Add support for decompressing .7z files

willstranton · willstranton · commit fa0051d43ad1 · 2025-10-15T12:01:13.000Z
This implementation follows the implementation of ZipDecompressor with the exception that .7z files do not handle symbolic links or preserving file permissions. The commit 9c98120 where .ar/.deb support was added, was used as a reference for what additional files to change. Closes #27231
diff --git a/src/main/java/com/google/devtools/build/lib/bazel/repository/decompressor/DecompressorValue.java b/src/main/java/com/google/devtools/build/lib/bazel/repository/decompressor/DecompressorValue.java
@@ -112,11 +112,13 @@ static Decompressor getDecompressor(Path archivePath) throws RepositoryFunctionE
       return TarBz2Function.INSTANCE;
     } else if (baseName.endsWith(".ar") || baseName.endsWith(".deb")) {
       return ArFunction.INSTANCE;
+    } else if (baseName.endsWith(".7z")) {
+      return SevenZDecompressor.INSTANCE;
     } else {
       throw new RepositoryFunctionException(
           Starlark.errorf(
               "Expected a file with a .zip, .jar, .war, .aar, .nupkg, .whl, .tar, .tar.gz, .tgz,"
-                  + " .tar.xz, , .tar.zst, .tzst, .tar.bz2, .tbz, .ar or .deb suffix (got %s)",
+                  + " .tar.xz, , .tar.zst, .tzst, .tar.bz2, .tbz, .ar, .deb or .7z suffix (got %s)",
               archivePath),
           Transience.PERSISTENT);
     }
diff --git a/src/main/java/com/google/devtools/build/lib/bazel/repository/decompressor/SevenZDecompressor.java b/src/main/java/com/google/devtools/build/lib/bazel/repository/decompressor/SevenZDecompressor.java
@@ -0,0 +1,105 @@
+package com.google.devtools.build.lib.bazel.repository.decompressor;
+
+import static java.nio.charset.StandardCharsets.UTF_8;
+
+import com.google.common.io.ByteStreams;
+import com.google.devtools.build.lib.bazel.repository.RepositoryFunctionException;
+import com.google.devtools.build.lib.bazel.repository.decompressor.DecompressorValue.Decompressor;
+import com.google.devtools.build.lib.vfs.Path;
+import com.google.devtools.build.lib.vfs.PathFragment;
+import org.apache.commons.compress.archivers.sevenz.SevenZArchiveEntry;
+import org.apache.commons.compress.archivers.sevenz.SevenZFile;
+
+import javax.annotation.Nullable;
+import java.io.IOException;
+import java.io.InputStream;
+import java.io.OutputStream;
+import java.util.HashSet;
+import java.util.Map;
+import java.util.Optional;
+import java.util.Set;
+
+/**
+ * Creates a repository by decompressing a 7-zip file. This implementation generally follows the
+ * logic from {@link ZipDecompressor} with the exception that the 7z format does not support file
+ * permissions or symbolic links.
