slaw 0.1.3 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 680301c5ade280eb7da5ea92c509f491631824f2
-  data.tar.gz: f2ddd5a99631121bf3693da5f229e38a6f590142
+  metadata.gz: 405a0b941536c74c13588e1bfb4350c566337626
+  data.tar.gz: 809e1fd9fd4ada655d3531b4eb31702398490a42
 SHA512:
-  metadata.gz: 844130f24fa5e4e7e2acd8bacc9381bbd043591676a4fd22e9f1deec87e99b813f3062e4c4ec7286aca4ec0fe2a17161c39d85f5a07c8819192c82cd6203e474
-  data.tar.gz: de11ab3cb747c7341209e79f131506f6e2fc44065a73d95bb936c2b36b348646644024b0b657252106cd4c6d9f1b792ca4f7884e8f45fff4bda453da0a736cb7
+  metadata.gz: a636be697e3589db697232bc01876a864ee8c02eb4548232b9db8addc2c3d9fb0a5004ffeb94f12494e88889b2954c3edd61100a865ce9329b5eddad7381fbe8
+  data.tar.gz: 57e78d5489aa950436b2e7dc3ebe7d19a639b86c3f4fa3b95add5edfa9adbb0185b6f191a012f74ba4fc35100432096ac1ea09b928327afa214aaedd1a2c070c

data/lib/slaw/act.rb

@@ -63,7 +63,14 @@ module Slaw
     # Parse the XML contained in the file-like object `io`
     # @param io [file-like] io object with XML
     def parse(io)
-      @doc = Nokogiri::XML(io)
+      self.doc = Nokogiri::XML(io)
+    end
+
+    # Set the XML document backing this bylaw.
+    #
+    # @param doc [Nokogiri::XML::Document] document
+    def doc=(doc)
+      @doc = doc
       @meta = @doc.at_xpath('/a:akomaNtoso/a:act/a:meta', a: NS)
       @body = @doc.at_xpath('/a:akomaNtoso/a:act/a:body', a: NS)
 

data/lib/slaw/parse/builder.rb

@@ -12,61 +12,77 @@ module Slaw
     # XML document.
     #
     # @example Parse some text into a well-formed document
-    #     builder = Slaw::Builder.new
+    #     builder = Slaw::Builder.new(parser: parser)
     #     xml = builder.parse_text(text)
     #     doc = builder.parse_xml(xml)
     #     builder.postprocess(doc)
     #
     # @example A quicker way to build a well-formed document
-    #     builder = Slaw::Builder.new
     #     doc = builder.parse_and_process_text(text)
     #
     class Builder
       include Slaw::Namespace
       include Slaw::Logging
 
-      Treetop.load(File.dirname(__FILE__) + "/bylaw.treetop")
+      @@parsers = {}
 
-      # [Hash] A Hash of options that are made available to the parser when parsing.
-      attr_accessor :parse_options
-
-      def initialize(parse_options={})
-        @parse_options = parse_options
+      # Create a new builder.
+      #
+      # Specify either `:parser` or `:grammar_file` and `:grammar_class`.
+      #
+      # @option opts [Treetop::Runtime::CompiledParser] :parser parser to use
+      # @option opts [String] :grammar_file grammar filename to load a parser from
+      # @option opts [String] :grammar_class name of the class that the grammar will generate
+      def initialize(opts={})
+        if opts[:parser]
+          @parser = opts[:parser]
+        elsif opts[:grammar_file] and opts[:grammar_class]
+          if @@parsers[opts[:grammar_class]]
+            # already compiled the grammar, just use it
+            @parser = @@parsers[opts[:grammar_class]]
+          else
+            # load the grammar
+            Treetop.load(opts[:grammar_file])
+            cls = eval(opts[:grammar_class])
+            @parser = cls.new
+          end
+        else
+          raise ArgumentError.new("Specify either :parser or :grammar_file and :grammar_class")
+        end
       end
 
       # Do all the work necessary to parse text into a well-formed XML document.
       #
       # @param text [String] the text to parse
-      # @param root [Symbol] the root element of the grammar
+      # @param parse_options [Hash] options to parse to the parser
       #
       # @return [Nokogiri::XML::Document] a well formed document
-      def parse_and_process_text(text, root=:bylaw)
-        postprocess(parse_xml(parse_text(text, root)))
+      def parse_and_process_text(text, parse_options={})
+        postprocess(parse_xml(parse_text(text, parse_options)))
       end
 
       # Parse text into XML. You should still run {#postprocess} on the
       # resulting XML to normalise it.
       #
       # @param text [String] the text to parse
-      # @param root [Symbol] the root element of the grammar
+      # @param parse_options [Hash] options to parse to the parser
       #
       # @return [String] an XML string
-      def parse_text(text, root=:bylaw)
-        tree = text_to_syntax_tree(text, root)
+      def parse_text(text, parse_options={})
+        tree = text_to_syntax_tree(text, parse_options)
         xml_from_syntax_tree(tree)
       end
 
       # Parse plain text into a syntax tree.
       #
       # @param text [String] the text to parse
-      # @param root [Symbol] the root element of the grammar
+      # @param parse_options [Hash] options to parse to the parser
       #
-      # @return [Object] the root of the resulting parse tree, usually a Treetop::Node object
-      def text_to_syntax_tree(text, root=:bylaw)
-        parser = Slaw::Parse::BylawParser.new
-        parser.options = @parse_options
-
-        tree = parser.parse(text, {root: root})
+      # @return [Object] the root of the resulting parse tree, usually a Treetop::Runtime::SyntaxNode object
+      def text_to_syntax_tree(text, parse_options={})
+        logger.info("Parsing...")
+        tree = @parser.parse(text, parse_options)
+        logger.info("Parsed!")
 
         if tree.nil?
           raise Slaw::Parse::ParseError.new(parser.failure_reason || "Couldn't match to grammar",
@@ -80,7 +96,7 @@ module Slaw
       # Generate an XML document from the given syntax tree. You should still
       # run {#postprocess} on the resulting XML to normalise it.
       #
-      # @param tree [Object] a Treetop::Node object
+      # @param tree [Object] a Treetop::Runtime::SyntaxNode object
       #
       # @return [String] an XML string
       def xml_from_syntax_tree(tree)

data/lib/slaw/version.rb

@@ -1,3 +1,3 @@
 module Slaw
-  VERSION = "0.1.3"
+  VERSION = "0.2.0"
 end

data/lib/slaw/{parse → za}/bylaw.treetop

@@ -3,7 +3,7 @@
 require 'slaw/parse/grammar_helpers'
 
 module Slaw
-  module Parse
+  module ZA
     grammar Bylaw
       include Slaw::Parse::GrammarHelpers
 

data/lib/slaw/za/bylaw_generator.rb

@@ -0,0 +1,41 @@
+require 'slaw/za/bylaw_nodes'
+
+module Slaw
+  # Support specifically for South Africa
+  module ZA
+
+    # Support class for generating South African bylaws
+    class BylawGenerator
+      Treetop.load(File.dirname(__FILE__) + "/bylaw.treetop")
+
+      # [Treetop::Runtime::CompiledParser] compiled bylaw parser
+      attr_accessor :parser
+
+      # [Slaw::Parse::Builder] builder used by the generator
+      attr_accessor :builder
+
+      def initialize
+        @parser = Slaw::ZA::BylawParser.new
+        @builder = Slaw::Parse::Builder.new(parser: @parser)
+        @cleanser = Slaw::Parse::Cleanser.new
+      end
+
+      # Generate a Slaw::Bylaw instance from plain text.
+      #
+      # @param text [String] plain text
+      #
+      # @return [Slaw::ByLaw] the resulting bylaw
+      def generate_from_text(text)
+        bylaw = Slaw::ByLaw.new
+        bylaw.doc = @builder.parse_and_process_text(cleanup(text))
+        bylaw
+      end
+
+      def cleanup(text)
+        text = @cleanser.cleanup(text)
+        text = @cleanser.reformat(text)
+        text
+      end
+    end
+  end
+end

data/lib/slaw/{parse/nodes.rb → za/bylaw_nodes.rb}

@@ -1,5 +1,5 @@
 module Slaw
-  module Parse
+  module ZA
     module Bylaw
       class Bylaw < Treetop::Runtime::SyntaxNode
         def to_xml(b)

data/lib/slaw.rb

@@ -16,8 +16,6 @@ require 'slaw/parse/blocklists'
 require 'slaw/parse/builder'
 require 'slaw/parse/cleanser'
 require 'slaw/parse/error'
-require 'slaw/parse/grammar_helpers'
-require 'slaw/parse/nodes'
 
 require 'slaw/extract/extractor'
 

data/spec/parse/builder_spec.rb

@@ -4,6 +4,9 @@ require 'spec_helper'
 require 'slaw'
 
 describe Slaw::Parse::Builder do
+  let(:parser) { double("parser") }
+  subject { Slaw::Parse::Builder.new(parser: parser) }
+
   describe '#nest_blocklists' do
     it 'should nest simple blocks' do
       doc = xml2doc(subsection(<<XML

data/spec/{parse → za}/bylaw_spec.rb

@@ -1,16 +1,18 @@
 # encoding: UTF-8
 
-require 'slaw'
 require 'builder'
 
-describe Slaw::Parse::Builder do
+require 'slaw'
+require 'slaw/za/bylaw_generator'
+
+describe Slaw::ZA::BylawGenerator do
   def parse(rule, s)
-    subject.text_to_syntax_tree(s, rule)
+    subject.builder.text_to_syntax_tree(s, {root: rule})
   end
 
   def should_parse(rule, s)
     s << "\n" unless s.end_with?("\n")
-    tree = subject.text_to_syntax_tree(s, rule)
+    tree = subject.builder.text_to_syntax_tree(s, {root: rule})
 
     if not tree
       raise Exception.new(subject.failure_reason || "Couldn't match to grammar") if tree.nil?
@@ -61,7 +63,7 @@ EOS
     end
 
     it 'should handle parts and odd section numbers' do
-      subject.parse_options = {section_number_after_title: false}
+      subject.parser.options = {section_number_after_title: false}
       node = parse :bylaw, <<EOS
 PART 1
 PREVENTION AND SUPPRESSION OF HEALTH NUISANCES
@@ -232,7 +234,7 @@ EOS
 
   context 'sections' do
     it 'should handle section numbers after title' do
-      subject.parse_options = {section_number_after_title: true}
+      subject.parser.options = {section_number_after_title: true}
       node = parse :bylaw, <<EOS
 Section
 1. (1) hello
@@ -244,7 +246,7 @@ EOS
     end
 
     it 'should handle section numbers before title' do
-      subject.parse_options = {section_number_after_title: false}
+      subject.parser.options = {section_number_after_title: false}
       node = parse :bylaw, <<EOS
 1. Section
 (1) hello
@@ -256,7 +258,7 @@ EOS
     end
 
     it 'should handle section numbers without a dot' do
-      subject.parse_options = {section_number_after_title: false}
+      subject.parser.options = {section_number_after_title: false}
       node = parse :bylaw, <<EOS
 1 A section
 (1) hello
@@ -274,7 +276,7 @@ EOS
     end
 
     it 'should handle sections without titles' do
-      subject.parse_options = {section_number_after_title: false}
+      subject.parser.options = {section_number_after_title: false}
       node = parse :bylaw, <<EOS
 1. No owner or occupier of any shop or business premises or vacant land, blah blah
 2. Notwithstanding the provision of any other By-law or legislation no person shall—
@@ -291,7 +293,7 @@ EOS
     end
 
     it 'should handle sections without titles and with subsections' do
-      subject.parse_options = {section_number_after_title: false}
+      subject.parser.options = {section_number_after_title: false}
       node = parse :bylaw, <<EOS
 10. (1) Transporters must remove medical waste.
 (2) Without limiting generality, stuff.
@@ -305,7 +307,7 @@ EOS
     end
 
     it 'should realise complex section titles are actually section content' do
-      subject.parse_options = {section_number_after_title: false}
+      subject.parser.options = {section_number_after_title: false}
       node = parse :bylaw, <<EOS
 10. The owner of any premises which is let or sublet to more than one tenant, shall maintain at all times in a clean and sanitary condition every part of such premises as may be used in common by more than one tenant.
 11. No person shall keep, cause or suffer to be kept any factory or trade premises so as to cause or give rise to smells or effluvia that constitute a health nuisance.

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: slaw
 version: !ruby/object:Gem::Version
-  version: 0.1.3
+  version: 0.2.0
 platform: ruby
 authors:
 - Greg Kempe
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-18 00:00:00.000000000 Z
+date: 2014-09-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -132,24 +132,25 @@ files:
 - lib/slaw/namespace.rb
 - lib/slaw/parse/blocklists.rb
 - lib/slaw/parse/builder.rb
-- lib/slaw/parse/bylaw.treetop
 - lib/slaw/parse/cleanser.rb
 - lib/slaw/parse/error.rb
 - lib/slaw/parse/grammar_helpers.rb
-- lib/slaw/parse/nodes.rb
 - lib/slaw/render/html.rb
 - lib/slaw/render/xsl/act.xsl
 - lib/slaw/render/xsl/elements.xsl
 - lib/slaw/render/xsl/fragment.xsl
 - lib/slaw/version.rb
 - lib/slaw/xml_support.rb
+- lib/slaw/za/bylaw.treetop
+- lib/slaw/za/bylaw_generator.rb
+- lib/slaw/za/bylaw_nodes.rb
 - slaw.gemspec
 - spec/extract/extractor_spec.rb
 - spec/parse/builder_spec.rb
-- spec/parse/bylaw_spec.rb
 - spec/parse/cleanser_spec.rb
 - spec/spec_helper.rb
 - spec/xml_helpers.rb
+- spec/za/bylaw_spec.rb
 homepage: ''
 licenses:
 - MIT
@@ -177,7 +178,7 @@ summary: A lightweight library for using Akoma Ntoso acts in Ruby.
 test_files:
 - spec/extract/extractor_spec.rb
 - spec/parse/builder_spec.rb
-- spec/parse/bylaw_spec.rb
 - spec/parse/cleanser_spec.rb
 - spec/spec_helper.rb
 - spec/xml_helpers.rb
+- spec/za/bylaw_spec.rb