add "archiveType" source.json property to set the http_archive's "typ…

…e" value.

Closes #17789.

PiperOrigin-RevId: 517408668
Change-Id: I7f89db0d2587cde3ff9d77c8657d162981cf32dc

site/en/external/registry.md

@@ -60,6 +60,10 @@ An index registry must follow the format below:
                 patch file names, and the values are the integrity checksum of
                 the patch files
             *   `patch_strip`: Same as the `--strip` argument of Unix `patch`.
+            *   `archive_type`: The archive type of the downloaded file (Same as `type` on `http_archive`).
+                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"`.
         *   The type can be changed to use a local path, representing a
             `local_repository` repo, with these fields:
             *   `type`: `local_path`
@@ -109,4 +113,4 @@ branch of the `my-org` fork, you would set
 `--registry=https://raw.githubusercontent.com/my-org/bazel-central-registry/main/`.
 
 Using the `--registry` flag stops the Bazel Central Registry from being used by
-default, but you can add it back by adding `--registry=https://bcr.bazel.build`.
+default, but you can add it back by adding `--registry=https://bcr.bazel.build`.

src/main/java/com/google/devtools/build/lib/bazel/bzlmod/ArchiveRepoSpecBuilder.java

@@ -15,6 +15,7 @@
 
 package com.google.devtools.build.lib.bazel.bzlmod;
 
+import com.google.common.base.Strings;
 import com.google.common.collect.ImmutableList;
 import com.google.common.collect.ImmutableMap;
 import com.google.errorprone.annotations.CanIgnoreReturnValue;
@@ -85,6 +86,14 @@ public ArchiveRepoSpecBuilder setRemotePatchStrip(int remotePatchStrip) {
     return this;
   }
 
+  @CanIgnoreReturnValue
+  public ArchiveRepoSpecBuilder setArchiveType(String archiveType) {
+    if (!Strings.isNullOrEmpty(archiveType)) {
+      attrBuilder.put("type", archiveType);
+    }
+    return this;
+  }
+
   public RepoSpec build() {
     return RepoSpec.builder()
         .setBzlFile("@bazel_tools//tools/build_defs/repo:http.bzl")

src/main/java/com/google/devtools/build/lib/bazel/bzlmod/IndexRegistry.java

@@ -110,6 +110,7 @@ private static class SourceJson {
     Map<String, String> patches;
     int patchStrip;
     String path;
+    String archiveType;
   }
 
   /**
@@ -244,6 +245,7 @@ private RepoSpec createArchiveRepoSpec(
         .setStripPrefix(Strings.nullToEmpty(sourceJson.get().stripPrefix))
         .setRemotePatches(remotePatches.buildOrThrow())
         .setRemotePatchStrip(sourceJson.get().patchStrip)
+        .setArchiveType(sourceJson.get().archiveType)
         .build();
   }
 

src/test/java/com/google/devtools/build/lib/bazel/bzlmod/IndexRegistryTest.java

@@ -290,4 +290,33 @@ public void testGetYankedVersion() throws Exception {
                 Version.parse("1.0"),
                 "red-pill 1.0 is yanked due to CVE-2000-101, please upgrade to 2.0"));
   }
+
+  @Test
+  public void testArchiveWithExplicitType() throws Exception {
+    server.serve(
+        "/modules/archive_type/1.0/source.json",
+        "{",
+        "  \"url\": \"https://mysite.com/thing?format=zip\",",
+        "  \"integrity\": \"sha256-blah\",",
+        "  \"archive_type\": \"zip\"",
+        "}");
+    server.start();
+
+    Registry registry = registryFactory.getRegistryWithUrl(server.getUrl());
+    assertThat(
+            registry.getRepoSpec(
+                createModuleKey("archive_type", "1.0"),
+                RepositoryName.create("archive_type_repo"),
+                reporter))
+        .isEqualTo(
+            new ArchiveRepoSpecBuilder()
+                .setRepoName("archive_type_repo")
+                .setUrls(ImmutableList.of("https://mysite.com/thing?format=zip"))
+                .setIntegrity("sha256-blah")
+                .setStripPrefix("")
+                .setArchiveType("zip")
+                .setRemotePatches(ImmutableMap.of())
+                .setRemotePatchStrip(0)
+                .build());
+  }
 }

src/test/py/bazel/bzlmod/bazel_module_test.py

@@ -1007,6 +1007,40 @@ def testWorkspaceItselfCanSeeRootModuleMappings(self):
     _, _, stderr = self.RunBazel(['build', ':a'], allow_failure=False)
     self.assertIn('I LUV U!', '\n'.join(stderr))
 
