[scripts] add Markdown to HTML document generator script

This commit is contained in:
Dmytro Meleshko 2019-12-05 22:50:38 +02:00
parent 38452b72b7
commit 6f3eb3b40e

188
scripts/markdown2htmldoc Executable file
View file

@ -0,0 +1,188 @@
#!/usr/bin/env ruby
require "commonmarker"
require "rouge"
ROUGE_THEME = Rouge::Themes::Pastie
COMMONMARKER_PARSE_OPTIONS = [:DEFAULT, :FOOTNOTES]
COMMONMARKER_RENDER_OPTIONS = [:UNSAFE]
COMMONMARKER_EXTENSIONS = [:tagfilter, :autolink, :table, :strikethrough, :tasklist]
# parse arguments {{{
def get_program_name()
return File.basename($PROGRAM_NAME)
end
def print_help()
$stderr.print <<EOF
usage: #{get_program_name()} [<input_file>] [<output_file>]
if <input_file> is not specified, input will be read from stdin
if <output_file> is not specified, output will be printed to stdout
EOF
exit(true)
end
def print_error(*args)
$stderr.puts("#{get_program_name()}: error:", *args)
exit(false)
end
input_file_path = nil
output_file_path = nil
reading_only_positional_args = false
positional_arg_index = 0
ARGV.each do |arg|
is_positional = reading_only_positional_args
if not is_positional
case arg
when "-h", "--help"
print_help()
when "--"
reading_only_positional_args = true
else
if arg.start_with?("-")
print_error("unknown option: #{arg}")
else
is_positional = true
end
end
end
if is_positional
case positional_arg_index
when 0 then input_file_path = arg
when 1 then output_file_path = arg
else print_error("unexpected argument: #{arg}")
end
positional_arg_index += 1
end
end
# }}}
# parse markdown document {{{
if input_file_path.nil?()
markdown = $stdin.read()
else
markdown = IO.read(input_file_path)
end
doc = CommonMarker.render_doc(markdown, COMMONMARKER_PARSE_OPTIONS, COMMONMARKER_EXTENSIONS)
# }}}
# add header anchors {{{
header_slug_occurences = {}
doc.walk do |node|
if node.type == :header
header_slug = ""
node.walk do |child_node|
if [:text, :html, :html_inline, :code, :code_block, :footnote_reference].include?(child_node.type)
header_slug += child_node.string_content
end
end
header_slug.downcase!()
header_slug.strip!()
# remove special characters
header_slug.delete!("\u2000-\u206F\u2E00-\u2E7F\\\\'!\"#$%&()*+,./:;<=>?@[]\\^`{|}~")
# remove emoji
header_slug.delete!("\u{1F600}-\u{1F6FF}")
# remove whitespace
header_slug.gsub!(/\s/, "-")
# make this slug unique
while header_slug_occurences.key?(header_slug)
occurences = header_slug_occurences[header_slug]
occurences += 1
header_slug_occurences[header_slug] = occurences
header_slug += "-" + occurences.to_s()
end
header_slug_occurences[header_slug] = 0
anchor_node = CommonMarker::Node.new(:inline_html)
anchor_node.string_content = "<a class=\"anchor\" name=\"#{header_slug}\" href=\"\##{header_slug}\"><span class=\"octicon octicon-link\"></span></a>"
node.prepend_child(anchor_node)
end
end
# }}}
# highlight code blocks {{{
rouge_formatter = Rouge::Formatters::HTML.new()
doc.walk do |node|
if node.type == :code_block
language = node.fence_info
if not language.empty?()
source = node.string_content
lexer = Rouge::Lexer.find_fancy(language) || Rouge::Lexers::PlainText.new()
highlighted_code = rouge_formatter.format(lexer.lex(source))
new_node = CommonMarker::Node.new(:html)
new_node.string_content = "<pre><code class=\"highlight highlight-source-#{new_node.html_escape_html(language)}\">#{highlighted_code}</code></pre>"
node.insert_after(new_node)
node.delete
end
end
end
# }}}
# render HTML {{{
rendered_html = <<EOF
<!DOCTYPE html>
<html>
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta http-equiv="X-UA-Compatible" content="ie=edge">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/github-markdown-css/3.0.1/github-markdown.min.css" integrity="sha256-HbgiGHMLxHZ3kkAiixyvnaaZFNjNWLYKD/QG6PWaQPc=" crossorigin="anonymous" />
<style>
html, body {
padding: 0;
margin: 0;
}
.markdown-body {
max-width: 882px;
margin: 0 auto;
padding: 32px;
}
.octicon-link {
font: normal normal 16px 'octicons-link';
line-height: 1;
-webkit-font-smoothing: antialiased;
-moz-osx-font-smoothing: grayscale;
-webkit-user-select: none;
-moz-user-select: none;
-ms-user-select: none;
user-select: none;
}
.octicon-link::before {
content: '\\f05c';
}
#{ROUGE_THEME.render()}
</style>
</head>
<body>
<article class="markdown-body">
#{doc.to_html(COMMONMARKER_RENDER_OPTIONS)}
</article>
</body>
</html>
EOF
if output_file_path.nil?()
$stdout.write(rendered_html)
else
IO.write(output_file_path, rendered_html)
end
# }}}