infoboxer 0.2.8 → 0.3.0.pre

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: ef407b6160d6d0ff8cfb874338320532d1b18906
-  data.tar.gz: fdc4ccf4b051c50958e6cc9720619ea5110b5ea5
+  metadata.gz: 67b07bcee67c15aeb316e2eccf75e85a26cb43ae
+  data.tar.gz: 91f3253151ff816a83b4ffb8f7937c878a5a3ed8
 SHA512:
-  metadata.gz: d5224104b02115c47f27c69926128023da3765d1d7089241190a6b3f5c8c3de79be91f65276c82536c671424d10000e2f5198413b3aa21e9802ccc9ca41fbdef
-  data.tar.gz: 0f1f8d9ee3a4e94d6529208a9678bcb0915a884525433d2397249afada141d96c88c74e0546ddd0576adffcc2785f96ad1da57d61284b84d96661fb49c0fd46d
+  metadata.gz: f7dd9aa6545853dbf827c1b0b7857b2b5654caf33dcbfcaf4c2194709379f7e73d0978776675fd6265cbf83d88a814332626945db1b45c9562e54a01fa486a55
+  data.tar.gz: fa70f671da0d9acd7a0a131d8119a775c1ff762d7d2b640096f448678b7547477fd0fe93df74c438cd846fa550e33e530999c811bb4b0d4afe347bf058c3321b

data/.yardopts

@@ -1,2 +1,3 @@
 --markup=markdown
+--markup-provider=redcarpet
 --no-private

data/CHANGELOG.md

@@ -1,5 +1,23 @@
 # Infoboxer's change log
 
+## 0.3.1.pre (2017-09-16)
+
+* Introduce interwiki links following (and proper handling of interwikis, in general);
+* Add `<gallery>` tag support;
+* Introduce `Navigation::Selector#===`;
+* Much more `Enumerable`'s methods supported by `Nodes`;
+* Lot of small simplifications, cleanups and bugfixes.
+
+TBH, it should be 0.4.0 or more, but it would be a shame to change versions so fast :) So, at least
+until it is `-pre`, let it be 0.3.1.
+
+## 0.3.0 (2017-07-23)
+
+* Change logic of navigation through templates; now templates contents aren't hidden from global
+  lookups. While sometimes leading to less impressive demos, this approach proved itself to be more
+  useful for production.
+* Introduce WikiPath query language as an alternative to series of lookups.
+
 ## 0.2.8 (2017-05-11)
 
 * Switch to MediaWiktory 0.1.0 + some subsequent cleanup of internal logic;

data/infoboxer.gemspec

@@ -32,7 +32,6 @@ Gem::Specification.new do |s|
   s.executables << 'infoboxer'
 
   s.add_dependency 'htmlentities'
-  s.add_dependency 'procme'
   s.add_dependency 'mediawiktory', '>= 0.1.0'
   s.add_dependency 'addressable'
   s.add_dependency 'terminal-table'

data/lib/infoboxer.rb

@@ -1,7 +1,5 @@
 # encoding: utf-8
 
-require 'procme'
-
 # Main client module for entire infoboxer functionality. If you're lucky,
 # there's no other classes/modules you need to instantiate or call
 # directly. You just do:
@@ -74,8 +72,8 @@ module Infoboxer
   end
 
   # Includeable version of {Infoboxer.wiki}
-  def wiki(api_url, options = {})
-    wikis[api_url] ||= MediaWiki.new(api_url, options || {})
+  def wiki(api_url, **options)
+    wikis[api_url] ||= MediaWiki.new(api_url, options)
   end
 
   class << self
@@ -170,7 +168,7 @@ module Infoboxer
   end
 
   WIKIMEDIA_PROJECTS.each do |name, domain|
-    define_method name do |lang = 'en', options = {}|
+    define_method name do |lang = 'en', **options|
       lang, options = 'en', lang if lang.is_a?(Hash)
 
       wiki("https://#{lang}.#{domain}/w/api.php", options)
@@ -180,7 +178,7 @@ module Infoboxer
   alias_method :wp, :wikipedia
 
   WIKIMEDIA_COMMONS.each do |name, domain|
-    define_method name do |options = {}|
+    define_method name do |**options|
       wiki("https://#{domain}/w/api.php", options)
     end
   end

data/lib/infoboxer/media_wiki.rb

@@ -47,15 +47,14 @@ module Infoboxer
     # for it, as well as shortcuts for some well-known wikis, like
     # {Infoboxer.wikipedia}.
     #
-    # @param api_base_url URL of `api.php` file in your MediaWiki
+    # @param api_base_url [String] URL of `api.php` file in your MediaWiki
     #   installation. Typically, its `<domain>/w/api.php`, but can vary
     #   in different wikis.
-    # @param options Only one option is currently supported:
-    #   * `:user_agent` (also aliased as `:ua`) -- custom User-Agent header.
-    def initialize(api_base_url, options = {})
+    # @param user_agent [String] (also aliased as `:ua`) Custom User-Agent header.
+    def initialize(api_base_url, ua: nil, user_agent: ua)
       @api_base_url = Addressable::URI.parse(api_base_url)
-      @client = MediaWiktory::Wikipedia::Api.new(api_base_url, user_agent: user_agent(options))
-      @traits = Traits.get(@api_base_url.host, namespaces: extract_namespaces)
+      @client = MediaWiktory::Wikipedia::Api.new(api_base_url, user_agent: user_agent(user_agent))
+      @traits = Traits.get(@api_base_url.host, siteinfo)
     end
 
