sunscraper.rb in sunscraper-1.2.0.beta1

- old
+ new

@@ -1,7 +1,7 @@
-if !defined?(RUBY_ENGINE) && RUBY_VERSION =~ /^1.8/
-  raise RuntimeError, "Sunscraper does not work on Ruby MRI 1.8.x."
+if RUBY_VERSION =~ /^1.8/
+  raise RuntimeError, "Sunscraper does not work on Ruby 1.8."
 end
 
 # Sunscraper loads an HTML page in a headless browser and waits for `Sunscraper.finish()`
 # method to be called. It blocks the calling thread, but is threadsafe, does
 # not acquire GIL and thus can be called from multiple threads simultaneously.
@@ -25,13 +25,13 @@
 
     # Scrape an inline HTML. The content is loaded without a particular base URL.
     # If your application depends on base URL being available, use {scrape_url}.
     #
     # @param [Integer] timeout timeout in milliseconds
-    def scrape_html(html, timeout=5000)
+    def scrape_html(html, url="about:blank", timeout=5000)
       scrape(timeout) do |worker, context|
-        worker.load_html context, html
+        worker.load_html context, html, url
       end
     end
 
     # Scrape an URL.
     #
@@ -45,22 +45,22 @@
     private
 
     def scrape(timeout)
       worker = load_worker
 
-      context = worker.create
-      yield worker, context
-      worker.wait(context, timeout)
+      begin
+        context = worker.create
 
-      data = worker.fetch(context)
+        yield worker, context
 
-      if data == "!SUNSCRAPER_TIMEOUT"
-        raise ScrapeTimeout, "Sunscraper has timed out waiting for the callback"
-      else
-        data
+        if worker.wait(context, timeout)
+          worker.fetch(context)
+        else
+          raise ScrapeTimeout, "Sunscraper has timed out waiting for the callback"
+        end
+      ensure
+        worker.finalize(context) if context
       end
-    ensure
-      worker.discard(context) if context
     end
 
     def load_worker
       case @worker
       when :standalone
\ No newline at end of file