scrape github release page for changelog

.gitea/scripts/update-appimage-nix.sh

@@ -70,29 +70,72 @@ release_tag="${release_tag%\}}"
 release_tag="${release_tag#\'}"
 release_tag="${release_tag%\'}"
 
-api_url="https://api.github.com/repos/${RELEASE_API_REPO}/releases/tags/${release_tag}"
+release_url="${LATEST_RELEASE_URL%/latest}/tag/${release_tag}"
-
+release_html=$(curl -fsSL "$release_url" || true)
-curl_headers=(
-  -H "Accept: application/vnd.github+json"
-  -H "X-GitHub-Api-Version: 2022-11-28"
-)
-if [ -n "${GITHUB_TOKEN:-}" ]; then
-  curl_headers+=( -H "Authorization: Bearer ${GITHUB_TOKEN}" )
-fi
-
-api_response=$(curl -sS -w '\n%{http_code}' "${curl_headers[@]}" "$api_url" || true)
-api_body=$(printf '%s\n' "$api_response" | sed '$d')
-api_code=$(printf '%s\n' "$api_response" | tail -n1)
 
 release_notes=""
-if [ "$api_code" = "200" ]; then
+if [ -n "$release_html" ]; then
-  release_notes=$(printf '%s' "$api_body" | "$PYTHON_BIN" -c 'import json,sys; d=json.load(sys.stdin); print((d.get("body") or "").strip())' || true)
+  release_notes=$(printf '%s' "$release_html" | "$PYTHON_BIN" -c '
+import re, sys
+from html.parser import HTMLParser
+
+html = sys.stdin.read()
+
+m = re.search(r"<div[^>]*data-test-selector=\"body-content\"[^>]*class=\"[^\"]*markdown-body[^\"]*\"[^>]*>(.*?)</div>", html, re.S)
+if not m:
+  print("")
+  raise SystemExit(0)
+
+fragment = m.group(1)
+
+class Extractor(HTMLParser):
+  def __init__(self):
+    super().__init__()
+    self.out = []
+    self.in_code = False
+
+  def handle_starttag(self, tag, attrs):
+    if tag in ("br",):
+      self.out.append("\n")
+    elif tag in ("p", "div", "h1", "h2", "h3", "h4", "h5", "h6", "li"):
+      if self.out and not self.out[-1].endswith("\n"):
+        self.out.append("\n")
+    elif tag in ("pre",):
+      if self.out and not self.out[-1].endswith("\n"):
+        self.out.append("\n")
+      self.out.append("```\n")
+      self.in_code = True
+    elif tag in ("code",) and not self.in_code:
+      self.out.append("`")
+
+  def handle_endtag(self, tag):
+    if tag in ("p", "div", "li", "h1", "h2", "h3", "h4", "h5", "h6"):
+      if not self.out or not self.out[-1].endswith("\n"):
+        self.out.append("\n")
+    elif tag in ("pre",):
+      if not self.out or not self.out[-1].endswith("\n"):
+        self.out.append("\n")
+      self.out.append("```\n")
+      self.in_code = False
+    elif tag in ("code",) and not self.in_code:
+      self.out.append("`")
+
+  def handle_data(self, data):
+    if data:
+      self.out.append(data)
+
+parser = Extractor()
+parser.feed(fragment)
+text = "".join(parser.out)
+text = re.sub(r"\n{3,}", "\n\n", text)
+print(text.strip())
+' || true)
 else
-  echo "warning: failed to fetch release notes from GitHub API (status=$api_code, url=$api_url)"
+  echo "warning: failed to fetch release page ${release_url}"
 fi
 
 if [ -z "$release_notes" ]; then
-  release_notes="_No changelog found in upstream release notes. Check ${LATEST_RELEASE_URL%/latest}/tag/${release_tag}._"
+  release_notes="_No changelog found on upstream release page. Check ${release_url}._"
 fi
 
 delimiter="CHANGELOG_$(date +%s%N)"