python/samba/ms_schema.py

   1 # create schema.ldif (as a string) from WSPP documentation
   2 #
   3 # based on minschema.py and minschema_wspp
   4 #
   5 # This program is free software; you can redistribute it and/or modify
   6 # it under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 3 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # This program is distributed in the hope that it will be useful,
  11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License
  16 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
  17
  18 """Generate LDIF from WSPP documentation."""
  19
  20 import re
  21 import base64
  22 import uuid
  23
  24 bitFields = {}
  25
  26 # ADTS: 2.2.9
  27 # bit positions as labeled in the docs
  28 bitFields["searchflags"] = {
  29     'fATTINDEX': 31,         # IX
  30     'fPDNTATTINDEX': 30,     # PI
  31     'fANR': 29,  # AR
  32     'fPRESERVEONDELETE': 28,         # PR
  33     'fCOPY': 27,     # CP
  34     'fTUPLEINDEX': 26,       # TP
  35     'fSUBTREEATTINDEX': 25,  # ST
  36     'fCONFIDENTIAL': 24,     # CF
  37     'fNEVERVALUEAUDIT': 23,  # NV
  38     'fRODCAttribute': 22,    # RO
  39
  40
  41     # missing in ADTS but required by LDIF
  42     'fRODCFilteredAttribute': 22,    # RO ?
  43     'fCONFIDENTAIL': 24, # typo
  44     'fRODCFILTEREDATTRIBUTE': 22 # case
  45     }
  46
  47 # ADTS: 2.2.10
  48 bitFields["systemflags"] = {
  49     'FLAG_ATTR_NOT_REPLICATED': 31, 'FLAG_CR_NTDS_NC': 31,     # NR
  50     'FLAG_ATTR_REQ_PARTIAL_SET_MEMBER': 30, 'FLAG_CR_NTDS_DOMAIN': 30,     # PS
  51     'FLAG_ATTR_IS_CONSTRUCTED': 29, 'FLAG_CR_NTDS_NOT_GC_REPLICATED': 29,     # CS
  52     'FLAG_ATTR_IS_OPERATIONAL': 28,     # OP
  53     'FLAG_SCHEMA_BASE_OBJECT': 27,     # BS
  54     'FLAG_ATTR_IS_RDN': 26,     # RD
  55     'FLAG_DISALLOW_MOVE_ON_DELETE': 6,     # DE
  56     'FLAG_DOMAIN_DISALLOW_MOVE': 5,     # DM
  57     'FLAG_DOMAIN_DISALLOW_RENAME': 4,     # DR
  58     'FLAG_CONFIG_ALLOW_LIMITED_MOVE': 3,     # AL
  59     'FLAG_CONFIG_ALLOW_MOVE': 2,     # AM
  60     'FLAG_CONFIG_ALLOW_RENAME': 1,     # AR
  61     'FLAG_DISALLOW_DELETE': 0     # DD
  62     }
  63
  64 # ADTS: 2.2.11
  65 bitFields["schemaflagsex"] = {
  66     'FLAG_ATTR_IS_CRITICAL': 31
  67     }
  68
  69 # ADTS: 3.1.1.2.2.2
  70 oMObjectClassBER = {
  71     '1.3.12.2.1011.28.0.702' : base64.b64encode('\x2B\x0C\x02\x87\x73\x1C\x00\x85\x3E'),
  72     '1.2.840.113556.1.1.1.12': base64.b64encode('\x2A\x86\x48\x86\xF7\x14\x01\x01\x01\x0C'),
  73     '2.6.6.1.2.5.11.29'      : base64.b64encode('\x56\x06\x01\x02\x05\x0B\x1D'),
  74     '1.2.840.113556.1.1.1.11': base64.b64encode('\x2A\x86\x48\x86\xF7\x14\x01\x01\x01\x0B'),
  75     '1.3.12.2.1011.28.0.714' : base64.b64encode('\x2B\x0C\x02\x87\x73\x1C\x00\x85\x4A'),
  76     '1.3.12.2.1011.28.0.732' : base64.b64encode('\x2B\x0C\x02\x87\x73\x1C\x00\x85\x5C'),
  77     '1.2.840.113556.1.1.1.6' : base64.b64encode('\x2A\x86\x48\x86\xF7\x14\x01\x01\x01\x06')
  78 }
  79
  80 # separated by commas in docs, and must be broken up
  81 multivalued_attrs = set(["auxiliaryclass","maycontain","mustcontain","posssuperiors",
  82                          "systemauxiliaryclass","systemmaycontain","systemmustcontain",
  83                          "systemposssuperiors"])
  84
  85 def __read_folded_line(f, buffer):
  86     """ reads a line from an LDIF file, unfolding it"""
  87     line = buffer
  88
  89     while True:
  90         l = f.readline()
  91
  92         if l[:1] == " ":
  93             # continued line
  94
  95             # cannot fold an empty line
  96             assert(line != "" and line != "\n")
  97
  98             # preserves '\n '
  99             line = line + l
 100         else:
 101             # non-continued line
 102             if line == "":
 103                 line = l
 104
 105                 if l == "":
 106                     # eof, definitely won't be folded
 107                     break
 108             else:
 109                 # marks end of a folded line
 110                 # line contains the now unfolded line
 111                 # buffer contains the start of the next possibly folded line
 112                 buffer = l
 113                 break
 114
 115     return (line, buffer)
 116
 117
 118 def __read_raw_entries(f):
 119     """reads an LDIF entry, only unfolding lines"""
 120
 121     # will not match options after the attribute type
 122     attr_type_re = re.compile("^([A-Za-z]+[A-Za-z0-9-]*):")
 123
 124     buffer = ""
 125
 126     while True:
 127         entry = []
 128
 129         while True:
 130             (l, buffer) = __read_folded_line(f, buffer)
 131
 132             if l[:1] == "#":
 133                 continue
 134
 135             if l == "\n" or l == "":
 136                 break
 137
 138             m = attr_type_re.match(l)
 139
 140             if m:
 141                 if l[-1:] == "\n":
 142                     l = l[:-1]
 143
 144                 entry.append(l)
 145             else:
 146                 print >>sys.stderr, "Invalid line: %s" % l,
 147                 sys.exit(1)
 148
 149         if len(entry):
 150             yield entry
 151
 152         if l == "":
 153             break
 154
 155
 156 def fix_dn(dn):
 157     """fix a string DN to use ${SCHEMADN}"""
 158
 159     # folding?
 160     if dn.find("<RootDomainDN>") != -1:
 161         dn = dn.replace("\n ", "")
 162         dn = dn.replace(" ", "")
 163         return dn.replace("CN=Schema,CN=Configuration,<RootDomainDN>", "${SCHEMADN}")
 164     else:
 165         return dn
 166
 167 def __convert_bitfield(key, value):
 168     """Evaluate the OR expression in 'value'"""
 169     assert(isinstance(value, str))
 170
 171     value = value.replace("\n ", "")
 172     value = value.replace(" ", "")
 173
 174     try:
 175         # some attributes already have numeric values
 176         o = int(value)
 177     except ValueError:
 178         o = 0
 179         flags = value.split("|")
 180         for f in flags:
 181             bitpos = bitFields[key][f]
 182             o = o | (1 << (31 - bitpos))
 183
 184     return str(o)
 185
 186 def __write_ldif_one(entry):
 187     """Write out entry as LDIF"""
 188     out = []
 189
 190     for l in entry:
 191         if isinstance(l[1], str):
 192             vl = [l[1]]
 193         else:
 194             vl = l[1]
 195
 196         if l[0].lower() == 'omobjectclass':
 197             out.append("%s:: %s" % (l[0], l[1]))
 198             continue
 199
 200         for v in vl:
 201             out.append("%s: %s" % (l[0], v))
 202
 203
 204     return "\n".join(out)
 205
 206 def __transform_entry(entry, objectClass):
 207     """Perform transformations required to convert the LDIF-like schema
 208        file entries to LDIF, including Samba-specific stuff."""
 209
 210     entry = [l.split(":", 1) for l in entry]
 211
 212     cn = ""
 213
 214     for l in entry:
 215         key = l[0].lower()
 216         l[1] = l[1].lstrip()
 217         l[1] = l[1].rstrip()
 218
 219         if not cn and key == "cn":
 220             cn = l[1]
 221
 222         if key in multivalued_attrs:
 223             # unlike LDIF, these are comma-separated
 224             l[1] = l[1].replace("\n ", "")
 225             l[1] = l[1].replace(" ", "")
 226
 227             l[1] = l[1].split(",")
 228
 229         if key in bitFields:
 230             l[1] = __convert_bitfield(key, l[1])
 231
 232         if key == "omobjectclass":
 233             l[1] = oMObjectClassBER[l[1].strip()]
 234
 235         if isinstance(l[1], str):
 236             l[1] = fix_dn(l[1])
 237
 238
 239     assert(cn)
 240     entry.insert(0, ["dn", "CN=%s,${SCHEMADN}" % cn])
 241     entry.insert(1, ["objectClass", ["top", objectClass]])
 242     entry.insert(2, ["cn", cn])
 243     entry.insert(2, ["objectGUID", str(uuid.uuid4())])
 244     entry.insert(2, ["adminDescription", cn])
 245     entry.insert(2, ["adminDisplayName", cn])
 246
 247     for l in entry:
 248         key = l[0].lower()
 249
 250         if key == "cn":
 251             entry.remove(l)
 252
 253     return entry
 254
 255 def __parse_schema_file(filename, objectClass):
 256     """Load and transform a schema file."""
 257
 258     out = []
 259
 260     f = open(filename, "rU")
 261     for entry in __read_raw_entries(f):
 262         out.append(__write_ldif_one(__transform_entry(entry, objectClass)))
 263
 264     return "\n\n".join(out)
 265
 266
 267 def read_ms_schema(attr_file, classes_file, dump_attributes = True, dump_classes = True, debug = False):
 268     """Read WSPP documentation-derived schema files."""
 269
 270     attr_ldif = ""
 271     classes_ldif = ""
 272
 273     if dump_attributes:
 274         attr_ldif =  __parse_schema_file(attr_file, "attributeSchema")
 275     if dump_classes:
 276         classes_ldif = __parse_schema_file(classes_file, "classSchema")
 277
 278     return attr_ldif + "\n\n" + classes_ldif + "\n\n"
 279
 280 if __name__ == '__main__':
 281     import sys
 282
 283     try:
 284         attr_file = sys.argv[1]
 285         classes_file = sys.argv[2]
 286     except IndexError:
 287         print >>sys.stderr, "Usage: %s attr-file.txt classes-file.txt" % (sys.argv[0])
 288         sys.exit(1)
 289
 290     print read_ms_schema(attr_file, classes_file)