJava() {
+    FmtStyle = getGoogleStyle(FormatStyle::LK_Java);
+    FmtStyle.JavaImportGroups = {"com.test", "org", "com"};
+    FmtStyle.SortIncludes = true;
+  }
+};
+
+TEST_F(SortImportsTestJava, StaticSplitFromNormal) {
+  EXPECT_EQ("import static org.b;\n"
+            "\n"
+            "import org.a;\n",
+            sort("import org.a;\n"
+                 "import static org.b;\n"));
+}
+
+TEST_F(SortImportsTestJava, CapitalBeforeLowercase) {
+  EXPECT_EQ("import org.Test;\n"
+            "import org.a.Test;\n"
+            "import org.b;\n",
+            sort("import org.a.Test;\n"
+                 "import org.Test;\n"
+                 "import org.b;\n"));
+}
+
+TEST_F(SortImportsTestJava, SplitGroupsWithNewline) {
+  EXPECT_EQ("import static com.test.a;\n"
+            "\n"
+            "import static org.a;\n"
+            "\n"
+            "import static com.a;\n"
+            "\n"
+            "import com.test.b;\n"
+            "\n"
+            "import org.b;\n"
+            "\n"
+            "import com.b;\n",
+            sort("import static com.test.a;\n"
+                 "import static org.a;\n"
+                 "import static com.a;\n"
+                 "import com.test.b;\n"
+                 "import org.b;\n"
+                 "import com.b;\n"));
+}
+
+TEST_F(SortImportsTestJava, UnspecifiedGroupAfterAllGroups) {
+  EXPECT_EQ("import com.test.a;\n"
+            "\n"
+            "import org.a;\n"
+            "\n"
+            "import com.a;\n"
+            "\n"
+            "import abc.a;\n"
+            "import xyz.b;\n",
+            sort("import com.test.a;\n"
+                 "import com.a;\n"
+                 "import xyz.b;\n"
+                 "import abc.a;\n"
+                 "import org.a;\n"));
+}
+
+TEST_F(SortImportsTestJava, NoSortOutsideRange) {
+  std::vector<tooling::Range> Ranges = {tooling::Range(27, 15)};
+  EXPECT_EQ("import org.b;\n"
+            "import org.a;\n"
+            "// comments\n"
+            "// that do\n"
+            "// nothing\n",
+            sort("import org.b;\n"
+                 "import org.a;\n"
+                 "// comments\n"
+                 "// that do\n"
+                 "// nothing\n",
+                 Ranges));
+}
+
+TEST_F(SortImportsTestJava, SortWhenRangeContainsOneLine) {
+  std::vector<tooling::Range> Ranges = {tooling::Range(27, 20)};
+  EXPECT_EQ("import org.a;\n"
+            "import org.b;\n"
+            "\n"
+            "import com.a;\n"
+            "// comments\n"
+            "// that do\n"
+            "// nothing\n",
+            sort("import org.b;\n"
+                 "import org.a;\n"
+                 "import com.a;\n"
+                 "// comments\n"
+                 "// that do\n"
+                 "// nothing\n",
+                 Ranges));
+}
+
+TEST_F(SortImportsTestJava, DeduplicateImports) {
+  EXPECT_EQ("import org.a;\n", sort("import org.a;\n"
+                                    "import org.a;\n"));
+}
+
+} // end namespace
+} // end namespace format
+} // end namespace clang