swissparser 0.8.1 → 0.9.0

data/CHANGELOG.rdoc

@@ -1,3 +1,8 @@
+== 0.9.0 / 2009-11-14
+
+* 1 new feature:
+  - SwissParser can now directly parse remote files using HTTP or FTP.
+
 == 0.8.1 / 2009-11-14
 
 * 1 bugfix:
@@ -6,8 +11,7 @@
 == 0.8.0 / 2009-11-14
 
 * 1 new feature:
-  - helper methods are now defined with the helper method. When the
-  parser is extended they can be overriden in a one per one basis.
+  - helper methods are now defined with the helper method. When the parser is extended they can be overriden in a one per one basis.
 
 == 0.7.0 / 2009-11-14
 
@@ -18,7 +22,7 @@
 
 == 0.6.0 / 2009-11-13
 
-* 2 new features
+* 2 new features:
   - Parsing parameters are now accessed thru the +params+ method
     and are accessible to parsing rules too.
   - Helpers methods are now defined in an helper block and are

data/README.rdoc

@@ -13,8 +13,9 @@ be stable enough to be used for bioinformatics research.
 
 * Defines parsers with a clear and compact declarative syntax.
 * The whole parsing workflow is configurable.
-* The user can create new parsers by extending existing parsers.
-* Parser have access to global parameters and user defined helper methods.
+* Able to parse remote files accessible from a web or an FTP server.
+* Users can create new parsers by extending existing parsers.
+* A parser have access to global parameters and user defined helper methods.
 
 == USAGE:
 

data/examples/parse_from_uri.rb

@@ -0,0 +1,88 @@
+=begin
+Copyright (C) 2009 Paradigmatic
+
+This file is part of SwissParser.
+
+SwissParser is free software: you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation, either version 3 of the License, or
+(at your option) any later version.
+
+SwissParser is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with SwissParser.  If not, see <http://www.gnu.org/licenses/>.
+=end
+
+#!/usr/bin/ruby -w
+
+require 'yaml'
+require 'swissparser.rb'
+
+class Protein
+
+  attr_accessor :id, :size, :species, :taxonomy, :sequence
+
+  def initialize
+    @taxonomy = []
+    @sequence = ""
+  end
+
+end
+
+module Uniprot
+
+  Parser = Swiss::Parser.define do
+
+    # Each entry must be stored in a Protein instance
+    new_entry do
+      Protein.new
+    end
+    
+    rules do 
+      
+      # Parse the uniprot id
+      with("ID") do |content,protein|
+        content =~ /([A-Z]\w+)\D+(\d+)/
+        protein.id = $1
+        protein.size = $2.to_i
+      end
+      
+      # Parse the organism
+      with("OS") do |content,protein|
+        content =~ /(\w+ \w+)/
+        protein.species = $1
+      end
+      
+      # Parse the complete taxonomy
+      with("OC") do |content,protein|
+        ary = content.gsub(".","").split("; ")
+        protein.taxonomy += ary
+      end
+      
+      # Parse the Sequence
+      with_text_after("SQ") do |content,protein|
+        seq = content.strip.gsub(" ","")
+        protein.sequence += seq
+      end
+      
+    end
+    
+  end
+  
+end
+  
+if $0 == __FILE__
+    
+  uri = ARGV.shift
+
+  entries = Uniprot::Parser.parse_URI( uri )
+
+  entries.each do |e|
+    puts e.to_yaml
+  end
+
+end

data/lib/swissparser.rb

@@ -17,9 +17,11 @@ You should have received a copy of the GNU General Public License
 along with SwissParser.  If not, see <http://www.gnu.org/licenses/>.
 =end
 
+require 'open-uri'
+
 module Swiss
 
-  VERSION = "0.8.1"
+  VERSION = "0.9.0"
 
   # This class defines parsing rules. Its methods
   # are accessible within the +rules+ section of
@@ -216,6 +218,26 @@ module Swiss
     # It returns the value specified in the +after+ block. By default, 
     # it returns an array containing _entry_ objects.
     def parse_file( filename, params={} )
+      File.open( filename, 'r' ) do |file|
+        parse( file, params )
+      end
+    end
+
+    # Parses a file specified by an +URI+. Both http and ftp are
+    # supported. An optional hash of arbitrary arguments (+params+)
+    # can be specified. It is passed to the workflow methods blocks
+    # (+before+, +new_entry+, ...)  It returns the value specified in
+    # the +after+ block. By default, it returns an array containing
+    # _entry_ objects.
+    def parse_URI( uri, params={} )
+      open( uri ) do |file|
+        parse( file, params )
+      end
+    end
+
+    private
+    
+    def parse( file, params )
       @ctx = ParsingContext.new( params )
       helperModule = Module.new
       @helpers.each do |name, proc|
@@ -223,21 +245,17 @@ module Swiss
       end
       @ctx.extend( helperModule )
       container = @ctx.instance_exec( &@before )
-      File.open( filename, 'r' ) do |file|
-        entry = @ctx.instance_exec( &@begin )
-        file.each_line do |line|
-          state = parse_line( line, entry )
-          if state == :end
-            @ctx.instance_exec( entry, container, &@end )
-            entry = @ctx.instance_exec( &@begin )
-          end
+      entry = @ctx.instance_exec( &@begin )
+      file.each_line do |line|
+        state = parse_line( line, entry )
+        if state == :end
+          @ctx.instance_exec( entry, container, &@end )
+          entry = @ctx.instance_exec( &@begin )
         end
       end
       @ctx.instance_exec( container, &@after )
     end
 
-    private
-
     PROTOTYPE = Parser.new
     PROTOTYPE.instance_eval do
       before { || [] }

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: swissparser
 version: !ruby/object:Gem::Version 
-  version: 0.8.1
+  version: 0.9.0
 platform: ruby
 authors: 
 - paradigmatic
@@ -41,6 +41,7 @@ files:
 - examples/data/kegg_enzyme_short.txt
 - examples/data/uniprot.txt
 - examples/kegg_demo.rb
+- examples/parse_from_uri.rb
 - examples/signal_demo.rb
 - examples/tutorial_1.rb
 - examples/uniprot_param_demo.rb