#!/usr/bin/ruby
# encoding: UTF-8
require 'set'

REGEX_HASHTAGS = /(?<=^| )#([\p{L}\p{N}]+)/
HASHTAG_URLS = {
	mastodon_social_profile: 'https://mastodon.social/@FiXato/tagged/%{keyword}',
	google_search: 'https://www.google.com/search?q=%%23%{keyword}&oq=%%23%{keyword}&ie=UTF-8'
}
referenced_hashtags = Set.new()
parsed_lines = []
ARGF.each_line do |line|
	line.gsub!(REGEX_HASHTAGS) do |s|
		hashtag = $1
		# Exclude hex-encoded colours from hashtags.
		# FIXME: I guess this could lead to for instance #deadbeef not being tagged, but it's the best I could come up with right now.
		if hashtag.match?(/[0-9a-fA-F]{3}([0-9a-fA-F]{3})?$/) || line.match?(/#{s}\s*[{><%]/)
			next s
		else
			referenced_hashtags << hashtag
			next "#{HASHTAG_URLS[:mastodon_social_profile]}[#%{keyword}]" % {keyword: hashtag}
		end
	end
	parsed_lines << line
end
end_of_header = parsed_lines.find_index{|line|line.strip == ''}
parsed_lines[end_of_header] = ":referenced_hashtags: #{referenced_hashtags.to_a.compact.sort.join(', ')}\n"

puts parsed_lines.join("")