test-2757

Committed 11 Jan 2026 09:04AM UTC coverage: 29.475% (-58.6%) from 88.026%

Build # test-2757

Build Type

Pull #1501

github

Committed by

web-flow

Commit Message

Merge 4229b955d into f9b5e1682

Pull Request Pull Request #1501: Fixed start program timeouts and other issues in listener tests; Enabled tests again

Run Details

59 of 116 new or added lines in 3 files covered. (50.86%)

3924 existing lines in 32 files now uncovered.

1953 of 6626 relevant lines covered (29.47%)

0.59 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

7.04

/pywbemtools/pywbemcli/_display_cimobjects.py

# (C) Copyright 2020 IBM Corp.
# (C) Copyright 2020 Inova Development Inc.
# All Rights Reserved
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#    http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
Common function to display cim objects in multiple formats.
display_cimobjects() is the function that should be used for all CIM
object display in pywbemcli.

It displays a list of CIM objects of a single type in multiple formats
defined by output_format (as CIM objects (mof, etc.) as one or more tables
or as text.
"""


import re

import click

from pywbem import CIMInstanceName, CIMInstance, CIMClass, \
    CIMQualifierDeclaration, CIMClassName, ValueMapping, siunit_obj, \
    CIMError, CIM_ERR_NOT_SUPPORTED
from pywbem._nocasedict import NocaseDict

from ._common import sort_cimobjects, to_wbem_uri_folded
from ._cimvalueformatter import cimvalue_to_fmtd_string
from .._utils import get_terminal_width
from .._output_formatting import DEFAULT_MAX_CELL_WIDTH, \
    output_format_is_table, format_table, fold_strings

INT_TYPE_PATTERN = re.compile(r'^[su]int(8|16|32|64)$')

# Minimum width for table view cell size. Below this width the columns become
# unreadable.  Note that this overrides the max_terminal width.
MIN_CELL_WIDTH = 10

####################################################################
#
#  Display of CIM objects in format defined by output_format
#
####################################################################


def display_cim_objects(context, cim_objects, output_format, summary=False,
                        property_list=None, quote_strings=True,
                        ignore_null_properties=True, object_order=False,
                        ctx_options=None):

    """
    Display CIM objects in form determined by input parameters.

    Input is either a list of cim objects or a single object. It may be
    any of the CIM types.  This is used to display:

      * CIMClass
      * CIMClassName:
      * CIMInstance
      * CIMInstanceName
      * CIMQualifierDeclaration
      * class associatiors (tuple of CIMClassName, CIMClass)
      * list of the above types
      * Dictionary of the above types where the keys are namespaces and the
        values are any of the above types. This is the return from commands
        that handle multiple namespaces in the command option --namespace.

    This function may override output type choice in cases where the output
    choice is not available for the object type.  Thus, for example,
    mof output makes no sense for class names. In that case, the output is
    the str with the name or namespace:name.

    Parameters:

      context (:class:`ContextObj`):
        Click context contained in ContextObj object.

      objects (iterable of :class:`~pywbem.CIMInstance`,
        :class:`~pywbem.CIMInstanceName`, :class:`~pywbem.CIMClass`,
        :class:`~pywbem.CIMClassName`,
        :class:`~pywbem.CIMQualifierDeclaration`, or tuple where the tuple
        consists of (CIMClassName, CIMClass) and is return from class
        associators or class references.
        Iterable of zero or more CIM objects to be displayed.  If the iterable
        is a dictionary it contains the objects for multiple namespaces where
        the keys are namespace names and the values are lists of objects in
        the namespace.

      output_format (:term:`string`):
        String defining the preferred output format. Must not be None since
        the correct output_format must have been selected before this call.
        Note that the output formats allowed may depend on a) whether
        summary is True, b)the specific type because we do not have a table
        output format for CIMClass.

      summary (:class:`py:bool`):
        Boolean that defines whether the data in objects should be displayed
        or just a summary of the objects (ex. count of number of objects).

      property_list (iterable of :term:`string`):
        List of property names to be displayed, in the desired order, when the
        output format is a table, or None to use the default of sorting
        the properties alphabetically within key and non-key groups.

      quote_strings (:class:`py.bool`):
        If False, strings are not encased by quote marks in the table view for
        instance displays. The default is True so that strings are encased in
        quotes in all views.

      ignore_null_properties (:class:`py.bool`):
        If True, cases where a property has Null value or does not exist in all
        of the instances to be displayed will not be included in table output.
        They will be included in CIM object(MOF) output. If False, The
        properties that do not have a value in any instance will be included in
        tables.  This allows table output to ignore properties that do not add
        value to the table display.

      object_order (:class:`py.bool`):
        If True sort the objects by classname/objectname rather than by
        namespace on so that objects of same class in different namespaces
        appear together.

