Switch to consuming JSON from GitHub for "latest commit touching file"

tianon · tianon · commit c007698b9f91 · 2024-01-09T11:27:09.000-08:00
For some reason, https://github.com/xxx/yyy/commits/HEAD/zzz.atom has recently changed from returning XML by default to returning JSON, which from https://docs.github.com/en/rest/activity/feeds?apiVersion=2022-11-28 seems like it's intentional, so this adds an explicit `Accept:` header (the only explicit JSON one that seems to work) and uses `jq` to determine the latest commit instead of hacking up the XML via `grep`+`awk`. See also docker-library/docker#472 (Additionally, this switches the file to use `wget` consistently instead of a mix of `wget` and `curl`)
diff --git a/versions.sh b/versions.sh
@@ -17,10 +17,12 @@ else
 fi
 versions=( "${versions[@]%/}" )
 
-getPipCommit="$(curl -fsSL 'https://github.com/pypa/get-pip/commits/main/public/get-pip.py.atom' | grep -E 'id.*Commit')"
-getPipCommit="$(awk <<<"$getPipCommit" -F '[[:space:]]*[<>/]+' '$2 == "id" && $3 ~ /Commit/ { print $4; exit }')"
+getPipCommit="$(
+	wget -qO- --header 'Accept: application/json' 'https://github.com/pypa/get-pip/commits/main/public/get-pip.py.atom' \
+		| jq -r '.payload | first(.commitGroups[].commits[].oid)'
+)"
 getPipUrl="https://github.com/pypa/get-pip/raw/$getPipCommit/public/get-pip.py"
-getPipSha256="$(curl -fsSL "$getPipUrl" | sha256sum | cut -d' ' -f1)"
+getPipSha256="$(wget -qO- "$getPipUrl" | sha256sum | cut -d' ' -f1)"
 export getPipCommit getPipUrl getPipSha256
 
 has_linux_version() {
@@ -64,7 +66,7 @@ for version in "${versions[@]}"; do
 				|| :
 
 			# this page has a very aggressive varnish cache in front of it, which is why we also scrape tags from GitHub
-			curl -fsSL 'https://www.python.org/ftp/python/' \
+			wget -qO- 'https://www.python.org/ftp/python/' \
 				| grep '<a href="'"$rcVersion." \
 				| sed -r 's!.*<a href="([^"/]+)/?".*!\1!' \
 				| grep $rcGrepV -E -- '[a-zA-Z]+' \
