Add rss trigger

Gemfile

@@ -16,6 +16,9 @@ gem "multi_json"
 # HTML/XML parsing
 gem "nokogiri"
 
+# RSS/Atom feed parsing
+gem "rss"
+
 # LLM integration
 gem "ruby_llm"
 

Gemfile.lock

@@ -398,6 +398,8 @@ GEM
       uber (< 0.2.0)
     retriable (3.4.1)
     rexml (3.4.4)
+    rss (0.3.2)
+      rexml
     rubocop (1.85.1)
       json (~> 2.3)
       language_server-protocol (~> 3.17.0.2)
@@ -520,6 +522,7 @@ DEPENDENCIES
   propshaft
   puma (>= 5.0)
   rails (~> 8.1.2)
+  rss
   rubocop-rails-omakase
   ruby_llm
   solid_cache
@@ -670,6 +673,7 @@ CHECKSUMS
   representable (3.2.0) sha256=cc29bf7eebc31653586849371a43ffe36c60b54b0a6365b5f7d95ec34d1ebace
   retriable (3.4.1) sha256=fb3f114b7d492121c158c01f3d5152b5a615c5b70d5877d0bc08c7ec3725c3bc
   rexml (3.4.4) sha256=19e0a2c3425dfbf2d4fc1189747bdb2f849b6c5e74180401b15734bc97b5d142
+  rss (0.3.2) sha256=3bd0446d32d832cda00ba07f4b179401f903b52ea1fdaac0f1f08de61a501efa
   rubocop (1.85.1) sha256=3dbcf9e961baa4c376eeeb2a03913dca5e3987033b04d38fa538aa1e7406cc77
   rubocop-ast (1.49.1) sha256=4412f3ee70f6fe4546cc489548e0f6fcf76cafcfa80fa03af67098ffed755035
   rubocop-performance (1.26.1) sha256=cd19b936ff196df85829d264b522fd4f98b6c89ad271fa52744a8c11b8f71834

lib/r3x/triggers/rss.rb

@@ -0,0 +1,122 @@
+require "rss"
+
+module R3x
+  module Triggers
+    class Rss < Base
+      include Concerns::CronSchedulable
+      include Concerns::ChangeDetecting
+
+      validates :url, presence: true
+      validates_with Validators::Url, url_field: :url
+
+      def initialize(url: nil, cron: nil, **options)
+        normalized_url = url.is_a?(String) ? url.strip : url
+        normalized_cron = cron.is_a?(String) ? cron.strip : cron
+        super(:rss, url: normalized_url, cron: normalized_cron, **options)
+      end
+
+      def url
+        options[:url]
+      end
+
+      def cron
+        options[:cron]
+      end
+
+      def unique_key
+        "rss:#{Digest::SHA256.hexdigest(url)[0..15]}"
+      end
+
+      def detect_changes(workflow_key:, state:)
+        response = Faraday.get(url)
+        raise Faraday::Error, "HTTP #{response.status}" unless response.success?
+
+        feed = RSS::Parser.parse(response.body, false)
+
+        current_links = extract_links(feed)
+        seen_links = (state[:seen_links] || []).map(&:to_s)
+        new_links = current_links - seen_links
+
+        if new_links.empty?
+          { changed: false, state: { seen_links: current_links }, payload: nil }
+        else
+          new_items = feed.items.select { |item|
+            link = extract_link(item)
+            new_links.include?(link)
+          }
+
+          {
+            changed: true,
+            state: { seen_links: current_links },
+            payload: {
+              feed_title: extract_feed_title(feed),
+              feed_url: url,
+              new_items: new_items.map { |item|
+                {
+                  title: extract_title(item),
+                  link: extract_link(item),
+                  published_at: extract_published(item),
+                  description: extract_description(item)
+                }
+              }
+            }
+          }
+        end
+      end
+
+      private
+
+      def extract_links(feed)
+        feed.items.map { |item| extract_link(item) }.compact
+      end
+
+      def extract_link(item)
+        if item.respond_to?(:link) && item.link.is_a?(RSS::Atom::Feed::Link)
+          item.link.href
+        elsif item.respond_to?(:link) && item.link.is_a?(String)
+          item.link.presence
+        elsif item.respond_to?(:guid) && item.guid
+          item.guid.content
+        end
+      end
+
+      def extract_feed_title(feed)
+        title = if feed.respond_to?(:channel) && feed.channel
+          feed.channel.title
+        else
+          feed.title
+        end
+        extract_text(title)
+      end
+
+      def extract_title(item)
+        extract_text(item.title)
+      end
+
+      def extract_text(value)
+        return nil if value.nil?
+        value.respond_to?(:content) ? value.content : value.to_s
+      end
+
+      def extract_published(item)
+        if item.respond_to?(:pubDate) && item.pubDate
+          item.pubDate.to_s
+        elsif item.respond_to?(:published) && item.published
+          item.published.to_s
+        elsif item.respond_to?(:updated) && item.updated
+          item.updated.to_s
+        end
+      end
+
+      def extract_description(item)
+        if item.respond_to?(:description) && item.description
+          item.description
+        elsif item.respond_to?(:content) && item.content
+          item.content
+        elsif item.respond_to?(:summary) && item.summary
+          item.summary
+        end
+      end
+    end
+  end
+end

test/fixtures/workflows/rss_test_workflow/workflow.rb

@@ -0,0 +1,9 @@
+module Workflows
+  class RssTestWorkflow < R3x::Workflow::Base
+    trigger :rss, url: "https://example.com/feed.xml", cron: "every 15 minutes"
+
+    def run(ctx)
+      ctx.trigger.payload
+    end
+  end
+end