     # Receive "raw" data from Wikipedia (without parsing or wrapping in
@@ -123,7 +122,9 @@ module Infoboxer
     #     and obtain meaningful results instead of `NoMethodError` or
     #     `SomethingNotFound`.
     #
-    def get(*titles, prop: [])
+    def get(*titles, prop: [], interwiki: nil)
+      return interwikis(interwiki).get(*titles, prop: prop) if interwiki
+
       pages = get_h(*titles, prop: prop).values.compact
       titles.count == 1 ? pages.first : Tree::Nodes[*pages]
     end
@@ -251,17 +252,26 @@ module Infoboxer
       [namespace, titl].join(':')
     end
 
-    def user_agent(options)
-      options[:user_agent] || options[:ua] || self.class.user_agent || UA
+    def user_agent(custom)
+      custom || self.class.user_agent || UA
+    end
+
+    def siteinfo
+      @siteinfo ||= @client.query.meta(:siteinfo).prop(:namespaces, :namespacealiases, :interwikimap).response.to_h
     end
 
-    def extract_namespaces
-      siteinfo = @client.query.meta(:siteinfo).prop(:namespaces, :namespacealiases).response
-      siteinfo['namespaces'].map do |_, namespace|
-        aliases =
-          siteinfo['namespacealiases'].select { |a| a['id'] == namespace['id'] }.map { |a| a['*'] }
-        namespace.merge('aliases' => aliases)
-      end
+    def interwikis(prefix)
+      @interwikis ||= Hash.new { |h, pre|
+        interwiki = siteinfo['interwikimap'].detect { |iw| iw['prefix'] == prefix } or
+          fail ArgumentError, "Undefined interwiki: #{prefix}"
+
+        # FIXME: fragile, but what can we do?..
+        m = interwiki['url'].match(%r{^(.+)/wiki/\$1$}) or
+          fail ArgumentError, "Interwiki #{interwiki} seems not to be a MediaWiki instance"
+        h[pre] = self.class.new("#{m[1]}/w/api.php") # TODO: copy useragent
+      }
+
+      @interwikis[prefix]
     end
   end
 end

data/lib/infoboxer/media_wiki/traits.rb

@@ -34,9 +34,8 @@ module Infoboxer
         end
 
         # @private
-        def get(domain, options = {})
-          cls = Traits.domains[domain]
-          cls ? cls.new(options) : Traits.new(options)
+        def get(domain, site_info = {})
+          (Traits.domains[domain] || Traits).new(site_info)
         end
 
         # @private
@@ -68,18 +67,27 @@ module Infoboxer
         alias_method :default, :new
       end
 
-      def initialize(options = {})
-        @options = options
-        @file_namespace =
-          [DEFAULTS[:file_namespace], namespace_aliases(options, 'File')]
-          .flatten.compact.uniq
-        @category_namespace =
-          [DEFAULTS[:category_namespace], namespace_aliases(options, 'Category')]
-          .flatten.compact.uniq
+      def initialize(site_info = {})
+        @site_info = site_info
+      end
+
+      def namespace?(prefix)
+        known_namespaces.include?(prefix)
+      end
+
+      def interwiki?(prefix)
+        known_interwikis.key?(prefix)
+      end
+
+      # @private
+      def file_namespace
+        @file_namespace ||= ns_aliases('File')
       end
 
       # @private
-      attr_reader :file_namespace, :category_namespace
+      def category_namespace
+        @category_namespace ||= ns_aliases('Category')
+      end
 
       # @private
       def templates
@@ -88,16 +96,54 @@ module Infoboxer
 
       private
 
-      def namespace_aliases(options, canonical)
-        namespace = (options[:namespaces] || []).detect { |v| v['canonical'] == canonical }
-        return nil unless namespace
-        [namespace['*'], *namespace['aliases']]
+      def known_namespaces
+        @known_namespaces ||=
+          if @site_info.empty?
+            STANDARD_NAMESPACES
+          else
+            (@site_info['namespaces'].values + @site_info['namespacealiases']).map { |n| n['*'] }
+          end
+      end
+
+      def known_interwikis
+        @known_interwikis ||=
+          if @site_info.empty?
+            {}
+          else
+            @site_info['interwikimap'].map { |iw| [iw['prefix'], iw] }.to_h
+          end
+      end
+
+      def ns_aliases(base)
+        return [base] if @site_info.empty?
+        main = @site_info['namespaces'].values.detect { |n| n['canonical'] == base }
+        [base, main['*']] +
+          @site_info['namespacealiases']
+          .select { |a| a['id'] == main['id'] }.flat_map { |n| n['*'] }
+          .compact.uniq
       end
 
