#
# (C) Copyright 2006-2007 Novell, Inc.
#
# This library is free software; you can redistribute it and/or
# modify it under the terms of the GNU Lesser General Public
# License as published by the Free Software Foundation; either
# version 2.1 of the License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
# Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public
# License along with this program; if not, write to the Free Software
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
#
# Author: Bart Whiteley <bwhiteley suse.de>
# Author: Ross Peoples <ross.peoples@gmail.com>
#
"""
The language in which CIM classes, CIM Instances, etc. are specified, is
called `MOF` (for Managed Object Format). It is defined in :term:`DSP0004`.
MOF compilers take MOF files as input, compile them and use the result
(CIM classes, instances, and/or qualifier declarations) to update a target
CIM repository. The repository may initially be empty, or may contain CIM
classes, instances, and/or qualifier declarations that are used to resolve
dependencies the new MOF compilation may have.
The pywbem package includes a MOF compiler that is provided in two forms:
* as an API (described in this chapter)
* as a command (described in section :ref:`mof_compiler`)
The pywbem MOF compiler will compile MOF files whose syntax complies with
:term:`DSP0004`, with some limitations:
1. Although there is no formal keyword list of illegal words
for property/parameter.etc. names , there is a list of mof syntax tokens
in :term:`DSP0004` section A.3. Generally these should not be used as
property names. The pywbem MOF compiler largely enforces this so that words
like 'indication' are not allowed as property/parameter/etc. names.
2. The key properties of instances with aliases must be initialized in the
instance specification, or their default values must be non-NULL.
(See `pywbem issue #1079 <https://github.com/pywbem/pywbem/issues/1079>`_).
3. An alias must be defined before it is used. In :term:`DSP0004`, no such
requirement is documented.
(See `pywbem issue #1079 <https://github.com/pywbem/pywbem/issues/1079>`_).
The MOF compiler API provides for invoking the MOF compiler, it supports
plugging in a user-provided CIM repository (see
:class:`~pywbem.BaseRepositoryConnection`), and it supports a rollback
capability that undoes compilation effects (see
:meth:`~pywbem.MOFCompiler.rollback`).
The MOF compiler API consists of the following parts, which are described in the
remaining sections of this chapter:
* :ref:`MOFCompiler Class` - Describes the :class:`~pywbem.MOFCompiler`
class, which allows invoking the MOF compiler programmatically.
* :ref:`Repository connections` - Describes the
:class:`~pywbem.BaseRepositoryConnection` class that defines
the interface for connecting to a CIM repository, and the
:class:`~pywbem.MOFWBEMConnection` class that is a connection
to an in-memory repository on top of an underlying repository, and is used
by the MOF compiler to provide rollback support.
* :ref:`Exceptions <MOF compiler exceptions>` - Describes the exceptions
that can be raised by the MOF compiler API.
"""
from __future__ import print_function, absolute_import
import sys
import os
import re
import tempfile
# import logging
from abc import ABCMeta, abstractmethod
try:
from collections import OrderedDict
except ImportError:
from ordereddict import OrderedDict
import six
from ply import yacc, lex
from ._nocasedict import NocaseDict
from .cim_obj import CIMInstance, CIMInstanceName, CIMClass, CIMProperty, \
CIMMethod, CIMParameter, CIMQualifier, CIMQualifierDeclaration, \
cimvalue
from .cim_operations import WBEMConnection
from ._server import WBEMServer
from .cim_constants import CIM_ERR_NOT_FOUND, CIM_ERR_FAILED, \
CIM_ERR_ALREADY_EXISTS, CIM_ERR_INVALID_NAMESPACE, \
CIM_ERR_INVALID_SUPERCLASS, CIM_ERR_INVALID_PARAMETER, \
CIM_ERR_NOT_SUPPORTED, CIM_ERR_INVALID_CLASS, _statuscode2string
from .exceptions import Error, CIMError
from ._utils import _format
__all__ = ['MOFParseError', 'MOFWBEMConnection', 'MOFCompiler',
'BaseRepositoryConnection']
# The following pylint is applied for the complete file because invalid
# names are used throughout the file and about 200 flags generated if
# this is not applied and at least some # may be part of ply rules.
# pylint: disable=invalid-name
_optimize = 1
_tabmodule = 'mofparsetab'
_lextab = 'moflextab'
# Directory for _tabmodule and _lextab
_tabdir = os.path.dirname(os.path.abspath(__file__))
# -----------------------------------------------------------------------------
#
# IMPORTANT NOTE:
#
# This MOF compiler implementation is based on the PLY Python package.
# This module here contains LEX & YACC rules in docstrings of its functions.
# The formatting of these docstrings is critical in that it defines the parser
# functionality. These docstrings are processed by the LEX & YACC in PLY.
# Changing the strings or even the formatting breaks the PLY rule generation!
#
# In the YACC of PLY 2.3 (included in pywbem up to 0.7), the requirement was
# that each choice of a YACC rule needed to be on a single line, and the first
# choice needed to be on the same line as the rule name. Not sure what the
# requirements of the current PLY version are.
#
# The LEX token functions and strings are all named "t_*".
# Note the order of LEX token processing, described in:
# https://www.dabeaz.com/ply/ply.html#ply_nn6
#
# The YACC parser functions are all named "p_*".
#
# -----------------------------------------------------------------------------
reserved = {
'any': 'ANY',
'as': 'AS',
'association': 'ASSOCIATION',
'class': 'CLASS',
'disableoverride': 'DISABLEOVERRIDE',
'boolean': 'DT_BOOL',
'char16': 'DT_CHAR16',
'datetime': 'DT_DATETIME',
'pragma': 'PRAGMA',
'real32': 'DT_REAL32',
'real64': 'DT_REAL64',
'sint16': 'DT_SINT16',
'sint32': 'DT_SINT32',
'sint64': 'DT_SINT64',
'sint8': 'DT_SINT8',
'string': 'DT_STR',
'uint16': 'DT_UINT16',
'uint32': 'DT_UINT32',
'uint64': 'DT_UINT64',
'uint8': 'DT_UINT8',
'enableoverride': 'ENABLEOVERRIDE',
'false': 'FALSE',
'flavor': 'FLAVOR',
'indication': 'INDICATION',
'instance': 'INSTANCE',
'method': 'METHOD',
'null': 'NULL',
'of': 'OF',
'parameter': 'PARAMETER',
'property': 'PROPERTY',
'qualifier': 'QUALIFIER',
'ref': 'REF',
'reference': 'REFERENCE',
'restricted': 'RESTRICTED',
'schema': 'SCHEMA',
'scope': 'SCOPE',
'tosubclass': 'TOSUBCLASS',
'toinstance': 'TOINSTANCE',
'translatable': 'TRANSLATABLE',
'true': 'TRUE',
} # noqa: E123
tokens = list(reserved.values()) + [
'IDENTIFIER',
'stringValue',
'floatValue',
'charValue',
'binaryValue',
'octalValue',
'decimalValue',
'hexValue',
]
literals = '#(){};[],$:='
# UTF-8 (from Unicode 4.0.0 standard):
# Table 3-6. Well-Formed UTF-8 Byte Sequences Code Points
# 1st Byte 2nd Byte 3rd Byte 4th Byte
# U+0000..U+007F 00..7F
# U+0080..U+07FF C2..DF 80..BF
# U+0800..U+0FFF E0 A0..BF 80..BF
# U+1000..U+CFFF E1..EC 80..BF 80..BF
# U+D000..U+D7FF ED 80..9F 80..BF
# U+E000..U+FFFF EE..EF 80..BF 80..BF
# U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
# U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
# U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
utf8_2 = r'[\xC2-\xDF][\x80-\xBF]'
utf8_3_1 = r'\xE0[\xA0-\xBF][\x80-\xBF]'
utf8_3_2 = r'[\xE1-\xEC][\x80-\xBF][\x80-\xBF]'
utf8_3_3 = r'\xED[\x80-\x9F][\x80-\xBF]'
utf8_3_4 = r'[\xEE-\xEF][\x80-\xBF][\x80-\xBF]'
utf8_4_1 = r'\xF0[\x90-\xBF][\x80-\xBF][\x80-\xBF]'
utf8_4_2 = r'[\xF1-\xF3][\x80-\xBF][\x80-\xBF][\x80-\xBF]'
utf8_4_3 = r'\xF4[\x80-\x8F][\x80-\xBF][\x80-\xBF]'
utf8Char = r'({0})|({1})|({2})|({3})|({4})|({5})|({6})|({7})'.format(
utf8_2, utf8_3_1, utf8_3_2, utf8_3_3, utf8_3_4, utf8_4_1, utf8_4_2,
utf8_4_3)
# pylint: disable=unused-argument
def t_COMMENT(t):
r'//.*'
return # discard token
def t_MCOMMENT(t):
r'/\*(.|\n)*?\*/'
t.lineno += t.value.count('\n')
return # discard token
# These simple tokens must also be defined as functions, in order to control
# the order of evaluation.
def t_floatValue(t):
r'[+-]?[0-9]*\.[0-9]+([eE][+-]?[0-9]+)?'
