Feature filter keep_keys (#8456)

* Add filter keep_keys. Implement feature request #8438

* Fix comment indentation.

* Fix regex reference.

* Fix indentation.

* Fix isinstance list.

* Update plugins/plugin_utils/keys_filter.py

Co-authored-by: Felix Fontein <felix@fontein.de>

* Update plugins/plugin_utils/keys_filter.py

Co-authored-by: Felix Fontein <felix@fontein.de>

* Update plugins/plugin_utils/keys_filter.py

Co-authored-by: Felix Fontein <felix@fontein.de>

* Update plugins/plugin_utils/keys_filter.py

Co-authored-by: Felix Fontein <felix@fontein.de>

* Update plugins/filter/keep_keys.py

Co-authored-by: Felix Fontein <felix@fontein.de>

* Update documentation, examples, and integration tests.

* _keys_filter_target_str returns tuple of unique target strings if
target is list. Update documentation, function comments, and error
messages.

* Sort maintainers.

* Update plugins/filter/keep_keys.py

Co-authored-by: Felix Fontein <felix@fontein.de>

* Update examples with explicit collection.

---------

Co-authored-by: Felix Fontein <felix@fontein.de>

.github/BOTMETA.yml

@@ -157,6 +157,8 @@ files:
   $filters/jc.py:
     maintainers: kellyjonbrazil
   $filters/json_query.py: {}
+  $filters/keep_keys.py:
+    maintainers: vbotka
   $filters/lists.py:
     maintainers: cfiehe
   $filters/lists_difference.yml:
@@ -1417,6 +1419,8 @@ files:
     ignore: matze
     labels: zypper
     maintainers: $team_suse
+  $plugin_utils/keys_filter.py:
+    maintainers: vbotka
   $plugin_utils/unsafe.py:
     maintainers: felixfontein
   $tests/a_module.py:

plugins/filter/keep_keys.py

@@ -0,0 +1,138 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) 2024 Vladimir Botka <vbotka@gmail.com>
+# Copyright (c) 2024 Felix Fontein <felix@fontein.de>
+# GNU General Public License v3.0+ (see LICENSES/GPL-3.0-or-later.txt or https://www.gnu.org/licenses/gpl-3.0.txt)
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+from __future__ import (absolute_import, division, print_function)
+__metaclass__ = type
+
+DOCUMENTATION = '''
+    name: keep_keys
+    short_description: Keep specific keys from dictionaries in a list
+    version_added: "9.1.0"
+    author:
+      - Vladimir Botka (@vbotka)
+      - Felix Fontein (@felixfontein)
+    description: This filter keeps only specified keys from a provided list of dictionaries.
+    options:
+      _input:
+        description:
+          - A list of dictionaries.
+          - Top level keys must be strings.
+        type: list
+        elements: dictionary
+        required: true
+      target:
+        description:
+          - A single key or key pattern to keep, or a list of keys or keys patterns to keep.
+          - If O(matching_parameter=regex) there must be exactly one pattern provided.
+        type: raw
+        required: true
+      matching_parameter:
+        description: Specify the matching option of target keys.
+        type: str
+        default: equal
+        choices:
+          equal: Matches keys of exactly one of the O(target) items.
+          starts_with: Matches keys that start with one of the O(target) items.
+          ends_with: Matches keys that end with one of the O(target) items.
+          regex:
+            - Matches keys that match the regular expresion provided in O(target).
+            - In this case, O(target) must be a regex string or a list with single regex string.
+'''
+
+EXAMPLES = '''
+  l:
+    - {k0_x0: A0, k1_x1: B0, k2_x2: [C0], k3_x3: foo}
+    - {k0_x0: A1, k1_x1: B1, k2_x2: [C1], k3_x3: bar}
+
+  # 1) By default match keys that equal any of the items in the target.
+  t: [k0_x0, k1_x1]
+  r: "{{ l | community.general.keep_keys(target=t) }}"
+
+  # 2) Match keys that start with any of the items in the target.
+  t: [k0, k1]
+  r: "{{ l | community.general.keep_keys(target=t, matching_parameter='starts_with') }}"
+
+  # 3) Match keys that end with any of the items in target.
+  t: [x0, x1]
+  r: "{{ l | community.general.keep_keys(target=t, matching_parameter='ends_with') }}"
+
+  # 4) Match keys by the regex.
+  t: ['^.*[01]_x.*$']
+  r: "{{ l | community.general.keep_keys(target=t, matching_parameter='regex') }}"
+
+  # 5) Match keys by the regex.
+  t: '^.*[01]_x.*$'
+  r: "{{ l | community.general.keep_keys(target=t, matching_parameter='regex') }}"
+
+  # The results of above examples 1-5 are all the same.
+  r:
+    - {k0_x0: A0, k1_x1: B0}
+    - {k0_x0: A1, k1_x1: B1}
+
+  # 6) By default match keys that equal the target.
+  t: k0_x0
+  r: "{{ l | community.general.keep_keys(target=t) }}"
+
+  # 7) Match keys that start with the target.
+  t: k0
+  r: "{{ l | community.general.keep_keys(target=t, matching_parameter='starts_with') }}"
+
+  # 8) Match keys that end with the target.
+  t: x0
+  r: "{{ l | community.general.keep_keys(target=t, matching_parameter='ends_with') }}"
+
+  # 9) Match keys by the regex.
+  t: '^.*0_x.*$'
+  r: "{{ l | community.general.keep_keys(target=t, matching_parameter='regex') }}"
+
+  # The results of above examples 6-9 are all the same.
+  r:
+    - {k0_x0: A0}
+    - {k0_x0: A1}
+'''
+
+RETURN = '''
+  _value:
+    description: The list of dictionaries with selected keys.
+    type: list
+    elements: dictionary
+'''
+
+from ansible_collections.community.general.plugins.plugin_utils.keys_filter import (
+    _keys_filter_params,
+    _keys_filter_target_str)
+
+
+def keep_keys(data, target=None, matching_parameter='equal'):
+    """keep specific keys from dictionaries in a list"""
+
+    # test parameters
+    _keys_filter_params(data, target, matching_parameter)
+    # test and transform target
+    tt = _keys_filter_target_str(target, matching_parameter)
+
+    if matching_parameter == 'equal':
+        def keep_key(key):
+            return key in tt
+    elif matching_parameter == 'starts_with':
+        def keep_key(key):
+            return key.startswith(tt)
+    elif matching_parameter == 'ends_with':
+        def keep_key(key):
+            return key.endswith(tt)
+    elif matching_parameter == 'regex':
+        def keep_key(key):
+            return tt.match(key) is not None
+
+    return [dict((k, v) for k, v in d.items() if keep_key(k)) for d in data]
+
+
+class FilterModule(object):
+
+    def filters(self):
+        return {
+            'keep_keys': keep_keys,
+        }

plugins/plugin_utils/keys_filter.py

@@ -0,0 +1,108 @@
+# Copyright (c) 2024 Vladimir Botka <vbotka@gmail.com>
+# Copyright (c) 2024 Felix Fontein <felix@fontein.de>
+# GNU General Public License v3.0+ (see LICENSES/GPL-3.0-or-later.txt or https://www.gnu.org/licenses/gpl-3.0.txt)
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+from __future__ import (absolute_import, division, print_function)
+__metaclass__ = type
+
+import re
+
+from ansible.errors import AnsibleFilterError
+from ansible.module_utils.six import string_types
+from ansible.module_utils.common._collections_compat import Mapping, Sequence
+
+
+def _keys_filter_params(data, target, matching_parameter):
+    """test parameters:
+    * data must be a list of dictionaries. All keys must be strings.
+    * target must be a non-empty sequence.
+    * matching_parameter is member of a list.
+    """
+
+    mp = matching_parameter
+    ml = ['equal', 'starts_with', 'ends_with', 'regex']
+
+    if not isinstance(data, Sequence):
+        msg = "First argument must be a list. %s is %s"
+        raise AnsibleFilterError(msg % (data, type(data)))
+
+    for elem in data:
+        if not isinstance(elem, Mapping):
+            msg = "The data items must be dictionaries. %s is %s"
+            raise AnsibleFilterError(msg % (elem, type(elem)))
+
+    for elem in data:
+        if not all(isinstance(item, string_types) for item in elem.keys()):
+            msg = "Top level keys must be strings. keys: %s"
+            raise AnsibleFilterError(msg % elem.keys())
+
+    if not isinstance(target, Sequence):
+        msg = ("The target must be a string or a list. target is %s.")
+        raise AnsibleFilterError(msg % target)
+
+    if len(target) == 0:
+        msg = ("The target can't be empty.")
+        raise AnsibleFilterError(msg)
+
+    if mp not in ml:
+        msg = ("The matching_parameter must be one of %s. matching_parameter is %s")
+        raise AnsibleFilterError(msg % (ml, mp))
+
+    return
+
+
+def _keys_filter_target_str(target, matching_parameter):
+    """test:
+    * If target is list all items are strings
+    * If matching_parameter=regex target is a string or list with single string
+    convert and return:
+    * tuple of unique target items, or
+    * tuple with single item, or
+    * compiled regex if matching_parameter=regex
+    """
+
+    if isinstance(target, list):
+        for elem in target:
+            if not isinstance(elem, string_types):
+                msg = "The target items must be strings. %s is %s"
+                raise AnsibleFilterError(msg % (elem, type(elem)))
+
+    if matching_parameter == 'regex':
+        if isinstance(target, string_types):
+            r = target
+        else:
+            if len(target) > 1:
+                msg = ("Single item is required in the target list if matching_parameter is regex.")
+                raise AnsibleFilterError(msg)
+            else:
+                r = target[0]
+        try:
+            tt = re.compile(r)
+        except re.error:
+            msg = ("The target must be a valid regex if matching_parameter is regex."
+                   " target is %s")
+            raise AnsibleFilterError(msg % r)
+    elif isinstance(target, string_types):
+        tt = (target, )
+    else:
+        tt = tuple(set(target))
+
+    return tt
+
+
+def _keys_filter_target_dict(target, matching_parameter):
+    """test:
+    * target is a list of dictionaries
+    * ...
+    """
+
+    # TODO: Complete and use this in filter replace_keys
+
+    if isinstance(target, list):
+        for elem in target:
+            if not isinstance(elem, Mapping):
+                msg = "The target items must be dictionary. %s is %s"
+                raise AnsibleFilterError(msg % (elem, type(elem)))
+
+    return

tests/integration/targets/filter_keep_keys/aliases

@@ -0,0 +1,5 @@
+# Copyright (c) Ansible Project
+# GNU General Public License v3.0+ (see LICENSES/GPL-3.0-or-later.txt or https://www.gnu.org/licenses/gpl-3.0.txt)
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+azp/posix/2

tests/integration/targets/filter_keep_keys/tasks/keep_keys.yml

@@ -0,0 +1,79 @@
+---
+# Copyright (c) Ansible Project
+# GNU General Public License v3.0+ (see LICENSES/GPL-3.0-or-later.txt or https://www.gnu.org/licenses/gpl-3.0.txt)
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+- name: Debug ansible_version
+  ansible.builtin.debug:
+    var: ansible_version
+  when: not quite_test | d(true) | bool
+  tags: ansible_version
+
+- name: Test keep keys equal (default)
+  ansible.builtin.assert:
+    that:
+      - (rr | difference(result1) | length) == 0
+    success_msg: |
+      [OK]  result:
+      {{ rr | to_yaml }}
+    fail_msg: |
+      [ERR] result:
+      {{ rr | to_yaml }}
+    quiet: "{{ quiet_test | d(true) | bool }}"
+  vars:
+    rr: "{{ list1 | community.general.keep_keys(target=tt) }}"
+    tt: [k0_x0, k1_x1]
+  tags: equal_default
+
+- name: Test keep keys regex string
+  ansible.builtin.assert:
+    that:
+      - (rr | difference(result1) | length) == 0
+    success_msg: |
+      [OK]  result:
+      {{ rr | to_yaml }}
+    fail_msg: |
+      [ERR] result:
+      {{ rr | to_yaml }}
+    quiet: "{{ quiet_test | d(true) | bool }}"
+  vars:
+    rr: "{{ list1 | community.general.keep_keys(target=tt, matching_parameter=mp) }}"
+    mp: regex
+    tt: '^.*[01]_x.*$'
+  tags: regex_string
+
+- name: Test keep keys targets1
+  ansible.builtin.assert:
+    that:
+      - (rr | difference(result1) | length) == 0
+    success_msg: |
+      [OK]  result:
+      {{ rr | to_yaml }}
+    fail_msg: |
+      [ERR] result:
+      {{ rr | to_yaml }}
+    quiet: "{{ quiet_test | d(true) | bool }}"
+  loop: "{{ targets1 }}"
+  loop_control:
+    label: "{{ item.mp }}: {{ item.tt }}"
+  vars:
+    rr: "{{ list1 | community.general.keep_keys(target=item.tt, matching_parameter=item.mp) }}"
+  tags: targets1
+
+- name: Test keep keys targets2
+  ansible.builtin.assert:
+    that:
+      - (rr | difference(result2) | length) == 0
+    success_msg: |
+      [OK]  result:
+      {{ rr | to_yaml }}
+    fail_msg: |
+      [ERR] result:
+      {{ rr | to_yaml }}
+    quiet: "{{ quiet_test | d(true) | bool }}"
+  loop: "{{ targets2 }}"
+  loop_control:
+    label: "{{ item.mp }}: {{ item.tt }}"
+  vars:
+    rr: "{{ list2 | community.general.keep_keys(target=item.tt, matching_parameter=item.mp) }}"
+  tags: targets2

tests/integration/targets/filter_keep_keys/tasks/main.yml

@@ -0,0 +1,7 @@
+---
+# Copyright (c) Ansible Project
+# GNU General Public License v3.0+ (see LICENSES/GPL-3.0-or-later.txt or https://www.gnu.org/licenses/gpl-3.0.txt)
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+- name: Test keep_keys
+  import_tasks: keep_keys.yml

tests/integration/targets/filter_keep_keys/vars/main.yml

@@ -0,0 +1,33 @@
+---
+# Copyright (c) Ansible Project
+# GNU General Public License v3.0+ (see LICENSES/GPL-3.0-or-later.txt or https://www.gnu.org/licenses/gpl-3.0.txt)
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+targets1:
+  - {mp: equal, tt: [k0_x0, k1_x1]}
+  - {mp: starts_with, tt: [k0, k1]}
+  - {mp: ends_with, tt: [x0, x1]}
+  - {mp: regex, tt: ['^.*[01]_x.*$']}
+  - {mp: regex, tt: '^.*[01]_x.*$'}
+
+list1:
+  - {k0_x0: A0, k1_x1: B0, k2_x2: [C0], k3_x3: foo}
+  - {k0_x0: A1, k1_x1: B1, k2_x2: [C1], k3_x3: bar}
+
+result1:
+   - {k0_x0: A0, k1_x1: B0}
+   - {k0_x0: A1, k1_x1: B1}
+
+targets2:
+  - {mp: equal, tt: k0_x0}
+  - {mp: starts_with, tt: k0}
+  - {mp: ends_with, tt: x0}
+  - {mp: regex, tt: '^.*0_x.*$'}
+
+list2:
+  - {k0_x0: A0, k1_x1: B0, k2_x2: [C0], k3_x3: foo}
+  - {k0_x0: A1, k1_x1: B1, k2_x2: [C1], k3_x3: bar}
+
+result2:
+   - {k0_x0: A0}
+   - {k0_x0: A1}