Lex processing instructions in chunks

Similar to comments (ea8b4aa92f) and CDATA
tags (8acc7fc743) processing instructions
are now lexed in separate chunks _with_ proper support for streaming
input.

Related issue: #93

lib/oga/xml/lexer.rb

@@ -344,6 +344,15 @@ module Oga
         add_token(:T_PROC_INS_NAME, value)
       end
 
+      ##
+      # Called on the body of a processing instruction.
+      #
+      # @param [String] value
+      #
+      def on_proc_ins_body(value)
+        add_token(:T_PROC_INS_BODY, value)
+      end
+
       ##
       # Called on the end of a processing instruction.
       #

lib/oga/xml/parser.rll

@@ -29,7 +29,7 @@
 %terminals T_CDATA_START T_CDATA_BODY T_CDATA_END;
 %terminals T_ELEM_START T_ELEM_NAME T_ELEM_NS T_ELEM_END T_ATTR T_ATTR_NS;
 %terminals T_XML_DECL_START T_XML_DECL_END;
-%terminals T_PROC_INS_START T_PROC_INS_NAME T_PROC_INS_END;
+%terminals T_PROC_INS_START T_PROC_INS_NAME T_PROC_INS_BODY T_PROC_INS_END;
 
 document
   = expressions { on_document(val[0]) }
@@ -117,12 +117,17 @@ comment_body
 # Processing Instructions
 
 proc_ins
-  = T_PROC_INS_START T_PROC_INS_NAME T_TEXT? T_PROC_INS_END
+  = T_PROC_INS_START T_PROC_INS_NAME proc_ins_body T_PROC_INS_END
     {
       on_proc_ins(val[1], val[2])
     }
   ;
 
+proc_ins_body
+  = T_PROC_INS_BODY proc_ins_body { val[0] + val[1] }
+  | _                             { '' }
+  ;
+
 # Elements
 
 element_name_ns