diff options
| author | Wayne Davison <wayne@opencoder.net> | 2020-06-09 15:21:38 -0700 |
|---|---|---|
| committer | Wayne Davison <wayne@opencoder.net> | 2020-06-09 17:02:41 -0700 |
| commit | 03fc62ad2fcc065bcbe67ad4968b331f52189e26 (patch) | |
| tree | 4554bde300ab4ecc3b10e1e972d28d5b2b73f46d /md2man | |
| parent | 68c865c9e6ddc8b7ab00787c4b0cf433b08b3dbc (diff) | |
| download | rsync-03fc62ad2fcc065bcbe67ad4968b331f52189e26.tar.gz | |
More man processing improvements
- Support the commonmark library in addition to cmarkgfm.
- Remove github-flavor from the markup.
- A few more html style improvements.
Diffstat (limited to 'md2man')
| -rwxr-xr-x | md2man | 84 |
1 files changed, 57 insertions, 27 deletions
@@ -1,15 +1,15 @@ #!/usr/bin/python3 -# This script takes a manpage written in github-flavored markdown and turns it -# into a html web page and a nroff man page. The input file must have the name -# of the program and the section in the format: NAME.NUM.md. The output files -# are written into the current directory named NAME.NUM.html and NAME.NUM. The -# input format has one extra extension: if a numbered list starts at 0, it is -# turned into a description list. The dl's dt tag is taken from the contents of -# the first tag inside the li, which is usually a p tag or a code tag. The -# cmarkgfm lib is used to transforms the input file into html. The html.parser -# is used as a state machine that both tweaks the html and outputs the nroff -# data based on the html tags. +# This script takes a manpage written in markdown and turns it into an html web +# page and a nroff man page. The input file must have the name of the program +# and the section in this format: NAME.NUM.md. The output files are written +# into the current directory named NAME.NUM.html and NAME.NUM. The input +# format has one extra extension: if a numbered list starts at 0, it is turned +# into a description list. The dl's dt tag is taken from the contents of the +# first tag inside the li, which is usually a p, code, or strong tag. The +# cmarkgfm or commonmark lib is used to transforms the input file into html. +# The html.parser is used as a state machine that both tweaks the html and +# outputs the nroff data based on the html tags. # # Copyright (C) 2020 Wayne Davison # @@ -23,16 +23,26 @@ CONSUMES_TXT = set('h1 h2 p li pre'.split()) HTML_START = """\ <html><head> <title>%s</title> -<link href="https://fonts.googleapis.com/css2?family=Roboto&display=swap" rel="stylesheet"> +<link href="https://fonts.googleapis.com/css2?family=Roboto&family=Roboto+Mono&display=swap" rel="stylesheet"> <style> body { - max-width: 40em; + max-width: 50em; margin: auto; font-size: 1.2em; +} +body, b, strong, u { font-family: 'Roboto', sans-serif; } +code { + font-family: 'Roboto Mono', monospace; + font-weight: bold; +} +pre code { + display: block; + font-weight: normal; +} blockquote pre code { - background: #eee; + background: #f1f1f1; } dd p:first-of-type { margin-block-start: 0em; @@ -47,7 +57,7 @@ HTML_END = """\ """ MAN_START = r""" -.TH "%s" "%s" "%s" "" "" +.TH "%s" "%s" "%s" "%s" "User Commands" """.lstrip() MAN_END = """\ @@ -57,6 +67,8 @@ NORM_FONT = ('\1', r"\fP") BOLD_FONT = ('\2', r"\fB") ULIN_FONT = ('\3', r"\fI") +md_parser = None + def main(): fi = re.match(r'^(?P<fn>(?P<srcdir>.+/)?(?P<name>(?P<prog>[^/]+)\.(?P<sect>\d+))\.md)$', args.mdfile) if not fi: @@ -93,10 +105,14 @@ def main(): break with open(fi.fn, 'r', encoding='utf-8') as fh: - txt = re.sub(r'@VERSION@', env_subs['VERSION'], fh.read()) - txt = re.sub(r'@LIBDIR@', env_subs['libdir'], txt) - fi.html_in = cmarkgfm.github_flavored_markdown_to_html(txt) - txt = None + txt = fh.read() + + txt = re.sub(r'@VERSION@', env_subs['VERSION'], txt) + txt = re.sub(r'@LIBDIR@', env_subs['libdir'], txt) + fi.html_in = md_parser(txt) + txt = None + + fi.man_headings = (fi.prog, fi.sect, fi.date, fi.prog + ' ' + env_subs['VERSION']) HtmlToManPage(fi) @@ -109,6 +125,13 @@ def main(): with open(fn, 'w', encoding='utf-8') as fh: fh.write(txt) +def html_via_cmarkgfm(txt): + return cmarkgfm.markdown_to_html(txt) + + +def html_via_commonmark(txt): + return commonmark.HtmlRenderer().render(commonmark.Parser().parse(txt)) + class HtmlToManPage(HTMLParser): def __init__(self, fi): @@ -122,7 +145,7 @@ class HtmlToManPage(HTMLParser): dt_from = None, in_pre = False, html_out = [ HTML_START % fi.title ], - man_out = [ MAN_START % (fi.prog, fi.sect, fi.date) ], + man_out = [ MAN_START % fi.man_headings ], txt = '', ) @@ -171,9 +194,10 @@ class HtmlToManPage(HTMLParser): st.man_out.append(st.p_macro + ".nf\n") elif tag == 'code' and not st.in_pre: st.txt += BOLD_FONT[0] - elif tag == 'strong' or tag == 'bold': + elif tag == 'strong' or tag == 'b': st.txt += BOLD_FONT[0] - elif tag == 'i' or tag == 'em': + elif tag == 'em' or tag == 'i': + tag = 'u' # Change it into underline to be more like the man page st.txt += ULIN_FONT[0] elif tag == 'ol': start = 1 @@ -236,11 +260,12 @@ class HtmlToManPage(HTMLParser): st.in_pre = False st.man_out.append(manify(txt) + "\n.fi\n") elif tag == 'code' and not st.in_pre: - add_to_txt = NORM_FONT[0] - elif tag == 'strong' or tag == 'bold': - add_to_txt = NORM_FONT[0] - elif tag == 'i' or tag == 'em': - add_to_txt = NORM_FONT[0] + add_to_txt = NORM_FONT[0] + elif tag == 'strong' or tag == 'b': + add_to_txt = NORM_FONT[0] + elif tag == 'em' or tag == 'i': + tag = 'u' # Change it into underline to be more like the man page + add_to_txt = NORM_FONT[0] elif tag == 'ol' or tag == 'ul': if st.list_state.pop() == 'dl': tag = 'dl' @@ -315,7 +340,12 @@ if __name__ == '__main__': try: import cmarkgfm + md_parser = html_via_cmarkgfm except: - die("The cmarkgfm library is not available for python3.") + try: + import commonmark + md_parser = html_via_commonmark + except: + die("Failed to find cmarkgfm or commonmark for python3.") main() |