feat: add "Multiple on PyPI" badge for SDK monorepos

azure-sdk-for-python and google-cloud-python were rendering "Not on
PyPI", which is misleading. Both do ship on PyPI, just as many
per-service packages (azure-identity, azure-storage-blob,
google-cloud-storage, etc.) rather than under the repo name.

pypi_name_overrides.json already recorded that distinction in its
reason field; those two entries now carry an optional "badge" value
that build.py reads into the PyPI Downloads column. The other sixteen
no-count entries (cpython, renpy, agent skill repos, etc.) keep
"Not on PyPI" since that remains accurate for them.

Co-Authored-By: Claude <noreply@anthropic.com>

website/build.py

@@ -11,7 +11,7 @@ from datetime import UTC, datetime
 from pathlib import Path
 from typing import TypedDict
 
-from fetch_pypi_downloads_via_clickpy import normalize
+from fetch_pypi_downloads_via_clickpy import OVERRIDES_FILE, normalize
 from jinja2 import Environment, FileSystemLoader
 from readme_parser import AlsoSee, ParsedGroup, ParsedSection, parse_readme, parse_sponsors, slugify
 
@@ -60,6 +60,7 @@ class TemplateEntry(TypedDict):
     last_commit_at: str | None
     source_type: str | None
     bundled: bool
+    badge: str | None
     also_see: list[AlsoSee]
 
 
@@ -101,6 +102,17 @@ def load_stars(path: Path) -> dict[str, dict]:
     return {}
 
 
+def load_pypi_badges() -> dict[str, str]:
+    """Load the badge each entry shows in place of a download count.
+
+    Keyed by normalized README name. Only entries whose missing count needs
+    explaining beyond "Not on PyPI" carry one, e.g. the SDK monorepos that do
+    ship on PyPI, just as many per-service packages.
+    """
+    raw = json.loads(OVERRIDES_FILE.read_text())
+    return {name: entry["badge"] for name, entry in raw.items() if entry.get("badge")}
+
+
 def load_downloads(path: Path) -> dict[str, int]:
     """Load last-30-day download counts from the TSV cache, keyed by normalized README name.
 
@@ -524,6 +536,7 @@ def extract_entries(
                     last_commit_at=None,
                     source_type=detect_source_type(entry["url"]),
                     bundled=bool(BUNDLED_PREFIX_RE.match(entry["description"])),
+                    badge=None,
                     also_see=entry["also_see"],
                 )
                 seen[key] = existing
@@ -576,6 +589,7 @@ def build(repo_root: Path) -> None:
 
     stars_data = load_stars(website / "data" / "github_stars.json")
     downloads_data = load_downloads(website / "data" / "pypi_downloads.tsv")
+    pypi_badges = load_pypi_badges()
 
     repo_self = stars_data.get("vinta/awesome-python", {})
     repo_stars = None
@@ -595,6 +609,7 @@ def build(repo_root: Path) -> None:
         # Stdlib entries would hit same-named PyPI backports (e.g. asyncio), not the stdlib.
         if entry.get("source_type") != BUILTIN_FILTER:
             entry["downloads"] = downloads_data.get(normalize(entry["name"]))
+        entry["badge"] = pypi_badges.get(normalize(entry["name"]))
 
     entries = sort_entries(entries)
     category_urls = {cat["name"]: category_path(cat) for cat in categories}

website/data/pypi_name_overrides.json

@@ -1,6 +1,6 @@
 {
   "autobahn-python": { "package": "autobahn", "reason": null },
-  "azure-sdk-for-python": { "package": null, "reason": "monorepo umbrella; ships as many azure-* packages, no single package represents it" },
+  "azure-sdk-for-python": { "package": null, "reason": "monorepo umbrella; ships as many azure-* packages, no single package represents it", "badge": "Multiple on PyPI" },
   "cloud-init": { "package": null, "reason": "installed via distro images, not pip" },
   "concurrent-futures": { "package": null, "reason": "stdlib module" },
   "cpython": { "package": null, "reason": "not pip-installable; the cpython PyPI name was squatted" },
@@ -12,7 +12,7 @@
   "django-rules": { "package": "rules", "reason": "PyPI django-rules is an abandoned unrelated fork; dfunckt's project publishes as rules" },
   "fasthtml": { "package": "python-fasthtml", "reason": null },
   "geodjango": { "package": null, "reason": "part of Django (django.contrib.gis), not a package" },
-  "google-cloud-python": { "package": null, "reason": "monorepo umbrella; ships as many google-cloud-* packages" },
+  "google-cloud-python": { "package": null, "reason": "monorepo umbrella; ships as many google-cloud-* packages", "badge": "Multiple on PyPI" },
   "graphify": { "package": null, "reason": "GitHub-only so far; young project, remap if it publishes" },
   "httpx-url": { "package": null, "reason": "class in the httpx package, not a package" },
   "jinja": { "package": "jinja2", "reason": "PyPI jinja is Jinja1; current releases publish as jinja2" },

website/fetch_pypi_downloads_via_clickpy.py

@@ -56,8 +56,10 @@ def normalize(name: str) -> str:
 def load_overrides() -> dict[str, str | None]:
     """Load curated overrides: normalized README name -> PyPI package to measure (None = never query).
 
-    Each file entry is {"package": str | null, "reason": str | null}; the reason is
-    documentation only — required for null packages, optional for remaps.
+    Each file entry is {"package": str | null, "reason": str | null} plus an
+    optional "badge" the website renders in place of a download count; the
+    reason is documentation only — required for null packages, optional for
+    remaps.
     """
     raw = json.loads(OVERRIDES_FILE.read_text())
     return {name: entry["package"] for name, entry in raw.items()}

website/templates/category.html

@@ -157,8 +157,9 @@
             {% if entry.downloads is not none %}{{
             "{:,}".format(entry.downloads) }}{% elif entry.source_type ==
             'Stdlib' %}<span class="source-badge">Stdlib</span>{% elif
-            entry.bundled %}<span class="source-badge">Bundled</span>{% else
-            %}<span class="source-badge">Not on PyPI</span>{% endif %}
+            entry.bundled %}<span class="source-badge">Bundled</span>{% elif
+            entry.badge %}<span class="source-badge">{{ entry.badge }}</span>{%
+            else %}<span class="source-badge">Not on PyPI</span>{% endif %}
           </td>
           <td class="col-stars">
             {% if entry.stars is not none %}{{ "{:,}".format(entry.stars) }}{%

website/templates/index.html

@@ -194,8 +194,9 @@
             {% if entry.downloads is not none %}{{
             "{:,}".format(entry.downloads) }}{% elif entry.source_type ==
             'Stdlib' %}<span class="source-badge">Stdlib</span>{% elif
-            entry.bundled %}<span class="source-badge">Bundled</span>{% else
-            %}<span class="source-badge">Not on PyPI</span>{% endif %}
+            entry.bundled %}<span class="source-badge">Bundled</span>{% elif
+            entry.badge %}<span class="source-badge">{{ entry.badge }}</span>{%
+            else %}<span class="source-badge">Not on PyPI</span>{% endif %}
           </td>
           <td class="col-stars">
             {% if entry.stars is not none %}{{ "{:,}".format(entry.stars) }}{%

website/tests/test_build.py

@@ -18,6 +18,7 @@ from build import (
     extract_entries,
     extract_github_repo,
     load_downloads,
+    load_pypi_badges,
     load_stars,
     sort_entries,
     subcategory_path,
@@ -1012,6 +1013,16 @@ class TestLoadStars:
 # ---------------------------------------------------------------------------
 
 
+class TestLoadPypiBadges:
+    def test_only_entries_with_a_badge_are_returned(self):
+        badges = load_pypi_badges()
+        assert badges["azure-sdk-for-python"] == "Multiple on PyPI"
+        assert badges["google-cloud-python"] == "Multiple on PyPI"
+        # Entries whose missing count needs no explaining stay off the map
+        assert "cpython" not in badges
+        assert "tomllib" not in badges
+
+
 def _template_entry(name: str, stars: int | None, source_type: str | None = None) -> TemplateEntry:
     return TemplateEntry(
         name=name,
@@ -1026,6 +1037,7 @@ def _template_entry(name: str, stars: int | None, source_type: str | None = None
         last_commit_at=None,
         source_type=source_type,
         bundled=False,
+        badge=None,
         also_see=[],
     )