t.value = float(t.value)
return t
def t_hexValue(t):
r'[+-]?0[xX][0-9a-fA-F]+'
t.value = int(t.value, 16)
return t
def t_binaryValue(t):
r'[+-]?[0-9]+[bB]'
# We must match [0-9], and then check the validity of the binary number.
# If we match [0-1], the invalid binary number "2b" would match
# 'decimalValue' 2 and 'IDENTIFIER 'b'.
if re.search(r'[2-9]', t.value) is not None:
msg = _format("Invalid binary number {0!A}", t.value)
t.lexer.last_msg = msg
t.type = 'error'
# Setting error causes the value to be automatically skipped
else:
t.value = int(t.value[0:-1], 2)
return t
def t_octalValue(t):
r'[+-]?0[0-9]+'
# We must match [0-9], and then check the validity of the octal number.
# If we match [0-7], the invalid octal number "08" would match
# 'decimalValue' 0 and 'decimalValue' 8.
if re.search(r'[8-9]', t.value) is not None:
msg = _format("Invalid octal number {0!A}", t.value)
t.lexer.last_msg = msg
t.type = 'error'
# Setting error causes the value to be automatically skipped
else:
t.value = int(t.value, 8)
return t
# Matching for decimal must be at the end of the other numerics because of
# the 0. If not at the end, 0 would match at the begin of e.g. an octal value.
def t_decimalValue(t):
r'[+-]?([1-9][0-9]*|0)'
t.value = int(t.value)
return t
simpleEscape = r"""[bfnrt'"\\]"""
hexEscape = r'[xX][0-9a-fA-F]{1,4}'
escapeSequence = r'[\\](({0})|({1}))'.format(simpleEscape, hexEscape)
cChar = r"[^'\\\n\r]|({0})".format(escapeSequence)
sChar = r'[^"\\\n\r]|({0})'.format(escapeSequence)
charvalue_re = r"'({0})'".format(cChar)
@lex.TOKEN(charvalue_re)
def t_charValue(t): # pylint: disable=missing-docstring
return t
stringvalue_re = r'"({0})*"'.format(sChar)
@lex.TOKEN(stringvalue_re)
def t_stringValue(t): # pylint: disable=missing-docstring
return t
identifier_re = r'([a-zA-Z_]|({0}))([0-9a-zA-Z_]|({1}))*'.format( # noqa E501 pylint: disable=W1308
utf8Char, utf8Char)
@lex.TOKEN(identifier_re)
def t_IDENTIFIER(t): # pylint: disable=missing-docstring
t.type = reserved.get(t.value.lower(), 'IDENTIFIER')
return t
def t_newline(t): # pylint: disable=missing-docstring
r'\n+'
t.lexer.lineno += len(t.value)
t.lexer.linestart = t.lexpos
return # discard token
t_ignore = ' \r\t'
def t_error(t):
""" Lexer error callback from PLY Lexer with token in error.
"""
msg = _format("Illegal character {0!A}", t.value[0])
t.lexer.last_msg = msg
t.lexer.skip(1)
return t # Return the error token for the YACC parser to handle
[docs]class MOFParseError(Error):
"""
Exception raised when MOF cannot be parsed correctly, e.g. for
syntax errors. Derived from :exc:`~pywbem.Error`.
"""
# pylint: disable=super-init-not-called
def __init__(self, parser_token=None, msg=None):
"""
Parameters:
parser_token:
PLY parser token for the error (that is, the ``p`` argument
of a PLY parser function). This token contains information on
the location of the error in the MOF file, which is copied
into this object, and is accessible via properties.
msg (:term:`string`):
Message text supplied by the creator of the error
"""
if parser_token is None:
self.args = (None, None, None, None)
else:
mof_ = parser_token.lexer.parser.mof
self.args = (parser_token.lineno,
_find_column(mof_, parser_token),
parser_token.lexer.parser.file,
_get_error_context(mof_, parser_token))
self._msg = msg
@property
def lineno(self):
"""
:term:`integer`: Line number in the MOF file or MOF string where the
error occurred (1-based).
"""
return self.args[0]
@property
def column(self):
"""
:term:`integer`: Position within the line in the MOF file or MOF string
where the error occurred (1-based).
"""
return self.args[1]
@property
def file(self):
"""
:term:`string`: File name of the MOF file where the error occurred.
`None` if the error occurred in a MOF string that was compiled.
"""
return self.args[2]
@property
def context(self):
"""
:term:`string`: MOF context, consisting of a first line that is the
MOF line in error, and a second line that uses the '^' character to
indicate the position of the token in error in the MOF line.
"""
return self.args[3]
@property
def msg(self):
"""
:term:`string`: Brief description of the problem.
"""
return self._msg
def __str__(self):
ret_str = 'MOFParseError:\n'
if self.lineno is not None:
ret_str += _format('{0}:{1}:{2} msg={3}\n{4}',
self.file, self.lineno, self.column,
self.msg, self.context)
else:
ret_str += _format("{0}", self.msg)
return ret_str
[docs] def get_err_msg(self):
"""
Return a multi-line error message for being printed, in the following
format. The text in angle brackets refers to the same-named properties
of the exception instance:
::
Syntax error:<file>:<lineno>: <msg>
<context - MOF line>
<context - position indicator line>
Returns:
:term:`string`: Multi-line error message.
"""
ret_str = 'Syntax error:'
disp_file = 'NoFile' if self.file is None else self.file
if self.lineno is not None:
ret_str += _format("{0}:{1}:{2}",
disp_file, self.lineno, self.column)
if self.msg:
ret_str += _format(" {0}", self.msg)
if self.context is not None:
ret_str += '\n'
ret_str += '\n'.join(self.context)
return ret_str
def p_error(p):
"""
YACC Error Callback from the parser. The parameter is the token
in error and contains information on the file and position of the
error. If p is `None`, PLY is returning eof error.
"""
if p is None:
raise MOFParseError(msg='Unexpected end of file')
msg = p.lexer.last_msg
p.lexer.last_msg = None
raise MOFParseError(parser_token=p, msg=msg)
# pylint: disable=unused-argument
def p_mofSpecification(p):
"""mofSpecification : mofProductionList"""
# pylint: disable=unused-argument
def p_mofProductionList(p):
"""mofProductionList : empty
| mofProductionList mofProduction
"""
# pylint: disable=unused-argument
def p_mofProduction(p):
"""mofProduction : compilerDirective
| mp_createClass
| mp_setQualifier
| mp_createInstance
"""
def p_mp_createClass(p):
"""mp_createClass : classDeclaration
"""
# pylint: disable=too-many-branches,too-many-statements,too-many-locals
ns = p.parser.handle.default_namespace
cc = p[1]
try:
fixedNS = fixedRefs = fixedSuper = False
while not fixedNS or not fixedRefs or not fixedSuper:
try:
if p.parser.verbose:
p.parser.log(
_format("Creating class {0!A}:{1!A}", ns, cc.classname))
p.parser.handle.CreateClass(cc)
if p.parser.verbose:
p.parser.log(
_format("Created class {0!A}:{1!A}", ns, cc.classname))
p.parser.classnames[ns].append(cc.classname.lower())
break
except CIMError as ce:
ce.file_line = (p.parser.file, p.lexer.lineno)
errcode = ce.status_code
if errcode == CIM_ERR_INVALID_NAMESPACE:
if fixedNS:
raise
if p.parser.verbose:
p.parser.log(
_format("Creating namespace {0!A}", ns))
p.parser.server.create_namespace(ns)
fixedNS = True
continue
if not p.parser.search_paths:
raise
if errcode == CIM_ERR_INVALID_SUPERCLASS:
if fixedSuper:
raise
moffile = p.parser.mofcomp.find_mof(cc.superclass)
if not moffile:
raise
p.parser.mofcomp.compile_file(moffile, ns)
fixedSuper = True
elif errcode in [CIM_ERR_INVALID_PARAMETER,
CIM_ERR_NOT_FOUND,
CIM_ERR_FAILED]:
if fixedRefs:
raise
if not p.parser.qualcache[ns]:
for fname in ['qualifiers', 'qualifiers_optional']:
qualfile = p.parser.mofcomp.find_mof(fname)
if qualfile:
p.parser.mofcomp.compile_file(qualfile, ns)
if not p.parser.qualcache[ns]:
# can't find qualifiers
raise
objects = list(cc.properties.values())
for meth in cc.methods.values():
objects += list(meth.parameters.values())
dep_classes = NocaseDict() # dict dep_class, ce
for obj in objects:
if obj.type not in ['reference', 'string']:
continue
if obj.type == 'reference':
if obj.reference_class not in dep_classes:
dep_classes[obj.reference_class] = ce
elif obj.type == 'string':
try:
embedded_inst = \
obj.qualifiers['embeddedinstance']
except KeyError:
continue
if embedded_inst.value not in dep_classes:
dep_classes[embedded_inst.value] = ce
continue
for cln, err in dep_classes.items():
if cln in p.parser.classnames[ns]:
continue
try:
# don't limit it with LocalOnly=True,
# PropertyList, IncludeQualifiers=False, ...
# because of caching in case we're using the
# special WBEMConnection subclass used for
# removing schema elements
p.parser.handle.GetClass(cln,
LocalOnly=False,
IncludeQualifiers=True)
p.parser.classnames[ns].append(cln)
except CIMError:
moffile = p.parser.mofcomp.find_mof(cln)
if not moffile:
raise err
try:
if p.parser.verbose:
p.parser.log(
_format("Class {0!A} namespace {1!A} "
"depends on class {2!A} which "
"is not in repository.",
cc.classname, ns, cln))
p.parser.mofcomp.compile_file(moffile, ns)
except CIMError as ce:
if ce.status_code == CIM_ERR_NOT_FOUND:
raise err
raise
p.parser.classnames[ns].append(cln)
fixedRefs = True
else:
raise
except CIMError as ce:
ce.file_line = (p.parser.file, p.lexer.lineno)
if ce.status_code != CIM_ERR_ALREADY_EXISTS:
raise
if p.parser.verbose:
p.parser.log(
_format("Class {0!A} already exist. Modifying...",
cc.classname))
try:
p.parser.handle.ModifyClass(cc, ns)
except CIMError as ce:
p.parser.log(
_format("Error modifying class {0!A}: {1}, {2}",
cc.classname, ce.status_code, ce.status_description))
def p_mp_createInstance(p):
"""mp_createInstance : instanceDeclaration"""
inst = p[1]
if p.parser.verbose:
p.parser.log(
_format("Creating instance of {0!A}.", inst.classname))
try:
p.parser.handle.CreateInstance(inst)
except CIMError as ce:
if ce.status_code == CIM_ERR_ALREADY_EXISTS:
if p.parser.verbose:
p.parser.log(
_format("Instance of class {0!A} already exist. "
"Modifying...", inst.classname))
try:
p.parser.handle.ModifyInstance(inst)
except CIMError as ce:
if ce.status_code == CIM_ERR_NOT_SUPPORTED:
if p.parser.verbose:
p.parser.log(
_format("ModifyInstance not supported. "
"Deleting instance of {0!A}: {1}",
inst.classname, inst.path))
p.parser.handle.DeleteInstance(inst.path)
if p.parser.verbose:
p.parser.log(
_format("Creating instance of {0!A}.",
inst.classname))
p.parser.handle.CreateInstance(inst)
else:
ce.file_line = (p.parser.file, p.lexer.lineno)
raise
def p_mp_setQualifier(p):
"""mp_setQualifier : qualifierDeclaration"""
qualdecl = p[1]
ns = p.parser.handle.default_namespace
if p.parser.verbose:
p.parser.log(
_format("Setting qualifier {0!A}", qualdecl.name))
try:
p.parser.handle.SetQualifier(qualdecl)
except CIMError as ce:
if ce.status_code == CIM_ERR_INVALID_NAMESPACE:
if p.parser.verbose:
p.parser.log(
_format("Creating namespace {0!A}", ns))
p.parser.server.create_namespace(ns)
if p.parser.verbose:
p.parser.log(
_format("Setting qualifier {0!A}", qualdecl.name))
p.parser.handle.SetQualifier(qualdecl)
elif ce.status_code == CIM_ERR_NOT_SUPPORTED:
if p.parser.verbose:
p.parser.log(
_format("Qualifier {0!A} already exists. Deleting...",
qualdecl.name))
p.parser.handle.DeleteQualifier(qualdecl.name)
if p.parser.verbose:
p.parser.log(
_format("Setting qualifier {0!A}", qualdecl.name))
p.parser.handle.SetQualifier(qualdecl)
else:
ce.file_line = (p.parser.file, p.lexer.lineno)
raise
p.parser.qualcache[ns][qualdecl.name] = qualdecl
def p_compilerDirective(p):
"""compilerDirective : '#' PRAGMA pragmaName '(' pragmaParameter ')'"""
directive = p[3].lower()
param = p[5]
if directive == 'include':
fname = param
if p.parser.file:
if os.path.dirname(p.parser.file):
fname = os.path.join(os.path.dirname(p.parser.file),
fname)
p.parser.mofcomp.compile_file(fname, p.parser.handle.default_namespace)
elif directive == 'namespace':
p.parser.handle.default_namespace = param
if param not in p.parser.qualcache:
p.parser.qualcache[param] = NocaseDict()
p[0] = None
def p_pragmaName(p):
"""pragmaName : identifier"""
p[0] = p[1]
def p_pragmaParameter(p):
"""pragmaParameter : stringValue"""
p[0] = _fixStringValue(p[1], p)
def p_classDeclaration(p):
# pylint: disable=line-too-long
"""classDeclaration : CLASS className '{' classFeatureList '}' ';'
| CLASS className superClass '{' classFeatureList '}' ';'
| CLASS className alias '{' classFeatureList '}' ';'
| CLASS className alias superClass '{' classFeatureList '}' ';'
| qualifierList CLASS className '{' classFeatureList '}' ';'
| qualifierList CLASS className superClass '{' classFeatureList '}' ';'
| qualifierList CLASS className alias '{' classFeatureList '}' ';'
| qualifierList CLASS className alias superClass '{' classFeatureList '}' ';'
""" # noqa: E501
superclass = None
alias = None
quals = []
if isinstance(p[1], six.string_types): # no class qualifiers
cname = p[2]
if p[3][0] == '$': # alias present
alias = p[3]
if p[4] == '{': # no superclass
cfl = p[5]
else: # superclass
superclass = p[4]
cfl = p[6]
else: # no alias
if p[3] == '{': # no superclass
cfl = p[4]
else: # superclass
superclass = p[3]
cfl = p[5]
else: # class qualifiers
quals = p[1]
cname = p[3]
if p[4][0] == '$': # alias present
alias = p[4]
if p[5] == '{': # no superclass
cfl = p[6]
else: # superclass
superclass = p[5]
cfl = p[7]
else: # no alias
if p[4] == '{': # no superclass
cfl = p[5]
else: # superclass
superclass = p[4]
cfl = p[6]
quals = OrderedDict([(x.name, x) for x in quals])
methods = OrderedDict()
props = OrderedDict()
for item in cfl:
item.class_origin = cname
if isinstance(item, CIMMethod):
methods[item.name] = item
else:
props[item.name] = item
p[0] = CIMClass(cname, properties=props, methods=methods,
superclass=superclass, qualifiers=quals)
if alias:
p.parser.aliases[alias] = p[0]
def p_classFeatureList(p):
"""classFeatureList : empty
| classFeatureList classFeature
"""
if len(p) == 2:
p[0] = []
else:
p[0] = p[1] + [p[2]]
def p_qualifierListEmpty(p):
"""qualifierListEmpty : empty
| qualifierListEmpty ',' qualifier
"""
if len(p) == 2:
p[0] = []
else:
p[0] = p[1] + [p[3]]
def p_className(p):
"""className : identifier"""
p[0] = p[1]
def p_alias(p):
"""alias : AS aliasIdentifier"""
p[0] = p[2]
def p_aliasIdentifier(p):
"""aliasIdentifier : '$' identifier"""
p[0] = '${0}'.format(p[2])
def p_superClass(p):
"""superClass : ':' className"""
p[0] = p[2]
def p_classFeature(p):
"""classFeature : propertyDeclaration
| methodDeclaration
| referenceDeclaration
"""
p[0] = p[1]
def p_qualifierList(p):
"""qualifierList : '[' qualifier qualifierListEmpty ']'"""
p[0] = [p[2]] + p[3]
def p_qualifier(p):
"""qualifier : qualifierName
| qualifierName ':' flavorList
| qualifierName qualifierParameter
| qualifierName qualifierParameter ':' flavorList
"""
# pylint: disable=too-many-branches
qname = p[1]
ns = p.parser.handle.default_namespace
qval = None
flavorlist = []
if len(p) == 3:
qval = p[2]
elif len(p) == 4:
flavorlist = p[3]
elif len(p) == 5:
qval = p[2]
flavorlist = p[4]
try:
qualdecl = p.parser.qualcache[ns][qname]
except KeyError:
try:
quals = p.parser.handle.EnumerateQualifiers()
except CIMError as ce:
if ce.status_code != CIM_ERR_INVALID_NAMESPACE:
ce.file_line = (p.parser.file, p.lexer.lineno)
raise
if p.parser.verbose:
p.parser.log(
_format("Creating namespace {0!A}", ns))
p.parser.server.create_namespace(ns)
quals = None
if quals:
for qual in quals:
p.parser.qualcache[ns][qual.name] = qual
else:
for fname in ['qualifiers', 'qualifiers_optional']:
qualfile = p.parser.mofcomp.find_mof(fname)
if qualfile:
p.parser.mofcomp.compile_file(qualfile, ns)
try:
qualdecl = p.parser.qualcache[ns][qname]
except KeyError:
ce = CIMError(
CIM_ERR_FAILED,
_format("Unknown Qualifier: {0!A}", qname))
ce.file_line = (p.parser.file, p.lexer.lineno)
raise ce
flavors = _build_flavors(p[0], flavorlist, qualdecl)
if qval is None:
if qualdecl.type == 'boolean':
qval = True
else:
qval = qualdecl.value # default value
else:
qval = cimvalue(qval, qualdecl.type)
p[0] = CIMQualifier(qname, qval, type=qualdecl.type, **flavors)
# Note: The propagated flag is not set because this is parsed MOF, which
# contains specified qualifiers and not propagated qualifiers.
def p_flavorList(p):
"""flavorList : flavor
| flavorList flavor
"""
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1] + [p[2]]
def p_qualifierParameter(p):
"""qualifierParameter : '(' constantValue ')'
| arrayInitializer
"""
if len(p) == 2:
p[0] = p[1]
else:
p[0] = p[2]