        Used only when the objects parameter is a dictionary ( i.e. multiple
        namespaces)
    """

    # Note: In the docstring above, the line for parameter 'objects' was too
    #       long. Since we are not putting it into docmentation, we folded it.
    context.spinner_stop()

    # Assure that there is output format exists, None not allowed.
    assert output_format

    # Sort valid only for dictionaries.  Single lists are always sorted by
    # class name
    if not isinstance(cim_objects, NocaseDict):
        assert not object_order

    # If dictionary of objects in namespaces and there is only one namespace
    # reduce to list of cim objects. Code deals with single namespace request
    # without identifying it in displays as we do for multiple namespace display
    if isinstance(cim_objects, NocaseDict) and len(cim_objects) == 1:
        ns1 = list(cim_objects.keys())
        ns = ns1[0]
        cim_objects = cim_objects[ns]

    # If summary flag, call the summary display function and return
    if summary:
        _display_cim_objects_summary(context, cim_objects, output_format)
        return

    # Terminate if no actual cim_objects returned for single item and list
    # before any further processing. Simplifies further processing so it
    # can ignore None conditions.
    if cim_objects is None or (isinstance(cim_objects, list) and
                               not cim_objects):
        if context.verbose:
            click.echo("No objects returned")
        return

    if isinstance(cim_objects, NocaseDict) and \
            not any(list(cim_objects.values())):
        if context.verbose:
            click.echo('No objects returned for namespace(s): '
                       f'{", ".join(cim_objects.keys())}')
        return

    # Process by type based on receiving dictionaries of ns:objects or lists
    # of objects or tuples from class associators/references

    if output_format_is_table(output_format):
        # Table format output is processed as a group in a single table both
        # for multiple and single namespaces.
        _display_as_table(context, cim_objects, output_format,
                          property_list, quote_strings,
                          ignore_null_properties,
                          object_order,
                          ctx_options)
        return

    # pylint: disable=too-many-function-args
    _display_as_cim_objects(cim_objects, output_format, object_order)


############################################################################
#
# Support methods for displaying CIM objects.  This includes multiple
# output formats (ie. MOF, TABLE, TEXT), table display and summary display.
#
############################################################################

def sort_display_objects(cim_objects):
    """
    Sort the list of cim_objects and return a sorted list of the objects.
    Includes QualdeclWrapper as sortable object
    This method requires that there be at least one object in cim_objects
    """
    assert isinstance(cim_objects, list)
    if isinstance(cim_objects[0], QualDeclWrapper):
        return sorted(cim_objects, key=lambda x: x.name())

    return sort_cimobjects(cim_objects)


def _order_display_objects(cim_objects, object_order):
    """
    Orders the objects in cim_objects in object_name, ns order if object order
    is set.  Otherwise it keeps the original order which is the order objects
    were inserted for the namespaces and sorts the order of the returned
    objects.
    Defines an iterator which is returned. That iterator
    will return namespace, cim_object for each cim object to be displayed
    in the order defined by object_order.
    """
    if object_order:
        # Reorder the objects so an iterator returns in object, ns order
        order_dict = NocaseDict()
        for ns, ns_objects in cim_objects.items():
            if ns_objects:
                if not isinstance(ns_objects, list):
                    ns_objects = [ns_objects]
                ns_objects = sort_display_objects(ns_objects)

                # Determine the object name and find all matching objects
                # in the list
                for obj in ns_objects:
                    if isinstance(obj, (CIMClass, CIMClassName)):
                        obj_name = obj.classname
                    # Instance name without namespace becomes name for ordering
                    elif isinstance(obj, CIMInstance):
                        objpath = obj.path.copy()
                        objpath.namespace = None
                        obj_name = objpath.to_wbem_uri(format="canonical")
                    elif isinstance(obj, CIMInstanceName):
                        objtmp = obj.copy()
                        objtmp.namespace = None
                        obj_name = objtmp.to_wbem_uri(format="canonical")
                    elif isinstance(obj, CIMQualifierDeclaration):
                        obj_name = obj.name
                    elif isinstance(obj, str):
                        obj_name = obj
                    else:
                        assert False

                    assert obj_name

                    # Add obj_name,ns,objects to order_dict
                    if obj_name not in order_dict:
                        order_dict[obj_name] = NocaseDict()
                    if ns not in order_dict[obj_name]:
                        order_dict[obj_name][ns] = [obj]
                    else:
                        order_dict[obj_name][ns].append(obj)

        for obj_name, namespaces in order_dict.items():
            for ns in namespaces:
                # pylint: disable=unnecessary-dict-index-lookup
                objs = order_dict[obj_name][ns]
                for obj in objs:
                    yield ns, obj

    # No reordering required. Return in ns, object name order
    # iterator returns namespace, object in namespace, object order
    else:
        for ns, ns_objects in cim_objects.items():
            if ns_objects:
                if not isinstance(ns_objects, list):
                    ns_objects = [ns_objects]
                for ns_object in sort_display_objects(ns_objects):
                    yield ns, ns_object


def _display_as_cim_objects(cim_objects, output_format, object_order):
    """
    Display the objects in the objects parameter using one of the individual
    object display formats (MOF, etc.). This parameter may be either a
    NocaseDictionary (multi-namespace type request) or a list (single namespace
    type request)
    """
    # CIM Object and text output formats are displayed as single objects.
    # Note: for class associaters/references each object is a tuple of
    # CIMClassName, CIMClass
    # If object_order is True, reorganize so the order of display
    # is class, namespace
    # Otherwise, the order of display is namespace, class

    # Starts with dict(namespace: list_of_objects)
    # Want dict (class: namespace: list of objects with this class, namespace)

    if isinstance(cim_objects, NocaseDict):
        for ns, ns_object in _order_display_objects(cim_objects, object_order):
            _display_one_cim_object(ns_object, output_format, namespace=ns)
        return

    # If not NocaseDict, it is list or single object. Sort by classname and
    # then call _display_one_cim_objects with each item from list
    if not isinstance(cim_objects, list):
        cim_objects = [cim_objects]

    for cim_obj in sort_display_objects(cim_objects):
        _display_one_cim_object(cim_obj, output_format)


def _display_as_table(context, cim_objects, output_format, property_list,
                      quote_strings, ignore_null_properties, object_order,
                      ctx_options):
    """
    Display the cim_objects as a table where the table rows are dependent on
    the cim object type being displayed. Calls a method for each type of object
    to display that object type information as a table.

    This function processes objects from multiple namespaces in a NocaseDict to
    produce a single table showing the requested output in one row per
    namespace/class or from a single list (objects from single namespace) to
    produce a table that does not show the namespaces
    """
    # If NocaseDic reduce cim_objects to a new list of objects for simpler
    # table processing. String and QualiferDeclaration objects are modified to
    # include the namespace with each object.
    if isinstance(cim_objects, NocaseDict):
        objects = []
        for ns, ns_object in _order_display_objects(cim_objects, object_order):
            if isinstance(ns_object, CIMQualifierDeclaration):
                objects.append(QualDeclWrapper(ns, ns_object))
            elif isinstance(ns_object, str):
                objects.append(CIMClassName(ns_object, namespace=ns))
            else:
                objects.append(ns_object)

        # pass ordered object list on to display function for tables
        _display_list_as_table(context, objects, output_format,
                               property_list, quote_strings,
                               ignore_null_properties,
                               use_namespace=True,
                               ctx_options=ctx_options)

    else:  # original input was a list sort but do not include namespace
        if not isinstance(cim_objects, (list)):
            cim_objects = [cim_objects]
        cim_objects = sort_display_objects(cim_objects)
        _display_list_as_table(context, cim_objects, output_format,
                               property_list, quote_strings,
                               ignore_null_properties,
                               use_namespace=False,
                               ctx_options=ctx_options)


def _display_list_as_table(context, cim_objects, output_format, property_list,
                           quote_strings, ignore_null_properties,
                           use_namespace=None, ctx_options=None):
    """
    Display the cim_objects parameter list as a table.  The parameter must be a
    list. Calls lower level display functions for each object type.

    cim_objects in form of a single list of the objects in the order that
    they will be displayed in the table.
    """
    assert isinstance(cim_objects, list)
    assert use_namespace is not None

    table_width = get_terminal_width()

    if cim_objects:  # This and the list test probably not necessary
        # Call table display method for each object type
        if isinstance(cim_objects[0], CIMInstance):
            _display_instances_as_table(
                cim_objects, table_width, output_format, context=context,
                property_list=property_list, quote_strings=quote_strings,
                ignore_null_properties=ignore_null_properties,
                namespace=use_namespace, ctx_options=ctx_options)
        elif isinstance(cim_objects[0], CIMClass):
            _display_classes_as_table(cim_objects, table_width, output_format)
        elif isinstance(cim_objects[0], QualDeclWrapper):
            _display_qual_decls_as_table(cim_objects, table_width,
                                         output_format)
        elif isinstance(cim_objects[0], CIMQualifierDeclaration):
            _display_qual_decls_as_table(cim_objects, table_width,
                                         output_format)
        elif isinstance(cim_objects[0], (CIMClassName, CIMInstanceName,
                                         (str,), tuple)):
            _display_paths_as_table(cim_objects, table_width, output_format,
                                    namespace=use_namespace)
        else:
            assert False


def _display_one_cim_object(cim_object, output_format, namespace=None):
    """
    Display a single CIM object.  This creates and outputs the display for one
    CIM object in  the output formats groups CIM object (MOF, XML, Text).

    NOTE: namespace defines the namespace in which the object exists. It is
    used in some cases to add namespace information to the display.
    Specifically it is used if the object is converted to mof, xml to add the
    # pragama statement and if the object is a string type (i.e. classname)
    """
    assert isinstance(
        cim_object, (CIMClass, CIMClassName, CIMInstance, CIMInstanceName,
                     CIMQualifierDeclaration, tuple, (str,)))

    # Display in the selected CIM object format (mof, xml, repr, txt)
    if output_format == 'mof':
        try:
            mofstr = cim_object.tomof()
            if namespace:
                click.echo(f"#pragma namespace (\"{namespace}\")")
            click.echo(mofstr)

        except AttributeError:
            # inserting NL between instance names for readability since the
            # display is always a single line per display
            if isinstance(cim_object, CIMInstanceName):
                click.echo(f"\n{cim_object}")
            elif isinstance(cim_object, CIMClassName):
                click.echo(cim_object)
            elif isinstance(cim_object, tuple):  # representation of class assoc
                assert isinstance(cim_object[0], CIMClassName)
                assert isinstance(cim_object[1], CIMClass)
                click.echo(cim_object[0])
                click.echo(cim_object[1].tomof())
            elif isinstance(cim_object, str):
                if namespace:
                    click.echo(f"{namespace}:{cim_object}")
                else:
                    click.echo(cim_object)
            else:
                raise click.ClickException(
                    f'output_format {output_format} invalid '
                    f'for {type(cim_object)}')
    elif output_format == 'xml':
        try:
            if isinstance(cim_object,
                          (CIMClass, CIMInstance, CIMQualifierDeclaration,
                           CIMInstanceName, CIMClassName)):
                if namespace:
                    click.echo(f"<!-- Namespace = {namespace} -->")
                click.echo(cim_object.tocimxmlstr(indent=4))
            elif isinstance(cim_object, tuple):
                if namespace:
                    click.echo(f"<!-- Namespace = {namespace} -->")
                assert isinstance(cim_object[0], CIMClassName)
                assert isinstance(cim_object[1], CIMClass)
                click.echo(cim_object[0].tocimxmlstr(indent=4))
                click.echo(cim_object[1].tocimxmlstr(indent=4))
            elif isinstance(cim_object, str):
                if namespace:
                    click.echo(f"{namespace}:{cim_object}")
                else:
                    click.echo(cim_object)
            else:
                assert False, f"Output_format {output_format} invalid " \
                              f"for {type(cim_object)}"

        except AttributeError:
            # no tocimxmlstr functionality
            raise click.ClickException(
                f'Output Format {output_format} not supported. '
                f'Default to\n{cim_object!r}')
    elif output_format == 'repr':
        try:
            click.echo(repr(cim_object))
        except AttributeError:
            raise click.ClickException(
                f'"repr" display of {cim_object!r} failed')

    elif output_format == 'txt':
        try:
            click.echo(cim_object)
        except AttributeError:
            raise click.ClickException(
                f'"txt" display of {cim_object!r} failed')
    else:
        raise click.ClickException(f'Invalid output format {output_format}')


class QualDeclWrapper():  # pylint: disable=too-few-public-methods
    """
    Convert qualifier declarations to instance of this class to be able to
    pass the namespace to the display function unambiguously.  This is only
    a wrapper within the display_cim_objects method and its sub-methods since
    the CIM_QualifierDeclaration has no namespace component.
    """
    def __init__(self, namespace, qual_decl_inst):
        self.qualdecl = qual_decl_inst
        self.namespace = namespace

    def name(self):
        """Return the qualifier declaration name."""
        return self.qualdecl.name


def _get_cimtype(objects):
    """
    Get the cim_type for any returned cim object.  Normally this is the
    name of the class name except that the classname return from
    getclass and enumerate class is just unicode string

    Parameters:
      objects (dict, list, object where object is any CIM object or tuple)

    Returns:
      The CIM type (string) for that object. If the object is a tuple, the
      type comes from the second object in the tuple

      If no objects are in thd dict or list None is returned

    """
    # associators and references return tuple
    if isinstance(objects, list) and objects:
        test_object = objects[0]
    elif objects:
        test_object = object
    else:
        return None

    if isinstance(test_object, tuple):
        # associator or reference class level return is tuple get type
        # from second item, it is the requested object
        cim_type = test_object[1].__class__.__name__
    else:
        cim_type = test_object.__class__.__name__

    # account for fact the enumerate class name operation returns uniicode.
    if isinstance(test_object, str):
        cim_type = 'CIMClassName'
    return cim_type


def _display_cim_objects_summary(context, objects, output_format):
    """
    Display a summary of the objects received. This displays the
    count of objects.
    """
    context.spinner_stop()

    cim_type = None
    if isinstance(objects, NocaseDict):
        headers = ['Namespace', 'Count', 'CIM Type']
        rows = []
        for ns, objlist in objects.items():
            if not cim_type:
                cim_type = _get_cimtype(objlist)
            objlistlen = len(objlist) if objlist else 0
            rows.append([ns, objlistlen, cim_type])
    else:
        headers = ['Count', 'CIM Type']
        if isinstance(objects, list):
            cim_type = _get_cimtype(objects)
            rows = [[len(objects), cim_type]]
        elif objects is not None:
            cim_type = _get_cimtype(objects)
            rows = [["1", cim_type]]
        else:
            rows = [[0, cim_type]]

    title = f'Summary of {cim_type}(s) returned'
    if output_format_is_table(output_format):
        click.echo(format_table(rows, headers, title=title,
                                table_format=output_format))
    else:
        for row in rows:  # must be type text
            # The following two are hangover from initial format output
            # "<number <object_name>(s)". Future, redo this whole output format
            cim_type_pos = len(row) - 1
            len_pos = len(row) - 2
            if row[len_pos] == 0:
                row[cim_type_pos] = "objects"
            else:
                row[cim_type_pos] = f"{row[cim_type_pos]}(s)"
            row[len_pos] = str(row[len_pos])
            click.echo(f'{" ".join(row)} returned')


#
# The following functions each display one object type as a table
#
def _display_classes_as_table(classes, table_width, table_format):
    """
    Display classes as a table.
    """
    # pylint: disable=unused-argument

    # ISSUE #249: Display classes as a table or mof rather than just as MOF.
    for class_ in classes:
        click.echo(class_.tomof())


def _display_paths_as_table(objects, table_width, table_format, namespace=None):
    # pylint: disable=unused-argument
    """
    Display paths as a table. This include CIMInstanceName, ClassPath,
    and unicode (the return type for enumerateClasses).
    """
    title = None
    if objects:
        # Strings only for single namespace requests for classnames.
        # No namespace in result
        if isinstance(objects[0], str):
            title = 'Classnames:'
            headers = ['Class Name']
            rows = [[obj] for obj in objects]
        # if obj is tuple, this is a tuple of namespace and classname
        elif isinstance(objects[0], tuple):
            assert False
            title = 'Classnames:'
            headers = ("namespace", 'class')
            rows = [[obj[0], obj[1]] for obj in objects]

        elif isinstance(objects[0], CIMClassName):
            title = 'Classnames'
            headers = ('host', 'namespace', 'class')
            rows = [[obj.host, obj.namespace, obj.classname] for obj in objects]

        elif isinstance(objects[0], CIMInstanceName):
            # Since there are cases where the list has instances with
            # different keys (i.e. enumerate CIM_ManagedElement), build
            # a dictionary for instances with the same key names.
            # Sort into dictionary for each set of key names.  If all keys are
            # the same, there will be one table.  This allows building
            # a table for each set of keynames
            objs_by_key_set = {}
            original_keys = {}

            for instname in objects:
                # key_names is a tuple of sorted key names in lower case so
                # all instances with the same set of keys ends up in a
                # single dictionary item. NOTE: objects must be tuple to be
                # a dictionary key, the key.
                test_key_names = tuple(
                    sorted(map(lambda x: x.lower(), instname.keys())))
                try:
                    objs_by_key_set[test_key_names].append(instname)
                except KeyError:
                    objs_by_key_set[test_key_names] = [instname]

                # Set the original keys into dict to recover correct key
                # case later
                if test_key_names not in original_keys:
                    original_keys[test_key_names] = instname.keys()

            # If multiple key_names we create multiple tables and add table
            # number to each table title.
            # FUTURE: This was a hack. by forcing classname into the table
            table_number = 0
            for key_names, inst_names in objs_by_key_set.items():
                # Build headers for this table with the common elements and key
                # names for each key in the object. We use inst_names[0] to
                # restore original case to key strings. We sort keys for
                # consistent table output.
                inst_keys = sorted(original_keys[key_names])

                rows = []
                for instname in inst_names:
                    # Build header row for this table
                    row = [instname.host, instname.namespace,
                           instname.classname]
                    for key in inst_keys:
                        if isinstance(instname[key], CIMInstanceName):
                            # If key is CIMInstanceName, fold the value
                            row.append(to_wbem_uri_folded(
                                instname[key], uri_format='standard',
                                max_len=30))
                        else:
                            row.append(instname[key])
                    rows.append(row)

                # If multiple tables, number them as hint to reader that there
                # are multiples.
                if len(objs_by_key_set) > 1:
                    table_number += 1
                    table_number_str = f", (table #{table_number})"
                else:
                    table_number_str = ''

                headers = ['host', 'namespace', 'class'] + \
                    [f"key=\n{kn}" for kn in inst_keys]

                title = f'InstanceNames: {inst_names[0].classname}' \
                        f'{table_number_str}'

                # Generate multiple tables, one for each key_name and
                # return local to this scope.
                click.echo(format_table(rows, headers, title=title,
                                        table_format=table_format))

            return  # Return to avoid following table output

        else:
            assert False

        click.echo(format_table(rows, headers, title=title,
                                table_format=table_format))


def _display_qual_decls_as_table(qual_decls, table_width, table_format):
    """
    Display the elements of qualifier declarations as a table with a
    row for each qualifier declaration and a column for each of the attributes
    of the qualifier declaration (name, type, Value, Array, Scopes, Flavors.

    qual_decls is a list of the class QualDeclWrapper where each instance is
    namespace, qualdecl ib multiple namespaces are to be didsplayed or a list
    of qualdecls if only one namespace is to be displayed

    The function displays all of the qualifier declarations in the the list
    if a qualdecls is a list of CIMQualifierdecl or or by namespace if
    qualdecls is a list of tuples
    """
    def build_row(qdecl, ns):
        """
        Build a single row representing a qualifier declaration
        """
        scopes = '\n'.join([key for key in qdecl.scopes if qdecl.scopes[key]])
        flavors = []
        flavors.append('EnableOverride' if qdecl.overridable
                       else 'DisableOverride')
        flavors.append('ToSubclass' if qdecl.tosubclass else 'Restricted')
        if qdecl.translatable:
            flavors.append('Translatable')
        sep = "\n" if sum(map(len, flavors)) >= max_column_width else ", "
        flavors = sep.join(flavors)
        row = [qdecl.name, qdecl.type, qdecl.value, qdecl.is_array, scopes,
               flavors]
        if ns:
            row.insert(0, ns)
        return row

    # Build header line.  Add Namespace column if namespace flag set.
    rows = []
    headers = ['Name', 'Type', 'Value', 'Array', 'Scopes', 'Flavors']

    # build the data rows
    max_column_width = int(table_width / len(headers)) - 4
    rows = []
    add_namespace = False
    for qdecl in qual_decls:
        if isinstance(qdecl, QualDeclWrapper):
            row = build_row(qdecl.qualdecl, qdecl.namespace)
            add_namespace = True
        else:
            row = build_row(qdecl, None)
        rows.append(row)

    if add_namespace:
        headers.insert(0, "namespace")

    click.echo(format_table(rows, headers, title='Qualifier Declarations',
                            table_format=table_format))


def _format_instances_as_rows(insts, max_cell_width, include_classnames=False,
                              context=None, prop_names=None,
                              quote_strings=True, namespace=None):
    """
    Format the list of instances properties into a list of the property
    values for each instance(a row of the table) gathered into a list of
    the rows.

    The prop_names parameter is the list of (originally cased) property names
    to be output, in the desired output order. It could be determined from
    the instances, but since it is determined already by the caller, it
    is passed in as an optimization. For test convenience, None is permitted
    and causes the properties to again be determined from the instances.

    Include_classesnames for each instance if True. Sets the classname as
    the first (or second if multiple namespaces) column.

    max_width if not None folds col entries longer than the defined
    max_cell_width. If max_width is None, the data length is ignored.

    The property values are formatted similar to MOF output. Properties that
    have a ValueMap qualifier (effectively, in the creation class of the
    instance) are shown with both the actual property value and the mapped
    value in parenthesis.

    NOTE: This is a separate function to allow testing of the table formatting
    independently of print output.

    Returns:
        list of strings where each string is a row in the table and each
        item in a row is a cell entry
    """
    conn = context.pywbem_server.conn if context else None

    # Avoid crash deeper in code if max_cell_width is None.
    # FUTURE: Should move this up to common place
    if max_cell_width is None:
        max_cell_width = DEFAULT_MAX_CELL_WIDTH
    rows = []

    # Build list of properties in any of the instances to display.
    if prop_names is None:
        prop_names = _sorted_prop_names(insts)

    # Cache of ValueMapping objects for integer-typed properties.
    # Key: classname.propertyname, both in lower case.
    # A value of None indicates the property does not have a value mapping.
    valuemappings = {}

    for inst in insts:
        assert isinstance(inst, CIMInstance), \
            f"Invalid CIM Type {type(inst)}"

        # Setup row list and set namespace if this is multi-namespace
        # Note: We use namespace as a flag here ignoring the actual values.
        row = [inst.path.namespace] if namespace else []

        # Insert classname before properties if flag set
        if include_classnames:
            if row:
                row.append(inst.classname)
            else:
                row = [inst.classname]

        # Get value for each property in this object
        for name in prop_names:
            # Account for possible instances without all properties
            # Outputs empty  string.  Note that instance with no value
            # results in same output as not instance name.
            if name not in inst.properties:
                val_str = ''
            else:
                value = inst.get(name)
                prop = inst.properties[name]

                # Cache value mappings for integer-typed properties
                if INT_TYPE_PATTERN.match(prop.type) and context:
                    vm_key = f'{inst.classname.lower()}.{name.lower()}'
                    try:
                        valuemapping = valuemappings[vm_key]
                    except KeyError:
                        try:
                            valuemapping = ValueMapping.for_property(
                                conn,
                                inst.path.namespace,
                                inst.classname,
                                name)
                        except ValueError:
                            # Property does not have a value mapping.
                            valuemapping = None
                        valuemappings[vm_key] = valuemapping
                else:
                    valuemapping = None

                if value is None:
                    val_str = ''
                else:
                    val_str, _ = cimvalue_to_fmtd_string(
                        prop.value, prop.type, indent=0, maxline=max_cell_width,
                        line_pos=0, end_space=0, avoid_splits=False,
                        valuemapping=valuemapping, quote_strings=quote_strings)
            row.append(val_str)
        rows.append(row)

    return rows


def _display_instances_as_table(insts, table_width, table_format,
                                include_classnames=False, context=None,
                                property_list=None, quote_strings=True,
                                ignore_null_properties=True,
                                namespace=None, ctx_options=None):
    """
    Print the properties of the instances defined in insts as a table where
    each row is an instance and each column is a property value.

    All properties in the instance are included.

    The header line consists of the property names.

    The property values are formatted similar to MOF output. Properties that
    have a ValueMap qualifier (effectively, in the creation class of the
    instance) are shown with both the actual property value and the mapped
    value in parenthesis.

    """
    conn = context.pywbem_server.conn if context else None

    if table_width is None:
        table_width = get_terminal_width()

    for inst in insts:
        assert isinstance(inst, CIMInstance)

    # Include classnames if multiple classes in display: See issue # 1145
    # If multiple classes, display classname in table, otherwise show in title
    classnames = [inst.classname.lower() for inst in insts]
    include_classnames = bool(len(set(classnames)) > 1)

    # Build list of properties, either all properties in a list or all
    # properties in all instances in list of instances
    if property_list:
        prop_names = _propertylist_prop_names(insts, property_list)
    else:
        prop_names = _sorted_prop_names(insts)

    if ignore_null_properties:
        # Get names of properties in any instances that have a value (i.e.
        # not None or empty list)
        props_with_value_dict = NocaseDict()
        for inst in insts:
            for propname, propvalue in inst.properties.items():
                if propvalue.value is not None:
                    props_with_value_dict[propname] = True
        # Rebuild prop names from dict in same order as original list
        prop_names = [pn for pn in prop_names if pn in props_with_value_dict]

    # Try to estimate max cell width from number of cols and properties
    # This allows folding long data. Further, the actual output
    # width of a column involves the tabulate outputter, output_format
    # so this is not deterministic.
    max_cell_width = int(table_width / len(prop_names)) \
        if prop_names else table_width

    # Sets a minimum size for cells so they are at least readable.
    # This means we can build tables wider than the terminal width.
    max_cell_width = max(max_cell_width, MIN_CELL_WIDTH)

    # ISSUE #953 Future: Decide whether and how the showing of units should be
    #            controlled. Or are we satisfied with always showing them?
    show_units = True

    # Retrieve all creation classes of the instances in order to get
    # to their PUnit and Units qualifiers.
    if show_units:
        class_objs = NocaseDict()  # CIMClass objects, by classname
        for inst in insts:
            classname = inst.classname
            if classname in class_objs:
                continue
            if inst.path is None:
                # The only operation returning instances without a path is
                # query execution. For now, we simply don't display units in
                # case the inst.path is None.
                # ISSUE #953: Pass the namespace for query execution to allow
                # class retrieval as part of getting  units information.
                show_units = False
                break
            try:
                class_obj = conn.GetClass(
                    classname, namespace=inst.path.namespace,
                    IncludeQualifiers=True, LocalOnly=False)
            except CIMError as exc:
                if exc.status_code == CIM_ERR_NOT_SUPPORTED:
                    # The WBEM Server does not support class operations. We
                    # silently give up showing units in this case.
                    show_units = False
                    break
            class_objs[classname] = class_obj

    # Construct the header line
    headers = []  # list of header strings
    # If we want to include namespace in table as a row
    if namespace:
        headers.append("namespace")
    if include_classnames:
        headers.append("classname")
    for pname in prop_names:
        hdr = pname
        if show_units:
            # In theory, two leaf classes from different vendors could have
            # introduced same-named properties with different unit definitions.
            # We account for that by showing a list of units in that case.
            siunits = []
            for class_obj in class_objs.values():
                if pname not in class_obj.properties:
                    # Not all classes may have the property, but one of them
                    # will.
                    continue
                prop_obj = class_obj.properties[pname]
                siunit = siunit_obj(prop_obj)
                if siunit is not None and siunit not in siunits:
                    siunits.append(siunit)
            for siunit in siunits:
                hdr += f" [{siunit}]"
        headers.append(hdr)

    rows = _format_instances_as_rows(
        insts, max_cell_width,
        include_classnames=include_classnames,
        context=context, prop_names=prop_names,
        quote_strings=quote_strings, namespace=namespace)

    # Fold the headers if necessary. Fold on either hypens or single word
    # too long because the headers are all single words
    disp_headers = []
    for header in headers:
        if len(header) > max_cell_width:
            disp_headers.append(fold_strings(header,
                                             max_cell_width,
                                             break_long_words=True,
                                             break_on_hyphens=True))
        else:
            disp_headers.append(header)

    # Display string if cmd option deep_inheritance exists and is True

    di = ""
    if ctx_options:   # This is just for test support
        di = "; deep-inheritance" if ctx_options.get('deep_inheritance') else ""
    title = f'Instances: {insts[0].classname}{di}'

    click.echo(format_table(rows, disp_headers, title=title,
                            table_format=table_format))


def _sorted_prop_names(insts):
    """
    Return the list of (originally cased) property names that is the superset
    of all properties in the  any of theinput instances.

    The returned list has the key properties first, followed by the non-key
    properties. Each group is sorted case insensitively.

    The key properties are determined from the instance paths, if present.
    The function tolerates it if only some of the instances have a path,
    and if instances of subclasses have additional keys.
    """

    all_props = NocaseDict()  # key: org prop name, value: lower prop name
    key_props = NocaseDict()  # key: org prop name, value: lower prop name
    for inst in insts:
        for pn in inst.properties:
            all_props[pn] = pn.lower()
            if inst.path and pn in inst.path.keybindings:
                key_props[pn] = pn.lower()

    nonkey_props = NocaseDict()  # key: org prop name, value: lower prop name
    for pn in all_props:
        if pn not in key_props:
            nonkey_props[pn] = all_props[pn]

    key_prop_list = sorted(key_props.keys(), key=lambda pn: key_props[pn])
    nonkey_prop_list = sorted(
        nonkey_props.keys(), key=lambda pn: nonkey_props[pn])
    key_prop_list.extend(nonkey_prop_list)
    return key_prop_list


def _propertylist_prop_names(insts, property_list):
    """
    Return the originally cased property list, based on the lexical case
    in the instances.

    If a property name is not in any instance, it is not returned.
    """
    prop_list = []
    for pname in property_list:
        for inst in insts:
            if pname in inst:
                prop_list.append(inst.properties[pname].name)
                break
    return prop_list

pywbem / pywbemtools / test-2757

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous