diff options
Diffstat (limited to 'lib')
| -rw-r--r-- | lib/banzai/querying.rb | 56 | ||||
| -rw-r--r-- | lib/banzai/reference_extractor.rb | 5 | ||||
| -rw-r--r-- | lib/banzai/reference_parser/base_parser.rb | 5 | ||||
| -rw-r--r-- | lib/banzai/reference_parser/directly_addressed_user_parser.rb | 8 | ||||
| -rw-r--r-- | lib/gitlab/reference_extractor.rb | 8 |
5 files changed, 73 insertions, 9 deletions
diff --git a/lib/banzai/querying.rb b/lib/banzai/querying.rb index 1e1b51e683e..fb2faae02bc 100644 --- a/lib/banzai/querying.rb +++ b/lib/banzai/querying.rb @@ -1,18 +1,64 @@ module Banzai module Querying + module_function + # Searches a Nokogiri document using a CSS query, optionally optimizing it # whenever possible. # - # document - A document/element to search. - # query - The CSS query to use. + # document - A document/element to search. + # query - The CSS query to use. + # reference_options - A hash with nodes filter options # - # Returns a Nokogiri::XML::NodeSet. - def self.css(document, query) + # Returns an array of Nokogiri::XML::Element objects if location is specified + # in reference_options. Otherwise it would a Nokogiri::XML::NodeSet. + def css(document, query, reference_options = {}) # When using "a.foo" Nokogiri compiles this to "//a[...]" but # "descendant::a[...]" is quite a bit faster and achieves the same result. xpath = Nokogiri::CSS.xpath_for(query)[0].gsub(%r{^//}, 'descendant::') + xpath = restrict_to_p_nodes_at_root(xpath) if filter_nodes_at_beginning?(reference_options) + nodes = document.xpath(xpath) + + filter_nodes(nodes, reference_options) + end + + def restrict_to_p_nodes_at_root(xpath) + xpath.gsub('descendant::', './p/') + end + + def filter_nodes(nodes, reference_options) + if filter_nodes_at_beginning?(reference_options) + filter_nodes_at_beginning(nodes) + else + nodes + end + end + + def filter_nodes_at_beginning?(reference_options) + reference_options && reference_options[:location] == :beginning + end + + # Selects child nodes if they are present in the beginning among other siblings. + # + # nodes - A Nokogiri::XML::NodeSet. + # + # Returns an array of Nokogiri::XML::Element objects. + def filter_nodes_at_beginning(nodes) + parents_and_nodes = nodes.group_by(&:parent) + filtered_nodes = [] + + parents_and_nodes.each do |parent, nodes| + children = parent.children + nodes = nodes.to_a + + children.each do |child| + next if child.text.blank? + node = nodes.shift + break unless node == child + filtered_nodes << node + end + end - document.xpath(xpath) + filtered_nodes end end end diff --git a/lib/banzai/reference_extractor.rb b/lib/banzai/reference_extractor.rb index b26a41a1f3b..8e3b0c4db79 100644 --- a/lib/banzai/reference_extractor.rb +++ b/lib/banzai/reference_extractor.rb @@ -16,6 +16,11 @@ module Banzai processor.process(html_documents) end + def reset_memoized_values + @html_documents = nil + @texts_and_contexts = [] + end + private def html_documents diff --git a/lib/banzai/reference_parser/base_parser.rb b/lib/banzai/reference_parser/base_parser.rb index d8a855ec1fe..2058a58d0ae 100644 --- a/lib/banzai/reference_parser/base_parser.rb +++ b/lib/banzai/reference_parser/base_parser.rb @@ -33,7 +33,7 @@ module Banzai # they have access to. class BaseParser class << self - attr_accessor :reference_type + attr_accessor :reference_type, :reference_options end # Returns the attribute name containing the value for every object to be @@ -182,9 +182,10 @@ module Banzai # the references. def process(documents) type = self.class.reference_type + reference_options = self.class.reference_options nodes = documents.flat_map do |document| - Querying.css(document, "a[data-reference-type='#{type}'].gfm").to_a + Querying.css(document, "a[data-reference-type='#{type}'].gfm", reference_options).to_a end gather_references(nodes) diff --git a/lib/banzai/reference_parser/directly_addressed_user_parser.rb b/lib/banzai/reference_parser/directly_addressed_user_parser.rb new file mode 100644 index 00000000000..77df9bbd024 --- /dev/null +++ b/lib/banzai/reference_parser/directly_addressed_user_parser.rb @@ -0,0 +1,8 @@ +module Banzai + module ReferenceParser + class DirectlyAddressedUserParser < UserParser + self.reference_type = :user + self.reference_options = { location: :beginning } + end + end +end diff --git a/lib/gitlab/reference_extractor.rb b/lib/gitlab/reference_extractor.rb index 11c0b01f0dc..437a339dd2b 100644 --- a/lib/gitlab/reference_extractor.rb +++ b/lib/gitlab/reference_extractor.rb @@ -1,13 +1,12 @@ module Gitlab # Extract possible GFM references from an arbitrary String for further processing. class ReferenceExtractor < Banzai::ReferenceExtractor - REFERABLES = %i(user issue label milestone merge_request snippet commit commit_range) + REFERABLES = %i(user issue label milestone merge_request snippet commit commit_range directly_addressed_user) attr_accessor :project, :current_user, :author def initialize(project, current_user = nil) @project = project @current_user = current_user - @references = {} super() @@ -21,6 +20,11 @@ module Gitlab super(type, project, current_user) end + def reset_memoized_values + @references = {} + super() + end + REFERABLES.each do |type| define_method("#{type}s") do @references[type] ||= references(type) |