]> git.madduck.net Git - etc/mutt.git/blob - .mutt/markdown2html

madduck's git repository

Every one of the projects in this repository is available at the canonical URL git://git.madduck.net/madduck/pub/<projectpath> — see each project's metadata for the exact URL.

All patches and comments are welcome. Please squash your changes to logical commits before using git-format-patch and git-send-email to patches@git.madduck.net. If you'd read over the Git project's submission guidelines and adhered to them, I'd be especially grateful.

SSH access, as well as push access can be individually arranged.

If you use my repositories frequently, consider adding the following snippet to ~/.gitconfig and using the third clone URL listed for each project:

[url "git://git.madduck.net/madduck/"]
  insteadOf = madduck:

7853ae59ca5448e86833f036553bd0094b8d4b4a
[etc/mutt.git] / .mutt / markdown2html
1 #!/usr/bin/python3
2 #
3 # markdown2html.py — simple Markdown-to-HTML converter for use with Mutt
4 #
5 # Mutt recently learnt [how to compose `multipart/alternative`
6 # emails][1]. This script assumes a message has been composed using Markdown
7 # (with a lot of pandoc extensions enabled), and translates it to `text/html`
8 # for Mutt to tie into such a `multipart/alternative` message.
9 #
10 # [1]: https://gitlab.com/muttmua/mutt/commit/0e566a03725b4ad789aa6ac1d17cdf7bf4e7e354)
11 #
12 # Configuration:
13 #   muttrc:
14 #     set send_multipart_alternative=yes
15 #     set send_multipart_alternative_filter=/path/to/markdown2html.py
16 #
17 # Optionally, Custom CSS styles will be read from `~/.mutt/markdown2html.css`,
18 # if present.
19 #
20 # Requirements:
21 #   - python3
22 #   - PyPandoc (and pandoc installed, or downloaded)
23 #   - Pynliner
24 #
25 # Optional:
26 #   - Pygments, if installed, then syntax highlighting is enabled
27 #
28 # Latest version:
29 #   https://git.madduck.net/etc/mutt.git/blob_plain/HEAD:/.mutt/markdown2html
30 #
31 # Copyright © 2019 martin f. krafft <madduck@madduck.net>
32 # Released under the GPL-2+ licence, just like Mutt itself.
33 #
34
35 import pypandoc
36 import pynliner
37 import re
38 import os
39 import sys
40
41 try:
42     from pygments.formatters import get_formatter_by_name
43     formatter = get_formatter_by_name('html', style='default')
44     DEFAULT_CSS = formatter.get_style_defs('.sourceCode')
45
46 except ImportError:
47     DEFAULT_CSS = ""
48
49
50 DEFAULT_CSS += '''
51 .quote {
52     padding: 0 0.5em;
53     margin: 0;
54     font-style: italic;
55     border-left: 2px solid #ccc;
56     color: #999;
57     font-size: 80%;
58 }
59 .quotelead {
60     font-style: italic;
61     margin-bottom: -1em;
62     color: #999;
63     font-size: 80%;
64 }
65 .quotechar { display: none; }
66 .footnote-ref, .footnote-back { text-decoration: none;}
67 .signature {
68     color: #999;
69     font-family: monospace;
70     white-space: pre;
71     margin: 1em 0 0 0;
72     font-size: 80%;
73 }
74 table, th, td {
75     border-collapse: collapse;
76     border: 1px solid #999;
77 }
78 th, td { padding: 0.5em; }
79 .header {
80     background: #eee;
81 }
82 .even { background: #eee; }
83 '''
84
85 STYLESHEET = os.path.join(os.path.expanduser('~/.mutt'),
86                           'markdown2html.css')
87 if os.path.exists(STYLESHEET):
88     DEFAULT_CSS += open(STYLESHEET).read()
89
90 HTML_DOCUMENT = '''<!DOCTYPE html>
91 <html><head>
92 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
93 <meta charset="utf-8"/>
94 <meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes"/>
95 <title>HTML E-Mail</title>
96 </head><body class="email">
97 {htmlbody}
98 </body></html>'''
99
100
101 SIGNATURE_HTML = \
102         '<div class="signature"><span class="leader">-- </span>{sig}</div>'
103
104
105 def _preprocess_markdown(mdwn):
106     '''
107     Preprocess Markdown for handling by the converter.
108     '''
109     # convert hard line breaks within paragraphs to 2 trailing spaces, which
110     # is the markdown way of representing hard line breaks. Note how the
111     # regexp will not match between paragraphs.
112     ret = re.sub(r'(\S)\n(\s*\S)', r'\g<1>  \n\g<2>', mdwn, flags=re.MULTILINE)
113
114     # Clients like Thunderbird need the leading '>' to be able to properly
115     # create nested quotes, so we duplicate the symbol, the first instance
116     # will tell pandoc to create a blockquote, while the second instance will
117     # be a <span> containing the character, along with a class that causes CSS
118     # to actually hide it from display. However, this does not work with the
119     # text-mode HTML2text converters, and so it's left commented for now.
120     #ret = re.sub(r'\n>', r'  \n>[>]{.quotechar}', ret, flags=re.MULTILINE)
121
122     return ret
123
124
125 def _identify_quotes_for_later(mdwn):
126     '''
127     Email quoting such as:
128
129     ```
130     On 1970-01-01, you said:
131     > The Flat Earth Society has members all around the globe.
132     ```
133
134     isn't really properly handled by Markdown, so let's do our best to
135     identify the individual elements, and mark them, using a syntax similar to
136     what pandoc uses already in some cases. As pandoc won't actually use these
137     data (yet?), we call `self._reformat_quotes` later to use these markers
138     to slap the appropriate classes on the HTML tags.
139     '''
140
141     def generate_lines_with_context(mdwn):
142         '''
143         Iterates the input string line-wise, returning a triplet of
144         previous, current, and next line, the first and last of which
145         will be None on the first and last line of the input data
146         respectively.
147         '''
148         prev = cur = nxt = None
149         lines = iter(mdwn.splitlines())
150         cur = next(lines)
151         for nxt in lines:
152             yield prev, cur, nxt
153             prev = cur
154             cur = nxt
155         yield prev, cur, None
156
157     ret = []
158     for prev, cur, nxt in generate_lines_with_context(mdwn):
159
160         # The lead-in to a quote is a single line immediately preceding the
161         # quote, and ending with ':'. Note that there could be multiple of
162         # these:
163         if re.match(r'^.+:\s*$', cur) and nxt.startswith('>'):
164             ret.append(f'{{.quotelead}}{cur.strip()}')
165             # pandoc needs an empty line before the blockquote, so
166             # we enter one for the purpose of HTML rendition:
167             ret.append('')
168             continue
169
170         # The first blockquote after such a lead-in gets marked as the
171         # "initial" quote:
172         elif prev and re.match(r'^.+:\s*$', prev) and cur.startswith('>'):
173             ret.append(re.sub(r'^(\s*>\s*)+(.+)',
174                               r'\g<1>{.quoteinitial}\g<2>',
175                               cur, flags=re.MULTILINE))
176
177         # All other occurrences of blockquotes get the "subsequent" marker:
178         elif cur.startswith('>') and prev and not prev.startswith('>'):
179             ret.append(re.sub(r'^((?:\s*>\s*)+)(.+)',
180                               r'\g<1>{.quotesubsequent}\g<2>',
181                               cur, flags=re.MULTILINE))
182
183         else: # pass through everything else.
184             ret.append(cur)
185
186     return '\n'.join(ret)
187
188
189 def _reformat_quotes(html):
190     '''
191     Earlier in the pipeline, we marked email quoting, using markers, which we
192     now need to turn into HTML classes, so that we can use CSS to style them.
193     '''
194     ret = html.replace('<p>{.quotelead}', '<p class="quotelead">')
195     ret = re.sub(r'<blockquote>\n((?:<blockquote>\n)*)<p>(?:\{\.quote(\w+)\})',
196                  r'<blockquote class="quote \g<2>">\n\g<1><p>', ret, flags=re.MULTILINE)
197     return ret
198
199
200
201 def _convert_with_pandoc(mdwn, inputfmt='markdown', outputfmt='html5',
202                          ext_enabled=None, ext_disabled=None,
203                          standalone=True, title="HTML E-Mail"):
204     '''
205     Invoke pandoc to do the actual conversion of Markdown to HTML5.
206     '''
207     if not ext_enabled:
208         ext_enabled = [ 'backtick_code_blocks',
209                        'line_blocks',
210                        'fancy_lists',
211                        'startnum',
212                        'definition_lists',
213                        'example_lists',
214                        'table_captions',
215                        'simple_tables',
216                        'multiline_tables',
217                        'grid_tables',
218                        'pipe_tables',
219                        'all_symbols_escapable',
220                        'intraword_underscores',
221                        'strikeout',
222                        'superscript',
223                        'subscript',
224                        'fenced_divs',
225                        'bracketed_spans',
226                        'footnotes',
227                        'inline_notes',
228                        'emoji',
229                        'tex_math_double_backslash',
230                       ]
231     if not ext_disabled:
232         ext_disabled = [ 'tex_math_single_backslash',
233                          'tex_math_dollars',
234                          'smart',
235                          'raw_html'
236                        ]
237
238     enabled = '+'.join(ext_enabled)
239     disabled = '-'.join(ext_disabled)
240     inputfmt = f'{inputfmt}+{enabled}-{disabled}'
241
242     args = []
243     if standalone:
244         args.append('--standalone')
245     if title:
246         args.append(f'--metadata=pagetitle:"{title}"')
247
248     return pypandoc.convert_text(mdwn, format=inputfmt, to=outputfmt,
249                                  extra_args=args)
250
251
252 def _apply_styling(html):
253     '''
254     Inline all styles defined and used into the individual HTML tags.
255     '''
256     return pynliner.Pynliner().from_string(html).with_cssString(DEFAULT_CSS).run()
257
258
259 def _postprocess_html(html):
260     '''
261     Postprocess the generated and styled HTML.
262     '''
263     return html
264
265
266 def convert_markdown_to_html(mdwn):
267     '''
268     Converts the input Markdown to HTML, handling separately the body, as well
269     as an optional signature.
270     '''
271     parts = re.split(r'^-- $', mdwn, 1, flags=re.MULTILINE)
272     body = parts[0]
273     if len(parts) == 2:
274         sig = parts[1]
275     else:
276         sig = None
277
278     html=''
279     if body:
280         body = _preprocess_markdown(body)
281         body = _identify_quotes_for_later(body)
282         html = _convert_with_pandoc(body, standalone=False)
283         html = _reformat_quotes(html)
284
285     if sig:
286         sig = _preprocess_markdown(sig)
287         html += SIGNATURE_HTML.format(sig='<br/>'.join(sig.splitlines()))
288
289     html = HTML_DOCUMENT.format(htmlbody=html)
290     html = _apply_styling(html)
291     html = _postprocess_html(html)
292
293     return html
294
295
296 def main():
297     '''
298     Convert text on stdin to HTML, and print it to stdout, like mutt would
299     expect.
300     '''
301     html = convert_markdown_to_html(sys.stdin.read())
302     if html:
303         # mutt expects the content type in the first line, so:
304         print(f'text/html\n\n{html}')
305
306
307 if __name__ == '__main__':
308     main()