filters: HTML info extraction is now a filter

diff --git a/data/rbot/plugins/url.rb b/data/rbot/plugins/url.rb
index b048d3f09e7a944cfaa191f701e7ba83607a6a6e..d08c24e837fea4fe2bb88b6829a7ba5e322fef8a 100644 (file)
--- a/data/rbot/plugins/url.rb
+++ b/data/rbot/plugins/url.rb
@@ -79,7 +79,7 @@ class UrlPlugin < Plugin
 
     begin
       debug "+ getting info for #{url.request_uri}"
-      info = Utils.get_html_info(url)
+      info = @bot.filter(:htmlinfo, url)
       debug info
       resp = info[:headers]
 

diff --git a/lib/rbot/core/filters_ui.rb b/lib/rbot/core/filters_ui.rb
index f4f0719e48921827f886d56aafddebe1513444d9..ed3e65d7cdb0ea8976f66d90f79b75993cddafde 100644 (file)
--- a/lib/rbot/core/filters_ui.rb
+++ b/lib/rbot/core/filters_ui.rb
@@ -12,6 +12,7 @@ class FiltersModule < CoreBotModule
   def initialize
     super
     @bot.clear_filters
+    @bot.register_filter(:htmlinfo) { |s| Utils.get_html_info(s.to_s, s) }
   end
 
   def help(plugin, topic="")

diff --git a/lib/rbot/core/utils/utils.rb b/lib/rbot/core/utils/utils.rb
index ba3a4f2431cb39f5ec735245e170e26af6daba80..25e5159b928edb472dcedb1e2dbef7431bfa028d 100644 (file)
--- a/lib/rbot/core/utils/utils.rb
+++ b/lib/rbot/core/utils/utils.rb
@@ -505,9 +505,9 @@ module ::Irc
       when Net::HTTPResponse
         Utils.get_resp_html_info(doc, opts)
       when URI
-        ret = Hash.new
+        ret = DataStream.new
         @@bot.httputil.get_response(doc) { |resp|
-          ret = Utils.get_resp_html_info(resp, opts)
+          ret.replace Utils.get_resp_html_info(resp, opts)
         }
         return ret
       else
@@ -521,10 +521,13 @@ module ::Irc
     # This method extracts title, content (first par) and extra
     # information from the given Net::HTTPResponse _resp_.
     #
-    # Currently, the only accepted option (in _opts_) is
+    # Currently, the only accepted options (in _opts_) are
     # uri_fragment:: the URI fragment of the original request
+    # full_body::    get the whole body instead of
+    #                @@bot.config['http.info_bytes'] bytes only
     #
-    # Returns a Hash with the following keys:
+    # Returns a DataStream with the following keys:
+    # text:: the (partial) body
     # title:: the title of the document (if any)
     # content:: the first paragraph of the document (if any)
     # headers::
@@ -533,17 +536,21 @@ module ::Irc
     #   header fields, and whose values are Arrays.
     #
     def Utils.get_resp_html_info(resp, opts={})
-      ret = Hash.new
       case resp
       when Net::HTTPSuccess
+        loc = URI.parse(resp['x-rbot-location'] || resp['location']) rescue nil
+        if loc and loc.fragment and not loc.fragment.empty?
+          opts[:uri_fragment] ||= loc.fragment
+        end
+        ret = DataStream.new(opts.dup)
         ret[:headers] = resp.to_hash
+        ret[:text] = partial = opts[:full_body] ? resp.body : resp.partial_body(@@bot.config['http.info_bytes'])
 
-        partial = resp.partial_body(@@bot.config['http.info_bytes'])
-        if resp['content-type'] =~ /^text\/|(?:x|ht)ml/
-          loc = URI.parse(resp['x-rbot-location'] || resp['location']) rescue nil
-          if loc and loc.fragment and not loc.fragment.empty?
-            opts[:uri_fragment] ||= loc.fragment
-          end
+        filtered = Utils.try_htmlinfo_filters(ret)
+
+        if filtered
+          return filtered
+        elsif resp['content-type'] =~ /^text\/|(?:x|ht)ml/
           ret.merge!(Utils.get_string_html_info(partial, opts))
         end
         return ret
@@ -552,6 +559,28 @@ module ::Irc
       end
     end
 
+    # This method runs an appropriately-crafted DataStream _ds_ through the
+    # filters in the :htmlinfo filter group, in order. If one of the filters
+    # returns non-nil, its results are merged in _ds_ and returned. Otherwise
+    # nil is returned.
+    #
+    # The input DataStream shuold have the downloaded HTML as primary key
+    # (:text) and possibly a :headers key holding the resonse headers.
+    #
+    def Utils.try_htmlinfo_filters(ds)
+      filters = @@bot.filter_names(:htmlinfo)
+      return nil if filters.empty?
+      cur = nil
+      # TODO filter priority
+      filters.each { |n|
+        debug "testing filter #{n}"
+        cur = @@bot.filter(@@bot.global_filter_name(n, :htmlinfo), ds)
+        debug "returned #{cur.pretty_inspect}"
+        break if cur
+      }
+      return ds.merge(cur) if cur
+    end
+
     # This method extracts title and content (first par)
     # from the given HTML or XML document _text_, using
     # standard methods (String#ircify_html_title,