Module: Zettacode::Parse

Defined in:
lib/zettacode/parse/parse.rb

Class Method Summary collapse

Class Method Details

.call(filepath) ⇒ Object 



12
13
14
15
16
17
 
# File 'lib/zettacode/parse/parse.rb', line 12

def self.call(filepath)
  raw_content, to_folder = read_content_from filepath
  dirty_examples = scrap_examples_from raw_content
  clean_examples = clean dirty_examples
  save_files clean_examples, to_folder
end

.clean(dirty_examples) ⇒ Object 



59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
 
# File 'lib/zettacode/parse/parse.rb', line 59

def self.clean(dirty_examples)
  examples = []
  dirty_examples.each do |example|
    # lang = /header\|([\w\d\s.\-_*\(\)]+)/.match(example[:lang])&.captures&.first
    lang = example[:lang][11, example[:lang].size - 15]
    filename = lang.tr(" ", "_").downcase
    filename.tr!("/", "-")
    syntax = nil
    lines = example[:code]
    raw = lines.join("\n")

    # Clean syntaxhighlight tag
    lines.each do |line|
      next if line == ""
      # <syntaxhighlight lang="LANG">
      filter = /<syntaxhighlight\s+lang="([\w\d]+)">/
      value = filter.match(line)&.captures&.first
      unless value.nil?
        syntax = value
        line.gsub!("<syntaxhighlight lang=\"#{value}\">", "")
      else
        line.gsub!("</text>", "")
        line.gsub!("</revision>", "")
        line.gsub!("</page>", "")
        line.gsub!("</mediawiki>", "")
        line = nil if line.strip.start_with? "<sha1>"
      end
    end

    code = lines.join("\n")
    code.gsub!("&lt;/syntaxhighlight&gt;", "")
    code.gsub!("&lt;pre&gt;", "")
    code.gsub!("&lt;/pre&gt;", "")
    code.gsub!("&lt;", "<")
    code.gsub!("&gt;", ">")
    code.gsub!("&lt;code&gt;", "'")
    code.gsub!("&lt;/code&gt;", "'")
    examples << {
      lang: lang,
      filename: filename,
      syntax: syntax,
      code: code,
      raw: raw
    }
  end
  examples
end

.echo(title, text) ⇒ Object 



7
8
9
10
 
# File 'lib/zettacode/parse/parse.rb', line 7

def self.echo(title, text)
  print title.rjust(12).colorize(:green)
  puts " #{text}"
end

.read_content_from(filepath) ⇒ Object 



19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
 
# File 'lib/zettacode/parse/parse.rb', line 19

def self.read_content_from(filepath)
  unless File.exist?(filepath)
    echo "ERROR", "File not found! (#{filepath})"
    exit 1
  end

  echo "Parsing", filepath
  content = File.read(filepath)
  title = /<title>([\w\d\s._-]+.[\w\d\s._-]+)<\/title>/
  name = title.match(content).captures.first
  name.tr!("/", ".")
  name.tr!(" ", "_")
  name.downcase!
  folder = File.join(Zettacode::OUTPUT_FOLDER, name)
  FileUtils.mkdir_p(folder) unless Dir.exist? folder
  [content, folder]
end

.save_files(examples, folder) ⇒ Object 



107
108
109
110
111
112
113
114
 
# File 'lib/zettacode/parse/parse.rb', line 107

def self.save_files(examples, folder)
  examples.each do |example|
    filepath = File.join(folder, "#{example[:filename]}.txt")
    File.open(filepath, "w") { |file| file.write(example[:code]) }
  end
  echo "Folder", folder
  echo "Saving", "#{examples.size} files"
end

.scrap_examples_from(content) ⇒ Object 



37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
 
# File 'lib/zettacode/parse/parse.rb', line 37

def self.scrap_examples_from(content)
  examples = []
  example = nil
  skip = true
  content.split("\n").each do |line|
    if line.start_with? "=={{header|"
      skip = false
      examples << example unless example.nil?
      example = {lang: line, code: []}
    elsif line.downcase.start_with? "{{out}}"
      skip = true
      next
    elsif skip
      next
    else
      example[:code] << line
    end
  end
  examples << example
  examples
end