+  def testArchiveWithArchiveType(self):
+    # make the archive without the .zip extension
+    self.main_registry.createCcModule(
+        'aaa', '1.2', archive_pattern='%s.%s', archive_type='zip'
+    )
+
+    self.ScratchFile(
+        'MODULE.bazel',
+        [
+            'bazel_dep(name = "aaa", version = "1.2")',
+        ],
+    )
+    self.ScratchFile(
+        'BUILD',
+        [
+            'cc_binary(',
+            '  name = "main",',
+            '  srcs = ["main.cc"],',
+            '  deps = ["@aaa//:lib_aaa"],',
+            ')',
+        ],
+    )
+    self.ScratchFile(
+        'main.cc',
+        [
+            '#include "aaa.h"',
+            'int main() {',
+            '    hello_aaa("main function");',
+            '}',
+        ],
+    )
+    _, stdout, _ = self.RunBazel(['run', '//:main'], allow_failure=False)
+    self.assertIn('main function => aaa@1.2', stdout)
+
 
 if __name__ == '__main__':
   unittest.main()

src/test/py/bazel/bzlmod/test_utils.py

@@ -63,6 +63,7 @@ def __init__(self, name, version):
     self.module_dot_bazel = None
     self.patches = []
     self.patch_strip = 0
+    self.archive_type = None
 
   def set_source(self, archive_url, strip_prefix=None):
     self.archive_url = archive_url
@@ -78,6 +79,10 @@ def set_patches(self, patches, patch_strip):
     self.patch_strip = patch_strip
     return self
 
+  def set_archive_type(self, archive_type):
+    self.archive_type = archive_type
+    return self
+
 
 class BazelRegistry:
   """A class to help create a Bazel module project from scatch and add it into the registry."""
@@ -186,9 +191,9 @@ def calc_repo_name_str(dep):
         ])
     return src_dir
 
-  def createArchive(self, name, version, src_dir):
+  def createArchive(self, name, version, src_dir, filename_pattern='%s.%s.zip'):
     """Create an archive with a given source directory."""
-    zip_path = self.archives.joinpath('%s.%s.zip' % (name, version))
+    zip_path = self.archives.joinpath(filename_pattern % (name, version))
     zip_obj = zipfile.ZipFile(str(zip_path), 'w')
     for foldername, _, filenames in os.walk(str(src_dir)):
       for filename in filenames:
@@ -225,24 +230,39 @@ def addModule(self, module):
         source['patches'][patch.name] = integrity(read(patch))
         shutil.copy(str(patch), str(patch_dir))
 
+    if module.archive_type:
+      source['archive_type'] = module.archive_type
+
     with module_dir.joinpath('source.json').open('w') as f:
       json.dump(source, f, indent=4, sort_keys=True)
 
-  def createCcModule(self,
-                     name,
-                     version,
-                     deps=None,
-                     repo_names=None,
-                     patches=None,
-                     patch_strip=0):
+  def createCcModule(
+      self,
+      name,
+      version,
+      deps=None,
+      repo_names=None,
+      patches=None,
+      patch_strip=0,
+      archive_pattern=None,
+      archive_type=None,
+  ):
     """Generate a cc project and add it as a module into the registry."""
     src_dir = self.generateCcSource(name, version, deps, repo_names)
-    archive = self.createArchive(name, version, src_dir)
+    if archive_pattern:
+      archive = self.createArchive(
+          name, version, src_dir, filename_pattern=archive_pattern
+      )
+    else:
+      archive = self.createArchive(name, version, src_dir)
     module = Module(name, version)
     module.set_source(archive.resolve().as_uri())
     module.set_module_dot_bazel(src_dir.joinpath('MODULE.bazel'))
     if patches:
       module.set_patches(patches, patch_strip)
+    if archive_type:
+      module.set_archive_type(archive_type)
+
     self.addModule(module)
     return self