docbrown/app/services/markdown_processor/fixer/base.rb
2021-08-26 10:01:08 -04:00

99 lines
3.3 KiB
Ruby

module MarkdownProcessor
module Fixer
# Services here in the Fixer module should inherit from this base class.
# #call is implemented here so other services only need to define a
# METHODS constant that is an Array of symbols referencing other methods
# found here.
#
# For example
# METHODS = %i[add_quotes_to_tile add_quotes_to_description]
class Base
FRONT_MATTER_DETECTOR = /-{3}.*?-{3}/m
def self.call(markdown)
return unless markdown
fix_methods.reduce(markdown) { |acc, elem| public_send(elem, acc) }
end
def self.fix_methods
self::METHODS
end
def self.add_quotes_to_title(markdown)
add_quotes_to_section(markdown, section: "title")
end
def self.add_quotes_to_description(markdown)
add_quotes_to_section(markdown, section: "description")
end
# This turns --- into ------- after the first two,
# because --- messes with front matter
def self.modify_hr_tags(markdown)
markdown.gsub(/^---/).with_index { |match, i| i > 1 ? "#{match}-----" : match }
end
def self.lowercase_published(markdown)
markdown.gsub(/-{3}.*?-{3}/m) do |front_matter|
front_matter.gsub(/^published: /i, "published: ")
end
end
def self.convert_new_lines(markdown)
markdown.gsub("\r\n", "\n")
end
def self.split_tags(markdown)
markdown.gsub(/\ntags:.*\n/) do |tags|
tags.split(" #").join(",").delete("#").gsub(":,", ": ")
end
end
def self.underscores_in_usernames(markdown)
return markdown unless markdown.match?(USERNAME_WITH_UNDERSCORE_REGEXP)
traverser = MarkdownProcessor::Traverser.new(markdown)
traverser.each do |line|
next if traverser.in_codeblock?
escape_underscored_username_in_line!(line)
end.join
end
# Match @_username_ that is not preceded by backtick
USERNAME_WITH_UNDERSCORE_REGEXP = /(?<!`)@_\w+_/
# Escapes underscored username that is not in code
def self.escape_underscored_username_in_line!(line)
line.scan(USERNAME_WITH_UNDERSCORE_REGEXP).each do |to_escape|
line.sub!(to_escape, to_escape.gsub("_", "\\_"))
end
line
end
def self.add_quotes_to_section(markdown, section:)
# Only add quotes to front matter, or text between triple dashes
markdown.sub(FRONT_MATTER_DETECTOR) do |front_matter|
front_matter.gsub(/#{section}: ?(?<content>.*?)(\r\n|\n)/m) do |target|
# `content` is the captured group (.*?)
captured_text = Regexp.last_match("content")
# The query below checks if the whole text is wrapped in
# either single or double quotes.
match = captured_text.scan(/(^".*"$|^'.*'$)/)
if match.empty?
# Double quotes that aren't already escaped will get esacped.
# Then the whole text get warped in double quotes.
parsed_text = captured_text.gsub(/(?<!\\)"/, "\\\"")
"#{section}: \"#{parsed_text}\"\n"
else
# if the text comes pre-warped in either single or double quotes,
# no more processing is done
target
end
end
end
end
end
end
end