1: #!/usr/bin/env python3
2:
3: # This script takes a manpage written in markdown and turns it into an html web
4: # page and a nroff man page. The input file must have the name of the program
5: # and the section in this format: NAME.NUM.md. The output files are written
6: # into the current directory named NAME.NUM.html and NAME.NUM. The input
7: # format has one extra extension: if a numbered list starts at 0, it is turned
8: # into a description list. The dl's dt tag is taken from the contents of the
9: # first tag inside the li, which is usually a p, code, or strong tag. The
10: # cmarkgfm or commonmark lib is used to transforms the input file into html.
11: # The html.parser is used as a state machine that both tweaks the html and
12: # outputs the nroff data based on the html tags.
13: #
14: # Copyright (C) 2020 Wayne Davison
15: #
16: # This program is freely redistributable.
17:
18: import sys, os, re, argparse, subprocess, time
19: from html.parser import HTMLParser
20:
21: CONSUMES_TXT = set('h1 h2 p li pre'.split())
22:
23: HTML_START = """\
24: <html><head>
25: <title>%s</title>
26: <link href="https://fonts.googleapis.com/css2?family=Roboto&family=Roboto+Mono&display=swap" rel="stylesheet">
27: <style>
28: body {
29: max-width: 50em;
30: margin: auto;
31: }
32: body, b, strong, u {
33: font-family: 'Roboto', sans-serif;
34: }
35: code {
36: font-family: 'Roboto Mono', monospace;
37: font-weight: bold;
38: white-space: pre;
39: }
40: pre code {
41: display: block;
42: font-weight: normal;
43: }
44: blockquote pre code {
45: background: #f1f1f1;
46: }
47: dd p:first-of-type {
48: margin-block-start: 0em;
49: }
50: </style>
51: </head><body>
52: """
53:
54: HTML_END = """\
55: <div style="float: right"><p><i>%s</i></p></div>
56: </body></html>
57: """
58:
59: MAN_START = r"""
60: .TH "%s" "%s" "%s" "%s" "User Commands"
61: """.lstrip()
62:
63: MAN_END = """\
64: """
65:
66: NORM_FONT = ('\1', r"\fP")
67: BOLD_FONT = ('\2', r"\fB")
68: UNDR_FONT = ('\3', r"\fI")
69: NBR_DASH = ('\4', r"\-")
70: NBR_SPACE = ('\xa0', r"\ ")
71:
72: md_parser = None
73:
74: def main():
75: fi = re.match(r'^(?P<fn>(?P<srcdir>.+/)?(?P<name>(?P<prog>[^/]+)\.(?P<sect>\d+))\.md)$', args.mdfile)
76: if not fi:
77: die('Failed to parse NAME.NUM.md out of input file:', args.mdfile)
78: fi = argparse.Namespace(**fi.groupdict())
79:
80: if not fi.srcdir:
81: fi.srcdir = './'
82:
83: fi.title = fi.prog + '(' + fi.sect + ') man page'
84: fi.mtime = 0
85:
86: git_dir = fi.srcdir + '.git'
87: if os.path.lexists(git_dir):
88: fi.mtime = int(subprocess.check_output(['git', '--git-dir', git_dir, 'log', '-1', '--format=%at']))
89:
90: env_subs = { 'prefix': os.environ.get('RSYNC_OVERRIDE_PREFIX', None) }
91:
92: if args.test:
93: env_subs['VERSION'] = '1.0.0'
94: env_subs['libdir'] = '/usr'
95: else:
96: for fn in (fi.srcdir + 'version.h', 'Makefile'):
97: try:
98: st = os.lstat(fn)
99: except:
100: die('Failed to find', fi.srcdir + fn)
101: if not fi.mtime:
102: fi.mtime = st.st_mtime
103:
104: with open(fi.srcdir + 'version.h', 'r', encoding='utf-8') as fh:
105: txt = fh.read()
106: m = re.search(r'"(.+?)"', txt)
107: env_subs['VERSION'] = m.group(1)
108:
109: with open('Makefile', 'r', encoding='utf-8') as fh:
110: for line in fh:
111: m = re.match(r'^(\w+)=(.+)', line)
112: if not m:
113: continue
114: var, val = (m.group(1), m.group(2))
115: if var == 'prefix' and env_subs[var] is not None:
116: continue
117: while re.search(r'\$\{', val):
118: val = re.sub(r'\$\{(\w+)\}', lambda m: env_subs[m.group(1)], val)
119: env_subs[var] = val
120: if var == 'srcdir':
121: break
122:
123: with open(fi.fn, 'r', encoding='utf-8') as fh:
124: txt = fh.read()
125:
126: txt = re.sub(r'@VERSION@', env_subs['VERSION'], txt)
127: txt = re.sub(r'@LIBDIR@', env_subs['libdir'], txt)
128:
129: fi.html_in = md_parser(txt)
130: txt = None
131:
132: fi.date = time.strftime('%d %b %Y', time.localtime(fi.mtime))
133: fi.man_headings = (fi.prog, fi.sect, fi.date, fi.prog + ' ' + env_subs['VERSION'])
134:
135: HtmlToManPage(fi)
136:
137: if args.test:
138: print("The test was successful.")
