refactor: deduplicate fetch_json, EOL filtering, and version sort

- Extract shared fetch_json into rhdh_lifecycle.redhat; remove local
  copies from check_aks_lifecycle.py, check_eks_lifecycle.py,
  check_gke_lifecycle.py, and pg.py
- Extract filter_supported_eol_entries helper to deduplicate the
  identical EOL cross-verification blocks in AKS and EKS scripts
- Replace all inline version sort lambdas with ver_sort_key from
  rhdh_lifecycle.redhat (lifecycle scripts) and rhdh_prow.__init__
  (prow scripts); add AGENTS.md sync rule for the prow copy

Assisted-by: OpenCode

Author: zdrapela

AGENTS.md

@@ -78,6 +78,8 @@ Bump all three when releasing.
 
 `skills/prow/scripts/rhdh_prow/repo.py` and `skills/prow/scripts/rhdh_prow/yaml.py` are copies of `skills/lifecycle/scripts/rhdh_lifecycle/repo.py` and `skills/lifecycle/scripts/rhdh_lifecycle/yaml.py`. The only difference is the internal import path (`rhdh_prow.repo` vs `rhdh_lifecycle.repo`). When modifying either copy, update both to keep them in sync.
 
+`ver_sort_key` in `skills/prow/scripts/rhdh_prow/__init__.py` is a copy of the same function in `skills/lifecycle/scripts/rhdh_lifecycle/redhat.py`. Keep both in sync.
+
 ## Agent skills
 
 ### Issue tracker

skills/lifecycle/scripts/check_aks_lifecycle.py

@@ -14,29 +14,17 @@
 import argparse
 import json
 import sys
-import urllib.error
-import urllib.request
 from datetime import datetime, timezone
 
 from rhdh_lifecycle.configured_versions import print_configured_versions
+from rhdh_lifecycle.redhat import fetch_json, filter_supported_eol_entries, ver_sort_key
 from rhdh_lifecycle.repo import resolve_repo_root
 
 AKS_API_URL = "https://releases.aks.azure.com/parsed_data.json"
 EOL_API_URL = "https://endoflife.date/api/azure-kubernetes-service.json"
 CONFIG_DIR = "ci-operator/config/redhat-developer/rhdh"
 
 
-def fetch_json(url):
-    """Fetch JSON from a URL."""
-    req = urllib.request.Request(url, headers={"User-Agent": "rhdh-skill"})
-    try:
-        with urllib.request.urlopen(req, timeout=30) as resp:
-            return json.loads(resp.read().decode("utf-8"))
-    except (urllib.error.URLError, OSError) as exc:
-        print(f"ERROR: Failed to fetch {url}: {exc}", file=sys.stderr)
-        return None
-
-
 def main(argv=None):
     parser = argparse.ArgumentParser(description="Check AKS K8s version lifecycle.")
     parser.add_argument("--mapt-ref", help="Path to MAPT ref YAML (repo-relative)")
@@ -83,7 +71,7 @@ def main(argv=None):
 
     sorted_versions = sorted(
         minor_versions.items(),
-        key=lambda x: [int(n) for n in x[0].split(".")],
+        key=lambda x: ver_sort_key(x[0]),
         reverse=True,
     )
 
@@ -98,23 +86,7 @@ def main(argv=None):
     # Cross-verify with endoflife.date
     today = datetime.now(timezone.utc).strftime("%Y-%m-%d")
     eol_data = fetch_json(EOL_API_URL)
-    eol_supported = []
-    if eol_data:
-        for entry in eol_data:
-            eol = entry.get("eol", "N/A")
-            ext = entry.get("extendedSupport", "N/A")
-            has_support = False
-            if eol == "N/A":
-                has_support = True
-            elif isinstance(eol, bool):
-                has_support = not eol
-            elif isinstance(eol, str) and eol > today:
-                has_support = True
-            if not has_support and isinstance(ext, str) and ext > today:
-                has_support = True
-            if has_support:
-                eol_supported.append(entry)
-        eol_supported.sort(key=lambda e: [int(x) for x in e["cycle"].split(".")], reverse=True)
+    eol_supported = filter_supported_eol_entries(eol_data, today) if eol_data else []
 
