[chef-inspec] Use GitHub releases instead of git tags (#367)

GitHub releases gives better dates than git tags.

src/chef-inspec.py

@@ -1,6 +1,5 @@
 from bs4 import BeautifulSoup
-from common import dates, http, releasedata
+from common import dates, github, http, releasedata
-from common.git import Git
 
 """Fetch released versions from docs.chef.io and retrieve their date from GitHub.
 docs.chef.io needs to be scraped because not all tagged versions are actually released.
@@ -13,12 +12,8 @@ with releasedata.ProductData("chef-inspec") as product_data:
     rn_soup = BeautifulSoup(rn_response.text, features="html5lib")
     released_versions = [h2.get('id') for h2 in rn_soup.find_all('h2', id=True) if h2.get('id')]
 
-    git = Git("https://github.com/inspec/inspec.git")
+    for release in github.fetch_releases("inspec/inspec"):
-    git.setup(bare=True)
+        sanitized_version = release.tag_name.replace("v", "")
-
-    versions = git.list_tags()
-    for version, date_str in versions:
-        sanitized_version = version.replace("v", "")
         if sanitized_version in released_versions:
-            date = dates.parse_date(date_str)
+            date = dates.parse_datetime(release.published_at)
             product_data.declare_version(sanitized_version, date)

src/common/github.py

@@ -0,0 +1,50 @@
+import json
+import logging
+import subprocess
+
+
+class Release:
+    def __init__(self, name: str, tag_name: str, published_at: str, is_prerelease: bool) -> None:
+        self.name: str = name
+        self.tag_name: str = tag_name
+        self.published_at: str = published_at
+        self.is_prerelease: bool = is_prerelease
+
+
+def fetch_releases(repo_id: str) -> list[Release]:
+    logging.info(f"fetching {repo_id} GitHub releases")
+    (owner, repo) = repo_id.split('/')
+    child = subprocess.run("""gh api graphql --paginate -f query='
+query($endCursor: String) {
+  repository(name: "%s", owner: "%s") {
+    releases(
+      orderBy: {field: NAME, direction: ASC}
+      first: 100
+      after: $endCursor
+    ) {
+      pageInfo { hasNextPage, endCursor }
+      edges {
+        node {
+          name
+          publishedAt
+          isPrerelease
+          tagName
+        }
+      }
+    }
+  }
+}'""" % (repo, owner), capture_output=True, timeout=300, check=True, shell=True)  # noqa: UP031
+    logging.info(f"fetched {repo_id} GitHub releases")
+
+    # splitting because response may contain multiple JSON objects on a single line
+    responses = child.stdout.decode("utf-8").strip().replace('}{', '}\n{').split("\n")
+    pages = [json.loads(response) for response in responses]
+
+    releases = []
+    for page in pages:
+        for edge in page['data']['repository']['releases']['edges']:
+            release_data = edge['node']
+            releases.append(Release(release_data['name'], release_data['tagName'], release_data['publishedAt'],
+                                    release_data['isPrerelease']))
+
+    return releases

src/github_releases.py

@@ -1,9 +1,6 @@
-import json
-import logging
-import subprocess
 import sys
 
-from common import dates, endoflife, releasedata
+from common import dates, endoflife, github, releasedata
 
 """Fetches versions from GitHub releases using the GraphQL API and the GitHub CLI.
 
@@ -13,47 +10,17 @@ Note: GraphQL API and GitHub CLI are used because it's simpler: no need to manag
 METHOD = "github_releases"
 
 
-def fetch_releases(repo_id: str) -> list[dict]:
-    logging.info(f"fetching {repo_id} GitHub releases")
-    (owner, repo) = repo_id.split('/')
-    child = subprocess.run("""gh api graphql --paginate -f query='
-query($endCursor: String) {
-  repository(name: "%s", owner: "%s") {
-    releases(
-      orderBy: {field: NAME, direction: ASC}
-      first: 100
-      after: $endCursor
-    ) {
-      pageInfo { hasNextPage, endCursor }
-      edges {
-        node {
-          name
-          publishedAt
-          isPrerelease
-        }
-      }
-    }
-  }
-}'""" % (repo, owner), capture_output=True, timeout=300, check=True, shell=True)  # noqa: UP031
-    logging.info(f"fetched {repo_id} GitHub releases")
-
-    # splitting because response may contain multiple JSON objects on a single line
-    responses = child.stdout.decode("utf-8").strip().replace('}{', '}\n{').split("\n")
-    return [json.loads(response) for response in responses]
-
-
 p_filter = sys.argv[1] if len(sys.argv) > 1 else None
 m_filter = sys.argv[2] if len(sys.argv) > 2 else None
 for config in endoflife.list_configs(p_filter, METHOD, m_filter):
     with releasedata.ProductData(config.product) as product_data:
-        for page in fetch_releases(config.url):
+        for release in github.fetch_releases(config.url):
-            releases = [edge['node'] for edge in (page['data']['repository']['releases']['edges'])]
+            if release.is_prerelease:
+                continue
 
-            for release in releases:
+            version_str = release.name
-                if not release['isPrerelease']:
-                    version_str = release['name']
             version_match = config.first_match(version_str)
             if version_match:
                 version = config.render(version_match)
-                        date = dates.parse_datetime(release['publishedAt'])
+                date = dates.parse_datetime(release.published_at)
                 product_data.declare_version(version, date)