139: return
140:
141: for fn, txt in ((fi.name + '.html', fi.html_out), (fi.name, fi.man_out)):
142: print("Wrote:", fn)
143: with open(fn, 'w', encoding='utf-8') as fh:
144: fh.write(txt)
145:
146:
147: def html_via_commonmark(txt):
148: return commonmark.HtmlRenderer().render(commonmark.Parser().parse(txt))
149:
150:
151: class HtmlToManPage(HTMLParser):
152: def __init__(self, fi):
153: HTMLParser.__init__(self, convert_charrefs=True)
154:
155: st = self.state = argparse.Namespace(
156: list_state = [ ],
157: p_macro = ".P\n",
158: at_first_tag_in_li = False,
159: at_first_tag_in_dd = False,
160: dt_from = None,
161: in_pre = False,
162: in_code = False,
163: html_out = [ HTML_START % fi.title ],
164: man_out = [ MAN_START % fi.man_headings ],
165: txt = '',
166: )
167:
168: self.feed(fi.html_in)
169: fi.html_in = None
170:
171: st.html_out.append(HTML_END % fi.date)
172: st.man_out.append(MAN_END)
173:
174: fi.html_out = ''.join(st.html_out)
175: st.html_out = None
176:
177: fi.man_out = ''.join(st.man_out)
178: st.man_out = None
179:
180:
181: def handle_starttag(self, tag, attrs_list):
182: st = self.state
183: if args.debug:
184: self.output_debug('START', (tag, attrs_list))
185: if st.at_first_tag_in_li:
186: if st.list_state[-1] == 'dl':
187: st.dt_from = tag
188: if tag == 'p':
189: tag = 'dt'
190: else:
191: st.html_out.append('<dt>')
192: elif tag == 'p':
193: st.at_first_tag_in_dd = True # Kluge to suppress a .P at the start of an li.
194: st.at_first_tag_in_li = False
195: if tag == 'p':
196: if not st.at_first_tag_in_dd:
197: st.man_out.append(st.p_macro)
198: elif tag == 'li':
199: st.at_first_tag_in_li = True
200: lstate = st.list_state[-1]
201: if lstate == 'dl':
202: return
203: if lstate == 'o':
204: st.man_out.append(".IP o\n")
205: else:
206: st.man_out.append(".IP " + str(lstate) + ".\n")
207: st.list_state[-1] += 1
208: elif tag == 'blockquote':
209: st.man_out.append(".RS 4\n")
210: elif tag == 'pre':
211: st.in_pre = True
212: st.man_out.append(st.p_macro + ".nf\n")
213: elif tag == 'code' and not st.in_pre:
214: st.in_code = True
215: st.txt += BOLD_FONT[0]
216: elif tag == 'strong' or tag == 'b':
217: st.txt += BOLD_FONT[0]
218: elif tag == 'em' or tag == 'i':
219: tag = 'u' # Change it into underline to be more like the man page
220: st.txt += UNDR_FONT[0]
221: elif tag == 'ol':
222: start = 1
223: for var, val in attrs_list:
224: if var == 'start':
225: start = int(val) # We only support integers.
226: break
227: if st.list_state:
228: st.man_out.append(".RS\n")
229: if start == 0:
230: tag = 'dl'
231: attrs_list = [ ]
232: st.list_state.append('dl')
233: else:
234: st.list_state.append(start)
235: st.man_out.append(st.p_macro)
236: st.p_macro = ".IP\n"
237: elif tag == 'ul':
238: st.man_out.append(st.p_macro)
239: if st.list_state:
240: st.man_out.append(".RS\n")
241: st.p_macro = ".IP\n"
242: st.list_state.append('o')
243: st.html_out.append('<' + tag + ''.join(' ' + var + '="' + htmlify(val) + '"' for var, val in attrs_list) + '>')
244: st.at_first_tag_in_dd = False
245:
246:
247: def handle_endtag(self, tag):
248: st = self.state
249: if args.debug:
250: self.output_debug('END', (tag,))
251: if tag in CONSUMES_TXT or st.dt_from == tag:
252: txt = st.txt.strip()
253: st.txt = ''
254: else:
255: txt = None
256: add_to_txt = None
257: if tag == 'h1':
258: st.man_out.append(st.p_macro + '.SH "' + manify(txt) + '"\n')
259: elif tag == 'h2':
260: st.man_out.append(st.p_macro + '.SS "' + manify(txt) + '"\n')
261: elif tag == 'p':
262: if st.dt_from == 'p':
263: tag = 'dt'
264: st.man_out.append('.IP "' + manify(txt) + '"\n')
265: st.dt_from = None
266: elif txt != '':
267: st.man_out.append(manify(txt) + "\n")
268: elif tag == 'li':
269: if st.list_state[-1] == 'dl':
270: if st.at_first_tag_in_li:
271: die("Invalid 0. -> td translation")
