1 files changed, 59 insertions, 0 deletions

@@ -0,0 +1,59 @@
+#!/bin/sh
+
+main() {
+        # get-feed.sh URL(str) => DIRECTIVE(feed_directive)
+        url="$1"
+        wp="$(mktemp /tmp/get-feed.XXXXXX)"
+        curl -sL "$url" >"$wp"
+        case "$url" in
+        *html) # We know it's a webpage
+                type=html
+                ;;
+        *xml) # We know it's a feed
+                type=xml
+                ;;
+        *) # Not sure
+                type="$(head -n1 "$wp")"
+                ;;
+        esac
+        case "$type" in
+        *xml*) # a feed
+                title="$(get_title_xml <"$wp")"
+                output_feed "$title" "$url"
+                ;;
+        *html*) # a webpage
+                cat "$wp" | sfeed_web | cut -f1 |
+                        while read u; do
+                                title="$(curl -sL "$u" | get_title_xml)"
+                                output_feed "$title" "$u"
+                        done
+                ;;
+        *)
+                echo >&2 "Don't know type \"$type\"."
+                exit 1
+                ;;
+        esac
+}
+
+output_feed() {
+        ## output_feed TITLE(str) URL(str) => FEED_DIRECTIVE(str)
+        printf "feed \"%s\" '%s'\n" "$1" "$2"
+}
+
+get_title_xml() {
+        ## get_title_xml < FILE => TITLE(str)
+        awk '
+/<channel>/ { channel = 1; }/<item>/{ channel = 0; }
+channel && $0 ~ /<title>/ { title = 1; }
+title {
+      if (match($0,/<\/title>/)) title = 0;
+      gsub(/<\/?title>/,"");
+      sub(/^[ \t]*/,"");
+      sub(/[ \t]*$/,"");
+      print;
+}
+channel && $0 ~ /<\/title>/ { title = 0; }
+'
+}
+
+main "$@"