]> 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:

md2html: introduce block class
[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 .block {
52     padding: 0 0.5em;
53     margin: 0;
54     border-left: 2px solid #eee;
55 }
56 .quote, blockquote {
57     padding: 0 0.5em;
58     margin: 0;
59     font-style: italic;
60     border-left: 2px solid #666;
61     color: #666;
62     font-size: 80%;
63 }
64 .quotelead {
65     margin-bottom: -1em;
66     font-size: 80%;
67 }
68 .quotechar { display: none; }
69 .footnote-ref, .footnote-back { text-decoration: none;}
70 .signature {
71     color: #999;
72     font-family: monospace;
73     white-space: pre;
74     margin: 1em 0 0 0;
75     font-size: 80%;
76 }
77 table, th, td {
78     border-collapse: collapse;
79     border: 1px solid #999;
80 }
81 th, td { padding: 0.5em; }
82 .header {
83     background: #eee;
84 }
85 .even { background: #eee; }
86 h1, h2, h3, h4, h5, h6 {
87     color: #666;
88     background-color: #eee;
89     padding-left: 0.5em
90 }
91 h1 { font-size: 130%; }
92 h2 { font-size: 120%; }
93 h3 { font-size: 110%; }
94 h4 { font-size: 107%; }
95 h5 { font-size: 103%; }
96 h6 { font-size: 100%; }
97 p { padding: 0 0.5em; }
98 '''
99
100 STYLESHEET = os.path.join(os.path.expanduser('~/.mutt'),
101                           'markdown2html.css')
102 if os.path.exists(STYLESHEET):
103     DEFAULT_CSS += open(STYLESHEET).read()
104
105 HTML_DOCUMENT = '''<!DOCTYPE html>
106 <html><head>
107 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
108 <meta charset="utf-8"/>
109 <meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes"/>
110 <title>HTML E-Mail</title>
111 </head><body class="email">
112 {htmlbody}
113 </body></html>'''
114
115
116 SIGNATURE_HTML = \
117         '<div class="signature"><span class="leader">-- </span>{sig}</div>'
118
119
120 def _preprocess_markdown(mdwn):
121     '''
122     Preprocess Markdown for handling by the converter.
123     '''
124     # convert hard line breaks within paragraphs to 2 trailing spaces, which
125     # is the markdown way of representing hard line breaks. Note how the
126     # regexp will not match between paragraphs.
127     ret = re.sub(r'(\S)\n(\s*\S)', r'\g<1>  \n\g<2>', mdwn, flags=re.MULTILINE)
128
129     # Clients like Thunderbird need the leading '>' to be able to properly
130     # create nested quotes, so we duplicate the symbol, the first instance
131     # will tell pandoc to create a blockquote, while the second instance will
132     # be a <span> containing the character, along with a class that causes CSS
133     # to actually hide it from display. However, this does not work with the
134     # text-mode HTML2text converters, and so it's left commented for now.
135     #ret = re.sub(r'\n>', r'  \n>[>]{.quotechar}', ret, flags=re.MULTILINE)
136
137     return ret
138
139
140 def _identify_quotes_for_later(mdwn):
141     '''
142     Email quoting such as:
143
144     ```
145     On 1970-01-01, you said:
146     > The Flat Earth Society has members all around the globe.
147     ```
148
149     isn't really properly handled by Markdown, so let's do our best to
150     identify the individual elements, and mark them, using a syntax similar to
151     what pandoc uses already in some cases. As pandoc won't actually use these
152     data (yet?), we call `self._reformat_quotes` later to use these markers
153     to slap the appropriate classes on the HTML tags.
154     '''
155
156     def generate_lines_with_context(mdwn):
157         '''
158         Iterates the input string line-wise, returning a triplet of
159         previous, current, and next line, the first and last of which
160         will be None on the first and last line of the input data
161         respectively.
162         '''
163         prev = cur = nxt = None
164         lines = iter(mdwn.splitlines())
165         cur = next(lines)
166         for nxt in lines:
167             yield prev, cur, nxt
168             prev = cur
169             cur = nxt
170         yield prev, cur, None
171
172     ret = []
173     for prev, cur, nxt in generate_lines_with_context(mdwn):
174
175         # The lead-in to a quote is a single line immediately preceding the
176         # quote, and ending with ':'. Note that there could be multiple of
177         # these:
178         if re.match(r'^.+:\s*$', cur) and nxt.startswith('>'):
179             ret.append(f'{{.quotelead}}{cur.strip()}')
180             # pandoc needs an empty line before the blockquote, so
181             # we enter one for the purpose of HTML rendition:
182             ret.append('')
183             continue
184
185         # The first blockquote after such a lead-in gets marked as the
186         # "initial" quote:
187         elif prev and re.match(r'^.+:\s*$', prev) and cur.startswith('>'):
188             ret.append(re.sub(r'^(\s*>\s*)+(.+)',
189                               r'\g<1>{.quoteinitial}\g<2>',
190                               cur, flags=re.MULTILINE))
191
192         # All other occurrences of blockquotes get the "subsequent" marker:
193         elif cur.startswith('>') and prev and not prev.startswith('>'):
194             ret.append(re.sub(r'^((?:\s*>\s*)+)(.+)',
195                               r'\g<1>{.quotesubsequent}\g<2>',
196                               cur, flags=re.MULTILINE))
197
198         else: # pass through everything else.
199             ret.append(cur)
200
201     return '\n'.join(ret)
202
203
204 def _reformat_quotes(html):
205     '''
206     Earlier in the pipeline, we marked email quoting, using markers, which we
207     now need to turn into HTML classes, so that we can use CSS to style them.
208     '''
209     ret = html.replace('<p>{.quotelead}', '<p class="quotelead">')
210     ret = re.sub(r'<blockquote>\n((?:<blockquote>\n)*)<p>(?:\{\.quote(\w+)\})',
211                  r'<blockquote class="quote \g<2>">\n\g<1><p>', ret, flags=re.MULTILINE)
212     return ret
213
214
215
216 def _convert_with_pandoc(mdwn, inputfmt='markdown', outputfmt='html5',
217                          ext_enabled=None, ext_disabled=None,
218                          standalone=True, title="HTML E-Mail"):
219     '''
220     Invoke pandoc to do the actual conversion of Markdown to HTML5.
221     '''
222     if not ext_enabled:
223         ext_enabled = [ 'backtick_code_blocks',
224                        'line_blocks',
225                        'fancy_lists',
226                        'startnum',
227                        'definition_lists',
228                        'example_lists',
229                        'table_captions',
230                        'simple_tables',
231                        'multiline_tables',
232                        'grid_tables',
233                        'pipe_tables',
234                        'all_symbols_escapable',
235                        'intraword_underscores',
236                        'strikeout',
237                        'superscript',
238                        'subscript',
239                        'fenced_divs',
240                        'bracketed_spans',
241                        'footnotes',
242                        'inline_notes',
243                        'emoji',
244                        'tex_math_double_backslash',
245                        'autolink_bare_uris'
246                       ]
247     if not ext_disabled:
248         ext_disabled = [ 'tex_math_single_backslash',
249                          'tex_math_dollars',
250                          'smart',
251                          'raw_html'
252                        ]
253
254     enabled = '+'.join(ext_enabled)
255     disabled = '-'.join(ext_disabled)
256     inputfmt = f'{inputfmt}+{enabled}-{disabled}'
257
258     args = []
259     if standalone:
260         args.append('--standalone')
261     if title:
262         args.append(f'--metadata=pagetitle:"{title}"')
263
264     return pypandoc.convert_text(mdwn, format=inputfmt, to=outputfmt,
265                                  extra_args=args)
266
267
268 def _apply_styling(html):
269     '''
270     Inline all styles defined and used into the individual HTML tags.
271     '''
272     return pynliner.Pynliner().from_string(html).with_cssString(DEFAULT_CSS).run()
273
274
275 def _postprocess_html(html):
276     '''
277     Postprocess the generated and styled HTML.
278     '''
279     return html
280
281
282 def convert_markdown_to_html(mdwn):
283     '''
284     Converts the input Markdown to HTML, handling separately the body, as well
285     as an optional signature.
286     '''
287     parts = re.split(r'^-- $', mdwn, 1, flags=re.MULTILINE)
288     body = parts[0]
289     if len(parts) == 2:
290         sig = parts[1]
291     else:
292         sig = None
293
294     html=''
295     if body:
296         body = _preprocess_markdown(body)
297         body = _identify_quotes_for_later(body)
298         html = _convert_with_pandoc(body, standalone=False)
299         html = _reformat_quotes(html)
300
301     if sig:
302         sig = _preprocess_markdown(sig)
303         html += SIGNATURE_HTML.format(sig='<br/>'.join(sig.splitlines()))
304
305     html = HTML_DOCUMENT.format(htmlbody=html)
306     html = _apply_styling(html)
307     html = _postprocess_html(html)
308
309     return html
310
311
312 def main():
313     '''
314     Convert text on stdin to HTML, and print it to stdout, like mutt would
315     expect.
316     '''
317     html = convert_markdown_to_html(sys.stdin.read())
318     if html:
319         # mutt expects the content type in the first line, so:
320         print(f'text/html\n\n{html}')
321
322
323 if __name__ == '__main__':
324     main()