class Capybara::Result
@see Capybara::Node::Element
* sample()
* values_at()
* empty?()
* last()
* first()
* length()
* count()
* size()
* at()
* each()
* []
collection similar to an Array because it implements Enumerable and offers the following Array methods through delegation:
A {Capybara::Result} represents a collection of {Capybara::Node::Element} on the page. It is possible to interact with this
#
def [](*args)
def [](*args) idx, length = args max_idx = case idx when Integer if idx.negative? nil else length.nil? ? idx : idx + length - 1 end when Range # idx.max is broken with beginless ranges # idx.end && idx.max # endless range will have end == nil max = idx.end max = nil if max&.negative? max -= 1 if max && idx.exclude_end? max end if max_idx.nil? full_results[*args] else load_up_to(max_idx + 1) @result_cache[*args] end end
def add_to_cache(elem)
def add_to_cache(elem) elem.allow_reload!(@result_cache.size) if @allow_reload @result_cache << elem end
def allow_reload!
- Api: - private
def allow_reload! @allow_reload = true self end
def compare_count
def compare_count return 0 unless @query count, min, max, between = @query.options.values_at(:count, :minimum, :maximum, :between) # Only check filters for as many elements as necessary to determine result if count && (count = Integer(count)) return load_up_to(count + 1) <=> count end return -1 if min && (min = Integer(min)) && (load_up_to(min) < min) return 1 if max && (max = Integer(max)) && (load_up_to(max + 1) > max) if between min, max = (between.begin && between.min) || 1, between.end max -= 1 if max && between.exclude_end? size = load_up_to(max ? max + 1 : min) return size <=> min unless between.include?(size) end 0 end
def each(&block)
def each(&block) return enum_for(:each) unless block @result_cache.each(&block) loop do next_result = @results_enum.next add_to_cache(next_result) yield next_result end self end
def empty?
def empty? !any? end
def failure_message
def failure_message message = @query.failure_message if count.zero? message << ' but there were no matches' else message << ", found #{count} #{Capybara::Helpers.declension('match', 'matches', count)}: " \ << full_results.map { |r| r.text.inspect }.join(', ') end unless rest.empty? elements = rest.map { |el| el.text rescue '<<ERROR>>' }.map(&:inspect).join(', ') # rubocop:disable Style/RescueModifier message << '. Also found ' << elements << ', which matched the selector but not all filters. ' message << @filter_errors.join('. ') if (rest.size == 1) && count.zero? end message end
def full_results
def full_results loop { @result_cache << @results_enum.next } @result_cache end
def initialize(elements, query)
def initialize(elements, query) @elements = elements @result_cache = [] @filter_errors = [] @results_enum = lazy_select_elements { |node| query.matches_filters?(node, @filter_errors) } @query = query @allow_reload = false end
def lazy_select_elements(&block)
https://github.com/teamcapybara/capybara/issues/2349
while JRuby >= 9.2.8.0 leaks threads when using lazy enumerators
https://github.com/jruby/jruby/issues/4212
causes a concurrency issue with network requests here
JRuby < 9.2.8.0 has an issue with lazy enumerators which
def lazy_select_elements(&block) @elements.select(&block).to_enum # non-lazy evaluation end
def lazy_select_elements(&block)
def lazy_select_elements(&block) @elements.lazy.select(&block) end
def load_up_to(num)
def load_up_to(num) loop do break if @result_cache.size >= num add_to_cache(@results_enum.next) end @result_cache.size end
def matches_count?
def matches_count? compare_count.zero? end
def negative_failure_message
def negative_failure_message failure_message.sub(/(to find)/, 'not \1') end
def rest
def rest @rest ||= @elements - full_results end
def unfiltered_size
def unfiltered_size @elements.length end