-      DEFAULTS = {
-        file_namespace: 'File',
-        category_namespace: 'Category'
-      }.freeze
+      # See https://www.mediawiki.org/wiki/Help:Namespaces#Standard_namespaces
+      STANDARD_NAMESPACES = [
+        'Media',            # Direct linking to media files.
+        'Special',          # Special (non-editable) pages.
+        '',                 # (Main)
+        'Talk',             # Article discussion.
+        'User',             #
+        'User talk',        #
+        'Project',          # Meta-discussions related to the operation and development of the wiki.
+        'Project talk',     #
+        'File',             # Metadata for images, videos, sound files and other media.
+        'File talk',        #
+        'MediaWiki',        # System messages and other important content.
+        'MediaWiki talk',   #
+        'Template',         # Templates: blocks of text or wikicode that are intended to be transcluded.
+        'Template talk',    #
+        'Help',             # Help files, instructions and "how-to" guides.
+        'Help talk',        #
+        'Category',         # Categories: dynamic lists of other pages.
+        'Category talk',    #
+      ].freeze
     end
   end
 end

data/lib/infoboxer/navigation.rb

@@ -38,6 +38,19 @@ module Infoboxer
   #
   # Look into {Shortcuts::Node} documentation for list of shortcuts.
   #
+  # ## Wikipath
+  #
+  # WikiPath is XPath-alike query language you can use to navigate the tree:
+  #
+  # ```ruby
+  # document.wikipath('//paragraph//wikilink[namespace=Category]')
+  # ```
+  #
+  # It can look more or less verbose than pure-ruby navigation, but the big advantage of WikiPath
+  # is it is pure data: you can store some paths in YAML file, for example.
+  #
+  # Look at {Wikipath#wikipath #wikipath} method docs for full reference.
+  #
   # ## Logical structure navigation
   #
   # MediaWiki page structure is flat, like HTML's (there's just sequence
@@ -62,7 +75,7 @@ module Infoboxer
   # {Sections::Node} for upwards.
   #
   module Navigation
-    %w[lookup shortcuts sections].each do |nav|
+    %w[lookup shortcuts sections wikipath].each do |nav|
       require_relative "navigation/#{nav}"
     end
 
@@ -70,12 +83,14 @@ module Infoboxer
       include Navigation::Lookup::Node
       include Navigation::Shortcuts::Node
       include Navigation::Sections::Node
+      include Navigation::Wikipath
     end
 
     class Tree::Nodes
       include Navigation::Lookup::Nodes
       include Navigation::Shortcuts::Nodes
       include Navigation::Sections::Nodes
+      include Navigation::Wikipath
     end
 
     class Tree::Document

data/lib/infoboxer/navigation/lookup.rb

@@ -98,9 +98,13 @@ module Infoboxer
         #   Selects matching nodes from current node's siblings, which
         #   are above current node in parents children list.
 
+        # @!method lookup_prev_sibling(*selectors, &block)
+        #   Selects first matching nodes from current node's siblings, which
+        #   are above current node in parents children list.
+
         # Underscored version of {#matches?}
         def _matches?(selector)
-          selector.matches?(self)
+          selector === self
         end
 
         # Underscored version of {#lookup}
@@ -136,6 +140,11 @@ module Infoboxer
           prev_siblings._find(selector)
         end
 
+        # Underscored version of {#lookup_prev_sibling}
+        def _lookup_prev_sibling(selector)
+          prev_siblings.reverse.detect { |n| selector === n }
+        end
+
         # Underscored version of {#lookup_next_siblings}
         def _lookup_next_siblings(selector)
           next_siblings._find(selector)
@@ -146,6 +155,7 @@ module Infoboxer
           lookup lookup_children lookup_parents
           lookup_siblings
           lookup_next_siblings lookup_prev_siblings
+          lookup_prev_sibling
         ]
           .map { |sym| [sym, :"_#{sym}"] }
           .each do |sym, underscored|

data/lib/infoboxer/navigation/sections.rb

@@ -81,6 +81,14 @@ module Infoboxer
           end
         end
 
+        def lookup_children(*arg)
+          if arg.include?(:Section)
+            sections.find(*(arg - [:Section]))
+          else
+            super
+          end
+        end
+
         private
 
         def make_sections
@@ -115,21 +123,25 @@ module Infoboxer
         #
         # @return {Tree::Nodes<Section>}
         def in_sections
-          main_node = parent.is_a?(Tree::Document) ? self : lookup_parents[-2]
+          return parent.in_sections unless parent.is_a?(Tree::Document)
+          return @in_sections if @in_sections
 
           heading =
-            if main_node.is_a?(Tree::Heading)
-              main_node.lookup_prev_siblings(Tree::Heading, level: main_node.level - 1).last
+            if is_a?(Tree::Heading)
+              lookup_prev_sibling(Tree::Heading, level: level - 1)
             else
-              main_node.lookup_prev_siblings(Tree::Heading).last
+              lookup_prev_sibling(Tree::Heading)
             end
-          return Tree::Nodes[] unless heading
+          unless heading
+            @in_sections = Tree::Nodes[]
+            return @in_sections
+          end
 
           body = heading.next_siblings
                         .take_while { |n| !n.is_a?(Tree::Heading) || n.level < heading.level }
 
           section = Section.new(heading, body)
-          Tree::Nodes[section, *heading.in_sections]
+          @in_sections = Tree::Nodes[section, *heading.in_sections]
         end
       end
 
@@ -145,6 +157,14 @@ module Infoboxer
             make_nodes(map { |n| n.send(sym, *args) })
           end
         end
+
+        def lookup_children(*arg)
+          if arg.include?(:Section)
+            sections.find(*(arg - [:Section]))
+          else
+            super
+          end
+        end
       end
 
