terom@247: #!/usr/bin/env python
terom@247: 
terom@247: """
terom@247:     Process zonefiles.
terom@247: """
terom@247: 
terom@247: import codecs
terom@253: import datetime
tero@639: import ipaddr
terom@247: import logging
terom@270: import math
terom@254: import os.path
tero@639: import pvl.dns.labels
terom@247: 
terom@247: log = logging.getLogger('pvl.dns.zone')
terom@247: 
terom@247: class ZoneError (Exception) :
terom@247:     pass
terom@247: 
terom@247: class ZoneLineError (ZoneError) :
terom@247:     """
terom@247:         ZoneLine-related error
terom@247:     """
terom@247: 
terom@247:     def __init__ (self, line, msg, *args, **kwargs) :
terom@247:         super(ZoneLineError, self).__init__("%s: %s" % (line, msg.format(*args, **kwargs)))
terom@247: 
tero@639: def zone_quote (field):
tero@456:     """
tero@639:         Quote a value for inclusion into TXT record.
tero@456: 
tero@639:         >>> print zone_quote("foo")
tero@639:         "foo"
tero@639:         >>> print zone_quote("foo\\bar")
tero@639:         "foo\\bar"
tero@639:         >>> print zone_quote("foo \"bar\" quux")
tero@639:         "foo \"bar\" quux"
tero@456:     """
tero@456: 
tero@639:     return u'"' + field.replace('\\', '\\\\').replace('"', '\\"') + u'"'
tero@456: 
tero@456: 
terom@247: class ZoneLine (object) :
terom@247:     """
terom@250:         A line parsed from a zonefile.
terom@247:     """
terom@247: 
terom@250:     @classmethod
tero@639:     def load (cls, file) :
terom@292:         """
tero@639:             Load (ZoneDirective or ZoneRecord) items from the given zonefile.
tero@639: 
tero@639:             Tracks $ORIGIN and $TTL state for ZoneRecords/ZoneDirectives.
terom@292:         """
terom@292: 
tero@639:         for line in cls.parse(file) :
tero@639:             if line.parts[0].startswith('$'):
tero@639:                 # control directive
tero@639:                 line_type = ZoneDirective
terom@292: 
terom@292:             else :
tero@639:                 # normal record
tero@639:                 line_type = ZoneRecord
terom@292: 
tero@639:             yield line_type.parse(line.parts,
tero@639:                     line        = line,
tero@639:                     comment     = line.comment,
tero@639:             )
terom@292: 
terom@292:     @classmethod
tero@635:     def parse (cls, file, filename=None):
terom@250:         """
terom@250:             Yield ZoneLines lexed from a file.
terom@250:         """
terom@250:         
terom@250:         if not filename :
terom@250:             filename = file.name
terom@250:         
terom@250:         multiline_start = None
terom@250:         multiline_parts = None
terom@250:         
terom@250:         for lineno, raw_line in enumerate(file) :
terom@292:             raw_line = raw_line.rstrip('\n')
terom@292: 
terom@250:             log.debug("%s:%d: %s", filename, lineno, raw_line)
terom@250:             
terom@250:             # capture indent from raw line
terom@250:             indent = raw_line.startswith(' ') or raw_line.startswith('\t')
terom@250:             line = raw_line.strip()
terom@250: 
tero@639:             if not line:
tero@639:                 # empty line
tero@639:                 continue
tero@639: 
terom@250:             # parse comment
terom@250:             if ';' in line:
terom@250:                 line, comment = line.split(';', 1)
terom@250: 
terom@250:                 line = line.strip()
terom@250:                 comment = comment.strip()
terom@250:             else :
terom@250:                 comment = None
terom@250:            
terom@250:             log.debug("%s:%d: indent=%r, line=%r, comment=%r", filename, lineno, indent, line, comment)
terom@250: 
terom@250:             # split (quoted) fields
tero@639:             if indent and not multiline_start:
tero@639:                 parts = ['']
tero@639:             else:
tero@639:                 parts = []
tero@639: 
terom@250:             if '"' in line :
terom@250:                 pre, data, post = line.split('"', 2)
tero@639:                 parts.extend(pre.split() + [data] + post.split())
terom@250:                
terom@250:             else :
tero@639:                 parts.extend(line.split())
terom@250: 
terom@250:             # handle multi-line statements...
terom@250:             if '(' in parts :
terom@250:                 assert not multiline_start
terom@250: 
terom@292:                 log.debug("%s:%d: Start of multi-line statement: %s", filename, lineno, line)
terom@250: 
tero@639:                 multiline_start = (lineno, comment)
terom@250:                 multiline_line = raw_line
terom@250:                 multiline_parts = []
terom@250: 
terom@250:             if multiline_start:
terom@292:                 log.debug("%s:%d: Multi-line statement: %s", filename, lineno, line)
terom@250:                 
terom@250:                 # XXX: some better way to do this
terom@250:                 multiline_parts.extend([part for part in parts if part not in set('()')])
terom@250:                 multiline_line += raw_line
terom@250: 
terom@250:             if ')' in parts :
terom@250:                 assert multiline_start
terom@250: 
terom@292:                 log.debug("%s:%d: End of multi-line statement: %s", filename, lineno, line)
terom@250:                 
tero@639:                 lineno, comment = multiline_start
terom@250:                 raw_line = multiline_line
terom@250:                 parts = multiline_parts
terom@250: 
terom@250:                 multiline_start = multiline_line = multiline_parts = None
terom@250:         
terom@250:             # parse
terom@250:             if multiline_start:
terom@250:                 pass
terom@250:             else:
tero@639:                 yield ZoneLine(parts,
tero@639:                         file    = filename,
tero@639:                         lineno  = lineno,
tero@639:                         line    = raw_line,
tero@639:                         comment = comment,
tero@639:                 )
terom@247: 
tero@639:     def __init__ (self, parts, file=None, lineno=None, line=None, comment=None) :
tero@639:         """
tero@639:             parts : [str]       - list of parsed (quoted) fields
tero@639:                                   if the line has a leading indent, the first field should be an empty string
tero@639:         """
terom@247: 
tero@639:         self.parts = parts
tero@639: 
terom@248:         # source
terom@247:         self.file = file
terom@247:         self.lineno = lineno
terom@248:         self.line = line
tero@639: 
tero@639:         # meta
terom@247:         self.comment = comment
terom@247: 
terom@316:     def __unicode__ (self) :
tero@639:         return u"{parts}{comment}".format(
terom@316:                 parts       = u'\t'.join(self.parts),
tero@639:                 comment     = u'\t; ' + self.comment if self.comment is not None else '',
terom@316:         )
terom@316: 
terom@316:     def __repr__ (self) :
terom@247:         return "{file}:{lineno}".format(file=self.file, lineno=self.lineno)
terom@247: 
tero@639: 
terom@323: class ZoneDirective (object) :
terom@323:     """
terom@323:         An $DIRECTIVE in a zonefile.
terom@323:     """
terom@323:     
terom@323:     # context
terom@323:     line    = None
terom@323:     origin  = None
terom@323: 
terom@323:     # fields
terom@323:     directive   = None
terom@323:     arguments   = None
terom@323:     
terom@323:     @classmethod
tero@639:     def parse (cls, parts, **opts) :
tero@639:         """
tero@639:             Parse from ZoneLine.parts
tero@639:         """
terom@323: 
tero@639:         directive = parts[0][1:].upper()
tero@639:         arguments = parts[1:]
tero@639: 
tero@639:         return cls(directive, arguments, **opts)
terom@323: 
terom@323:     @classmethod
tero@639:     def build (cls, directive, *arguments, **opts):
tero@639:         """
tero@639:             Build directive from optional parts
tero@639:         """
terom@323: 
tero@639:         return cls(directive, arguments, **opts)
tero@639: 
tero@639:     def __init__ (self, directive, arguments, comment=None, line=None, origin=None):
tero@639:         """
tero@639:             directive       - uppercase directive name, withtout leading $
tero@639:             arguments [str] - list of directive arguments
tero@639:             comment         - optional trailing comment
tero@639:             line            - optional associated ZoneLine
tero@639:             origin          - context origin
tero@639:         """
tero@639: 
terom@323:         self.directive  = directive
terom@323:         self.arguments  = arguments
terom@323: 
tero@639:         self.comment = comment
terom@323:         self.line = line
terom@323:         self.origin = origin
terom@323: 
tero@639:     def __unicode__ (self):
terom@323:         """
terom@323:             Construct a zonefile-format line...
terom@323:         """
terom@323: 
tero@639:         if self.comment:
terom@323:             comment = '\t; ' + self.comment
terom@323:         else :
terom@323:             comment = ''
terom@323:             
terom@323:         return u"${directive}\t{arguments}{comment}".format(
terom@323:                 directive   = self.directive,
terom@323:                 arguments   = '\t'.join(self.arguments),
terom@323:                 comment     = comment,
terom@323:         )
terom@323: 
terom@323:     def __repr__ (self) :
terom@323:         return '%s(%s)' % (self.__class__.__name__, ', '.join(repr(arg) for arg in (
tero@639:             (self.directive, ) + tuple(self.arguments)
terom@323:         )))
terom@323: 
terom@323: 
tero@639: def process_generate (line, range, parts, **opts) :
tero@639:     """
tero@639:         Parse a
tero@639:             $GENERATE <start>-<stop>[/<step>] lhs [ttl] [class] type rhs [comment]
tero@639:         directive and yield expanded ZoneRecords.
tero@639: 
tero@639:         Raises ZoneLineError
tero@639:     """
tero@639:         
tero@639: 
tero@639:     try:
tero@639:         range = pvl.dns.generate.parse_generate_range(range)
tero@639:         
tero@639:         lhs_func = pvl.dns.generate.parse_generate_field(parts[0], line=line)
tero@639:         body = parts[1:-1]
tero@639:         rhs_func = pvl.dns.generate.parse_generate_field(parts[-1], line=line)
tero@639: 
tero@639:     except ValueError as error:
tero@639:         raise ZoneLineError(line, "{error}", error=error)
tero@639:                     
tero@639:     log.info("%s: generate %s: %s ... %s", line, range, lhs_func, rhs_func)
tero@639: 
tero@639:     for i in range :
tero@639:         # build
tero@639:         parts = [lhs_func(i)] + body + [rhs_func(i)]
tero@639: 
tero@639:         log.debug(" %03d: %r", i, parts)
tero@639: 
tero@639:         yield ZoneRecord.parse(parts, line=line, **opts)
tero@639: 
tero@639: def process_include (line, origin, include_filename, include_origin=None, **opts):
tero@639:     """
tero@639:         Parse a
tero@639:             $INCLUDE filename [ origin ]
tero@639:         directive and yield expanded ZoneRecords.
tero@639: 
tero@639:         Raises ZoneLineError.
tero@639:     """
tero@639: 
tero@639:     path = os.path.join(os.path.dirname(file.name), include_include)
tero@639: 
tero@639:     if include_origin:
tero@639:         origin = pvl.dns.labels.join(include_origin, origin)
tero@639: 
tero@639:     for record in ZoneRecord.load(open(path), origin, **opts) :
tero@639:         yield line, record
tero@639: 
terom@247: class ZoneRecord (object) :
terom@247:     """
terom@247:         A record from a zonefile.
terom@247:     """
terom@254: 
terom@247:     @classmethod
tero@639:     def load (cls, file, origin,
tero@639:             ttl=None,
tero@639:     ):
terom@247:         """
tero@639:             Yield ZoneRecords from a file. Processes any ZoneDirectives.
tero@639:         """
tero@639:         
tero@639:         name = None
terom@274: 
tero@639:         for line in ZoneLine.parse(file):
tero@639:             if line.parts[0].startswith('$'):
tero@639:                 directive = ZoneDirective.parse(line.parts,
tero@639:                         origin      = origin,
tero@639:                         line        = line,
tero@639:                         comment     = line.comment,
tero@639:                 )
tero@639: 
tero@639:                 log.debug("%s: %s", line, directive)
tero@639: 
tero@639:                 if directive.directive == 'ORIGIN':
tero@639:                     directive_origin, = directive.arguments
tero@639:                     
tero@639:                     log.info("%s: $ORIGIN %s -> %s", file, origin, directive_origin)
tero@639:                     
tero@639:                     origin = pvl.dns.labels.join(origin, directive_origin)
tero@639: 
tero@639:                 elif directive.directive == 'TTL' :
tero@639:                     directive_ttl, = directive.arguments
tero@639:                     
tero@639:                     log.info("%s: $TTL %d -> %s", file, ttl, directive_ttl)
tero@639:                     
tero@639:                     ttl = int(directive_ttl)
tero@639: 
tero@639:                 elif directive.directive == 'GENERATE' :
tero@639:                     for record in process_generate(line, directive.arguments,
tero@639:                             name        = name,
tero@639:                             ttl         = ttl,
tero@639:                             origin      = origin,
tero@639:                     ) :
tero@639:                         yield record
tero@639: 
tero@639:                 elif directive.directive == 'INCLUDE' :
tero@639:                     for record in process_include(line, origin, 
tero@639:                             ttl         = ttl,
tero@639:                             *directive.arguments
tero@639:                     ):
tero@639:                         yield record
tero@639: 
tero@639:                 else :
tero@639:                     log.warn("%s: skip unknown control record: %r", line, directive)
tero@639:                     yield line, None
tero@639: 
tero@639:             else:
tero@639:                 record = ZoneRecord.parse(line.parts,
tero@639:                         name        = name,
tero@639:                         ttl         = ttl,
tero@639:                         comment     = line.comment,
tero@639:                         line        = line,
tero@639:                         origin      = origin,
tero@639:                 )
tero@639:                 
tero@639:                 log.debug("%s: %s", line, record)
tero@639: 
tero@639:                 # keep name across lines
tero@639:                 name = record.name
tero@639:                 
tero@639:                 yield record
tero@639: 
tero@639:     @classmethod
tero@639:     def parse (cls, parts, name=None, ttl=None, line=None, **opts) :
tero@639:         """
tero@639:             Build a ZoneRecord from ZoneLine.parts.
tero@639: 
tero@639:                 name    - context for name, if continuing previous line
tero@639:                 ttl     - context for ttl, if using $TTL
terom@274: 
terom@274:             Return: (name, ZoneRecord)
terom@247:         """
terom@247:         
tero@639:         parts = list(parts)
terom@247: 
tero@639:         # first field is either name or leading whitespace
tero@639:         leading = parts.pop(0)
terom@247:         
tero@639:         if leading:
tero@639:             name = leading
tero@639: 
terom@247:         if len(parts) < 2 :
tero@639:             raise ZoneLineError(line, "Too few parts to parse: {parts!r}", parts=parts)
terom@247: 
terom@247:         # parse ttl/cls/type
terom@254:         _cls = None
terom@247: 
terom@247:         if parts and parts[0][0].isdigit() :
tero@639:             ttl = int(parts.pop(0))
terom@247: 
terom@247:         if parts and parts[0].upper() in ('IN', 'CH') :
tero@639:             _cls = parts.pop(0).upper()
terom@247: 
terom@247:         # always have type
tero@639:         type = parts.pop(0).upper()
terom@247: 
terom@247:         # remaining parts are data
terom@247:         data = parts
terom@247: 
terom@247:         log.debug("  ttl=%r, cls=%r, type=%r, data=%r", ttl, _cls, type, data)
terom@247: 
tero@639:         return cls(name, ttl, _cls, type, data, line=line, **opts)
terom@252:     
terom@252:     @classmethod
tero@639:     def build (_cls, name, type, *data, **opts):
tero@639:         """
tero@639:             Simple interface to build ZoneRecord from required parts. All optional fields must be given as keyword arguments.
terom@247: 
tero@639:             Normalizes all fields to strs.
tero@639:         """
tero@639:         
tero@639:         # keyword-only arguments
tero@639:         ttl = opts.pop('ttl', None)
tero@639:         cls = opts.pop('cls', None)
terom@308: 
tero@639:         if name:
tero@639:             name = str(name)
tero@639:         else:
tero@639:             name = None
tero@639: 
tero@639:         if ttl or ttl == 0:
tero@639:             ttl = int(ttl)
tero@639:         else:
tero@639:             ttl = None
tero@639: 
tero@639:         if cls:
tero@639:             cls = cls.upper()
tero@639:         else:
tero@639:             cls = None
tero@639: 
tero@639:         type = type.upper()
tero@639: 
tero@639:         data = [unicode(item) for item in data]
tero@639: 
tero@639:         return _cls(name, ttl, cls, type, data, **opts)
terom@267: 
terom@267:     @classmethod
tero@639:     def A (_cls, name, ip4, **opts):
tero@639:         """
tero@639:             Build from ipaddr.IPv4Address.
tero@639:         """
tero@639: 
tero@639:         return _cls.build(name, 'A', ipaddr.IPv4Address(ip4), **opts)
terom@267: 
terom@267:     @classmethod
tero@639:     def AAAA (_cls, name, ip6, **opts):
tero@639:         """
tero@639:             Build from ipaddr.IPv6Address.
tero@639:         """
tero@639: 
tero@639:         return _cls.build(name, 'AAAA', ipaddr.IPv6Address(ip6), **opts)
terom@247: 
terom@247:     @classmethod
tero@639:     def CNAME (_cls, name, alias, **opts):
tero@639:         return _cls.build(name, 'CNAME', alias, **opts)
terom@247: 
terom@247:     @classmethod
tero@639:     def TXT (_cls, name, text, **opts):
tero@639:         """
tero@639:             Build from quoted (unicode) value.
tero@639:         """
terom@247: 
tero@639:         return _cls.build(name, 'TXT', zone_quote(unicode(text)), **opts)
tero@639: 
tero@639:     @classmethod
tero@639:     def PTR (_cls, name, ptr, **opts):
tero@639:         return _cls.build(name, 'PTR', ptr, **opts)
tero@639: 
tero@639:     @classmethod
tero@639:     def MX (_cls, name, priority, mx, **opts):
tero@639:         """
tero@639:                 priority    : int
tero@639:                 mx          : str       - hostname
tero@639:         """
tero@639: 
tero@639:         return _cls.build(name, 'MX', int(priority), str(mx), **opts)
tero@639: 
tero@639:     def __init__ (self, name, ttl, cls, type, data, comment=None, origin=None, line=None):
tero@639:         """
tero@639:             Using strict field ordering.
tero@639: 
tero@639:                 name            - local label with respect to $ORIGIN
tero@639:                                   may also be @ to refer to $ORIGIN
tero@639:                 ttl             - int TTL for record, default to implict $TTL
tero@639:                 cls             - uppercase class for record, default to IN
tero@639:                 type            - uppercase type for record, required
tero@639:                 data [...]      - list of data fields, interpretation varies by type
tero@639:                 comment         - optional comment to include in zone
tero@639:                 origin          - track implicit $ORIGIN or XXX: previous record state
tero@639:                 line            - associated ZoneLine
tero@639:         """
tero@639: 
terom@247:         self.name = name
tero@639:         self.ttl = ttl
tero@639:         self.cls = cls
terom@247:         self.type = type
terom@247:         self.data = data
terom@247:         
tero@639:         self.comment = comment
tero@639:         self.origin = origin
terom@294:         self.line = line
terom@247: 
terom@247:     def __unicode__ (self) :
terom@247:         """
terom@247:             Construct a zonefile-format line..."
terom@247:         """
terom@247: 
terom@247:         if self.comment :
terom@378:             comment = '\t; ' + self.comment
terom@247:         else :
terom@247:             comment = ''
terom@247:             
terom@247:         return u"{name:25} {ttl:4} {cls:2} {type:5} {data}{comment}".format(
tero@639:                 name    = '' if self.name is None else self.name,
tero@639:                 ttl     = '' if self.ttl is None else self.ttl,
tero@639:                 cls     = '' if self.cls is None else self.cls,
terom@247:                 type    = self.type,
terom@247:                 data    = ' '.join(unicode(data) for data in self.data),
terom@247:                 comment = comment,
terom@247:         )
terom@247: 
terom@247:     def __repr__ (self) :
terom@247:         return '%s(%s)' % (self.__class__.__name__, ', '.join(repr(arg) for arg in (
tero@639:             self.name, self.ttl, self.cls, self.type, self.data
terom@247:         )))
terom@247: 
terom@252: class SOA (ZoneRecord) :
terom@252:     @classmethod
terom@294:     def build (cls, line, name, ttl, _cls, type, data, **opts) :
terom@252:         assert name == '@'
terom@252: 
terom@294:         return cls(*data,
terom@252:             ttl     = ttl,
terom@252:             cls     = cls,
terom@252:             line    = line,
terom@252:             **opts
terom@252:         )
terom@252: 
terom@252:     def __init__ (self, master, contact, serial, refresh, retry, expire, nxttl, **opts) :
terom@252:         super(SOA, self).__init__('@', 'SOA',
terom@252:             [master, contact, serial, refresh, retry, expire, nxttl],
terom@252:             **opts
terom@252:         )
terom@252: 
terom@252:         self.master = master
terom@252:         self.contact = contact
terom@252:         self.serial = serial
terom@252:         self.refresh = refresh
terom@252:         self.retry = retry
terom@252:         self.expire = expire
terom@252:         self.nxttl = nxttl
terom@252: