archieml 0.2.0 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: e1769d7385a90df46fb88685e1be8cef959943c1
-  data.tar.gz: c92bea5829cc1729197ee1035f7398256b3205c8
+  metadata.gz: db51dfa0cf2524c0161ff7106ef2ab0d6dc38193
+  data.tar.gz: 6caa7995d8f4ed9e281c3c4e48f323b0eacaf69a
 SHA512:
-  metadata.gz: 045a8a57c8676be985fd6194c708f09eb157b93ea97ef032d3a88bfe0e4d6431f94924c25e66e0e9265bef9fd54ce07a47a3a2bdbc87e94bbee8d395abc81d00
-  data.tar.gz: 7bbd0d896a7ae3348814ab36ec6cbe795739ab9543b1971542671d21b629b26b1c15f32b4e7594ff83898c80dec090e3b8bd12e317b92a1023d0748d9742b19d
+  metadata.gz: e83fe41f68351de40718cca29cbc4651be79287961f5dd5a299f23d1d777f01f006af80bc5faf07211f55a987ed75dc17832ff7a6cf5e98f99d6aa1b5befe377
+  data.tar.gz: 87fa405bd6107a62922da75d67a5eb4c7e4029c0064e94ff8d9ae40ec3f260b23725397ba5e8f48826ab23c69ee00f851cd186eea495823525bb15fb615b0847

data/README.md

@@ -4,7 +4,7 @@ Parse Archie Markup Language (ArchieML) documents into Ruby Hashes.
 
 Read about the ArchieML specification at [archieml.org](http://archieml.org).
 
-The current version is `v0.2.0`.
+The current version is `v0.3.0`.
 
 ## Installation
 
@@ -58,7 +58,7 @@ client.authorization = flow.authorize
 
 Log into your Google account and authorize the application to access your Google Drive files.
 
-Now that you have an authenticated `client`, you can make an API call to a document saved in Drive. Create a document with some basic AML inside (such as "key: value"), save it, and note the long string of characters at the end of the URL: 
+Now that you have an authenticated `client`, you can make an API call to a document saved in Drive. Create a document with some basic AML inside (such as "key: value"), save it, and note the long string of characters at the end of the URL:
 
 `https://docs.google.com/a/nytimes.com/document/d/[FILE_ID]/edit`
 
@@ -169,10 +169,13 @@ aml = Archieml.load(html_aml)
 
 ## Tests
 
+Test examples are stored in a submodule. You may need to run `git submodule update --init` to fetch them.
+
 There is a full test suite using rspec. `bundle install`, and then `rspec` to execute them.
 
 ## Changelog
 
+* `0.3.0` - Freeform arrays type and unicode.
 * `0.2.0` - Updated to support an updated ArchieML spec: [2015-05-09](http://archieml.org/spec/1.0/CR-20150509.html). Adds support for nested arrays.
 * `0.1.1` - More consistent handling of newlines. Fixed bugs around detecting the scope of multi-line values.
 * `0.1.0` - Initial release supporting the first version of the ArchieML spec, published [2015-03-06](http://archieml.org/spec/1.0/CR-20150306.html).

data/lib/archieml/loader.rb

@@ -1,11 +1,13 @@
 module Archieml
   class Loader
 
-    NEXT_LINE     = /.*((\r|\n)+)/
-    START_KEY     = /^\s*([A-Za-z0-9\-_\.]+)[ \t\r]*:[ \t\r]*(.*(?:\n|\r|$))/
+    WHITESPACE_PATTERN = "\u0000\u0009\u000A\u000B\u000C\u000D\u0020\u00A0\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200A\u200B\u2028\u2029\u202F\u205F\u3000\uFEFF"
+    SLUG_BLACKLIST     = "#{WHITESPACE_PATTERN}\u005B\u005C\u005D\u007B\u007D\u003A"
+
+    START_KEY     = /^\s*([^#{Regexp.escape(SLUG_BLACKLIST)}]+)[ \t\r]*:[ \t\r]*(.*(?:\n|\r|$))/
     COMMAND_KEY   = /^\s*:[ \t\r]*(endskip|ignore|skip|end)(.*(?:\n|\r|$))/i
     ARRAY_ELEMENT = /^\s*\*[ \t\r]*(.*(?:\n|\r|$))/
-    SCOPE_PATTERN = /^\s*(\[|\{)[ \t\r]*([A-Za-z0-9\-_\.]*)[ \t\r]*(?:\]|\}).*?(\n|\r|$)/
+    SCOPE_PATTERN = /^\s*(\[|\{)[ \t\r]*([\+\.]*)[ \t\r]*([^#{Regexp.escape(SLUG_BLACKLIST)}]*)[ \t\r]*(?:\]|\}).*?(\n|\r|$)/
 
     def initialize(options = {})
       @data = @scope = {}
@@ -40,10 +42,10 @@ module Archieml
           self.parse_array_element(match[1])
 
         elsif !@is_skipping && match = line.match(SCOPE_PATTERN)
-          self.parse_scope(match[1], match[2])
+          self.parse_scope(match[1], match[2], match[3])
 
         else
-          @buffer_string += line
+          self.parse_text(line)
         end
       end
 
@@ -56,11 +58,12 @@ module Archieml
 
       self.increment_array_element(key)
 
+      key = 'value' if (@stack_scope && @stack_scope[:flags].match(/\+/))
+
       @buffer_key = key
       @buffer_string = rest_of_line
 
       self.flush_buffer_into(key, replace: true)
-      @buffer_key = key
     end
 
     def parse_array_element(value)
@@ -68,13 +71,10 @@ module Archieml
 
       @stack_scope[:array_type] ||= :simple
 
-      # Ignore simple array elements inside complex arrays
-      return if @stack_scope[:array_type] == :complex
-
       @stack_scope[:array] << ''
+      @buffer_key = @stack_scope[:array]
       @buffer_string = value
       self.flush_buffer_into(@stack_scope[:array], replace: true)
-      @buffer_key = @stack_scope[:array]
     end
 
     def parse_command_key(command)
@@ -100,47 +100,59 @@ module Archieml
       self.flush_buffer!
     end
 
-    def parse_scope(scope_type, scope_key)
+    def parse_scope(scope_type, flags, scope_key)
       self.flush_buffer!
 
       if scope_key == ''
-        case scope_type
-        when '{'
-          @scope = @data
-          @stack_scope = nil
-          @stack = []
-        when '['
-          # Move up a level
-          if last_stack_item = @stack.pop
-            @scope = last_stack_item[:scope] || @data
-            @stack_scope = @stack.last
-          end
-        end
+        last_stack_item = @stack.pop
+        @scope = (last_stack_item ? last_stack_item[:scope] : @data) || @data
+        @stack_scope = @stack.last
 
       elsif %w([ {).include?(scope_type)
         nesting = false
         key_scope = @data
 
-        if scope_key.match(/^\./)
-          scope_key = scope_key[1..-1]
+        if flags.match(/^\./)
           self.increment_array_element(scope_key)
           nesting = true
           key_scope = @scope if @stack_scope
+        else
+          @scope = @data
+          @stack = []
         end
 
-        key_bits  = scope_key.split('.')
-        key_bits[0...-1].each do |bit|
-          key_scope = key_scope[bit] ||= {}
+        # Within freeforms, the `type` of nested objects and arrays is taken
+        # verbatim from the `keyScope`.
+        if @stack_scope && @stack_scope[:flags].match(/\+/)
+          parsed_scope_key = scope_key
+
+        # Outside of freeforms, dot-notation interpreted as nested data.
+        else
+          key_bits = scope_key.split('.')
+          key_bits[0...-1].each do |bit|
+            key_scope = key_scope[bit] ||= {}
+          end
+          parsed_scope_key = key_bits.last
         end
 
-        if scope_type == '['
-          stack_scope_item = {
-            array: key_scope[key_bits.last] = [],
-            array_type: nil,
-            array_first_key: nil,
-            scope: @scope
-          }
+        # Content of nested scopes within a freeform should be stored under "value."
+        if (@stack_scope && @stack_scope[:flags].match(/\+/) && flags.match(/\./))
+          if scope_type == '['
+            parsed_scope_key = 'value'
+          elsif scope_type == '{'
+            @scope = @scope[:value] = {}
+          end
+        end
 
+        stack_scope_item = {
+          array: nil,
+          array_type: nil,
+          array_first_key: nil,
+          flags: flags,
+          scope: @scope
+        }
+        if scope_type == '['
+          stack_scope_item[:array] = key_scope[parsed_scope_key] = []
           if nesting
             @stack << stack_scope_item
           else
@@ -149,11 +161,25 @@ module Archieml
           @stack_scope = @stack.last
 
         elsif scope_type == '{'
-          @scope = key_scope[key_bits.last] = key_scope[key_bits.last].is_a?(Hash) ? key_scope[key_bits.last] : {}
+          if nesting
+            @stack << stack_scope_item
+          else
+            @scope = key_scope[parsed_scope_key] = key_scope[parsed_scope_key].is_a?(Hash) ? key_scope[parsed_scope_key] : {}
+            @stack = [stack_scope_item]
+          end
+          @stack_scope = @stack.last
         end
       end
     end
 
+    def parse_text(text)
+      if @stack_scope && @stack_scope[:flags].match(/\+/) && text.match(/[^\n\r\s]/)
+        @stack_scope[:array] << { "type" => "text", "value" => text.gsub(/(^\s*)|(\s*$)/, '') }
+      else
+        @buffer_string += text
+      end
+    end
+
     def increment_array_element(key)
       # Special handling for arrays. If this is the start of the array, remember
       # which key was encountered first. If this is a duplicate encounter of
@@ -168,7 +194,12 @@ module Archieml
         if @stack_scope[:array_first_key] == nil || @stack_scope[:array_first_key] == key
           @stack_scope[:array] << (@scope = {})
         end
-        @stack_scope[:array_first_key] ||= key
+        if (@stack_scope[:flags].match(/\+/))
+          @scope[:type] = key
+          # key = 'content'
+        else
+          @stack_scope[:array_first_key] ||= key
+        end
       end
     end
 
@@ -180,11 +211,13 @@ module Archieml
     end
 
     def flush_buffer_into(key, options = {})
+      existing_buffer_key = @buffer_key
       value = self.flush_buffer!
 
       if options[:replace]
         value = self.format_value(value, :replace).sub(/^\s*/, '')
         @buffer_string = value.match(/\s*\Z/)[0]
+        @buffer_key = existing_buffer_key
       else
         value = self.format_value(value, :append)
       end

data/lib/archieml/version.rb

@@ -1,3 +1,3 @@
 module Archieml
-  VERSION = '0.2.0'
+  VERSION = '0.3.0'
 end

data/spec/lib/archieml/loader_spec.rb

@@ -14,6 +14,8 @@ describe Archieml::Loader do
     aml.delete('test')
     aml.delete('result')
 
+    aml = JSON.parse(JSON.dump(aml))
+
     it "#{slug}.#{idx} #{test}" do
       aml.should == result
     end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: archieml
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors:
 - Michael Strickland
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-05-11 00:00:00.000000000 Z
+date: 2016-06-19 00:00:00.000000000 Z
 dependencies: []
 description: Parse Archie Markup Language documents
 email: