Class: MarkdownLint::Doc

Inherits:
Object
  • Object
show all
Defined in:
lib/mdl/doc.rb

Overview

Representation of the markdown document passed to rule checks

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(text, ignore_front_matter = false) ⇒ Doc

Create a new document given a string containing the markdown source


29
30
31
32
33
34
35
36
37
38
39
40
41
# File 'lib/mdl/doc.rb', line 29

def initialize(text, ignore_front_matter = false)
  regex = /^---\n(.*?)---\n\n?/m
  if ignore_front_matter && regex.match(text)
    @offset = regex.match(text).to_s.split("\n").length
    text.sub!(regex, '')
  else
    @offset = 0
  end
  @lines = text.split(/\R/)
  @parsed = Kramdown::Document.new(text, :input => 'MarkdownLint')
  @elements = @parsed.root.children
  add_annotations(@elements)
end

Instance Attribute Details

#elementsObject (readonly)

A list of raw markdown source lines. Note that the list is 0-indexed, while line numbers in the parsed source are 1-indexed, so you need to subtract 1 from a line number to get the correct line. The element_line* methods take care of this for you.


14
15
16
# File 'lib/mdl/doc.rb', line 14

def elements
  @elements
end

#linesObject (readonly)

A list of raw markdown source lines. Note that the list is 0-indexed, while line numbers in the parsed source are 1-indexed, so you need to subtract 1 from a line number to get the correct line. The element_line* methods take care of this for you.


14
15
16
# File 'lib/mdl/doc.rb', line 14

def lines
  @lines
end

#offsetObject (readonly)

A list of raw markdown source lines. Note that the list is 0-indexed, while line numbers in the parsed source are 1-indexed, so you need to subtract 1 from a line number to get the correct line. The element_line* methods take care of this for you.


14
15
16
# File 'lib/mdl/doc.rb', line 14

def offset
  @offset
end

#parsedObject (readonly)

A list of raw markdown source lines. Note that the list is 0-indexed, while line numbers in the parsed source are 1-indexed, so you need to subtract 1 from a line number to get the correct line. The element_line* methods take care of this for you.


14
15
16
# File 'lib/mdl/doc.rb', line 14

def parsed
  @parsed
end

Class Method Details

.new_from_file(filename, ignore_front_matter = false) ⇒ Object

Alternate 'constructor' passing in a filename


46
47
48
49
50
51
52
# File 'lib/mdl/doc.rb', line 46

def self.new_from_file(filename, ignore_front_matter = false)
  if filename == '-'
    new($stdin.read, ignore_front_matter)
  else
    new(File.read(filename, :encoding => 'UTF-8'), ignore_front_matter)
  end
end

Instance Method Details

#element_line(element) ⇒ Object

Returns the actual source line for a given element. You can pass in an element object or an options hash here. This is useful if you need to examine the source line directly for your rule to make use of information that isn't present in the parsed document.


133
134
135
# File 'lib/mdl/doc.rb', line 133

def element_line(element)
  @lines[element_linenumber(element) - 1]
end

#element_linenumber(element) ⇒ Object

Returns the line number a given element is located on in the source file. You can pass in either an element object or an options hash here.


122
123
124
125
# File 'lib/mdl/doc.rb', line 122

def element_linenumber(element)
  element = element.options if element.is_a?(Kramdown::Element)
  element[:location]
end

#element_linenumbers(elements) ⇒ Object

Returns a list of line numbers for all elements passed in. You can pass in a list of element objects or a list of options hashes here.


141
142
143
# File 'lib/mdl/doc.rb', line 141

def element_linenumbers(elements)
  elements.map { |e| element_linenumber(e) }
end

#element_lines(elements) ⇒ Object

Returns the actual source lines for a list of elements. You can pass in a list of elements objects or a list of options hashes here.


149
150
151
# File 'lib/mdl/doc.rb', line 149

def element_lines(elements)
  elements.map { |e| element_line(e) }
end

#extract_text(element, prefix = '', restore_whitespace = true) ⇒ Object

Extracts the text from an element whose children consist of text elements and other things


243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
# File 'lib/mdl/doc.rb', line 243

def extract_text(element, prefix = '', restore_whitespace = true)
  quotes = {
    :rdquo => '"',
    :ldquo => '"',
    :lsquo => "'",
    :rsquo => "'",
  }
  # If anything goes amiss here, e.g. unknown type, then nil will be
  # returned and we'll just not catch that part of the text, which seems
  # like a sensible failure mode.
  lines = element.children.map do |e|
    if e.type == :text
      e.value
    elsif %i{strong em p codespan}.include?(e.type)
      extract_text(e, prefix, restore_whitespace).join("\n")
    elsif e.type == :smart_quote
      quotes[e.value]
    end
  end.join.split("\n")
  # Text blocks have whitespace stripped, so we need to add it back in at
  # the beginning. Because this might be in something like a blockquote,
  # we optionally strip off a prefix given to the function.
  lines[0] = element_line(element).sub(prefix, '') if restore_whitespace
  lines
end

#find_type(type, nested = true) ⇒ Object

Find all elements of a given type, returning their options hash. The options hash has most of the useful data about an element and often you can just use this in your rules.

