145 lines
3.2 KiB
Ruby
145 lines
3.2 KiB
Ruby
class Markdowner
|
|
MAX_BARE_LINK = 50
|
|
|
|
def self.h(str)
|
|
# .to_str is needed because this becomes a SafeBuffer, which breaks gsub
|
|
# https://github.com/rails/rails/issues/1555
|
|
ERB::Util.h(str).to_str.gsub(/<(\/?)(em|u|strike)>/, '<\1\2>')
|
|
end
|
|
|
|
def self.markdown(string)
|
|
lines = string.rstrip.split(/\r?\n/)
|
|
|
|
out = "<p>"
|
|
inpre = false
|
|
lines.each do |line|
|
|
# [ ][ ]blah -> <pre> blah</pre>
|
|
if line.match(/^( |\t)/)
|
|
if !inpre
|
|
out << "<p><pre>"
|
|
end
|
|
|
|
out << ERB::Util.h(line) << "\n"
|
|
inpre = true
|
|
next
|
|
elsif inpre
|
|
out << "</pre></p>\n<p>"
|
|
inpre = false
|
|
end
|
|
|
|
line = self.h(line)
|
|
|
|
# lines starting with > are quoted
|
|
if m = line.match(/^>(.*)/)
|
|
line = "<blockquote> " << m[1] << " </blockquote>"
|
|
end
|
|
|
|
lead = '\A|\s|[><]'
|
|
trail = '[<>]|\z|\s'
|
|
|
|
# *text* -> <em>text</em>
|
|
line.gsub!(/(#{lead}|_|~)\*([^\* \t][^\*]*)\*(#{trail}|_|~)/) do |m|
|
|
"#{$1}<em>" << self.h($2) << "</em>#{$3}"
|
|
end
|
|
|
|
# _text_ -> <u>text</u>
|
|
line.gsub!(/(#{lead}|~)_([^_ \t][^_]*)_(#{trail}|~)/) do |m|
|
|
"#{$1}<u>" << self.h($2) << "</u>#{$3}"
|
|
end
|
|
|
|
# ~~text~~ -> <strike>text</strike> (from reddit)
|
|
line.gsub!(/(#{lead})\~\~([^~ \t][^~]*)\~\~(#{trail})/) do |m|
|
|
"#{$1}<strike>" << self.h($2) << "</strike>#{$3}"
|
|
end
|
|
|
|
# [link text](http://url) -> <a href="http://url">link text</a>
|
|
line.gsub!(/(#{lead})\[([^\]]+)\]\((http(s?):\/\/[^\)]+)\)(#{trail})/i) do |m|
|
|
"#{$1}<a href=\"" << self.h($3) << "\" rel=\"nofollow\">" <<
|
|
self.h($2) << "</a>#{$5}"
|
|
end
|
|
|
|
# find bare links that are not inside tags
|
|
|
|
# http://blah -> <a href=...>
|
|
chunk = ""
|
|
intag = false
|
|
outline = ""
|
|
line.each_byte do |n|
|
|
c = n.chr
|
|
|
|
if intag
|
|
outline << c
|
|
|
|
if c == ">"
|
|
intag = false
|
|
next
|
|
end
|
|
else
|
|
if c == "<"
|
|
if chunk != ""
|
|
outline << Markdowner._linkify_text(chunk)
|
|
end
|
|
|
|
chunk = ""
|
|
intag = true
|
|
outline << c
|
|
else
|
|
chunk << c
|
|
end
|
|
end
|
|
end
|
|
|
|
if chunk != ""
|
|
outline << Markdowner._linkify_text(chunk)
|
|
end
|
|
|
|
out << outline << "<br>\n"
|
|
end
|
|
|
|
if inpre
|
|
out << "</pre>"
|
|
end
|
|
|
|
out << "</p>"
|
|
|
|
# multiple br's into a p
|
|
out.gsub!(/<br>\n?<br>\n?/, "</p><p>")
|
|
|
|
# collapse things
|
|
out.gsub!(/<br>\n?<\/p>/, "</p>\n")
|
|
out.gsub!(/<p>\n?<br>\n?/, "<p>")
|
|
out.gsub!(/<p>\n?<\/p>/, "\n")
|
|
out.gsub!(/<p>\n?<p>/, "\n<p>")
|
|
out.gsub!(/<\/p><p>/, "</p>\n<p>")
|
|
|
|
out.strip
|
|
end
|
|
|
|
def self._linkify_text(chunk)
|
|
chunk.gsub(/
|
|
(\A|\s|[:,])
|
|
(http(s?):\/\/|www\.)
|
|
([^\s]+)/ix) do |m|
|
|
pre = $1
|
|
host_and_path = "#{$2 == "www." ? $2 : ""}#{$4}"
|
|
post = $5
|
|
|
|
# remove some chars that might be with a url at the end but aren't
|
|
# actually part of the url
|
|
if m = host_and_path.match(/(.*)([,\?;\!\.]+)\z/)
|
|
host_and_path = m[1]
|
|
post = "#{m[2]}#{post}"
|
|
end
|
|
|
|
url = "http#{$3}://#{host_and_path}"
|
|
url_text = host_and_path
|
|
|
|
if url_text.length > 50
|
|
url_text = url_text[0 ... 50] << "..."
|
|
end
|
|
|
|
"#{pre}<a href=\"#{url}\" rel=\"nofollow\">#{url_text}</a>#{post}"
|
|
end
|
|
end
|
|
end
|