+ */
+public class SevenZDecompressor implements Decompressor {
+  public static final Decompressor INSTANCE = new SevenZDecompressor();
+
+  /** Decompresses the file to directory {@link DecompressorDescriptor#destinationPath()} */
+  @Override
+  @Nullable
+  public Path decompress(DecompressorDescriptor descriptor)
+      throws IOException, RepositoryFunctionException, InterruptedException {
+    Path destinationDirectory = descriptor.destinationPath();
+    Optional<String> prefix = descriptor.prefix();
+    Map<String, String> renameFiles = descriptor.renameFiles();
+    boolean foundPrefix = false;
+
+    try (SevenZFile sevenZFile =
+        SevenZFile.builder().setFile(descriptor.archivePath().getPathFile()).get()) {
+      Iterable<SevenZArchiveEntry> entries = sevenZFile.getEntries();
+      for (SevenZArchiveEntry entry : entries) {
+        String entryName = entry.getName();
+        entryName = renameFiles.getOrDefault(entryName, entryName);
+        StripPrefixedPath entryPath =
+            StripPrefixedPath.maybeDeprefix(entryName.getBytes(UTF_8), prefix);
+        foundPrefix = foundPrefix || entryPath.foundPrefix();
+        if (entryPath.skip()) {
+          continue;
+        }
+        extract7zEntry(sevenZFile, entry, destinationDirectory, entryPath.getPathFragment());
+      }
+
+      if (prefix.isPresent() && !foundPrefix) {
+        Set<String> prefixes = new HashSet<>();
+        for (SevenZArchiveEntry entry : entries) {
+          StripPrefixedPath entryPath =
+              StripPrefixedPath.maybeDeprefix(entry.getName().getBytes(UTF_8), Optional.empty());
+          CouldNotFindPrefixException.maybeMakePrefixSuggestion(entryPath.getPathFragment())
+              .ifPresent(prefixes::add);
+        }
+        throw new CouldNotFindPrefixException(prefix.get(), prefixes);
+      }
+    }
+    return destinationDirectory;
+  }
+
+  private static void extract7zEntry(
+      SevenZFile sevenZFile,
+      SevenZArchiveEntry entry,
+      Path destinationDirectory,
+      PathFragment strippedRelativePath)
+      throws IOException, InterruptedException {
+    if (strippedRelativePath.isAbsolute()) {
+      throw new IOException(
+          String.format(
+              "Failed to extract %s, 7-zipped paths cannot be absolute", strippedRelativePath));
+    }
+    // Sanity/security check - at this point, uplevel references (..) should be resolved.
+    // There shouldn't be any remaining uplevel references, otherwise, the extracted file could
+    // "escape" the destination directory.
+    if (strippedRelativePath.containsUplevelReferences()) {
+      throw new IOException(
+          String.format(
+              "Failed to extract %s, 7-zipped entry contains uplevel references (..)",
+              strippedRelativePath));
+    }
+    Path outputPath = destinationDirectory.getRelative(strippedRelativePath);
+    outputPath.getParentDirectory().createDirectoryAndParents();
+    boolean isDirectory = entry.isDirectory();
+    if (isDirectory) {
+      outputPath.createDirectoryAndParents();
+    } else {
+      try (InputStream input = sevenZFile.getInputStream(entry);
+          OutputStream output = outputPath.getOutputStream()) {
+        ByteStreams.copy(input, output);
+        if (Thread.interrupted()) {
+          throw new InterruptedException();
+        }
+      }
+      outputPath.setLastModifiedTime(entry.getLastModifiedTime().toMillis());
+    }
+  }
+}
diff --git a/src/main/java/com/google/devtools/build/lib/bazel/repository/starlark/StarlarkBaseExternalContext.java b/src/main/java/com/google/devtools/build/lib/bazel/repository/starlark/StarlarkBaseExternalContext.java
@@ -906,7 +906,7 @@ public Object download(
                 determined from the file extension of the URL. If the file has no \
                 extension, you can explicitly specify either "zip", "jar", "war", \
                 "aar", "nupkg", "whl", "tar", "tar.gz", "tgz", "tar.xz", "txz", ".tar.zst", \
-                ".tzst", "tar.bz2", ".tbz", ".ar", or ".deb" here.
+                ".tzst", "tar.bz2", ".tbz", ".ar", ".deb", or ".7z" here.
