python/samba/ms_forest_updates_markdown.py

   1 # Create forest updates ldif from Github markdown
   2 #
   3 # Each update is converted to an ldif then gets written to a corresponding
   4 # .LDF output file or stored in a dictionary.
   5 #
   6 # Only add updates can generally be applied.
   7 #
   8 # Copyright (C) Andrew Bartlett <abartlet@samba.org> 2017
   9 #
  10 # This program is free software; you can redistribute it and/or modify
  11 # it under the terms of the GNU General Public License as published by
  12 # the Free Software Foundation; either version 3 of the License, or
  13 # (at your option) any later version.
  14 #
  15 # This program is distributed in the hope that it will be useful,
  16 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 # GNU General Public License for more details.
  19 #
  20 # You should have received a copy of the GNU General Public License
  21 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
  22
  23 from __future__ import print_function
  24 """Generate LDIF from Github documentation."""
  25
  26 import re
  27 import os
  28 import markdown
  29 import xml.etree.ElementTree as ET
  30
  31
  32 # Display specifier updates or otherwise (ignored in forest_update.py)
  33 def noop(description, attributes, sd):
  34     return (None, None, [], None)
  35
  36
  37 # ACE addition updates (ignored in forest_update.py)
  38 def parse_grant(description, attributes, sd):
  39     return ('modify', None, [], sd if sd.lower() != 'n/a' else None)
  40
  41
  42 # Addition of new objects to the directory (most are applied in forest_update.py)
  43 def parse_add(description, attributes, sd):
  44     dn = extract_dn(description)
  45     return ('add', dn, extract_attrib(dn, attributes), sd if sd.lower() != 'n/a' else None)
  46
  47
  48 # Set of a particular attribute (ignored in forest_update.py)
  49 def parse_set(description, attributes, sd):
  50     return ('modify', extract_dn_or_none(description),
  51             extract_replace_attrib(attributes),
  52             sd if sd.lower() != 'n/a' else None)
  53
  54
  55 # Set of a particular ACE (ignored in forest_update.py)
  56 # The general issue is that the list of DNs must be generated dynamically
  57 def parse_ace(description, attributes, sd):
  58
  59     def extract_dn_ace(text):
  60         if 'Sam-Domain' in text:
  61             return ('${DOMAIN_DN}', 'CN=Sam-Domain,${SCHEMA_DN}')
  62         elif 'Domain-DNS' in text:
  63             return ('${...}', 'CN=Domain-DNS,${SCHEMA_DN}')
  64
  65         return None
  66
  67     return [('modify', extract_dn_ace(description)[0],
  68              ['replace: nTSecurityDescriptor',
  69               'nTSecurityDescriptor: ${DOMAIN_SCHEMA_SD}%s' % sd], None),
  70             ('modify', extract_dn_ace(description)[1],
  71              ['replace: defaultSecurityDescriptor',
  72               'defaultSecurityDescriptor: ${OLD_SAMBA_SD}%s' % sd], None)]
  73
  74
  75 # We are really only interested in 'Created' items
  76 operation_map = {
  77     # modify
  78     'Granting': parse_grant,
  79     # add
  80     'Created': parse_add,
  81     # modify
  82     'Set': parse_set,
  83     # modify
  84     'Added ACE': parse_ace,
  85     # modify
  86     'Updated': parse_set,
  87     # unknown
  88     'Call': noop
  89 }
  90
  91
  92 def extract_dn(text):
  93     """
  94     Extract a DN from the textual description
  95     :param text:
  96     :return: DN in string form
  97     """
  98     text = text.replace(' in the Schema partition.', ',${SCHEMA_DN}')
  99     text = text.replace(' in the Configuration partition.', ',${CONFIG_DN}')
 100     dn = re.search('([CDO][NCU]=.*?,)*([CDO][NCU]=.*)', text).group(0)
 101
 102     # This should probably be also fixed upstream
 103     if dn == 'CN=ad://ext/AuthenticationSilo,CN=Claim Types,CN=Claims Configuration,CN=Services':
 104         return 'CN=ad://ext/AuthenticationSilo,CN=Claim Types,CN=Claims Configuration,CN=Services,${CONFIG_DN}'
 105
 106     return dn
 107
 108
 109 def extract_dn_or_none(text):
 110     """
 111     Same as above, but returns None if it doesn't work
 112     :param text:
 113     :return: DN or None
 114     """
 115     try:
 116         return extract_dn(text)
 117     except:
 118         return None
 119
 120
 121 def save_ldif(filename, answers, out_folder):
 122     """
 123     Save ldif to disk for each updates
 124     :param filename: filename use ([OPERATION NUM]-{GUID}.ldif)
 125     :param answers: array of tuples generated with earlier functions
 126     :param out_folder: folder to prepend
 127     """
 128     path = os.path.join(out_folder, filename)
 129     with open(path, 'w') as ldif:
 130         for answer in answers:
 131             change, dn, attrib, sd = answer
 132             ldif.write('dn: %s\n' % dn)
 133             ldif.write('changetype: %s\n' % change)
 134             if len(attrib) > 0:
 135                 ldif.write('\n'.join(attrib) + '\n')
 136             if sd is not None:
 137                 ldif.write('nTSecurityDescriptor: D:%s\n' % sd)
 138             ldif.write('-\n\n')
 139
 140
 141 def save_array(guid, answers, out_dict):
 142     """
 143     Save ldif to an output dictionary
 144     :param guid: GUID to store
 145     :param answers: array of tuples generated with earlier functions
 146     :param out_dict: output dictionary
 147     """
 148     ldif = ''
 149     for answer in answers:
 150         change, dn, attrib, sd = answer
 151         ldif += 'dn: %s\n' % dn
 152         ldif += 'changetype: %s\n' % change
 153         if len(attrib) > 0:
 154             ldif += '\n'.join(attrib) + '\n'
 155         if sd is not None:
 156             ldif += 'nTSecurityDescriptor: D:%s\n' % sd
 157         ldif += '-\n\n'
 158
 159     out_dict[guid] = ldif
 160
 161
 162 def extract_attrib(dn, attributes):
 163     """
 164     Extract the attributes as an array from the attributes column
 165     :param dn: parsed from markdown
 166     :param attributes: from markdown
 167     :return: attribute array (ldif-type format)
 168     """
 169     attrib = [x.lstrip('- ') for x in attributes.split('-   ') if x.lower() != 'n/a' and x != '']
 170     attrib = [x.replace(': True', ': TRUE') if x.endswith(': True') else x for x in attrib]
 171     attrib = [x.replace(': False', ': FALSE') if x.endswith(': False') else x for x in attrib]
 172     # We only have one such value, we may as well skip them all consistently
 173     attrib = [x for x in attrib if not x.lower().startswith('msds-claimpossiblevalues')]
 174
 175     return attrib
 176
 177
 178 def extract_replace_attrib(attributes):
 179     """
 180     Extract the attributes as an array from the attributes column
 181     (for replace)
 182     :param attributes: from markdown
 183     :return: attribute array (ldif-type format)
 184     """
 185     lines = [x.lstrip('- ') for x in attributes.split('-   ') if x.lower() != 'n/a' and x != '']
 186     lines = [('replace: %s' % line.split(':')[0], line) for line in lines]
 187     lines = [line for pair in lines for line in pair]
 188     return lines
 189
 190
 191 def innertext(tag):
 192     return (tag.text or '') + \
 193         ''.join(innertext(e) for e in tag) + \
 194         (tag.tail or '')
 195
 196
 197 def read_ms_markdown(in_file, out_folder=None, out_dict={}):
 198     """
 199     Read Github documentation to produce forest wide udpates
 200     :param in_file: Forest-Wide-Updates.md
 201     :param out_folder: output folder
 202     :param out_dict: output dictionary
 203     """
 204
 205     with open(in_file) as update_file:
 206         # There is a hidden ClaimPossibleValues in this md file
 207         html = markdown.markdown(re.sub(r'CN=<forest root domain.*?>',
 208                                         '${FOREST_ROOT_DOMAIN}',
 209                                         update_file.read()),
 210                                  output_format='xhtml')
 211
 212     html = html.replace('CN=Schema,%ws', '${SCHEMA_DN}')
 213
 214     tree = ET.fromstring('<root>' + html + '</root>')
 215
 216     for node in tree:
 217         if node.text and node.text.startswith('|Operation'):
 218             # Strip first and last |
 219             updates = [x[1:len(x) - 1].split('|') for x in
 220                        ET.tostring(node, method='text').splitlines()]
 221             for update in updates[2:]:
 222                 output = re.match('Operation (\d+): {(.*)}', update[0])
 223                 if output:
 224                     # print output.group(1), output.group(2)
 225                     guid = output.group(2)
 226                     filename = "%s-{%s}.ldif" % (output.group(1).zfill(4), guid)
 227
 228                 found = False
 229
 230                 if update[3].startswith('Created') or update[1].startswith('Added ACE'):
 231                     # Trigger the security descriptor code
 232                     # Reduce info to just the security descriptor
 233                     update[3] = update[3].split(':')[-1]
 234
 235                     result = parse_ace(update[1], update[2], update[3])
 236
 237                     if filename and out_folder is not None:
 238                         save_ldif(filename, result, out_folder)
 239                     else:
 240                         save_array(guid, result, out_dict)
 241
 242                     continue
 243
 244                 for operation in operation_map:
 245                     if update[1].startswith(operation):
 246                         found = True
 247
 248                         result = operation_map[operation](update[1], update[2], update[3])
 249
 250                         if filename and out_folder is not None:
 251                             save_ldif(filename, [result], out_folder)
 252                         else:
 253                             save_array(guid, [result], out_dict)
 254
 255                         break
 256
 257                 if not found:
 258                     raise Exception(update)
 259
 260             # print ET.tostring(node, method='text')
 261
 262
 263 if __name__ == '__main__':
 264     import sys
 265
 266     out_folder = ''
 267
 268     if len(sys.argv) == 0:
 269         print("Usage: %s <Forest-Wide-Updates.md> [<output folder>]" % (sys.argv[0]), file=sys.stderr)
 270         sys.exit(1)
 271
 272     in_file = sys.argv[1]
 273     if len(sys.argv) > 2:
 274         out_folder = sys.argv[2]
 275
 276     read_ms_markdown(in_file, out_folder)