       # Virtual node, representing logical section of the document.
@@ -175,6 +195,10 @@ module Infoboxer
           false
         end
 
+        def inspect
+          "#<#{descr}: #{children.count} nodes>"
+        end
+
         include Container
       end
     end

data/lib/infoboxer/navigation/selector.rb

@@ -7,8 +7,6 @@ module Infoboxer
       #
       # See {Lookup::Node Lookup::Node} for detailed explanation of available selectors.
       class Selector
-        include ProcMe
-
         def initialize(*arg, &block)
           @arg = [arg, block].flatten.compact.map(&method(:sym_to_class))
           @arg.each do |a|
@@ -26,8 +24,8 @@ module Infoboxer
           "#<Selector(#{@arg.map(&:to_s).join(', ')})>"
         end
 
-        def matches?(node)
-          @arg.all? { |a| arg_matches?(a, node) }
+        def ===(other)
+          @arg.all? { |a| arg_matches?(a, other) }
         end
 
         private
@@ -45,13 +43,24 @@ module Infoboxer
           when Proc
             check.call(node)
           when Hash
-            check.all? { |attr, value| node.respond_to?(attr) && value === node.send(attr) }
+            check.all? { |attr, value|
+              node.respond_to?(attr) && value_matches?(value, node.send(attr)) ||
+                node.params.key?(attr) && value_matches?(value, node.params[attr])
+            }
           when Symbol
             node.respond_to?(check) && node.send(check)
           else
             check === node
           end
         end
+
+        def value_matches?(matcher, value)
+          if matcher.is_a?(String) && value.is_a?(String)
+            matcher.casecmp(value).zero?
+          else
+            matcher === value
+          end
+        end
       end
     end
   end

data/lib/infoboxer/navigation/wikipath.rb

@@ -0,0 +1,32 @@
+# encoding: utf-8
+
+require_relative '../wiki_path'
+
+module Infoboxer
+  module Navigation
+    module Wikipath
+      # Search nodes inside current by XPath alike query language.
+      #
+      # This feature is experimental, but should work for most of the useful cases.
+      #
+      # Examples of WikiPath:
+      #
+      # ```
+      # /paragraph # direct child of current node, being paragraph
+      # //paragraph # any node in current node's subtree, being paragraph
+      # //template[name=Infobox] # template node in subtree, with name attribute equal to Infobox
+      # //template[name="Infobox country"] # optional quotes are allowed
+      # //template[name=/^Infobox/] # regexes are supported
+      # //wikilink[italic] # node predicates are supported (the same as `lookup(:Wikilink, :italic?)`
+      # //*[italic] # type wildcards are supported
+      # //template[name=/^Infobox/]/var[name=birthday] # series of lookups work
+      # ```
+      #
+      # @param string [String] WikiPath to lookup
+      # @return [Nodes]
+      def wikipath(string)
+        Infoboxer::WikiPath.parse(string).call(self)
+      end
+    end
+  end
+end

data/lib/infoboxer/parser.rb

@@ -1,7 +1,6 @@
 # encoding: utf-8
 
 require 'ostruct'
-require 'procme'
 require 'logger'
 
 module Infoboxer

data/lib/infoboxer/parser/inline.rb

@@ -83,7 +83,7 @@ module Infoboxer
 
       private
 
-      def inline_formatting(match)
+      def inline_formatting(match) # rubocop:disable Metrics/MethodLength, Metrics/CyclomaticComplexity, Metrics/AbcSize
         case match
         when "'''''"
           BoldItalic.new(short_inline(/'''''/))
@@ -109,6 +109,8 @@ module Infoboxer
           reference(Regexp.last_match(1))
         when /<math>/
           math
+        when /<gallery([^>]*)>/
+          gallery(Regexp.last_match(1))
         when '<'
           html || Text.new(match) # it was not HTML, just accidental <
         else
@@ -126,8 +128,18 @@ module Infoboxer
           caption = inline(/\]\]/)
           @context.pop_eol_sign
         end
+        name, namespace = link.split(':', 2).reverse
+        lnk, params =
+          if @context.traits.namespace?(namespace)
+            [link, {namespace: namespace}]
+          elsif @context.traits.interwiki?(namespace)
+            [name, {interwiki: namespace}]
+          else
+            [link, {}]
+          end
 
-        Wikilink.new(link, caption)
+        puts @context.rest if lnk.nil?
+        Wikilink.new(lnk, caption, **params)
       end
 