                 """),
         @Param(
             name = "strip_prefix",
diff --git a/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/BUILD b/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/BUILD
@@ -17,6 +17,7 @@ java_library(
     name = "DecompressorTests_lib",
     srcs = glob(["*.java"]),
     data = [
+        "test_decompress_archive.7z",
         "test_decompress_archive.tar.gz",
         "test_decompress_archive.zip",
         "test_files.ar",
@@ -27,6 +28,7 @@ java_library(
         "//src/main/java/com/google/devtools/build/lib/clock",
         "//src/main/java/com/google/devtools/build/lib/unix",
         "//src/main/java/com/google/devtools/build/lib/util:os",
+        "//src/main/java/com/google/devtools/build/lib/util:string_encoding",
         "//src/main/java/com/google/devtools/build/lib/vfs",
         "//src/main/java/com/google/devtools/build/lib/vfs:pathfragment",
         "//src/main/java/com/google/devtools/build/lib/vfs/inmemoryfs",
diff --git a/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/DecompressorValueTest.java b/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/DecompressorValueTest.java
@@ -67,6 +67,8 @@ public void testKnownFileExtensionsDoNotThrow() throws Exception {
     unused = DecompressorValue.getDecompressor(path);
     path = fs.getPath("/foo/.external-repositories/some-repo/bar.baz.deb");
     unused = DecompressorValue.getDecompressor(path);
+    path = fs.getPath("/foo/.external-repositories/some-repo/bar.baz.7z");
+    unused = DecompressorValue.getDecompressor(path);
   }
 
   @Test
diff --git a/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/SevenZDecompressorTest.java b/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/SevenZDecompressorTest.java
@@ -0,0 +1,110 @@
+package com.google.devtools.build.lib.bazel.repository.decompressor;
+
+import com.google.devtools.build.lib.util.StringEncoding;
+import com.google.devtools.build.lib.vfs.Path;
+import org.junit.Before;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.junit.runners.JUnit4;
+
+import java.util.HashMap;
+
+import static com.google.common.truth.Truth.assertThat;
+import static com.google.devtools.build.lib.bazel.repository.decompressor.TestArchiveDescriptor.INNER_FOLDER_NAME;
+import static com.google.devtools.build.lib.bazel.repository.decompressor.TestArchiveDescriptor.ROOT_FOLDER_NAME;
+
+/** Tests .7z decompression. */
+@RunWith(JUnit4.class)
+public class SevenZDecompressorTest {
+  /**
+   * .7z file, created with two files:
+   *
+   * <ul>
+   *   <li>root_folder/another_folder/regular_file
+   *   <li>root_folder/another_folder/ünïcödëFïlë.txt
+   * </ul>
+   *
+   * Compressed with command "7zz a test_decompress_archive.7z root_folder"
+   */
+  private static final String ARCHIVE_NAME = "test_decompress_archive.7z";
+
+  private static final String UNICODE_FILENAME = "ünïcödëFïlë.txt";
+
+  private TestArchiveDescriptor archiveDescriptor;
+
+  @Before
+  public void setUpFs() throws Exception {
+    archiveDescriptor =
+        new TestArchiveDescriptor(
+            ARCHIVE_NAME, /* outDirName= */ "out", /* withHardLinks= */ false);
+  }
+
+  /** Test decompressing a .7z file without stripping a prefix */
+  @Test
+  public void testDecompressWithoutPrefix() throws Exception {
+    Path outputDir = decompress(archiveDescriptor.createDescriptorBuilder().build());
+
+    archiveDescriptor.assertOutputFiles(outputDir, ROOT_FOLDER_NAME, INNER_FOLDER_NAME);
+  }
+
+  /** Test decompressing a .7z file and stripping a prefix. */
+  @Test
+  public void testDecompressWithPrefix() throws Exception {
+    DecompressorDescriptor.Builder descriptorBuilder =
+        archiveDescriptor.createDescriptorBuilder().setPrefix(ROOT_FOLDER_NAME);
+    Path outputDir = decompress(descriptorBuilder.build());
+
+    archiveDescriptor.assertOutputFiles(outputDir, INNER_FOLDER_NAME);
+  }
+
+  /** Test decompressing a .7z with entries being renamed during the extraction process. */