     # JSON output
     if args.json_output:

skills/lifecycle/scripts/check_eks_lifecycle.py

@@ -20,6 +20,7 @@
 from datetime import datetime, timezone
 
 from rhdh_lifecycle.configured_versions import print_configured_versions
+from rhdh_lifecycle.redhat import fetch_json, filter_supported_eol_entries
 from rhdh_lifecycle.repo import resolve_repo_root
 
 EKS_DOCS_URL = (
@@ -41,17 +42,6 @@ def fetch_text(url):
         return None
 
 
-def fetch_json(url):
-    """Fetch JSON from a URL."""
-    req = urllib.request.Request(url, headers={"User-Agent": "rhdh-skill"})
-    try:
-        with urllib.request.urlopen(req, timeout=30) as resp:
-            return json.loads(resp.read().decode("utf-8"))
-    except (urllib.error.URLError, OSError) as exc:
-        print(f"ERROR: Failed to fetch {url}: {exc}", file=sys.stderr)
-        return None
-
-
 def parse_supported_versions(docs):
     """Extract supported versions and their tiers from the AsciiDoc."""
     section = ""
@@ -125,23 +115,7 @@ def main(argv=None):
     # Cross-verify with endoflife.date
     today = datetime.now(timezone.utc).strftime("%Y-%m-%d")
     eol_data = fetch_json(EOL_API_URL)
-    eol_supported = []
-    if eol_data:
-        for entry in eol_data:
-            eol = entry.get("eol", "N/A")
-            ext = entry.get("extendedSupport", "N/A")
-            has_support = False
-            if eol == "N/A":
-                has_support = True
-            elif isinstance(eol, bool):
-                has_support = not eol
-            elif isinstance(eol, str) and eol > today:
-                has_support = True
-            if not has_support and isinstance(ext, str) and ext > today:
-                has_support = True
-            if has_support:
-                eol_supported.append(entry)
-        eol_supported.sort(key=lambda e: [int(x) for x in e["cycle"].split(".")], reverse=True)
+    eol_supported = filter_supported_eol_entries(eol_data, today) if eol_data else []
 
     # JSON output
     if args.json_output:

skills/lifecycle/scripts/check_gke_lifecycle.py

@@ -15,22 +15,11 @@
 import argparse
 import json
 import sys
-import urllib.error
-import urllib.request
 from datetime import datetime, timezone
 
-API_URL = "https://endoflife.date/api/google-kubernetes-engine.json"
-
+from rhdh_lifecycle.redhat import fetch_json, ver_sort_key
 
-def fetch_api():
-    """Fetch GKE lifecycle data from endoflife.date."""
-    req = urllib.request.Request(API_URL, headers={"User-Agent": "rhdh-skill"})
-    try:
-        with urllib.request.urlopen(req, timeout=30) as resp:
-            return json.loads(resp.read().decode("utf-8"))
-    except (urllib.error.URLError, OSError) as exc:
-        print(f"ERROR: Failed to fetch {API_URL}: {exc}", file=sys.stderr)
-        sys.exit(1)
+API_URL = "https://endoflife.date/api/google-kubernetes-engine.json"
 
 
 def is_supported(entry, today):
@@ -66,10 +55,12 @@ def main(argv=None):
     args = parser.parse_args(argv)
 
     today = datetime.now(timezone.utc).strftime("%Y-%m-%d")
-    data = fetch_api()
+    data = fetch_json(API_URL)
+    if not data:
+        sys.exit(1)
 
     supported = [e for e in data if is_supported(e, today)]
-    supported.sort(key=lambda e: [int(x) for x in e["cycle"].split(".")], reverse=True)
+    supported.sort(key=lambda e: ver_sort_key(e["cycle"]), reverse=True)
 
     if args.json_output:
         result = []

skills/lifecycle/scripts/rhdh_lifecycle/ocp.py

@@ -12,7 +12,7 @@
 
 import re
 
-from rhdh_lifecycle.redhat import is_date, to_date
+from rhdh_lifecycle.redhat import is_date, to_date, ver_sort_key
 
 
 def classify_ocp_versions(api_data, today):
@@ -100,5 +100,5 @@ def classify_ocp_versions(api_data, today):
         )
 
     # Sort by version
-    results.sort(key=lambda v: [int(x) for x in v["version"].split(".")])
+    results.sort(key=lambda v: ver_sort_key(v["version"]))
     return results

skills/lifecycle/scripts/rhdh_lifecycle/pg.py

@@ -12,10 +12,7 @@
 
 from __future__ import annotations
 
-import json
-import sys
-import urllib.error
-import urllib.request
+from rhdh_lifecycle.redhat import fetch_json
 
 PROVIDERS = {
     "upstream": "https://endoflife.date/api/postgresql.json",
@@ -24,17 +21,6 @@
 }
 
 
-def _fetch_json(url):
-    """Fetch JSON from a URL, return None on failure."""
-    req = urllib.request.Request(url, headers={"User-Agent": "rhdh-skill"})
-    try:
-        with urllib.request.urlopen(req, timeout=30) as resp:
-            return json.loads(resp.read().decode("utf-8"))
-    except (urllib.error.URLError, OSError) as exc:
-        print(f"WARNING: Failed to fetch {url}: {exc}", file=sys.stderr)
-        return None
-
-
 def _normalize_eol(val):
     """Normalize EOL value to a date string or 'N/A'."""
     if val is None or val == "N/A":
@@ -56,7 +42,7 @@ def fetch_pg_lifecycle(today=None):
     # Fetch all providers
     provider_data = {}
     for provider, url in PROVIDERS.items():
-        data = _fetch_json(url)
+        data = fetch_json(url)
         if data:
             provider_data[provider] = {str(e["cycle"]): e for e in data}
         else:

skills/lifecycle/scripts/rhdh_lifecycle/redhat.py

@@ -66,11 +66,51 @@ def ver_sort_key(version_str):
         return [0]
 
 
+def filter_supported_eol_entries(eol_data, today):
+    """Filter endoflife.date entries to those still supported.
+
+    Considers both ``eol`` and ``extendedSupport`` fields. Returns the
+    filtered list sorted by cycle version (newest first).
+    """
+    supported = []
+    for entry in eol_data:
+        eol = entry.get("eol", "N/A")
+        ext = entry.get("extendedSupport", "N/A")
+        has_support = False
+        if eol == "N/A":
+            has_support = True
+        elif isinstance(eol, bool):
+            has_support = not eol
+        elif isinstance(eol, str) and eol > today:
+            has_support = True
+        if not has_support and isinstance(ext, str) and ext > today:
+            has_support = True
+        if has_support:
+            supported.append(entry)
+    supported.sort(key=lambda e: ver_sort_key(e["cycle"]), reverse=True)
+    return supported
+
+
 def resolve_product_name(product):
     """Resolve a product alias to the full API product name."""
     return PRODUCT_ALIASES.get(product.lower(), product)
 
 
+def fetch_json(url):
+    """Fetch JSON from a URL.
+
+    Returns the parsed JSON, or None on failure. Shared by all lifecycle
+    scripts that consume external APIs (endoflife.date, AKS, EKS, etc.).
+    """
+    req = urllib.request.Request(url, headers={"User-Agent": "rhdh-skill"})
+    try:
+        with urllib.request.urlopen(req, timeout=30) as resp:
+            return json.loads(resp.read().decode("utf-8"))
+    except (urllib.error.URLError, OSError) as exc:
+        print(f"ERROR: Failed to fetch {url}: {exc}", file=sys.stderr)
+        return None
+
+
 def fetch_api(product_name):
     """Fetch raw lifecycle data from the Red Hat Product Life Cycles API.
 

skills/prow/scripts/analyze_coverage.py

@@ -24,6 +24,7 @@
 from datetime import datetime, timezone
 from pathlib import Path
 
+from rhdh_prow import ver_sort_key
 from rhdh_prow.repo import resolve_repo_root
 from rhdh_prow.yaml import extract_branch, fetch_yaml, list_yaml_files
 
@@ -32,11 +33,6 @@
 LIFECYCLE_API_URL = "https://access.redhat.com/product-life-cycles/api/v1/products"
 
 
-def ver_key(v):
-    """Sort key for version strings like '4.16'."""
-    return [int(x) for x in v.split(".")]
-
-
 def _fetch_lifecycle_json(script_name):
     """Run a lifecycle script with --json and return parsed output.
 
