Add category override mapping and use it in library update method (#1411)

Fixes #1034

- Adds a `CATEGORY_OVERRIDES` mapping
- Uses the override in `update_libraries` management command

### Manual testing
- Ran `./manage.py import_versions`
- Ran `./manage.py update_libraries`
- Static String had the correct category
<img width="440" alt="Screenshot 2024-11-04 at 5 24 05 PM"
src="https://github.com/user-attachments/assets/4b0cf4cb-15e2-4337-9284-7659afd40643">

### Post-deploy tasks
- Ensure `update_libraries` is run
- Verify **Container** category is orphaned
- Delete **Container** category

Co-authored-by: Greg Kaleka <gkaleka@energy-solution.com>

docs/syncing_data_with_github.md

@@ -21,15 +21,15 @@ Not all tags are official GitHub **Releases**, however, and this impacts where w
 To retrieve releases and tags, run:
 
 ```bash
-./manage.py import_releases
+./manage.py import_versions
 ```
 
-This will:
+Note the command enqueues a celery task rather than running synchronously. The task will:
 
-- Delete existing Versions and LibraryVersions
+- Delete existing Versions and LibraryVersions if you pass `--delete-versions` to the command
 - Retrieve tags and releases from the Boost GitHub repo
 - Create new Versions for each tag and release that is not a beta or rc release
-- Create a new LibraryVersion for each Library **but not for historical versions**
+- Create a new LibraryVersion for each Library (including for historical versions unless you pass `--new`)
 
 
 ## Library data

libraries/constants.py

@@ -288,6 +288,14 @@ LIBRARY_DOCS_EXCEPTIONS = {
     ],
 }
 
+# Mapping for duplicate library categories. If a library has one of the keys
+# in this mapping as its category, the associated canonical category should be used.
+# key: Duplicate category's Category.slug
+# value: Category.slug of the canonical category
+CATEGORY_OVERRIDES = {
+    "Container": "Containers",
+}
+
 
 # This constant is for library-version docs that we know are missing
 LIBRARY_DOCS_MISSING = {

libraries/github.py

@@ -17,6 +17,7 @@ from django.db import transaction
 from django.utils import dateparse, timezone
 
 from versions.models import Version
+from .constants import CATEGORY_OVERRIDES
 from .models import (
     Category,
     Commit,
@@ -163,10 +164,7 @@ class LibraryUpdater:
     """
 
     def __init__(self, client=None, token=None):
-        if client:
-            self.client = client
-        else:
-            self.client = GithubAPIClient()
+        self.client = client or GithubAPIClient()
         self.api = self.client.initialize_api(token=token)
         self.parser = GithubDataParser()
         self.logger = structlog.get_logger()
@@ -279,6 +277,7 @@ class LibraryUpdater:
 
         obj.categories.clear()
         for cat_name in categories:
+            cat_name = CATEGORY_OVERRIDES.get(cat_name, cat_name)
             cat, _ = Category.objects.get_or_create(name=cat_name)
             obj.categories.add(cat)
 

libraries/tests/test_github.py

@@ -175,6 +175,14 @@ def test_update_categories(library, library_updater):
     assert Category.objects.filter(name="Test").exists()
     assert library.categories.filter(name="Test").exists()
 
+    # Ensure category overrides respected
+    assert Category.objects.filter(name="Containers").exists() is False
+    assert Category.objects.filter(name="Container").exists() is False
+    library_updater.update_categories(library, ["Container"])  # overridden
+    library.refresh_from_db()
+    assert Category.objects.filter(name="Containers").exists() is True
+    assert Category.objects.filter(name="Container").exists() is False
+
 
 def test_update_issues_new(
     tp, library, github_api_repo_issues_response, library_updater

versions/management/commands/import_versions.py

@@ -23,6 +23,8 @@ def command(
         verbose (bool): Enable verbose output (show logging statements)
         delete_versions (bool): If True, deletes all existing Version instances before
             importing.
+        new (bool): If True, only imports versions that do not already exist in
+            the database.
         token (str): Github API token, if you need to use something other than the
             setting.
     """