+  @Test
+  public void testDecompressWithRenamedFiles() throws Exception {
+    String innerDirName = ROOT_FOLDER_NAME + "/" + INNER_FOLDER_NAME;
+
+    HashMap<String, String> renameFiles = new HashMap<>();
+    renameFiles.put(innerDirName + "/regular_file", innerDirName + "/renamedFile");
+    DecompressorDescriptor.Builder descriptorBuilder =
+        archiveDescriptor.createDescriptorBuilder().setRenameFiles(renameFiles);
+    Path outputDir = decompress(descriptorBuilder.build());
+
+    Path innerDir = outputDir.getRelative(ROOT_FOLDER_NAME).getRelative(INNER_FOLDER_NAME);
+    assertThat(innerDir.getRelative("renamedFile").exists()).isTrue();
+  }
+
+  /** Test that entry renaming is applied prior to prefix stripping. */
+  @Test
+  public void testDecompressWithRenamedFilesAndPrefix() throws Exception {
+    String innerDirName = ROOT_FOLDER_NAME + "/" + INNER_FOLDER_NAME;
+
+    HashMap<String, String> renameFiles = new HashMap<>();
+    renameFiles.put(innerDirName + "/regular_file", innerDirName + "/renamedFile");
+    DecompressorDescriptor.Builder descriptorBuilder =
+        archiveDescriptor
+            .createDescriptorBuilder()
+            .setPrefix(ROOT_FOLDER_NAME)
+            .setRenameFiles(renameFiles);
+    Path outputDir = decompress(descriptorBuilder.build());
+
+    Path innerDir = outputDir.getRelative(INNER_FOLDER_NAME);
+    assertThat(innerDir.getRelative("renamedFile").exists()).isTrue();
+  }
+
+  /** Test that Unicode filenames are handled. **/
+  @Test
+  public void testUnicodeFilename() throws Exception {
+    Path outputDir = decompress(archiveDescriptor.createDescriptorBuilder().build());
+
+    Path unicodeFile =
+        outputDir
+            .getRelative(ROOT_FOLDER_NAME)
+            .getRelative(INNER_FOLDER_NAME)
+            .getRelative(StringEncoding.unicodeToInternal(UNICODE_FILENAME));
+    assertThat(unicodeFile.exists()).isTrue();
+    assertThat(unicodeFile.getFileSize()).isNotEqualTo(0);
+  }
+
+  private Path decompress(DecompressorDescriptor descriptor) throws Exception {
+    return new SevenZDecompressor().decompress(descriptor);
+  }
+}
diff --git a/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/test_decompress_archive.7z b/src/test/java/com/google/devtools/build/lib/bazel/repository/decompressor/test_decompress_archive.7z
diff --git a/src/test/tools/bzlmod/MODULE.bazel.lock b/src/test/tools/bzlmod/MODULE.bazel.lock
diff --git a/tools/build_defs/repo/http.bzl b/tools/build_defs/repo/http.bzl
@@ -337,7 +337,8 @@ repository. Files are symlinked after remote files are downloaded and patches (`
 By default, the archive type is determined from the file extension of the
 URL. If the file has no extension, you can explicitly specify one of the
 following: `"zip"`, `"jar"`, `"war"`, `"aar"`, `"tar"`, `"tar.gz"`, `"tgz"`,
-`"tar.xz"`, `"txz"`, `"tar.zst"`, `"tzst"`, `"tar.bz2"`, `"ar"`, or `"deb"`.""",
+`"tar.xz"`, `"txz"`, `"tar.zst"`, `"tzst"`, `"tar.bz2"`, `"ar"`, `"deb"`, or
+`"7z"`.""",
     ),
     "patches": attr.label_list(
         default = [],
@@ -450,7 +451,7 @@ and makes its targets available for binding.
 
 It supports the following file extensions: `"zip"`, `"jar"`, `"war"`, `"aar"`, `"tar"`,
 `"tar.gz"`, `"tgz"`, `"tar.xz"`, `"txz"`, `"tar.zst"`, `"tzst"`, `tar.bz2`, `"ar"`,
-or `"deb"`.
+`"deb"`, or `"7z"`.
 
 Examples:
   Suppose the current repository contains the source code for a chat program,

Original file line number	Diff line number	Diff line change
`@@ -67,6 +67,8 @@ public void testKnownFileExtensionsDoNotThrow() throws Exception {`
`67`	`67`	`unused = DecompressorValue.getDecompressor(path);`
`68`	`68`	`path = fs.getPath("/foo/.external-repositories/some-repo/bar.baz.deb");`
`69`	`69`	`unused = DecompressorValue.getDecompressor(path);`
	`70`	`+ path = fs.getPath("/foo/.external-repositories/some-repo/bar.baz.7z");`
	`71`	`+ unused = DecompressorValue.getDecompressor(path);`
`70`	`72`	`}`
`71`	`73`
`72`	`74`	`@Test`