Add support for base_url

conda_index/cli/__init__.py

@@ -83,6 +83,14 @@
         will keep python 3.8.X and 3.9.Y in the current_index.json, instead of only the very latest python version.
         """,
 )
+@click.option(
+    "--base-url",
+    help="""
+        If packages should be served separately from repodata.json, URL of the
+        directory tree holding packages. Generates repodata_version=2 which is
+        incompatible with existing versions of conda.
+        """,
+)
 @click.option("--threads", default=MAX_THREADS_DEFAULT, show_default=True)
 @click.option(
     "--verbose",
@@ -107,6 +115,7 @@ def cli(
     rss=False,
     run_exports=False,
     compact=True,
+    base_url=None,
 ):
     logutil.configure()
     if verbose:
@@ -125,6 +134,7 @@ def cli(
         threads=threads,
         write_run_exports=run_exports,
         compact_json=compact,
+        base_url=base_url,
     )
 
     current_index_versions = None

conda_index/index/__init__.py

@@ -480,14 +480,15 @@ class ChannelIndex:
     :param output_root: Path to write repodata.json etc; defaults to channel_root.
     :param channel_url: fsspec URL where package files live. If provided, channel_root will only be used for cache and index output.
     :param fs: ``MinimalFS`` instance to be used with channel_url. Wrap fsspec AbstractFileSystem with ``conda_index.index.fs.FsspecFS(fs)``.
+    :param base_url: Add ``base_url/<subdir>`` to repodata.json to be able to host packages separate from repodata.json
     """
 
     fs: MinimalFS | None = None
     channel_url: str | None = None
 
     def __init__(
         self,
-        channel_root: Path,
+        channel_root: Path | str,
         channel_name: str | None,
         subdirs: Iterable[str] | None = None,
         threads: int | None = MAX_THREADS_DEFAULT,
@@ -499,8 +500,10 @@ def __init__(
         write_zst=False,
         write_run_exports=False,
         compact_json=True,
+        *,
         channel_url: str | None = None,
         fs: MinimalFS | None = None,
+        base_url: str | None = None,
     ):
         if threads is None:
             threads = MAX_THREADS_DEFAULT
@@ -526,6 +529,7 @@ def __init__(
         self.write_zst = write_zst
         self.write_run_exports = write_run_exports
         self.compact_json = compact_json
+        self.base_url = base_url
 
     def index(
         self,
@@ -748,6 +752,11 @@ def index_subdir(self, subdir, verbose=False, progress=False):
             "removed": [],  # can be added by patch/hotfix process
         }
 
+        if self.base_url:
+            # per https://github.com/conda-incubator/ceps/blob/main/cep-15.md
+            new_repodata["info"]["base_url"] = f"{self.base_url.rstrip('/')}/{subdir}/"
+            new_repodata["repodata_version"] = 2
+
         return new_repodata
 
     def cache_for_subdir(self, subdir):

conda_index/index/convert_cache.py

@@ -153,9 +153,7 @@ def basename(path):
             return path
         return path.rsplit("/")[-1]
 
-    conn.create_function(
-        "migrate_basename", narg=1, func=basename, deterministic=True
-    )
+    conn.create_function("migrate_basename", narg=1, func=basename, deterministic=True)
 
     for table in TABLE_NAMES + ["stat"]:
         conn.execute(

news/base-url

@@ -0,0 +1,20 @@
+### Enhancements
+
+* Add support for CEP-15 ``base_url`` to host packages separate from repodata.
+  (#150)
+
+### Bug fixes
+
+* <news item>
+
+### Deprecations
+
+* <news item>
+
+### Docs
+
+* <news item>
+
+### Other
+
+* <news item>

tests/test_index.py

@@ -3,6 +3,7 @@
 import os
 import shutil
 import tarfile
+import urllib.parse
 from logging import getLogger
 from os.path import dirname, isdir, isfile, join
 from pathlib import Path
@@ -1303,3 +1304,34 @@ def test_bad_patch_version(index_data):
         channel_index._create_patch_instructions(
             "noarch", {"packages": {}}, patch_generator=str(instructions)
         )
+
+
+def test_base_url(index_data):
+    """
+    conda-index should be able to add base_url to repodata.json.
+    """
+    pkg_dir = Path(index_data, "packages")
+
+    # compact json
+    channel_index = conda_index.index.ChannelIndex(
+        pkg_dir,
+        None,
+        write_bz2=False,
+        write_zst=False,
+        compact_json=True,
+        threads=1,
+        base_url="https://example.org/somechannel/",
+    )
+
+    channel_index.index(None)
+
+    osx = json.loads((pkg_dir / "osx-64" / "repodata.json").read_text())
+    noarch = json.loads((pkg_dir / "noarch" / "repodata.json").read_text())
+
+    assert osx["repodata_version"] == 2
+
+    assert osx["info"]["base_url"] == "https://example.org/somechannel/osx-64/"
+    assert noarch["info"]["base_url"] == "https://example.org/somechannel/noarch/"
+
+    package_url = urllib.parse.urljoin(osx["info"]["base_url"], "package-1.0.conda")
+    assert package_url == "https://example.org/somechannel/osx-64/package-1.0.conda"