272: tag = 'dd'
273: if txt != '':
274: st.man_out.append(manify(txt) + "\n")
275: st.at_first_tag_in_li = False
276: elif tag == 'blockquote':
277: st.man_out.append(".RE\n")
278: elif tag == 'pre':
279: st.in_pre = False
280: st.man_out.append(manify(txt) + "\n.fi\n")
281: elif (tag == 'code' and not st.in_pre):
282: st.in_code = False
283: add_to_txt = NORM_FONT[0]
284: elif tag == 'strong' or tag == 'b':
285: add_to_txt = NORM_FONT[0]
286: elif tag == 'em' or tag == 'i':
287: tag = 'u' # Change it into underline to be more like the man page
288: add_to_txt = NORM_FONT[0]
289: elif tag == 'ol' or tag == 'ul':
290: if st.list_state.pop() == 'dl':
291: tag = 'dl'
292: if st.list_state:
293: st.man_out.append(".RE\n")
294: else:
295: st.p_macro = ".P\n"
296: st.at_first_tag_in_dd = False
297: st.html_out.append('</' + tag + '>')
298: if add_to_txt:
299: if txt is None:
300: st.txt += add_to_txt
301: else:
302: txt += add_to_txt
303: if st.dt_from == tag:
304: st.man_out.append('.IP "' + manify(txt) + '"\n')
305: st.html_out.append('</dt><dd>')
306: st.at_first_tag_in_dd = True
307: st.dt_from = None
308: elif tag == 'dt':
309: st.html_out.append('<dd>')
310: st.at_first_tag_in_dd = True
311:
312:
313: def handle_data(self, txt):
314: st = self.state
315: if args.debug:
316: self.output_debug('DATA', (txt,))
317: if st.in_pre:
318: html = htmlify(txt)
319: else:
320: txt = re.sub(r'\s--(\s)', NBR_SPACE[0] + r'--\1', txt).replace('--', NBR_DASH[0]*2)
321: txt = re.sub(r'(^|\W)-', r'\1' + NBR_DASH[0], txt)
322: html = htmlify(txt)
323: if st.in_code:
324: txt = re.sub(r'\s', NBR_SPACE[0], txt)
325: html = html.replace(NBR_DASH[0], '-').replace(NBR_SPACE[0], ' ') # <code> is non-breaking in CSS
326: st.html_out.append(html.replace(NBR_SPACE[0], ' ').replace(NBR_DASH[0], '-⁠'))
327: st.txt += txt
328:
329:
330: def output_debug(self, event, extra):
331: import pprint
332: st = self.state
333: if args.debug < 2:
334: st = argparse.Namespace(**vars(st))
335: if len(st.html_out) > 2:
336: st.html_out = ['...'] + st.html_out[-2:]
337: if len(st.man_out) > 2:
338: st.man_out = ['...'] + st.man_out[-2:]
339: print(event, extra)
340: pprint.PrettyPrinter(indent=2).pprint(vars(st))
341:
342:
343: def manify(txt):
344: return re.sub(r"^(['.])", r'\&\1', txt.replace('\\', '\\\\')
345: .replace(NBR_SPACE[0], NBR_SPACE[1])
346: .replace(NBR_DASH[0], NBR_DASH[1])
347: .replace(NORM_FONT[0], NORM_FONT[1])
348: .replace(BOLD_FONT[0], BOLD_FONT[1])
349: .replace(UNDR_FONT[0], UNDR_FONT[1]), flags=re.M)
350:
351:
352: def htmlify(txt):
353: return txt.replace('&', '&').replace('<', '<').replace('>', '>').replace('"', '"')
354:
355:
356: def warn(*msg):
357: print(*msg, file=sys.stderr)
358:
359:
360: def die(*msg):
361: warn(*msg)
362: sys.exit(1)
363:
364:
365: if __name__ == '__main__':
366: parser = argparse.ArgumentParser(description='Transform a NAME.NUM.md markdown file into a NAME.NUM.html web page & a NAME.NUM man page.', add_help=False)
367: parser.add_argument('--test', action='store_true', help='Test if we can parse the input w/o updating any files.')
368: parser.add_argument('--debug', '-D', action='count', default=0, help='Output copious info on the html parsing. Repeat for even more.')
369: parser.add_argument("--help", "-h", action="help", help="Output this help message and exit.")
370: parser.add_argument('mdfile', help="The NAME.NUM.md file to parse.")
371: args = parser.parse_args()
372:
373: try:
374: import cmarkgfm
375: md_parser = cmarkgfm.markdown_to_html
376: except:
377: try:
378: import commonmark
379: md_parser = html_via_commonmark
380: except:
381: die("Failed to find cmarkgfm or commonmark for python3.")
382:
383: main()
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to md2man CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / rsync