iankelling.org Git - iankelling.org/blob - b.rb

   1 # encoding: utf-8
   2 # Copyright (C) 2016 Ian Kelling
   3
   4 # This program is free software: you can redistribute it and/or modify
   5 # it under the terms of the GNU General Public License as published by
   6 # the Free Software Foundation, either version 2 of the License, or
   7 # (at your option) any later version.
   8
   9 # This program is distributed in the hope that it will be useful,
  10 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 # GNU General Public License for more details.
  13
  14 # You should have received a copy of the GNU General Public License
  15 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
  16 module B # blog module
  17   require 'fileutils'
  18   require 'time'
  19   require 'safe_yaml'
  20   require 'pygments'
  21   require 'sqlite3'
  22   require 'redcarpet'
  23   JS_INFO = "<p>All JavaScript has <a href=\"https://www.gnu.org/software/librejs/index.html\">LibreJS</a> support.</p>"
  24
  25   DAY = 60*60*24
  26   DN = 'iankelling'
  27   FQDN = DN + '.org'
  28   DURL = 'https://' + FQDN
  29   DESCRIPTION = "Ian Kelling's personal site and blog on software"
  30   DATE_LEN = 'YYYY-MM-DD'.length
  31   NOW = Time.now.to_f
  32   WAIT_DATE = NOW - 60*60*24*1
  33
  34   def db_init
  35     SQLite3::Database.new('../proposed-comments/comments.sqlite')
  36   end
  37
  38   # from the redcarpet readme, then a bunch of googling to figure
  39   # out what to do on exception.
  40   class HTMLwithPygments < Redcarpet::Render::HTML
  41     def block_code(code, language)
  42       begin
  43         Pygments.highlight(code, lexer: language)
  44       rescue MentosError
  45         # when language detection fails
  46         Pygments.highlight(code, lexer: 'text')
  47       end
  48     end
  49   end
  50
  51   def fwrite(output_path, string)
  52     output_path = File.join('./', output_path)
  53     FileUtils.mkdir_p(File.dirname(output_path))
  54     File.write(output_path, string)
  55   end
  56
  57   def fskel(rel_path, title, content, o={})
  58     head = <<EOF
  59 <link rel="canonical" href="#{DURL}/#{rel_path}">
  60 EOF
  61     if rel_path =~ %r{^/blog/|^blog.html}
  62       head += <<EOF
  63 <link rel="alternate" type="application/atom+xml" title="#{DN}" href="#{DURL}/feed.xml">
  64 EOF
  65     end
  66     o[:head] = head
  67     fwrite(rel_path, skel(title, content, o))
  68   end
  69   def skel(title, content, o={})
  70     # got meta viewport from jekyll's default later. It's for better
  71     # mobile viewing.
  72     output = <<EOF
  73 <!DOCTYPE html>
  74 <html lang="en-US">
  75   <head>
  76     <meta charset="utf-8">
  77     <title>#{title}</title>
  78     <link rel="stylesheet" href="/css/main.css">
  79     <link rel="shortcut icon" href="/assets/favicon.png" />
  80     <meta name="description" content="#{o[:description] || DESCRIPTION}">
  81 #{o[:head]}
  82     <meta name="viewport" content="width=device-width, initial-scale=1">
  83   </head>
  84   <body>
  85     <header class="page_header">
  86       <h3><a href="/">iankelling.org</a>#{o[:header]}</h3>
  87     </header>
  88     <div class="main-content-stripe">
  89       <div class="#{o[:prose] ? "prose" : "content"}">
  90         #{content}
  91       </div>
  92     </div>
  93     <div class="comment-stripe">
  94 #{o[:comments]}
  95     </div>
  96     <footer>
  97 #{o[:footer]}
  98       <p>Sources in <a href="/git/?p=iankelling.org;a=summary">git</a>. Mostly markdown files. Default is <a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/"><img class="cc-by-sa" alt="Creative Commons License" src="/assets/cc-by-sa-4.0-80x15.png" /></a></p>
  99       <p><address><a href="mailto:ian@iankelling.org">ian@iankelling.org</a> let me know what you think</address></p>
 100     </footer>
 101   </body>
 102 </html>
 103 EOF
 104     output
 105   end
 106
 107   def stdpage(page_name, content)
 108     fskel("#{page_name}.html",
 109           "#{DN}/#{page_name}",
 110           content,
 111           header: " / <a href=\"/#{page_name}.html\">#{page_name}</a>")
 112   end
 113
 114   def md_to_html(md)
 115     # Using redcarpet over kramdown because syntax highlighting is
 116     # simpler. kramdown uses some crap highlighter by default,
 117     # supports using rouge, but then the classes are all screwy
 118     # for what pygments css expects, rouge has a pygments compatibility mode,
 119     # but that is a pita to get working, then it doesn't even work right.
 120     # kramdown is jekyll's default markdown parser, but it doesn't use
 121     # it for code blocks, it strips them out using custom templating
 122     # extension class, then uses rouge, then wraps it in some
 123     # custom html for pygments compatibility. It's a complicated mess.
 124     Redcarpet::Markdown.new(HTMLwithPygments, fenced_code_blocks: true).render(md)
 125   end
 126
 127   def comment_html(comment, date)
 128     # I tried putting the time,  %I:%M %p UTC, but it looks kinda
 129     # clunky, going against my simple theme.
 130     user_input = Redcarpet::Markdown.new(Redcarpet::Render::Safe,
 131                                          fenced_code_blocks: true).render(comment)
 132     <<EOF
 133 <div class="comment">
 134   #{user_input}
 135   <p class="comment-date">#{Time.at(date).strftime("%b %-d '%y")}</p>
 136 </div>
 137 EOF
 138   end
 139
 140
 141   def techpost(file)
 142
 143     b = File.basename(file,'.md')
 144     # double dash for one dash, single dash for space
 145     title = b.gsub(/--|-/, '--' => '-', '-' => ' ')
 146
 147     md = File.read(file)
 148     page_html = "<h1>#{title}</h1><b>Contents</b>"
 149     renderer = Redcarpet::Render::HTML_TOC.new(nesting_level: 2)
 150     page_html += Redcarpet::Markdown.new(renderer, fenced_code_blocks: true).render(md)
 151
 152     renderer = HTMLwithPygments.new(with_toc_data: true)
 153     page_html += Redcarpet::Markdown.new(renderer, fenced_code_blocks: true).render(md)
 154
 155     header_rel = ' / <a href="/technical-notes.html">technical notes</a> /'
 156     fskel("/technical-notes/#{b}.html", title, page_html,
 157           header: header_rel,
 158           prose: true)
 159
 160     technotes_index_entry = "<li><a href=\"/technical-notes/#{b}.html\">#{title}</a></li>"
 161     return technotes_index_entry
 162
 163   end
 164
 165   def post(file, build_time=false)
 166     content = File.read(file)
 167     content =~ %r{\A(---\s*\n.*?\n?)^((---)\s*$\n?)}m # yaml front matter
 168     # stuff after last match. jekyll uses $POSTMATCH,
 169     # but it's nil for me, I don't know what magic they are using.,
 170     # but only $' is listed here http://ruby-doc.org/core-2.3.1/doc/globals_rdoc.html,
 171     content = $'
 172
 173     front = SafeYAML.load(Regexp.last_match(1))
 174     title = front['title']
 175     $page_title = "#{title} | #{DN}"
 176     header_rel = ' / <a href="/blog.html">blog</a> /'
 177
 178     footer_extra = <<-EOF
 179 <p><a class="icon-rss" href="/feed.xml">Subscribe</a></p>
 180 EOF
 181     footer_extra += JS_INFO if content =~ /<script/
 182
 183
 184
 185     b = File.basename(file,'.md')
 186     # date is in the format: YYYY-MM-DD-
 187     date = Time.parse(b[0..DATE_LEN])
 188     rel_path = "/blog/#{b[(DATE_LEN + 1)..-1]}.html"
 189     comments = $db.execute <<-SQL, [WAIT_DATE]
 190       select comment, date from c
 191       where page = '#{rel_path}' and (
 192        state = 'picked' or state = 'known' or state = 'timed'
 193        or (state = 'waiting' and date < ?))
 194 SQL
 195     # get earliest comment. earlier ones stored in git will also be
 196     # published. This get's us easily sharable comments, and allows us
 197     # to expire unpublished comments and ip addresses which are PII and
 198     # should never be kept around indefinitely.
 199     sql_start_date = $db.execute('select min(date) from c')[0][0] || NOW
 200     comment_file_dir = "../comments/#{rel_path}"
 201     old_comments = Dir["#{comment_file_dir}/*"].reduce([]) do |memo, f|
 202       dt = File.basename(f).to_f
 203       if dt < sql_start_date
 204         memo << [File.read(f), dt]
 205       else
 206         FileUtils.rm(f) if build_time
 207         memo
 208       end
 209     end
 210     if build_time
 211       FileUtils.mkdir_p comment_file_dir
 212       comments.each do |c, c_date|
 213         # fyi: there is an extremely small chance of 2 comments having
 214         # the same floating point time and thus overwriting each other.
 215         # Small enough that it won't happen at my site's scale.
 216         File.write(File.join(comment_file_dir, c_date.to_s), c)
 217       end
 218       # Im slow at updating this site, it gets low traffic,
 219       # https://piwik.org/docs/privacy/ suggests 3-6 months, so
 220       # this cant be too bad.
 221       $db.execute("delete from c where date < #{NOW - DAY*180}")
 222     end
 223     comments = old_comments + comments
 224     pending_comments = $db.execute(<<-SQL, [WAIT_DATE])[0][0]
 225       select count(*) from c
 226       where page = '#{rel_path}' and
 227        (state = 'waiting' and date > ? or state = 'suspect')
 228 SQL
 229
 230     feed_html = md_to_html(content)
 231     page_html = <<-EOF
 232       <header class="post-header">
 233         <h1 class="post-title">#{title}</h1>
 234         <p class="post-date">#{date.strftime("%b %-d, %Y")}</p>
 235       </header>
 236       #{feed_html}
 237 EOF
 238     com_list = ''
 239     comments.each { |c, date| com_list += comment_html(c, date) }
 240     if pending_comments > 0
 241       if pending_comments >= 2
 242         text = "are #{pending_comments} new comments"
 243       else
 244         text = 'is 1 new comment'
 245       end
 246       com_list +=
 247         comment_html("Note: there #{text} pending approval.", NOW)
 248     end
 249     com_section = <<-EOF
 250       <form action="/cgi/comment" method="post">
 251         <input class="misc-comment-input" type="text" name="url">
 252         <input name="goto" type="hidden" value="#{rel_path}">
 253         <textarea rows="10" name="comment" maxlength="1000"></textarea>
 254         <input type="submit" value="Add a comment (markdown works)">
 255       </form>
 256       <div id="comments">
 257          #{com_list}
 258       </div>
 259 EOF
 260     links = front['comment_links']
 261     if links
 262       link_html = links.map { |name,url| "<a href=\"#{url}\">#{name}</a>" }
 263                     .join(', ')
 264       com_section += (<<EOF)
 265 <p>More comments at #{link_html}</p>
 266 EOF
 267     end
 268
 269     blog_toc_entry = "<li><a href=\"#{rel_path}\">#{title}</a></li>"
 270
 271     com_section = <<EOF
 272 <div id="comment-section" class="comment-section">
 273   #{com_section}
 274 </div>
 275 EOF
 276
 277
 278     if front['description']
 279       description = front['description']
 280     else
 281       # the first 300 saves ~ 1 ms
 282       # regex for striping html from liquid template src
 283       description = feed_html[0..300].gsub(/<script.*?<\/script>/m, '').
 284                       gsub(/<!--.*?-->/m, '').gsub(/<style.*?<\/style>/m, '').
 285                       gsub(/<.*?>/m, '')
 286       if description.length > 160
 287         description = description[0..156] + '...'
 288       end
 289     end
 290
 291     head = <<EOF
 292 EOF
 293
 294     fskel(rel_path, title, page_html,
 295           header: header_rel,
 296           footer: footer_extra,
 297           # We call the main content prose because it can contain
 298           # code, so we want to left justify it, so we dont have
 299           # to cram all the code into the middle and can use
 300           # a wider area.
 301           prose: true,
 302           comments: com_section,
 303           description: description)
 304     url="#{DURL}#{rel_path}"
 305
 306
 307     # following from https://creativecommons.org/choose,
 308     # with the addition of "unless otherwise noted", for js licenses.
 309     feed_copyright = <<-EOF
 310 <a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/"><img alt="Creative Commons License" style="border-width:0" src="https://i.creativecommons.org/l/by-sa/4.0/88x31.png" /></a><br /><span xmlns:dct="http://purl.org/dc/terms/" href="http://purl.org/dc/dcmitype/Text" property="dct:title" rel="dct:type">#{title}</span> by <a xmlns:cc="http://creativecommons.org/ns#" href="#{url}" property="cc:attributionName" rel="cc:attributionURL">Ian Kelling</a> unless otherwise noted is licensed under a <a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/">Creative Commons Attribution-ShareAlike 4.0 International License</a>.
 311 EOF
 312
 313     feed_entry = <<EOF
 314 <entry>
 315   <title>#{title}</title>
 316   <link rel="alternate" href="#{url}"/>
 317   <id>#{url}</id>
 318   <updated>#{date.to_datetime.rfc3339}</updated>
 319   <content type="html" xml:lang="en-us" xml:base="#{DURL}/blog">
 320     <![CDATA[
 321     #{feed_html}
 322     ]]>
 323   </content>
 324   <rights>
 325     #{feed_copyright}
 326   </rights>
 327 </entry>
 328 EOF
 329     return [feed_entry, blog_toc_entry]
 330   end
 331 end