X-Git-Url: https://git.netwichtig.de/gitweb/?a=blobdiff_plain;f=lib%2Frbot%2Fcore%2Futils%2Futils.rb;h=7b316ffe28cd3df6ffc6ecd0d1707063142dd1bb;hb=bf9734ff89a238c5a63015b68eabd8d0ef9d1308;hp=c47601c1baf89386d0801fb5d5458670866e7e24;hpb=00f2fa9d625b19d6c8b33c62ea41460924e43634;p=user%2Fhenk%2Fcode%2Fruby%2Frbot.git

diff --git a/lib/rbot/core/utils/utils.rb b/lib/rbot/core/utils/utils.rb
index c47601c1..7b316ffe 100644
--- a/lib/rbot/core/utils/utils.rb
+++ b/lib/rbot/core/utils/utils.rb
@@ -23,6 +23,7 @@ rescue LoadError
     'raquo' => '»',
     'quot' => '"',
     'apos' => '\'',
+    'deg' => '°',
     'micro' => 'µ',
     'copy' => '©',
     'trade' => '™',
@@ -32,6 +33,7 @@ rescue LoadError
     'gt' => '>',
     'hellip' => '…',
     'nbsp' => ' ',
+    'ndash' => '–',
     'Agrave' => 'À',
     'Aacute' => 'Á',
     'Acirc' => 'Â',
@@ -125,7 +127,7 @@ rescue LoadError
 
         # Some blogging and forum platforms use spans or divs with a 'body' or 'message' or 'text' in their class
         # to mark actual text
-        AFTER_PAR1_REGEX = /<\w+\s+[^>]*(?:body|message|text)[^>]*>.*?<\/?(?:p|div|html|body|table|td|tr)(?:\s+[^>]*)?>/im
+        AFTER_PAR1_REGEX = /<\w+\s+[^>]*(?:body|message|text|post)[^>]*>.*?<\/?(?:p|div|html|body|table|td|tr)(?:\s+[^>]*)?>/im
 
         # At worst, we can try stuff which is comprised between two <br>
         AFTER_PAR2_REGEX = /<br(?:\s+[^>]*)?\/?>.*?<\/?(?:br|p|div|html|body|table|td|tr)(?:\s+[^>]*)?\/?>/im
@@ -275,24 +277,47 @@ module ::Irc
     # Execute an external program, returning a String obtained by redirecting
     # the program's standards errors and output
     #
+    # TODO: find a way to expose some common errors (e.g. Errno::NOENT)
+    # to the caller
     def Utils.safe_exec(command, *args)
-      IO.popen("-") { |p|
+      output = IO.popen("-") { |p|
         if p
-          return p.readlines.join("\n")
+          break p.readlines.join("\n")
         else
           begin
             $stderr.reopen($stdout)
             exec(command, *args)
           rescue Exception => e
-            puts "exec of #{command} led to exception: #{e.pretty_inspect}"
-            Kernel::exit! 0
+            puts "exception #{e.pretty_inspect} trying to run #{command}"
+            Kernel::exit! 1
           end
           puts "exec of #{command} failed"
-          Kernel::exit! 0
+          Kernel::exit! 1
         end
       }
+      raise "safe execution of #{command} returned #{$?}" unless $?.success?
+      return output
     end
 
+    # Try executing an external program, returning true if the run was successful
+    # and false otherwise
+    def Utils.try_exec(command, *args)
+      IO.popen("-") { |p|
+        if p.nil?
+          begin
+            $stderr.reopen($stdout)
+            exec(command, *args)
+          rescue Exception => e
+            Kernel::exit! 1
+          end
+          Kernel::exit! 1
+        else
+          debug p.readlines
+        end
+      }
+      debug $?
+      return $?.success?
+    end
 
     # Safely (atomically) save to _file_, by passing a tempfile to the block
     # and then moving the tempfile to its final location when done.
@@ -313,14 +338,27 @@ module ::Irc
     # Decode HTML entities in the String _str_, using HTMLEntities if the
     # package was found, or UNESCAPE_TABLE otherwise.
     #
-    def Utils.decode_html_entities(str)
-      if defined? ::HTMLEntities
-        return HTMLEntities.decode_entities(str)
+
+    if defined? ::HTMLEntities
+      if ::HTMLEntities.respond_to? :decode_entities
+        def Utils.decode_html_entities(str)
+          return HTMLEntities.decode_entities(str)
+        end
       else
-        str.gsub(/(&(.+?);)/) {
+        @@html_entities = HTMLEntities.new
+        def Utils.decode_html_entities(str)
+          return @@html_entities.decode str
+        end
+      end
+    else
+      def Utils.decode_html_entities(str)
+        return str.gsub(/(&(.+?);)/) {
           symbol = $2
           # remove the 0-paddng from unicode integers
-          if symbol =~ /^#(\d+)$/
+          case symbol
+          when /^#x([0-9a-fA-F]+)$/
+            symbol = $1.to_i(16).to_s
+          when /^#(\d+)$/
             symbol = $1.to_i.to_s
           end
 
@@ -458,7 +496,11 @@ module ::Irc
 
     # HTML first par grabber without hpricot
     def Utils.ircify_first_html_par_woh(xml_org, opts={})
-      xml = xml_org.gsub(/<!--.*?-->/m, '').gsub(/<script(?:\s+[^>]*)?>.*?<\/script>/im, "").gsub(/<style(?:\s+[^>]*)?>.*?<\/style>/im, "")
+      xml = xml_org.gsub(/<!--.*?-->/m,
+                         "").gsub(/<script(?:\s+[^>]*)?>.*?<\/script>/im,
+                         "").gsub(/<style(?:\s+[^>]*)?>.*?<\/style>/im,
+                         "").gsub(/<select(?:\s+[^>]*)?>.*?<\/select>/im,
+                         "")
 
       strip = opts[:strip]
       strip = Regexp.new(/^#{Regexp.escape(strip)}/) if strip.kind_of?(String)