# Note: TOINSTANCE is deprecated in DSP0201 and is not specified in DSP004.
# Pywbem supports TOINSTANCE as deprecated, for historical reasons.
def p_flavor(p):
"""flavor : ENABLEOVERRIDE
| DISABLEOVERRIDE
| RESTRICTED
| TOSUBCLASS
| TOINSTANCE
| TRANSLATABLE
"""
p[0] = p[1].lower()
def p_propertyDeclaration(p):
"""propertyDeclaration : propertyDeclaration_1
| propertyDeclaration_2
| propertyDeclaration_3
| propertyDeclaration_4
| propertyDeclaration_5
| propertyDeclaration_6
| propertyDeclaration_7
| propertyDeclaration_8
"""
p[0] = p[1]
def p_propertyDeclaration_1(p):
"""propertyDeclaration_1 : dataType propertyName ';'"""
p[0] = CIMProperty(p[2], None, type=p[1])
def p_propertyDeclaration_2(p):
"""propertyDeclaration_2 : dataType propertyName defaultValue ';'"""
p[0] = CIMProperty(p[2], p[3], type=p[1])
def p_propertyDeclaration_3(p):
"""propertyDeclaration_3 : dataType propertyName array ';'"""
p[0] = CIMProperty(p[2], None, type=p[1], is_array=True,
array_size=p[3])
def p_propertyDeclaration_4(p):
"""propertyDeclaration_4 : dataType propertyName array defaultValue ';'"""
p[0] = CIMProperty(p[2], p[4], type=p[1], is_array=True,
array_size=p[3])
def p_propertyDeclaration_5(p):
"""propertyDeclaration_5 : qualifierList dataType propertyName ';'"""
quals = OrderedDict([(x.name, x) for x in p[1]])
p[0] = CIMProperty(p[3], None, type=p[2], qualifiers=quals)
def p_propertyDeclaration_6(p):
# pylint: disable=line-too-long
"""propertyDeclaration_6 : qualifierList dataType propertyName defaultValue ';'""" # noqa: E501
quals = OrderedDict([(x.name, x) for x in p[1]])
p[0] = CIMProperty(p[3], cimvalue(p[4], p[2]),
type=p[2], qualifiers=quals)
def p_propertyDeclaration_7(p):
"""propertyDeclaration_7 : qualifierList dataType propertyName array ';'"""
quals = OrderedDict([(x.name, x) for x in p[1]])
p[0] = CIMProperty(p[3], None, type=p[2], qualifiers=quals,
is_array=True, array_size=p[4])
def p_propertyDeclaration_8(p):
# pylint: disable=line-too-long
"""propertyDeclaration_8 : qualifierList dataType propertyName array defaultValue ';'""" # noqa: E501
quals = OrderedDict([(x.name, x) for x in p[1]])
p[0] = CIMProperty(p[3], cimvalue(p[5], p[2]),
type=p[2], qualifiers=quals, is_array=True,
array_size=p[4])
def p_referenceDeclaration(p):
# pylint: disable=line-too-long
"""referenceDeclaration : objectRef referenceName ';'
| objectRef referenceName defaultValue ';'
| qualifierList objectRef referenceName ';'
| qualifierList objectRef referenceName defaultValue ';'
""" # noqa: E501
quals = []
dv = None
if isinstance(p[1], list): # qualifiers
quals = p[1]
cname = p[2]
pname = p[3]
if len(p) == 6:
dv = p[4]
else:
cname = p[1]
pname = p[2]
if len(p) == 5:
dv = p[3]
quals = OrderedDict([(x.name, x) for x in quals])
p[0] = CIMProperty(pname, dv, type='reference',
reference_class=cname, qualifiers=quals)
def p_methodDeclaration(p):
# pylint: disable=line-too-long
"""methodDeclaration : dataType methodName '(' ')' ';'
| dataType methodName '(' parameterList ')' ';'
| qualifierList dataType methodName '(' ')' ';'
| qualifierList dataType methodName '(' parameterList ')' ';'
""" # noqa: E501
paramlist = []
quals = []
if isinstance(p[1], six.string_types): # no quals
dt = p[1]
mname = p[2]
if p[4] != ')':
paramlist = p[4]
else: # quals present
quals = p[1]
dt = p[2]
mname = p[3]
if p[5] != ')':
paramlist = p[5]
params = OrderedDict([(param.name, param) for param in paramlist])
quals = OrderedDict([(q.name, q) for q in quals])
p[0] = CIMMethod(mname, return_type=dt, parameters=params,
qualifiers=quals)
# Note: class_origin is set when adding method to class.
# Note: The propagated flag is not set because this is parsed MOF, which
# contains specified methods and not any inherited methods.
def p_propertyName(p):
"""propertyName : identifier"""
p[0] = p[1]
def p_referenceName(p):
"""referenceName : identifier"""
p[0] = p[1]
def p_methodName(p):
"""methodName : identifier"""
p[0] = p[1]
def p_dataType(p):
"""dataType : DT_UINT8
| DT_SINT8
| DT_UINT16
| DT_SINT16
| DT_UINT32
| DT_SINT32
| DT_UINT64
| DT_SINT64
| DT_REAL32
| DT_REAL64
| DT_CHAR16
| DT_STR
| DT_BOOL
| DT_DATETIME
"""
p[0] = p[1].lower()
def p_objectRef(p):
"""objectRef : className REF"""
p[0] = p[1]
def p_parameterList(p):
"""parameterList : parameter
| parameterList ',' parameter
"""
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1] + [p[3]]
def p_parameter(p):
"""parameter : parameter_1
| parameter_2
| parameter_3
| parameter_4
"""
p[0] = p[1]
def p_parameter_1(p):
"""parameter_1 : dataType parameterName
| dataType parameterName array
"""
args = {}
if len(p) == 4:
args['is_array'] = True
args['array_size'] = p[3]
p[0] = CIMParameter(p[2], p[1], **args)
def p_parameter_2(p):
"""parameter_2 : qualifierList dataType parameterName
| qualifierList dataType parameterName array
"""
args = {}
if len(p) == 5:
args['is_array'] = True
args['array_size'] = p[4]
quals = OrderedDict([(x.name, x) for x in p[1]])
p[0] = CIMParameter(p[3], p[2], qualifiers=quals, **args)
def p_parameter_3(p):
"""parameter_3 : objectRef parameterName
| objectRef parameterName array
"""
args = {}
if len(p) == 4:
args['is_array'] = True
args['array_size'] = p[3]
p[0] = CIMParameter(p[2], 'reference', reference_class=p[1], **args)
def p_parameter_4(p):
"""parameter_4 : qualifierList objectRef parameterName
| qualifierList objectRef parameterName array
"""
args = {}
if len(p) == 5:
args['is_array'] = True
args['array_size'] = p[4]
quals = OrderedDict([(x.name, x) for x in p[1]])
p[0] = CIMParameter(p[3], 'reference', qualifiers=quals,
reference_class=p[2], **args)
def p_parameterName(p):
"""parameterName : identifier"""
p[0] = p[1]
def p_array(p):
"""array : '[' ']'
| '[' integerValue ']'
"""
if len(p) == 3:
p[0] = None
else:
p[0] = p[2]
def p_defaultValue(p):
"""defaultValue : '=' initializer"""
p[0] = p[2]
def p_initializer(p):
"""initializer : constantValue
| arrayInitializer
| referenceInitializer
"""
p[0] = p[1]
def p_arrayInitializer(p):
"""arrayInitializer : '{' constantValueList '}'
| '{' '}'
"""
if len(p) == 3:
p[0] = []
else:
p[0] = p[2]
def p_constantValueList(p):
"""constantValueList : constantValue
| constantValueList ',' constantValue
"""
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1] + [p[3]]
def _fixStringValue(s, p):
"""Clean up string value including special characters, etc."""
# pylint: disable=too-many-branches
s = s[1:-1]
rv = ''
esc = False
i = -1
while i < len(s) - 1:
i += 1
ch = s[i]
if ch == '\\' and not esc:
esc = True
continue
if not esc:
rv += ch
continue
if ch == '"':
rv += '"'
elif ch == 'n':
rv += '\n'
elif ch == 't':
rv += '\t'
elif ch == 'b':
rv += '\b'
elif ch == 'f':
rv += '\f'
elif ch == 'r':
rv += '\r'
elif ch == '\\':
rv += '\\'
elif ch in ['x', 'X']:
hexc = 0
j = 0
i += 1
while j < 4:
c = s[i + j]
c = c.upper()
if not c.isdigit() and c not in 'ABCDEF':
break
hexc <<= 4
if c.isdigit():
hexc |= ord(c) - ord('0')
else:
hexc |= ord(c) - ord('A') + 0XA
j += 1
if j == 0:
# DSP0004 requires 1..4 hex chars - we have 0
raise MOFParseError(
parser_token=p,
msg="Unicode escape sequence (e.g. '\\x12AB') requires "
"at least one hex character")
rv += six.unichr(hexc)
i += j - 1
esc = False
return rv
def p_stringValueList(p):
"""stringValueList : stringValue
| stringValueList stringValue
"""
if len(p) == 2:
p[0] = _fixStringValue(p[1], p)
else:
p[0] = p[1] + _fixStringValue(p[2], p)
def p_constantValue(p):
"""constantValue : integerValue
| floatValue
| charValue
| stringValueList
| booleanValue
| nullValue
"""
# The lexer functions (t_floatValue(), etc.) return a properly typed value.
p[0] = p[1]
def p_integerValue(p):
"""integerValue : binaryValue
| octalValue
| decimalValue
| hexValue
"""
# The lexer functions (t_binaryValue(), etc.) return a properly typed value.
p[0] = p[1]
def p_referenceInitializer(p):
"""referenceInitializer : objectHandle
| aliasIdentifier
"""
if p[1][0] == '$':
try:
p[0] = p.parser.aliases[p[1]]
except KeyError:
ce = CIMError(
CIM_ERR_FAILED,
_format("Unknown alias: {0!A}", p[1]))
ce.file_line = (p.parser.file, p.lexer.lineno)
raise ce
else:
p[0] = p[1]
def p_objectHandle(p):
"""objectHandle : identifier"""
p[0] = p[1]
def p_qualifierDeclaration(p):
# pylint: disable=line-too-long
"""qualifierDeclaration : QUALIFIER qualifierName qualifierType scope ';'
| QUALIFIER qualifierName qualifierType scope defaultFlavor ';'
""" # noqa: E501
qualtype = p[3]
dt, is_array, array_size, value = qualtype
qualname = p[2]
scopes = p[4]
if len(p) == 5:
flist = []
else:
flist = p[5]
flavors = _build_flavors(p[0], flist)
p[0] = CIMQualifierDeclaration(
qualname, dt, value=value, is_array=is_array, array_size=array_size,
scopes=scopes, **flavors)
def _build_flavors(p, flist, qualdecl=None):
"""
Build and return a dictionary defining the flavors from the
flist argument.
This function maps from the input keyword definitions for the flavors
(ex. EnableOverride) to the PyWBEM internal definitions
(ex. overridable)
Uses the qualdecl argument as a basis if it exists. This is to define
qualifier flavors if qualfier declaractions exist.
This applies the values from the qualifierDecl to the the qualifier
flavor list.
This function and the defaultflavor function insure that all
flavors are defined in the created dictionary that is returned. This
is important because the PyWBEM classes allow `None` as a flavor
definition.
"""
flavors = {}
if ('disableoverride' in flist and 'enableoverride' in flist) \
or \
('restricted' in flist and 'tosubclass' in flist): # noqa: E125
raise MOFParseError(parser_token=p, msg="Conflicting flavors are"
"invalid")
if qualdecl is not None:
flavors = {'overridable': qualdecl.overridable,
'translatable': qualdecl.translatable,
'tosubclass': qualdecl.tosubclass,
'toinstance': qualdecl.toinstance}
if 'disableoverride' in flist:
flavors['overridable'] = False
if 'enableoverride' in flist:
flavors['overridable'] = True
if 'translatable' in flist:
flavors['translatable'] = True
if 'restricted' in flist:
flavors['tosubclass'] = False
if 'tosubclass' in flist:
flavors['tosubclass'] = True
if 'toinstance' in flist:
flavors['toinstance'] = True
# issue #193 ks 5/16 removed tosubclass & set toinstance.
return flavors
# The ASSOCIATION and INDICATION alternates are required because ASSOCIATION
# and INDICATION are reserved words as defined in the DMTF spec but also
# keywords in this LEX definition and used as part of the scope definition
def p_qualifierName(p):
"""qualifierName : identifier
| ASSOCIATION
| INDICATION
"""
p[0] = p[1]
def p_qualifierType(p):
"""qualifierType : qualifierType_1
| qualifierType_2
"""
p[0] = p[1]
def p_qualifierType_1(p):
"""qualifierType_1 : ':' dataType array
| ':' dataType array defaultValue
"""
dv = None
if len(p) == 5:
dv = p[4]
p[0] = (p[2], True, p[3], dv)
def p_qualifierType_2(p):
"""qualifierType_2 : ':' dataType
| ':' dataType defaultValue
"""
dv = None
if len(p) == 4:
dv = p[3]
p[0] = (p[2], False, None, dv)
def p_scope(p):
"""scope : ',' SCOPE '(' metaElementList ')'"""
slist = p[4]
scopes = OrderedDict()
for i in ('CLASS',
'ASSOCIATION',
'INDICATION',
'PROPERTY',
'REFERENCE',
'METHOD',
'PARAMETER',
'ANY'):
scopes[i] = i in slist
p[0] = scopes
def p_metaElementList(p):
"""metaElementList : metaElement
| metaElementList ',' metaElement
"""
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1] + [p[3]]
def p_metaElement(p):
"""metaElement : SCHEMA
| CLASS
| ASSOCIATION
| INDICATION
| QUALIFIER
| PROPERTY
| REFERENCE
| METHOD
| PARAMETER
| ANY
"""
p[0] = p[1].upper()
def p_defaultFlavor(p):
"""defaultFlavor : ',' FLAVOR '(' flavorListWithComma ')'"""
flist = p[4]
# Create dictionary of default flavors based on DSP0004 definition
# of defaults for flavors. This insures that all possible flavors keywords
# are defined in the created dictionary.
flavors = {'ENABLEOVERRIDE': True,
'TOSUBCLASS': True,
'TOINSTANCE': False,
'DISABLEOVERRIDE': False,
'RESTRICTED': False,
'TRANSLATABLE': False}
for i in flist:
flavors[i] = True
p[0] = flavors
def p_flavorListWithComma(p):
"""flavorListWithComma : flavor
| flavorListWithComma ',' flavor
"""
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1] + [p[3]]
def p_instanceDeclaration(p):
# pylint: disable=line-too-long
"""instanceDeclaration : INSTANCE OF className '{' valueInitializerList '}' ';'
| INSTANCE OF className alias '{' valueInitializerList '}' ';'
| qualifierList INSTANCE OF className '{' valueInitializerList '}' ';'
| qualifierList INSTANCE OF className alias '{' valueInitializerList '}' ';'
""" # noqa: E501
# pylint: disable=too-many-locals,too-many-branches,too-many-statements
alias = None
quals = OrderedDict()
ns = p.parser.handle.default_namespace
if isinstance(p[1], six.string_types): # no qualifiers
cname = p[3]
if p[4] == '{':
props = p[5]
else:
props = p[6]
alias = p[4]
else:
cname = p[4]
# quals = p[1] # qualifiers on instances are deprecated -- rightly so.
if p[5] == '{':
props = p[6]
else:
props = p[7]
alias = p[5]
try:
cc = p.parser.handle.GetClass(cname, LocalOnly=False,
IncludeQualifiers=True)
p.parser.classnames[ns].append(cc.classname.lower())
except CIMError as ce:
ce.file_line = (p.parser.file, p.lexer.lineno)
if ce.status_code == CIM_ERR_NOT_FOUND:
file_ = p.parser.mofcomp.find_mof(cname)
if p.parser.verbose:
p.parser.log(
_format("Class {0!A} does not exist", cname))
if file_:
p.parser.mofcomp.compile_file(file_, ns)
cc = p.parser.handle.GetClass(cname, LocalOnly=False,
IncludeQualifiers=True)
else:
if p.parser.verbose:
p.parser.log("Can't find file to satisfy class")
ce = CIMError(CIM_ERR_INVALID_CLASS, cname)
ce.file_line = (p.parser.file, p.lexer.lineno)
raise ce
else:
raise
path = CIMInstanceName(cname, namespace=ns)
inst = CIMInstance(cname, qualifiers=quals, path=path)
keybindings = NocaseDict() # dictionary to build kb if alias exists
for prop in props:
pname = prop[1]
pval = prop[2]
try:
cprop = cc.properties[pname]
except KeyError:
ce = CIMError(
CIM_ERR_INVALID_PARAMETER,
_format("Invalid property. Not in class: {0!A}", pname))
ce.file_line = (p.parser.file, p.lexer.lineno)
raise ce
# confirm property name not duplicated.
if pname in inst.properties:
ce = CIMError(
CIM_ERR_INVALID_PARAMETER,
_format("Duplicate property: {0!A}", pname))
ce.file_line = (p.parser.file, p.lexer.lineno)
raise ce
try:
# build instance property from class property but without
# qualifiers, default value,
pprop = cprop.copy()
pprop.qualifiers = NocaseDict(None)
pprop.value = cimvalue(pval, cprop.type)
inst.properties[pname] = pprop
# if alias and this is key property, add keybinding
if alias and 'key' in cprop.qualifiers:
keybindings[pname] = pprop.value
except ValueError as ve:
ce = CIMError(
CIM_ERR_INVALID_PARAMETER,
_format("Invalid value for property {0!A}: {1}", pname, ve))
ce.file_line = (p.parser.file, p.lexer.lineno)
raise ce
if alias:
if keybindings:
inst.path.keybindings = keybindings
p.parser.aliases[alias] = inst.path
p[0] = inst
def p_valueInitializerList(p):
"""valueInitializerList : valueInitializer
| valueInitializerList valueInitializer
"""
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1] + [p[2]]
def p_valueInitializer(p):
"""valueInitializer : identifier defaultValue ';'
| qualifierList identifier defaultValue ';'
"""
if len(p) == 4:
id_ = p[1]
val = p[2]
quals = []
else:
quals = p[1]
id_ = p[2]
val = p[3]
p[0] = (quals, id_, val)
def p_booleanValue(p):
"""booleanValue : FALSE
| TRUE
"""
p[0] = p[1].lower() == 'true'
def p_nullValue(p):
"""nullValue : NULL"""
p[0] = None
def p_identifier(p):
"""identifier : IDENTIFIER
| ANY
| AS
| CLASS
| DISABLEOVERRIDE
| dataType
| ENABLEOVERRIDE
| FLAVOR
| INSTANCE
| METHOD
| OF
| PARAMETER
| PRAGMA
| PROPERTY
| QUALIFIER
| REFERENCE
| RESTRICTED
| SCHEMA
| SCOPE
| TOSUBCLASS
| TOINSTANCE
| TRANSLATABLE
"""
p[0] = p[1]
def p_empty(p):
'empty :'
pass
def _find_column(input_, token):
"""
Find the column in file where error occured. This is taken from
token.lexpos converted to the position on the current line by
finding the previous EOL.
"""
i = token.lexpos
while i > 0:
if input_[i] == '\n':
break
i -= 1
column = token.lexpos - i - 1
return column
def _get_error_context(input_, token):
"""
Build a context string that defines where on the line the defined
error occurs. This consists of the characters ^ at the position
and for the length defined by the lexer position and token length
"""
try:
line = input_[token.lexpos: input_.index('\n', token.lexpos)]
except ValueError:
line = input_[token.lexpos:]
i = input_.rfind('\n', 0, token.lexpos)
if i < 0:
i = 0
line = input_[i:token.lexpos] + line
lines = [line.strip('\r\n')]
col = token.lexpos - i
while len(lines) < 5 and i > 0:
end = i
i = input_.rfind('\n', 0, i)
if i < 0:
i = 0
lines.insert(0, input_[i:end].strip('\r\n'))
pointer = ''
for dummy_ch in str(token.value):
pointer += '^'
pointline = ''
i = 0
while i < col - 1:
if lines[-1][i].isspace():
pointline += lines[-1][i]
# otherwise, tabs complicate the alignment
else:
pointline += ' '
i += 1
lines.append(pointline + pointer)
return lines
[docs]@six.add_metaclass(ABCMeta)
class BaseRepositoryConnection(object):
"""
An abstract base class for implementing CIM repository connections (or an
entire CIM repository) for use by the MOF compiler. This class defines the
interface that is used by the :class:`~pywbem.MOFCompiler` class when it
interacts with its associated CIM repository.
Class :class:`~pywbem.MOFCompiler` invokes only the WBEM operations that
are defined as methods on this class:
* :meth:`EnumerateInstanceNames` - Enumerate the paths of CIM instances in
the repository.
* :meth:`CreateInstance` - Create a CIM instance in the repository.
* :meth:`ModifyInstance` - Modify a CIM instance in the repository.
* :meth:`DeleteInstance` - Delete a CIM instance in the repository.
* :meth:`GetClass` - Retrieve a CIM class from the repository.
* :meth:`ModifyClass` - Modify a CIM class in the repository.
* :meth:`CreateClass` - Create a CIM class in the repository.
* :meth:`DeleteClass` - Delete a CIM class in the repository.
* :meth:`EnumerateQualifiers` - Enumerate CIM qualifier types in the
repository.
* :meth:`GetQualifier` - Retrieve a CIM qualifier type from the repository.
* :meth:`SetQualifier` - Create or modify a CIM qualifier type in the
repository.
* :meth:`DeleteQualifier` - Delete a qualifier type from the repository.
Raises:
: Implementation classes should raise only exceptions derived from
:exc:`~pywbem.Error`. Other exceptions are considered programming
errors.
"""
# See below
def _getns(self):
"""get namespace not implemented for BaseRepositoryConnection."""
raise NotImplementedError
# See below
def _setns(self, value):
"""
Function to set namespace. Not implemented for
BaseRepositoryConnection
"""
raise NotImplementedError
# Ideally this property would be created via abstractproperty(), but then
# Sphinx does not generate documentation for it. So we create it via
# property() and raise NotImplementedError exceptions in the default
# implementations of the getter/setter methods.
default_namespace = property(
_getns, _setns, None,
"""
:term:`string`: The default repository namespace.
This property is settable.
"""
)
[docs] @abstractmethod
def EnumerateInstanceNames(self, *args, **kwargs):
"""
Enumerate the instance paths of CIM instances in a namespace of the
CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.EnumerateInstanceNames`.
"""
raise NotImplementedError
[docs] @abstractmethod
def CreateInstance(self, *args, **kwargs):
"""
Create a CIM instance in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.CreateInstance`.
"""
raise NotImplementedError
[docs] @abstractmethod
def ModifyInstance(self, *args, **kwargs):
"""
Modify a CIM instance in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.ModifyInstance`.
"""
raise NotImplementedError
[docs] @abstractmethod
def DeleteInstance(self, *args, **kwargs):
"""
Delete a CIM instance in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.DeleteInstance`.
"""
raise NotImplementedError
[docs] @abstractmethod
def GetClass(self, *args, **kwargs):
"""
Retrieve a CIM class in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.GetClass`.
"""
raise NotImplementedError
[docs] @abstractmethod
def ModifyClass(self, *args, **kwargs):
"""
Modify a CIM class in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.ModifyClass`.
"""
raise NotImplementedError
[docs] @abstractmethod
def CreateClass(self, *args, **kwargs):
"""
Create a CIM class in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.CreateClass`.
"""
raise NotImplementedError
[docs] @abstractmethod
def DeleteClass(self, *args, **kwargs):
"""
Delete a CIM class in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.DeleteClass`.
"""
raise NotImplementedError
[docs] @abstractmethod
def EnumerateQualifiers(self, *args, **kwargs):
"""
Enumerate the CIM qualifier types in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.EnumerateQualifiers`.
"""
raise NotImplementedError
[docs] @abstractmethod
def GetQualifier(self, *args, **kwargs):
"""
Retrieve a CIM qualifier type in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.GetQualifier`.
"""
raise NotImplementedError
[docs] @abstractmethod
def SetQualifier(self, *args, **kwargs):
"""
Create or modify a CIM qualifier type in a namespace of the CIM
repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.SetQualifier`.
"""
raise NotImplementedError
[docs] @abstractmethod
def DeleteQualifier(self, *args, **kwargs):
"""
Delete a CIM qualifier type in a namespace of the CIM repository.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.DeleteQualifier`.
"""
raise NotImplementedError
BaseRepositoryConnection.register(WBEMConnection) # pylint: disable=no-member
[docs]class MOFWBEMConnection(BaseRepositoryConnection):
"""
A CIM repository connection to an in-memory repository on top of an
underlying repository, that is used by the MOF compiler to provide rollback
support.
This class implements the
:class:`~pywbem.BaseRepositoryConnection` interface.
Raises:
: The methods of this class may raise any exceptions described for
class :class:`~pywbem.WBEMConnection`.
"""
def __init__(self, conn=None):
"""
Parameters:
conn (BaseRepositoryConnection):
The underlying repository connection.
`None` means that there is no underlying repository and all
operations performed through this object will fail.
"""
self.conn = conn
self.conn_id = conn.conn_id if conn is not None else None
self.class_names = {}
self.qualifiers = {}
self.instances = {}
self.classes = {}
self.compile_ordered_classnames = []
if conn is None:
# This attribute is used only to make get/set
# of 'default_namespace' behave as it should, in the case
# of conn=None.
self.__default_namespace = 'root/cimv2'
def _getns(self):
"""
:term:`string`: Return the default repository namespace to be used.
This method exists for compatibility. Use the :attr:`default_namespace`
property instead.
"""
if self.conn is not None:
return self.conn.default_namespace
return self.__default_namespace
def _setns(self, value):
"""
Set the default repository namespace to be used.
This method exists for compatibility. Use the :attr:`default_namespace`
property instead.
"""
if self.conn is not None:
self.conn.default_namespace = value
else:
self.__default_namespace = value
getns = _getns # for compatibility
setns = _setns # for compatibility
default_namespace = property(
_getns, _setns, None,
"""
:term:`string`: The default repository namespace to be used.
The default repository namespace is the default namespace of the
underlying repository connection if there is such an underlying
connection, or the default namespace of this object.
Initially, the default namespace of this object is 'root/cimv2'.
This property is settable. Setting it will cause the default namespace
of the underlying repository connection to be updated if there is such
an underlying connection, or the default namespace of this object.
"""
)
[docs] def EnumerateInstanceNames(self, *args, **kwargs):
"""This method is used by the MOF compiler only when it creates a
namespace in the course of handling CIM_ERR_NAMESPACE_NOT_FOUND.
Because the operations of this class silently create every namespace
that is needed and never return that error, this method is never
called, and is therefore not implemented.
"""
raise CIMError(
CIM_ERR_FAILED, 'This should not happen!',
conn_id=self.conn_id)
[docs] def ModifyInstance(self, *args, **kwargs):
"""This method is used by the MOF compiler only in the course of
handling CIM_ERR_ALREADY_EXISTS after trying to create an instance.
Because :meth:`CreateInstance` overwrites existing instances, this
method is never called, and is therefore not implemented.
"""
raise CIMError(
CIM_ERR_FAILED, 'This should not happen!',
conn_id=self.conn_id)
[docs] def CreateInstance(self, *args, **kwargs):
"""Create a CIM instance in the local repository of this class.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.CreateInstance`.
"""
inst = args[0] if args else kwargs['NewInstance']
try:
self.instances[self.default_namespace].append(inst)
except KeyError:
self.instances[self.default_namespace] = [inst]
return inst.path
[docs] def DeleteInstance(self, *args, **kwargs):
"""This method is only invoked by :meth:`rollback` (on the underlying
repository), and never by the MOF compiler, and is therefore not
implemented."""
raise CIMError(
CIM_ERR_FAILED, 'This should not happen!',
conn_id=self.conn_id)
[docs] def GetClass(self, *args, **kwargs):
"""Retrieve a CIM class from the local repository of this class.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.GetClass`.
"""
cname = args[0] if args else kwargs['ClassName']
try:
cc = self.classes[self.default_namespace][cname]
except KeyError:
if self.conn is None:
ce = CIMError(CIM_ERR_NOT_FOUND, cname)
raise ce
cc = self.conn.GetClass(*args, **kwargs)
try:
self.classes[self.default_namespace][cc.classname] = cc
except KeyError:
self.classes[self.default_namespace] = \
NocaseDict({cc.classname: cc})
if 'LocalOnly' in kwargs and not kwargs['LocalOnly']:
if cc.superclass:
try:
del kwargs['ClassName']
except KeyError:
pass
if args:
args = args[1:]
super_ = self.GetClass(cc.superclass, *args, **kwargs)
for prop in super_.properties.values():
if prop.name not in cc.properties:
cc.properties[prop.name] = prop
for meth in super_.methods.values():
if meth.name not in cc.methods:
cc.methods[meth.name] = meth
return cc
[docs] def ModifyClass(self, *args, **kwargs): # pylint: disable=no-self-use
"""This method is used by the MOF compiler only in the course of
handling CIM_ERR_ALREADY_EXISTS after trying to create a class.
Because :meth:`CreateClass` overwrites existing classes, this method
is never called, and is therefore not implemented.
"""
raise CIMError(
CIM_ERR_FAILED, 'This should not happen!',
conn_id=self.conn_id)
[docs] def CreateClass(self, *args, **kwargs):
"""Create a CIM class in the local repository of this class.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.CreateClass`.
"""
cc = args[0] if args else kwargs['NewClass']
if cc.superclass:
try:
_ = self.GetClass(cc.superclass, LocalOnly=True, # noqa: F841
IncludeQualifiers=False)
except CIMError as ce:
if ce.status_code == CIM_ERR_NOT_FOUND:
ce.args = (CIM_ERR_INVALID_SUPERCLASS, cc.superclass)
raise
raise
try:
self.compile_ordered_classnames.append(cc.classname)
# The following generates an exception for each new ns
self.classes[self.default_namespace][cc.classname] = cc
except KeyError:
self.classes[self.default_namespace] = \
NocaseDict({cc.classname: cc})
objects = list(cc.properties.values())
for meth in cc.methods.values():
objects += list(meth.parameters.values())
for obj in objects:
# Validate that reference_class exists in repo
if obj.type == 'reference':
try:
self.GetClass(obj.reference_class, LocalOnly=True,
IncludeQualifiers=True)
except CIMError as ce:
if ce.status_code == CIM_ERR_NOT_FOUND:
raise CIMError(
CIM_ERR_INVALID_PARAMETER,
_format("Class {0!A} referenced by element {1!A} "
"of class {2!A} in namespace {3!A} does "
"not exist",
obj.reference_class, obj.name,
cc.classname, self.getns()),
conn_id=self.conn_id)
raise
elif obj.type == 'string':
if 'EmbeddedInstance' in obj.qualifiers:
eiqualifier = obj.qualifiers['EmbeddedInstance']
try:
self.GetClass(eiqualifier.value, LocalOnly=True,
IncludeQualifiers=False)
except CIMError as ce:
if ce.status_code == CIM_ERR_NOT_FOUND:
raise CIMError(
CIM_ERR_INVALID_PARAMETER,
_format("Class {0!A} specified by "
"EmbeddInstance qualifier on element "
"{1!A} of class {2!A} in namespace "
"{3!A} does not exist",
eiqualifier.value, obj.name,
cc.classname, self.getns()),
conn_id=self.conn_id)
raise
# Issue #991: CreateClass should reject if the class already exists
try:
self.class_names[self.default_namespace].append(cc.classname)
except KeyError:
self.class_names[self.default_namespace] = [cc.classname]
[docs] def DeleteClass(self, *args, **kwargs):
"""This method is only invoked by :meth:`rollback` (on the underlying
repository), and never by the MOF compiler, and is therefore not
implemented."""
raise CIMError(
CIM_ERR_FAILED, 'This should not happen!',
conn_id=self.conn_id)
[docs] def EnumerateQualifiers(self, *args, **kwargs):
"""Enumerate the qualifier types in the local repository of this class.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.EnumerateQualifiers`.
"""
if self.conn is not None:
rv = self.conn.EnumerateQualifiers(*args, **kwargs)
else:
rv = []
try:
rv += list(self.qualifiers[self.default_namespace].values())
except KeyError:
pass
return rv
[docs] def GetQualifier(self, *args, **kwargs):
"""Retrieve a qualifier type from the local repository of this class.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.GetQualifier`.
"""
qualname = args[0] if args else kwargs['QualifierName']
try:
qual = self.qualifiers[self.default_namespace][qualname]
except KeyError:
if self.conn is None:
raise CIMError(
CIM_ERR_NOT_FOUND, qualname, conn_id=self.conn_id)
qual = self.conn.GetQualifier(*args, **kwargs)
return qual
[docs] def SetQualifier(self, *args, **kwargs):
"""Create or modify a qualifier type in the local repository of this
class.
For a description of the parameters, see
:meth:`pywbem.WBEMConnection.SetQualifier`.
"""
qual = args[0] if args else kwargs['QualifierDeclaration']
try:
self.qualifiers[self.default_namespace][qual.name] = qual
except KeyError:
self.qualifiers[self.default_namespace] = \
NocaseDict({qual.name: qual})
[docs] def DeleteQualifier(self, *args, **kwargs):
"""This method is only invoked by :meth:`rollback` (on the underlying
repository), and never by the MOF compiler, and is therefore not
implemented."""
raise CIMError(
CIM_ERR_FAILED, 'This should not happen!',
conn_id=self.conn_id)
[docs] def rollback(self, verbose=False):
"""
Remove classes and instances from the underlying repository, that have
been created in the local repository of this class.
Limitation: At this point, only classes and instances will be removed,
but not qualifiers.
"""
for ns, insts in self.instances.items():
insts.reverse()
for inst in insts:
try:
if verbose:
print(_format("Deleting instance {0}", inst.path))
self.conn.DeleteInstance(inst.path)
except CIMError as ce:
print(_format("Error deleting instance {0}", inst.path))
print(_format(" {0} {1}",
ce.status_code, ce.status_description))
for ns, cnames in self.class_names.items():
self.default_namespace = ns
cnames.reverse()
for cname in cnames:
try:
if verbose:
print(_format("Deleting class {0!A}:{1!A}",
ns, cname))
self.conn.DeleteClass(cname)
except CIMError as ce:
print(_format("Error deleting class {0!A}:{1!A}",
ns, cname))
print(_format(" {0} {1}",
ce.status_code, ce.status_description))
# Issue #990: Also roll back changes to qualifier declarations
def _print_logger(msg):
"""Print the `msg` parameter to stdout."""
print(msg)
[docs]class MOFCompiler(object):
"""
A MOF compiler. See :ref:`MOF Compiler API` for an explanation of MOF
compilers in general.
A MOF compiler may be associated with one CIM repository. The repository is
used for looking up dependent CIM elements (e.g. the superclass specified
in a class whose MOF definition is being compiled), and it is also updated
with the result of the compilation. A repository contains CIM namespaces,
and the namespaces contain CIM classes, instances and qualifier types.
The association of a MOF compiler with a CIM repository is established when
creating an object of this class. The interactions with the CIM repository
are defined in the abstract base class
:class:`~pywbem.BaseRepositoryConnection`.
"""
def __init__(self, handle, search_paths=None, verbose=False,
log_func=_print_logger):
# pylint: disable=line-too-long
"""
Parameters:
handle (BaseRepositoryConnection or :class:`~pywbem.WBEMConnection`):
A handle identifying the CIM repository that will be associated
with the MOF compiler.
If the provided object is a repository connection (i.e. derived
from :class:`BaseRepositoryConnection`, typically that would be a
:class:`~pywbem.MOFWBEMConnection` object), it is directly used
by the MOF compiler to interface with the repository.
If the provided object is a WBEM connection (i.e.
:class:`~pywbem.WBEMConnection` or
:class:`~pywbem_mock.FakedWBEMConnection`), a
:class:`~pywbem.MOFWBEMConnection` object is created from the
provided object, and used by the MOF compiler to interface with
the repository.
`None` means that no CIM repository will be associated. In this
case, the MOF compiler can only process standalone MOF that does
not depend on existing CIM elements in the repository.
search_paths (:term:`py:iterable` of :term:`string` or :term:`string`):
Directory path name(s) where the MOF compiler will search for MOF
dependent files if the MOF element they define is not in the target
namespace of the CIM repository. The compiler searches the
specified directories and their subdirectories.
MOF dependent files are:
* The MOF file defining the superclass of a class that is compiled.
The MOF file searched for is '<classname>.mof'.
* The MOF file defining the qualifier type of a qualifier that
is specified on a MOF element that is compiled.
The MOF files searched for are 'qualifiers.mof' and
'qualifiers_optional.mof'.
* The MOF file of a class specified in a reference property or
in the `EmbeddedInstance` qualifier that is compiled.
The MOF file searched for is '<classname>.mof'.
This is only partly implemented, see issue #1138.
Note that MOF include files are not searched for; they are specified
in the ``pragma include`` directive as a file path relative to the
including file.
verbose (:class:`py:bool`):
Indicates whether to issue more detailed compiler messages.
log_func (:term:`callable`):
A logger function that is invoked for each compiler message.
The logger function must take one parameter of string type.
The default logger function prints to stdout.
""" # noqa: E501
if isinstance(handle, WBEMConnection):
conn = handle
handle = MOFWBEMConnection(conn)
elif handle is None:
conn = None
elif isinstance(handle, BaseRepositoryConnection):
conn = getattr(handle, 'conn', None)
if conn and not isinstance(conn, WBEMConnection):
raise TypeError(
_format("If the handle parameter is a CIM repository "
"connection its conn attribute must be None "
"or WBEMConnection, but it is: {0}", type(conn)))
else:
raise TypeError(
_format("The handle parameter must be either a CIM repository "
"connection (derived from BaseRepositoryConnection) "
"or a WBEM connection (WBEMConnection), but is: {0}",
type(handle)))
server = WBEMServer(conn) if conn else None
if search_paths is None:
search_paths = []
elif isinstance(search_paths, six.string_types):
search_paths = [search_paths]
elif not isinstance(search_paths, (list, tuple)):
raise TypeError(
_format("search_paths parameter must be list or tuple, but "
"is: {0}", type(search_paths)))
self.parser = _yacc(verbose)
self.parser.search_paths = search_paths
self.handle = handle
self.parser.handle = handle
self.server = server
self.parser.server = server
self.lexer = _lex(verbose)
self.lexer.parser = self.parser
self.lexer.last_msg = None
self.parser.qualcache = {}
self.parser.classnames = {}
if handle:
default_namespace = handle.default_namespace
self.parser.qualcache[default_namespace] = NocaseDict()
self.parser.classnames[default_namespace] = []
self.parser.mofcomp = self
self.parser.verbose = verbose
self.parser.log = log_func
self.parser.aliases = {}
[docs] def compile_string(self, mof, ns, filename=None):
"""
Compile a string of MOF statements into a namespace of the associated
CIM repository.
Parameters:
mof (:term:`string`):
The string of MOF statements to be compiled.
ns (:term:`string`):
The name of the CIM namespace in the associated CIM repository
that is used for lookup of any dependent CIM elements, and that
is also the target of the compilation.
filename (:term:`string`):
The path name of the file that the MOF statements were read from.
This information is used only in compiler messages.
Raises:
IOError: MOF file not found.
MOFParseError: Syntax error in the MOF.
: Any exceptions that are raised by the repository connection class.
"""
lexer = self.lexer.clone()
lexer.parser = self.parser
try:
oldfile = self.parser.file
except AttributeError:
oldfile = None
self.parser.file = filename
try:
oldmof = self.parser.mof
except AttributeError:
oldmof = None
self.parser.mof = mof
self.parser.handle.default_namespace = ns
if ns not in self.parser.qualcache:
self.parser.qualcache[ns] = NocaseDict()
if ns not in self.parser.classnames:
self.parser.classnames[ns] = []
try:
# Call the parser. To generate detailed output of states
# add debug=... to following line where debug may be a
# constant (ex. 1) or may be a log definition, ex..
# log = logging.getLogger()
# logging.basicConfig(level=logging.DEBUG)
rv = self.parser.parse(mof, lexer=lexer)
self.parser.file = oldfile
self.parser.mof = oldmof
return rv
except MOFParseError as pe:
# Generate the error message into log and reraise error
self.parser.log(pe.get_err_msg())
raise
except CIMError as ce:
if hasattr(ce, 'file_line'):
self.parser.log(
_format("Fatal Error: {0}:{1}",
ce.file_line[0], ce.file_line[1]))
else:
self.parser.log("Fatal Error:")
description = _format(":{0}", ce.status_description) if \
ce.status_description else ""
self.parser.log(
_format("{0}{1}",
_statuscode2string(ce.status_code), description))
raise
[docs] def compile_file(self, filename, ns):
"""
Compile a MOF file into a namespace of the associated CIM repository.
Parameters:
filename (:term:`string`):
The path name of the MOF file containing the MOF statements to be
compiled.
ns (:term:`string`):
The name of the CIM namespace in the associated CIM repository
that is used for lookup of any dependent CIM elements, and that
is also the target of the compilation.
Raises:
IOError: MOF file not found.
MOFParseError: Syntax error in the MOF.
: Any exceptions that are raised by the repository connection class.
"""
if self.parser.verbose:
self.parser.log(
_format("Compiling file {0!A}", filename))
if not os.path.exists(filename):
# try to find in search path
rfilename = self.find_mof(os.path.basename(filename[:-4]).lower())
if rfilename is None:
raise IOError(
_format("No such file: {0!A}", filename))
filename = rfilename
with open(filename, "r") as f:
mof = f.read()
return self.compile_string(mof, ns, filename=filename)
[docs] def find_mof(self, classname):
"""
Find the MOF file that defines a particular CIM class, in the search
path of the MOF compiler.
The MOF file is found based on its file name: It is assumed that the
base part of the file name is the CIM class name.
Example: The class "CIM_ComputerSystem" is expected to be in a file
"CIM_ComputerSystem.mof".
Parameters:
classame (:term:`string`):
The name of the CIM class to look up.
Returns:
:term:`string`: Path name of the MOF file defining the CIM class, if
it was found. `None`, if it was not found.
"""
classname = classname.lower()
for search in self.parser.search_paths:
for root, dummy_dirs, files in os.walk(search):
for file_ in files:
if file_.endswith('.mof') and \
file_[:-4].lower() == classname:
return os.path.join(root, file_)
return None
[docs] def rollback(self, verbose=False):
"""
Rollback any changes to the CIM repository that were performed by
compilations using this MOF compiler object, since the object was
created.
"""
self.handle.rollback(verbose=verbose)
def _build(verbose=False, out_dir=_tabdir):
"""
Build the LEX and YACC table modules for the MOF compiler, if they do
not exist yet, or if their table versions do not match the installed
version of the `ply` package.
"""
if verbose:
print(
_format("Building LEX/YACC modules for MOF compiler in: {0}",
out_dir))
_yacc(verbose, out_dir=out_dir)
_lex(verbose, out_dir=out_dir)
def _yacc(verbose=False, out_dir=None):
"""
Return YACC parser object for the MOF compiler.
Parameters:
verbose (bool): Print messages while creating the parser object.
out_dir (string): Path name of the directory in which the YACC table
module source file (mofparsetab.py) for the MOF compiler will be
generated. If None, that file will not be generated.
Returns:
yacc.Parser: YACC parser object for the MOF compiler.
"""
# The write_tables argument controls whether the YACC parser writes
# the YACC table module file.
write_tables = (out_dir is not None)
# In yacc(), the 'debug' parameter controls the main error
# messages to the 'errorlog' in addition to the debug messages
# to the 'debuglog'. Because we want to see the error messages,
# we enable debug but set the debuglog to the NullLogger.
# To enable debug logging, set debuglog to some other logger
# (ex. PlyLogger(sys.stdout) to generate log output.
return yacc.yacc(optimize=_optimize,
tabmodule=_tabmodule,
outputdir=out_dir,
write_tables=write_tables,
debug=True,
debuglog=yacc.NullLogger(),
errorlog=yacc.PlyLogger(sys.stdout))
def _lex(verbose=False, out_dir=None):
"""
Return LEX analyzer object for the MOF Compiler.
Parameters:
verbose (bool): Print messages while creating the parser object.
out_dir (string): Path name of the directory in which the LEX table
module source file (moflextab.py) for the MOF compiler will be
generated. If None, that file will not be generated.
Returns:
lex.Lexer: LEX analyzer object for the MOF compiler.
"""
# Unfortunately, lex() does not support a write_tables argument. It
# always tries to write the tables if optimize=True, so we supply a dummy
# output directory. Always setting optimize=False is also not a good
# solution because that causes the input table not to be used.
if out_dir is None:
out_dir = tempfile.gettempdir()
# To debug lex you may set debug=True and enable the debuglog statement.
# or other logger definition.
return lex.lex(optimize=_optimize,
lextab=_lextab,
outputdir=out_dir,
debug=False,
# debuglog = lex.PlyLogger(sys.stdout),
errorlog=lex.PlyLogger(sys.stdout))