       # http://en.wikipedia.org/wiki/Help:Link#External_links
@@ -159,6 +171,34 @@ module Infoboxer
           Text.new(@context.scan_continued_until(%r{</nowiki>}))
         end
       end
+
+      def gallery(tag_rest)
+        params = parse_params(tag_rest)
+        images = []
+        guarded_loop do
+          @context.next! if @context.eol?
+          path = @context.scan_until(%r{</gallery>|\||$})
+          attrs = @context.matched == '|' ? gallery_image_attrs : {}
+          unless path.empty?
+            images << Tree::Image.new(path.sub(/^#{re.file_namespace}/, ''), attrs)
+          end
+          break if @context.matched == '</gallery>'
+        end
+        Gallery.new(images, params)
+      end
+
+      def gallery_image_attrs
+        nodes = []
+
+        guarded_loop do
+          nodes << short_inline(%r{\||</gallery>})
+          break if @context.eol? || @context.matched?(%r{</gallery>})
+        end
+
+        nodes.map(&method(:image_attr))
+             .inject(&:merge)
+             .reject { |_k, v| v.nil? || v.empty? }
+      end
     end
 
     require_relative 'image'

data/lib/infoboxer/parser/paragraphs.rb

@@ -14,7 +14,7 @@ module Infoboxer
 
           @context.next!
         end
-        nodes.flow_templates
+        nodes
       end
 
       private

data/lib/infoboxer/parser/template.rb

@@ -29,8 +29,8 @@ module Infoboxer
 
         guarded_loop do
           @context.next! while @context.eol?
-          if @context.check(/\s*([^ =}|<]+)\s*=\s*/)
-            name = @context.scan(/\s*([^ =]+)/).strip
+          if @context.check(/\s*([^=}|<]+)\s*=\s*/)
+            name = @context.scan(/\s*([^=]+)/).strip
             @context.skip(/\s*=\s*/)
           else
             name = num
@@ -38,7 +38,7 @@ module Infoboxer
           end
           log "Variable #{name} found"
 
-          value = long_inline(/\||}}/)
+          value = sanitize_value(long_inline(/\||}}/))
 
           # it was just empty line otherwise
           res << Var.new(name.to_s, value) unless value.empty? && name.is_a?(Numeric)
@@ -50,6 +50,11 @@ module Infoboxer
         end
         res
       end
+
+      def sanitize_value(nodes)
+        nodes.pop if (nodes.last.is_a?(Pre) || nodes.last.is_a?(Text)) && nodes.last.text =~ /^\s*$/ # FIXME: dirty!
+        nodes
+      end
     end
   end
 end

data/lib/infoboxer/parser/util.rb

@@ -12,6 +12,7 @@ module Infoboxer
         \[[a-z]+://   |     # external link
         <nowiki[^>]*> |     # nowiki
         <ref[^>]*>    |     # reference
+        <gallery[^>]*>|     # gallery
         <math>        |     # math
         <                   # HTML tag
       ))x

data/lib/infoboxer/templates/base.rb

@@ -35,9 +35,11 @@ module Infoboxer
     #
     # Used for {Set} definitions.
     class Show < Base
-      alias_method :children, :unnamed_variables
+      def text
+        unnamed_variables.map(&:text).join(children_separator)
+      end
 
-      protected
+      private
 
       def children_separator
         ' '

data/lib/infoboxer/tree.rb

@@ -63,7 +63,7 @@ module Infoboxer
     require_relative 'tree/nodes'
 
     %w[text compound inline
-       image html paragraphs list template table ref math
+       image gallery html paragraphs list template table ref math
        document].each do |type|
       require_relative "tree/#{type}"
     end

data/lib/infoboxer/tree/compound.rb

@@ -4,7 +4,7 @@ module Infoboxer
   module Tree
     # Base class for all nodes with children.
     class Compound < Node
-      def initialize(children = Nodes.new, params = {})
+      def initialize(children = Nodes.new, **params)
         super(params)
         @children = Nodes[*children]
         @children.each { |c| c.parent = self }
@@ -40,8 +40,7 @@ module Infoboxer
         if children.count == 1 && children.first.is_a?(Text)
           "#{indent(level)}#{children.first.text} <#{descr}>\n"
         else
-          "#{indent(level)}<#{descr}>\n" +
-            children.map(&call(to_tree: level + 1)).join
+          "#{indent(level)}<#{descr}>\n" + children.map { |c| c.to_tree(level + 1) }.join
         end
       end
 

data/lib/infoboxer/tree/gallery.rb

@@ -0,0 +1,12 @@
+# encoding: utf-8
+
+module Infoboxer
+  module Tree
+    # Represents gallery of images (contents of `<gallery>` special tag).
+    #
+    # See [Wikipedia Tutorial](https://en.wikipedia.org/wiki/Help:Gallery_tag)
+    # for explanation of attributes.
+    class Gallery < Compound
+    end
+  end
+end

data/lib/infoboxer/tree/image.rb

@@ -7,8 +7,8 @@ module Infoboxer
     # See [Wikipedia Tutorial](https://en.wikipedia.org/wiki/Wikipedia:Extended_image_syntax)
     # for explanation of attributes.
     class Image < Node
-      def initialize(path, params = {})
-        @caption = params.delete(:caption)
+      def initialize(path, caption: nil, **params)
+        @caption = caption
         super({path: path}.merge(params))
       end
 
@@ -44,7 +44,7 @@ module Infoboxer
         super(level) +
           if caption && !caption.empty?
             indent(level + 1) + "caption:\n" +
-            caption.children.map(&call(to_tree: level + 2)).join
+            caption.children.map { |c| c.to_tree(level + 2) }.join
           else
             ''
           end

data/lib/infoboxer/tree/inline.rb

@@ -17,8 +17,8 @@ module Infoboxer
 
     # Base class for internal/external links,
     class Link < Compound
-      def initialize(link, label = nil)
-        super(label || Nodes.new([Text.new(link)]), link: link)
+      def initialize(link, label = nil, **attr)
+        super(label || Nodes.new([Text.new(link)]), link: link, **attr)
       end
 
       # @!attribute [r] link

data/lib/infoboxer/tree/linkable.rb

@@ -15,7 +15,7 @@ module Infoboxer
       # * {Tree::Nodes#follow} for extracting multiple links at once;
       # * {MediaWiki#get} for basic information on page extraction.
       def follow
-        client.get(link)
+        client.get(link, interwiki: interwiki)
       end
 
       # Human-readable page URL
@@ -28,6 +28,9 @@ module Infoboxer
 
       protected
 
+      # redefined in {Wikilink}
+      def interwiki; end
+
       def page
         lookup_parents(MediaWiki::Page).first or fail('Not in a page from real source')
       end

data/lib/infoboxer/tree/math.rb

@@ -4,9 +4,6 @@ module Infoboxer
     #
     # See also: https://en.wikipedia.org/wiki/Help:Displaying_a_formula
     class Math < Text
-      def text
-        "<math>#{super}</math>"
-      end
     end
   end
 end

data/lib/infoboxer/tree/node.rb

@@ -11,9 +11,7 @@ module Infoboxer
     # you will receive it from tree and use for navigations.
     #
     class Node
-      include ProcMe
-
-      def initialize(params = {})
+      def initialize(**params)
         @params = params
       end
 
@@ -42,6 +40,10 @@ module Infoboxer
         parent ? parent.index_of(self) : 0
       end
 
+      def first?
+        index.zero?
+      end
+
       # List of all sibling nodes (children of same parent)
       def siblings
         parent ? parent.children - [self] : Nodes[]
@@ -152,7 +154,7 @@ module Infoboxer
       end
 
       def show_params(prms = nil)
-        (prms || params).map { |k, v| "#{k}: #{v.inspect}" }.join(', ')
+        (prms || params).reject { |_, v| v.nil? }.map { |k, v| "#{k}: #{v.inspect}" }.join(', ')
       end
 
       def indent(level)

data/lib/infoboxer/tree/nodes.rb

@@ -38,10 +38,19 @@ module Infoboxer
       # @!method compact
       #    Just like Array#compact, but returns Nodes
 
+      # @!method grep(pattern)
+      #    Just like Array#grep, but returns Nodes
+
+      # @!method grep_v(pattern)
+      #    Just like Array#grep_v, but returns Nodes
+
       # @!method -(other)
       #    Just like Array#-, but returns Nodes
 
-      %i[select reject sort_by flatten compact -].each do |sym|
+      # @!method +(other)
+      #    Just like Array#+, but returns Nodes
+
+      %i[select reject sort_by flatten compact grep grep_v - +].each do |sym|
         define_method(sym) do |*args, &block|
           Nodes[*super(*args, &block)]
         end
@@ -75,6 +84,21 @@ module Infoboxer
         end
       end
 
+      # Just like Array#flat_map, but returns Nodes, **if** all map results are Node
+      def flat_map
+        res = super
+        if res.all? { |n| n.is_a?(Node) || n.is_a?(Nodes) }
+          Nodes[*res]
+        else
+          res
+        end
+      end
+
+      # Just like Array#group, but returns hash with `{<grouping variable> => Nodes}`
+      def group_by
+        super.map { |title, group| [title, Nodes[*group]] }.to_h
+      end
+
       # @!method prev_siblings
       #   Previous siblings (flat list) of all nodes inside.
 
@@ -139,12 +163,14 @@ module Infoboxer
       # @return [Nodes<MediaWiki::Page>] It is still `Nodes`, so you
       #   still can process them uniformely.
       def follow
-        links = select { |n| n.respond_to?(:link) }.map(&:link)
+        links = grep(Linkable)
         return Nodes[] if links.empty?
         page = first.lookup_parents(MediaWiki::Page).first or
           fail('Not in a page from real source')
         page.client or fail('MediaWiki client not set')
-        page.client.get(*links)
+        pages = links.group_by(&:interwiki)
+                     .flat_map { |iw, ls| page.client.get(*ls.map(&:link), interwiki: iw) }
+        pages.count == 1 ? pages.first : Nodes[*pages]
       end
 
       # @private
@@ -173,7 +199,9 @@ module Infoboxer
       # @private
       # Internal, used by {Parser}
       def flow_templates
-        make_nodes(map { |n| n.is_a?(Paragraph) ? n.to_templates? : n })
+        # TODO: will it be better?..
+        # make_nodes(map { |n| n.is_a?(Paragraph) ? n.to_templates? : n })
+        self
       end
 
       private

data/lib/infoboxer/tree/paragraphs.rb

@@ -75,7 +75,7 @@ module Infoboxer
       # @private
       # Internal, used by {Parser}
       def to_templates
-        children.select(&filter(itself: Template))
+        children.grep(Template)
       end
 
       # @private

data/lib/infoboxer/tree/table.rb

@@ -13,12 +13,12 @@ module Infoboxer
 
       # All table rows.
       def rows
-        children.select(&fltr(itself: TableRow))
+        children.grep(TableRow)
       end
 
       # Table caption, if exists.
       def caption
-        children.detect(&fltr(itself: TableCaption))
+        children.grep(TableCaption).first
       end
 
