| 1 | #!/usr/bin/env python2
|
| 2 | """
|
| 3 | split_doc.py
|
| 4 | """
|
| 5 | from __future__ import print_function
|
| 6 |
|
| 7 | import json
|
| 8 | import optparse
|
| 9 | import re
|
| 10 | import sys
|
| 11 |
|
| 12 |
|
| 13 | DATE_RE = re.compile(
|
| 14 | r'(\d\d\d\d) / (\d\d) / (\d\d)', re.VERBOSE)
|
| 15 |
|
| 16 | META_RE = re.compile(
|
| 17 | r'(\S+): [ ]* (.*)', re.VERBOSE)
|
| 18 |
|
| 19 |
|
| 20 | def SplitDocument(default_vals, entry_f, meta_f, content_f, strict=False):
|
| 21 | """Split a document into metadata JSON and content Markdown.
|
| 22 |
|
| 23 | Used for blog posts and index.md / cross-ref.md.
|
| 24 | """
|
| 25 | first_line = entry_f.readline()
|
| 26 | if strict and first_line.strip() != '---':
|
| 27 | raise RuntimeError("Document should start with --- (got %r)" % first_line)
|
| 28 |
|
| 29 | meta = {}
|
| 30 |
|
| 31 | # TODO: if first_line is ---, then read metadata in key: value format.
|
| 32 | if first_line.strip() == '---':
|
| 33 | while True:
|
| 34 | line = entry_f.readline().strip()
|
| 35 | if line == '---':
|
| 36 | break
|
| 37 | m = META_RE.match(line)
|
| 38 | if not m:
|
| 39 | raise RuntimeError('Invalid metadata line %r' % line)
|
| 40 | name, value = m.groups()
|
| 41 |
|
| 42 | if name == 'date':
|
| 43 | m2 = DATE_RE.match(value)
|
| 44 | if not m2:
|
| 45 | raise RuntimeError('Invalid date %r' % value)
|
| 46 | year, month, day = m2.groups()
|
| 47 | meta['year'] = int(year)
|
| 48 | meta['month'] = int(month)
|
| 49 | meta['day'] = int(day)
|
| 50 |
|
| 51 | elif name == 'updated_date':
|
| 52 | m2 = DATE_RE.match(value)
|
| 53 | if not m2:
|
| 54 | raise RuntimeError('Invalid date %r' % value)
|
| 55 | year, month, day = m2.groups()
|
| 56 | meta['updated_year'] = int(year)
|
| 57 | meta['updated_month'] = int(month)
|
| 58 | meta['updated_day'] = int(day)
|
| 59 |
|
| 60 | else:
|
| 61 | meta[name] = value
|
| 62 |
|
| 63 | #print('line = %r' % line, file=sys.stderr)
|
| 64 | while True:
|
| 65 | first_nonempty = entry_f.readline()
|
| 66 | if first_nonempty.strip() != '':
|
| 67 | break
|
| 68 |
|
| 69 | else:
|
| 70 | if first_line:
|
| 71 | first_nonempty = first_line
|
| 72 | else:
|
| 73 | while True:
|
| 74 | first_nonempty = entry_f.readline()
|
| 75 | if first_nonempty.strip() != '':
|
| 76 | break
|
| 77 |
|
| 78 | # Invariant: we've read the first non-empty line here. Now we need to see if
|
| 79 | # it's the title.
|
| 80 |
|
| 81 | #print('first_nonempty = %r' % first_nonempty, file=sys.stderr)
|
| 82 |
|
| 83 | line_two = entry_f.readline()
|
| 84 | if re.match('=+', line_two):
|
| 85 | meta['title'] = first_nonempty.strip()
|
| 86 |
|
| 87 | # Fill in defaults after parsing all values.
|
| 88 | for name, value in default_vals.iteritems():
|
| 89 | if name not in meta:
|
| 90 | meta[name] = value
|
| 91 |
|
| 92 | json.dump(meta, meta_f, indent=2)
|
| 93 |
|
| 94 | # Read the rest of the file and write it
|
| 95 | contents = entry_f.read()
|
| 96 |
|
| 97 | content_f.write(first_nonempty)
|
| 98 | content_f.write(line_two)
|
| 99 |
|
| 100 | content_f.write(contents)
|
| 101 |
|
| 102 | comments_url = meta.get('comments_url', '')
|
| 103 | if comments_url:
|
| 104 | content_f.write("""
|
| 105 | [comments-url]: %s
|
| 106 |
|
| 107 | """ % comments_url)
|
| 108 |
|
| 109 |
|
| 110 |
|
| 111 | def Options():
|
| 112 | """Returns an option parser instance."""
|
| 113 | p = optparse.OptionParser('split_doc.py [options] input_file out_prefix')
|
| 114 | # Like awk -v
|
| 115 | p.add_option(
|
| 116 | '-v', dest='default_vals', action='append', default=[],
|
| 117 | help="If the doc's own metadata doesn't define 'name', set it to this value")
|
| 118 | p.add_option(
|
| 119 | '-s', '--strict', dest='strict', action='store_true', default=False,
|
| 120 | help="Require metadata")
|
| 121 | return p
|
| 122 |
|
| 123 |
|
| 124 | def main(argv):
|
| 125 | o = Options()
|
| 126 | opts, argv = o.parse_args(argv)
|
| 127 |
|
| 128 | entry_path = argv[1] # e.g. blog/2016/11/01.md
|
| 129 | out_prefix = argv[2] # e.g _site/blog/2016/11/01
|
| 130 |
|
| 131 | meta_path = out_prefix + '_meta.json'
|
| 132 | content_path = out_prefix + '_content.md'
|
| 133 |
|
| 134 | default_vals = {}
|
| 135 | for pair in opts.default_vals:
|
| 136 | name, value = pair.split('=', 1)
|
| 137 | default_vals[name] = value
|
| 138 |
|
| 139 | with \
|
| 140 | open(entry_path) as entry_f, \
|
| 141 | open(meta_path, 'w') as meta_f, \
|
| 142 | open(content_path, 'w') as content_f:
|
| 143 | SplitDocument(default_vals, entry_f, meta_f, content_f, strict=opts.strict)
|
| 144 |
|
| 145 |
|
| 146 | if __name__ == '__main__':
|
| 147 | try:
|
| 148 | main(sys.argv)
|
| 149 | except RuntimeError as e:
|
| 150 | print('FATAL: %s' % e, file=sys.stderr)
|
| 151 | sys.exit(1)
|