# Returns [ { :location => 1, :element_level => 2 }, ... ]
elements = find_type(:li)

If nested is set to false, this returns only top level elements of a given type.


65
66
67
# File 'lib/mdl/doc.rb', line 65

def find_type(type, nested = true)
  find_type_elements(type, nested).map(&:options)
end

#find_type_elements(type, nested = true, elements = @elements) ⇒ Object

Find all elements of a given type, returning a list of the element objects themselves.

Instead of a single type, a list of types can be provided instead to find all types.

If nested is set to false, this returns only top level elements of a given type.


79
80
81
82
83
84
85
86
87
88
89
# File 'lib/mdl/doc.rb', line 79

def find_type_elements(type, nested = true, elements = @elements)
  results = []
  type = [type] if type.class == Symbol
  elements.each do |e|
    results.push(e) if type.include?(e.type)
    if nested && !e.children.empty?
      results.concat(find_type_elements(type, nested, e.children))
    end
  end
  results
end

#find_type_elements_except(type, nested_except = [], elements = @elements) ⇒ Object

A variation on find_type_elements that allows you to skip drilling down into children of specific element types.

Instead of a single type, a list of types can be provided instead to find all types.

Unlike find_type_elements, this method will always search for nested elements, and skip the element types given to nested_except.


101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
# File 'lib/mdl/doc.rb', line 101

def find_type_elements_except(
  type, nested_except = [], elements = @elements
)
  results = []
  type = [type] if type.class == Symbol
  nested_except = [nested_except] if nested_except.class == Symbol
  elements.each do |e|
    results.push(e) if type.include?(e.type)
    next if nested_except.include?(e.type) || e.children.empty?

    results.concat(
      find_type_elements_except(type, nested_except, e.children),
    )
  end
  results
end

#header_style(header) ⇒ Object

Returns the header 'style' - :atx (hashes at the beginning), :atx_closed (atx header style, but with hashes at the end of the line also), :setext (underlined). You can pass in the element object or an options hash here.


159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
# File 'lib/mdl/doc.rb', line 159

def header_style(header)
  if header.type != :header
    raise 'header_style called with non-header element'
  end

  line = element_line(header)
  if line.start_with?('#')
    if line.strip.end_with?('#')
      :atx_closed
    else
      :atx
    end
  else
    :setext
  end
end

#indent_for(line) ⇒ Object

Returns how much a given line is indented. Hard tabs are treated as an indent of 8 spaces. You need to pass in the raw string here.


204
205
206
# File 'lib/mdl/doc.rb', line 204

def indent_for(line)
  line.match(/^\s*/)[0].gsub("\t", ' ' * 8).length
end

#list_style(item) ⇒ Object

Returns the list style for a list: :asterisk, :plus, :dash, :ordered or :ordered_paren depending on which symbol is used to denote the list item. You can pass in either the element itself or an options hash here.


181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
# File 'lib/mdl/doc.rb', line 181

def list_style(item)
  raise 'list_style called with non-list element' if item.type != :li

  line = element_line(item).strip.gsub(/^>\s+/, '')
  if line.start_with?('*')
    :asterisk
  elsif line.start_with?('+')
    :plus
  elsif line.start_with?('-')
    :dash
  elsif line.match('[0-9]+\.')
    :ordered
  elsif line.match('[0-9]+\)')
    :ordered_paren
  else
    :unknown
  end
end

#matching_lines(regex) ⇒ Object

Returns line numbers for lines that match the given regular expression


211
212
213
214
215
216
# File 'lib/mdl/doc.rb', line 211

def matching_lines(regex)
  @lines.each_with_index.select { |text, _linenum| regex.match(text) }
        .map do |i|
          i[1] + 1
        end
end

#matching_text_element_lines(regex, exclude_nested = [:a]) ⇒ Object

Returns line numbers for lines that match the given regular expression. Only considers text inside of 'text' elements (i.e. regular markdown text and not code/links or other elements).


222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
# File 'lib/mdl/doc.rb', line 222

def matching_text_element_lines(regex, exclude_nested = [:a])
  matches = []
  find_type_elements_except(:text, exclude_nested).each do |e|
    first_line = e.options[:location]
    # We'll error out if kramdown doesn't have location information for
    # the current element. It's better to just not match in these cases
    # rather than crash.
    next if first_line.nil?

    lines = e.value.split("\n")
    lines.each_with_index do |l, i|
      matches << first_line + i if regex.match(l)
    end
  end
  matches
end