catstodon/app/lib/extractor.rb
luigi eb51e43fb4
Optimize some regex matching (#15528)
* Use Regex#match?

* Replace =~ too

* Avoid to call match? from Nil

* Keep value of Regexp.last_match
2021-01-22 10:09:08 +01:00

64 lines
1.7 KiB
Ruby

# frozen_string_literal: true
module Extractor
extend Twitter::Extractor
module_function
# :yields: username, list_slug, start, end
def extract_mentions_or_lists_with_indices(text)
return [] unless Twitter::Regex[:at_signs].match?(text)
possible_entries = []
text.to_s.scan(Account::MENTION_RE) do |screen_name, _|
match_data = $LAST_MATCH_INFO
after = $'
unless Twitter::Regex[:end_mention_match].match?(after)
start_position = match_data.char_begin(1) - 1
end_position = match_data.char_end(1)
possible_entries << {
screen_name: screen_name,
indices: [start_position, end_position],
}
end
end
if block_given?
possible_entries.each do |mention|
yield mention[:screen_name], mention[:indices].first, mention[:indices].last
end
end
possible_entries
end
def extract_hashtags_with_indices(text, **)
return [] unless /#/.match?(text)
tags = []
text.scan(Tag::HASHTAG_RE) do |hash_text, _|
match_data = $LAST_MATCH_INFO
start_position = match_data.char_begin(1) - 1
end_position = match_data.char_end(1)
after = $'
if %r{\A://}.match?(after)
hash_text.match(/(.+)(https?\Z)/) do |matched|
hash_text = matched[1]
end_position -= matched[2].char_length
end
end
tags << {
hashtag: hash_text,
indices: [start_position, end_position],
}
end
tags.each { |tag| yield tag[:hashtag], tag[:indices].first, tag[:indices].last } if block_given?
tags
end
def extract_cashtags_with_indices(_text)
[] # always returns empty array
end
end