@@ -97,7 +93,7 @@ def _get_rhdh_lifecycle():
                 "ocp_versions": ocp_versions,
             }
         )
-    results.sort(key=lambda v: ver_key(v["version"]) if "." in v["version"] else [0])
+    results.sort(key=lambda v: ver_sort_key(v["version"]) if "." in v["version"] else [0])
     return results
 
 
@@ -166,7 +162,7 @@ def _to_date(val):
                 "phase": current_phase,
             }
         )
-    results.sort(key=lambda v: ver_key(v["version"]))
+    results.sort(key=lambda v: ver_sort_key(v["version"]))
     return results
 
 
@@ -228,15 +224,15 @@ def main(argv=None):
                 for t in data["tests"]
                 if t.get("cluster_claim", {}).get("version")
             },
-            key=ver_key,
+            key=ver_sort_key,
         )
         if versions:
             branch_versions[branch] = versions
             all_test_versions.extend(versions)
             print(f"  {branch}: {' '.join(versions)}")
     print()
 
-    unique_test_versions = sorted(set(all_test_versions), key=ver_key)
+    unique_test_versions = sorted(set(all_test_versions), key=ver_sort_key)
 
     # ------- 3. RHDH lifecycle -------
     print("--- RHDH Lifecycle ---")
@@ -254,7 +250,7 @@ def main(argv=None):
 
     rhdh_supported_ocp = sorted(
         {ocp for v in rhdh_data if v["supported"] for ocp in v.get("ocp_versions", [])},
-        key=ver_key,
+        key=ver_sort_key,
     )
     print()
     print(f"  OCP versions supported by active RHDH releases: {' '.join(rhdh_supported_ocp)}")
@@ -285,14 +281,14 @@ def main(argv=None):
 
     # Compute "main" branch OCP support
     if latest_rhdh_ocp:
-        max_rhdh = max(latest_rhdh_ocp, key=ver_key)
-        max_parts = ver_key(max_rhdh)
+        max_rhdh = max(latest_rhdh_ocp, key=ver_sort_key)
+        max_parts = ver_sort_key(max_rhdh)
         main_ocp = list(latest_rhdh_ocp)
         for ocp_ver in ocp_supported:
-            ocp_parts = ver_key(ocp_ver)
+            ocp_parts = ver_sort_key(ocp_ver)
             if ocp_parts > max_parts and ocp_ver not in main_ocp:
                 main_ocp.append(ocp_ver)
-        rhdh_branch_ocp["main"] = sorted(main_ocp, key=ver_key)
+        rhdh_branch_ocp["main"] = sorted(main_ocp, key=ver_sort_key)
 
     # ------- 5. OCP version matrix -------
     print("--- OCP Version Matrix ---")
@@ -306,7 +302,7 @@ def main(argv=None):
 
     all_relevant = sorted(
         set(pool_versions + unique_test_versions + rhdh_supported_ocp + ocp_supported),
-        key=ver_key,
+        key=ver_sort_key,
     )
 
     ocp_phase_map = {v["version"]: v for v in ocp_lifecycle}
@@ -379,7 +375,7 @@ def main(argv=None):
     print("--- RHDH-Supported OCP Versions Missing Pools (ADD) ---")
     all_rhdh_ocp = sorted(
         {ver for ocp_list in rhdh_branch_ocp.values() for ver in ocp_list},
-        key=ver_key,
+        key=ver_sort_key,
     )
     for ver in all_rhdh_ocp:
         if ver in ocp_eol:

skills/prow/scripts/list_cluster_pools.py

@@ -11,6 +11,7 @@
 import sys
 from pathlib import Path
 
+from rhdh_prow import ver_sort_key
 from rhdh_prow.repo import resolve_repo_root
 from rhdh_prow.yaml import fetch_yaml, list_yaml_files
 
@@ -63,7 +64,7 @@ def main(argv=None):
         rows.append((ver, pool_name, size, max_size, running, image_set, filename))
 
     # Sort by version
-    rows.sort(key=lambda r: [int(x) for x in r[0].split(".")])
+    rows.sort(key=lambda r: ver_sort_key(r[0]))
     for ver, pool_name, size, max_size, running, image_set, filename in rows:
         print(
             f"  {ver:<8s}  {pool_name:<50s}  {size:<5s}  {max_size:<5s}  "