       # For now, returns first table row, if it consists only of
@@ -26,12 +26,12 @@ module Infoboxer
       #
       # FIXME: it can easily be several table heading rows
       def heading_row
-        rows.first if rows.first && rows.first.children.all?(&call(matches?: TableHeading))
+        rows.first if rows.first && rows.first.children.all? { |c| c.is_a?(TableHeading) }
       end
 
       # For now, returns all table rows except {#heading_row}
       def body_rows
-        if rows.first && rows.first.children.all?(&call(matches?: TableHeading))
+        if rows.first && rows.first.children.all? { |c| c.is_a?(TableHeading) }
           rows[1..-1]
         else
           rows
@@ -39,19 +39,11 @@ module Infoboxer
       end
 
       def text
-        table = Terminal::Table.new
-        table.title = caption.text.sub(/\n+\Z/, '') if caption
-
-        if heading_row
-          table.headings = heading_row.children.map(&:text)
-                                      .map(&call(sub: [/\n+\Z/, '']))
-        end
-
-        table.rows = body_rows.map { |r|
-          r.children.map(&:text)
-           .map(&call(sub: [/\n+\Z/, '']))
-        }
-        table.to_s + "\n\n"
+        Terminal::Table.new.tap { |table|
+          table.title = caption.text.sub(/\n+\Z/, '') if caption
+          table.headings = heading_row.children.map(&:text_) if heading_row
+          table.rows = body_rows.map { |r| r.children.map(&:text_) }
+        }.to_s + "\n\n"
       end
     end
 

data/lib/infoboxer/tree/template.rb

@@ -17,10 +17,15 @@ module Infoboxer
       end
 
       # Internal, used by {Parser}
+      # Means even children-less Var should not be removed from parser tree.
       def empty?
         false
       end
 
+      def named?
+        name !~ /^\d+$/
+      end
+
       protected
 
       def descr
@@ -54,6 +59,7 @@ module Infoboxer
     # values.
     #
     # ### On variables naming
+    #
     # MediaWiki templates can contain _named_ and _unnamed_ variables.
     # Example:
     #
@@ -104,12 +110,16 @@ module Infoboxer
       # See {Var} class to understand what you can do with them.
       #
       # @return [Nodes<Var>]
-      attr_reader :variables
+      # attr_reader :variables
+      alias_method :variables, :children
 
       def initialize(name, variables = Nodes[])
-        super(Nodes[], extract_params(variables))
+        super(variables, extract_params(variables))
         @name = name
-        @variables = Nodes[*variables].each { |v| v.parent = self }
+      end
+
+      def text
+        ''
       end
 
       # See {Node#to_tree}
@@ -133,7 +143,7 @@ module Infoboxer
       #
       # @return [Nodes<Var>]
       def unnamed_variables
-        variables.find(name: /^\d+$/)
+        variables.reject(&:named?)
       end
 
       # Fetches template variable(s) by name(s) or patterns.
@@ -236,7 +246,7 @@ module Infoboxer
       def extract_params(vars)
         vars
           .select { |v| v.children.count == 1 && v.children.first.is_a?(Text) }
-          .map { |v| [v.name, v.children.first.raw_text] }.to_h
+          .map { |v| [v.name.to_sym, v.children.first.raw_text] }.to_h
       end
 
       def inspect_variables(depth)

data/lib/infoboxer/tree/text.rb

@@ -15,7 +15,7 @@ module Infoboxer
       # Text fragment without decodint of HTML entities.
       attr_accessor :raw_text
 
-      def initialize(text, params = {})
+      def initialize(text, **params)
         super(params)
         @raw_text = text
       end

data/lib/infoboxer/tree/wikilink.rb

@@ -12,14 +12,23 @@ module Infoboxer
     # Note, that Wikilink is {Linkable}, so you can {Linkable#follow #follow}
     # it to obtain linked pages.
     class Wikilink < Link
-      def initialize(*)
-        super
-        parse_link!
+      def initialize(link, label = nil, namespace: nil, interwiki: nil)
+        super(link, label, namespace: namespace, interwiki: interwiki)
+        @namespace = namespace || ''
+        @interwiki = interwiki
+        parse_name!
       end
 
       # "Clean" wikilink name, for ex., `Cities` for `[Category:Cities]`
       attr_reader :name
 
+      # Interwiki identifier. For example, `[[wikt:Argentina]]`
+      # will have `"Argentina"` as its {#name} and `"wikt"` (wiktionary) as an
+      # interwiki. TODO: how to use it.
+      #
+      # See [Wikipedia docs](https://en.wikipedia.org/wiki/Help:Interwiki_linking) for details.
+      attr_reader :interwiki
+
       # Wikilink namespace, `Category` for `[Category:Cities]`, empty
       # string (not `nil`!) for just `[Cities]`
       attr_reader :namespace
@@ -46,10 +55,8 @@ module Infoboxer
 
       private
 
-      def parse_link!
-        @name, @namespace = link.split(':', 2).reverse
-        @namespace ||= ''
-
+      def parse_name!
+        @name = namespace.empty? ? link : link.sub(/^#{namespace}:/, '')
         @name, @anchor = @name.split('#', 2)
         @anchor ||= ''
 

data/lib/infoboxer/version.rb

@@ -2,7 +2,8 @@
 
 module Infoboxer
   MAJOR = 0
-  MINOR = 2
-  PATCH = 8
-  VERSION = [MAJOR, MINOR, PATCH].join('.')
+  MINOR = 3
+  PATCH = 0
+  PRE = 'pre'.freeze # set to `nil` for normal releases
+  VERSION = [MAJOR, MINOR, PATCH, PRE].compact.join('.')
 end

data/lib/infoboxer/wiki_path.rb

@@ -0,0 +1,94 @@
+module Infoboxer
+  # @private
+  class WikiPath
+    ParseError = Class.new(ArgumentError)
+
+    class << self
+      def _parse(string)
+        scanner = StringScanner.new(string)
+        res = []
+        loop do
+          res << scan_step(scanner)
+          break if scanner.eos?
+        end
+        res
+      end
+
+      def parse(string)
+        new(_parse(string))
+      end
+
+      private
+
+      def scan_step(scanner) # rubocop:disable Metrics/PerceivedComplexity
+        op = scanner.scan(%r{//?}) or unexpected(scanner, '/')
+        type = scanner.scan(/[A-Za-z_]*/)
+        attrs = {}
+        while scanner.scan(/\[/)
+          attr = scanner.scan(/[-a-z_0-9]+/) or unexpected(scanner, 'attribute name')
+          if scanner.scan(/\]/)
+            (attrs[:predicates] ||= []) << "#{attr}?".to_sym
+            next
+          end
+          scanner.scan(/\s*=\s*/) or unexpected(scanner, '= or ]')
+          value = scanner.scan(/[^\]]+/) # TODO: probably, should do a proper [] counting?..
+          scanner.scan(/\]/) or unexpected(scanner, ']')
+          attrs[attr.to_sym] = process_value(value)
+        end
+        res = op == '//' ? {op: :lookup} : {}
+        res[:type] = process_type(type) unless type.empty?
+        res.merge(attrs) # TODO: raise if empty selector
+      end
+
+      def process_value(value)
+        case value
+        when /^'(.*)'$/, /^"(.*)"$/
+          Regexp.last_match(1)
+        when %r{^/(.+)/$}
+          Regexp.new(Regexp.last_match(1))
+        else
+          value
+        end
+      end
+
+      def process_type(type)
+        type.gsub(/(?:^|_)([a-z])/, &:upcase).tr('_', '').to_sym
+            .tap { |t| valid_type?(t) or fail(ParseError, "Unrecognized node type: #{type}") }
+      end
+
+      def valid_type?(t)
+        t == :Section || Infoboxer::Tree.const_defined?(t)
+      end
+
+      def unexpected(scanner, expected)
+        place = scanner.eos? ? 'end of pattern' : scanner.rest.inspect
+        fail ParseError, "Unexpected #{place}, expecting #{expected}"
+      end
+    end
+
+    def initialize(path)
+      @path = path
+    end
+
+    def call(node)
+      @path.inject(node) { |res, step| apply_step(res, step) }
+    end
+
+    private
+
+    def apply_step(node, step)
+      # TODO: "compile" the op/args sequences at WikiPath initialization
+      step = step.dup
+      op = step.delete(:op) || :lookup_children
+      args = []
+      if (t = step.delete(:type))
+        args << t
+      end
+      if (pred = step.delete(:predicates))
+        args.concat(pred)
+      end
+      args << step unless step.empty?
+      node.send(op, *args)
+    end
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: infoboxer
 version: !ruby/object:Gem::Version
-  version: 0.2.8
+  version: 0.3.0.pre
 platform: ruby
 authors:
 - Victor Shepelev
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-05-11 00:00:00.000000000 Z
+date: 2017-09-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: htmlentities
@@ -24,20 +24,6 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-- !ruby/object:Gem::Dependency
-  name: procme
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :runtime
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
 - !ruby/object:Gem::Dependency
   name: mediawiktory
   requirement: !ruby/object:Gem::Requirement
@@ -115,6 +101,7 @@ files:
 - lib/infoboxer/navigation/sections.rb
 - lib/infoboxer/navigation/selector.rb
 - lib/infoboxer/navigation/shortcuts.rb
+- lib/infoboxer/navigation/wikipath.rb
 - lib/infoboxer/parser.rb
 - lib/infoboxer/parser/context.rb
 - lib/infoboxer/parser/html.rb
@@ -130,6 +117,7 @@ files:
 - lib/infoboxer/tree.rb
 - lib/infoboxer/tree/compound.rb
 - lib/infoboxer/tree/document.rb
+- lib/infoboxer/tree/gallery.rb
 - lib/infoboxer/tree/html.rb
 - lib/infoboxer/tree/image.rb
 - lib/infoboxer/tree/inline.rb
@@ -145,6 +133,7 @@ files:
 - lib/infoboxer/tree/text.rb
 - lib/infoboxer/tree/wikilink.rb
 - lib/infoboxer/version.rb
+- lib/infoboxer/wiki_path.rb
 - profile/out/.gitkeep
 - profile/pages/argentina.txt
 - profile/pages/canada.wiki
@@ -177,9 +166,9 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: 2.1.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - ">"
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 1.3.1
 requirements: []
 rubyforge_project: 
 rubygems_version: 2.6.10