]> 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:

fba46066e259fd278ee46e0aaedbeda79f0a9a7c
[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
75 STYLESHEET = os.path.join(os.path.expanduser('~/.mutt'),
76                           'markdown2html.css')
77 if os.path.exists(STYLESHEET):
78     DEFAULT_CSS += open(STYLESHEET).read()
79
80 HTML_DOCUMENT = '''<!DOCTYPE html>
81 <html><head>
82 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
83 <meta charset="utf-8"/>
84 <meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes"/>
85 <title>HTML E-Mail</title>
86 </head><body class="email">
87 {htmlbody}
88 </body></html>'''
89
90
91 SIGNATURE_HTML = \
92         '<div class="signature"><span class="leader">-- </span>{sig}</div>'
93
94
95 def _preprocess_markdown(mdwn):
96     '''
97     Preprocess Markdown for handling by the converter.
98     '''
99     # convert hard line breaks within paragraphs to 2 trailing spaces, which
100     # is the markdown way of representing hard line breaks. Note how the
101     # regexp will not match between paragraphs.
102     ret = re.sub(r'(\S)\n(\s*\S)', r'\g<1>  \n\g<2>', mdwn, flags=re.MULTILINE)
103
104     # Clients like Thunderbird need the leading '>' to be able to properly
105     # create nested quotes, so we duplicate the symbol, the first instance
106     # will tell pandoc to create a blockquote, while the second instance will
107     # be a <span> containing the character, along with a class that causes CSS
108     # to actually hide it from display. However, this does not work with the
109     # text-mode HTML2text converters, and so it's left commented for now.
110     #ret = re.sub(r'\n>', r'  \n>[>]{.quotechar}', ret, flags=re.MULTILINE)
111
112     return ret
113
114
115 def _identify_quotes_for_later(mdwn):
116     '''
117     Email quoting such as:
118
119     ```
120     On 1970-01-01, you said:
121     > The Flat Earth Society has members all around the globe.
122     ```
123
124     isn't really properly handled by Markdown, so let's do our best to
125     identify the individual elements, and mark them, using a syntax similar to
126     what pandoc uses already in some cases. As pandoc won't actually use these
127     data (yet?), we call `self._reformat_quotes` later to use these markers
128     to slap the appropriate classes on the HTML tags.
129     '''
130
131     def generate_lines_with_context(mdwn):
132         '''
133         Iterates the input string line-wise, returning a triplet of
134         previous, current, and next line, the first and last of which
135         will be None on the first and last line of the input data
136         respectively.
137         '''
138         prev = cur = nxt = None
139         lines = iter(mdwn.splitlines())
140         cur = next(lines)
141         for nxt in lines:
142             yield prev, cur, nxt
143             prev = cur
144             cur = nxt
145         yield prev, cur, None
146
147     ret = []
148     for prev, cur, nxt in generate_lines_with_context(mdwn):
149
150         # The lead-in to a quote is a single line immediately preceding the
151         # quote, and ending with ':'. Note that there could be multiple of
152         # these:
153         if re.match(r'^.+:\s*$', cur) and nxt.startswith('>'):
154             ret.append(f'{{.quotelead}}{cur.strip()}')
155             # pandoc needs an empty line before the blockquote, so
156             # we enter one for the purpose of HTML rendition:
157             ret.append('')
158             continue
159
160         # The first blockquote after such a lead-in gets marked as the
161         # "initial" quote:
162         elif prev and re.match(r'^.+:\s*$', prev) and cur.startswith('>'):
163             ret.append(re.sub(r'^(\s*>\s*)+(.+)',
164                               r'\g<1>{.quoteinitial}\g<2>',
165                               cur, flags=re.MULTILINE))
166
167         # All other occurrences of blockquotes get the "subsequent" marker:
168         elif cur.startswith('>') and prev and not prev.startswith('>'):
169             ret.append(re.sub(r'^((?:\s*>\s*)+)(.+)',
170                               r'\g<1>{.quotesubsequent}\g<2>',
171                               cur, flags=re.MULTILINE))
172
173         else: # pass through everything else.
174             ret.append(cur)
175
176     return '\n'.join(ret)
177
178
179 def _reformat_quotes(html):
180     '''
181     Earlier in the pipeline, we marked email quoting, using markers, which we
182     now need to turn into HTML classes, so that we can use CSS to style them.
183     '''
184     ret = html.replace('<p>{.quotelead}', '<p class="quotelead">')
185     ret = re.sub(r'<blockquote>\n((?:<blockquote>\n)*)<p>(?:\{\.quote(\w+)\})',
186                  r'<blockquote class="quote \g<2>">\n\g<1><p>', ret, flags=re.MULTILINE)
187     return ret
188
189
190
191 def _convert_with_pandoc(mdwn, inputfmt='markdown', outputfmt='html5',
192                          ext_enabled=None, ext_disabled=None,
193                          standalone=True, title="HTML E-Mail"):
194     '''
195     Invoke pandoc to do the actual conversion of Markdown to HTML5.
196     '''
197     if not ext_enabled:
198         ext_enabled = [ 'backtick_code_blocks',
199                        'line_blocks',
200                        'fancy_lists',
201                        'startnum',
202                        'definition_lists',
203                        'example_lists',
204                        'table_captions',
205                        'simple_tables',
206                        'multiline_tables',
207                        'grid_tables',
208                        'pipe_tables',
209                        'all_symbols_escapable',
210                        'intraword_underscores',
211                        'strikeout',
212                        'superscript',
213                        'subscript',
214                        'fenced_divs',
215                        'bracketed_spans',
216                        'footnotes',
217                        'inline_notes',
218                        'emoji',
219                        'tex_math_double_backslash',
220                       ]
221     if not ext_disabled:
222         ext_disabled = [ 'tex_math_single_backslash',
223                          'tex_math_dollars',
224                          'raw_html'
225                        ]
226
227     enabled = '+'.join(ext_enabled)
228     disabled = '-'.join(ext_disabled)
229     inputfmt = f'{inputfmt}+{enabled}-{disabled}'
230
231     args = []
232     if standalone:
233         args.append('--standalone')
234     if title:
235         args.append(f'--metadata=pagetitle:"{title}"')
236
237     return pypandoc.convert_text(mdwn, format=inputfmt, to=outputfmt,
238                                  extra_args=args)
239
240
241 def _apply_styling(html):
242     '''
243     Inline all styles defined and used into the individual HTML tags.
244     '''
245     return pynliner.Pynliner().from_string(html).with_cssString(DEFAULT_CSS).run()
246
247
248 def _postprocess_html(html):
249     '''
250     Postprocess the generated and styled HTML.
251     '''
252     return html
253
254
255 def convert_markdown_to_html(mdwn):
256     '''
257     Converts the input Markdown to HTML, handling separately the body, as well
258     as an optional signature.
259     '''
260     parts = re.split(r'^-- $', mdwn, 1, flags=re.MULTILINE)
261     body = parts[0]
262     if len(parts) == 2:
263         sig = parts[1]
264     else:
265         sig = None
266
267     html=''
268     if body:
269         body = _preprocess_markdown(body)
270         body = _identify_quotes_for_later(body)
271         html = _convert_with_pandoc(body, standalone=False)
272         html = _reformat_quotes(html)
273
274     if sig:
275         sig = _preprocess_markdown(sig)
276         html += SIGNATURE_HTML.format(sig='<br/>'.join(sig.splitlines()))
277
278     html = HTML_DOCUMENT.format(htmlbody=html)
279     html = _apply_styling(html)
280     html = _postprocess_html(html)
281
282     return html
283
284
285 def main():
286     '''
287     Convert text on stdin to HTML, and print it to stdout, like mutt would
288     expect.
289     '''
290     html = convert_markdown_to_html(sys.stdin.read())
291     if html:
292         # mutt expects the content type in the first line, so:
293         print(f'text/html\n\n{html}')
294
295
296 if __name__ == '__main__':
297     main()