Include vars updated to work with directories (#17207)

* New features for include_vars

include_vars.py now allows you to include an entire directory and its nested directories of variable files.

Added Features..

* Ignore by default *.md, *.py, and *.pyc
* Ignore any list of files.
* Only include files nested by depth (default=unlimited)
* Match only files matching (valid regex)
* Sort files alphabetically and load in that order.
* Sort directories alphabetically and load in that order.

```
    - include_vars: 'vars/all.yml'

    - name: include all.yml
      include_vars:
        file: 'vars/all.yml'

    - name: include all yml files in vars/all and all nested directories
      include_vars:
        dir: 'vars/all'

    - name: include all yml files in vars/all and all nested directories and save the output in test.
      include_vars:
        dir: 'vars/all'
        name: test

    - name: include all yml files in vars/services
      include_vars:
        dir: 'vars/services'
        depth: 1

    - name: include only bastion.yml files
      include_vars:
        dir: 'vars'
        files_matching: 'bastion.yml'

    - name: include only all yml files exception bastion.yml
      include_vars:
        dir: 'vars'
        ignore_files: 'bastion.yml'
```

* Added whitelist for file extensisions (yaml, yml, json)

* Removed unit tests in favor of integration tests

8 files changed, 365 insertions, 29 deletions

diff --git a/lib/ansible/plugins/action/include_vars.py b/lib/ansible/plugins/action/include_vars.py
index e3b843ad97..b1867d48f5 100644
--- a/lib/ansible/plugins/action/include_vars.py
+++ b/lib/ansible/plugins/action/include_vars.py
@@ -1,4 +1,4 @@
-# (c) 2013-2014, Benno Joy <benno@ansible.com>
+# (c) 2016, Allen Sanabria <asanabria@linuxdynasty.org>
 #
 # This file is part of Ansible
 #
@@ -14,53 +14,281 @@
 #
 # You should have received a copy of the GNU General Public License
 # along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
+
 from __future__ import (absolute_import, division, print_function)
 __metaclass__ = type
 
+from os import path, walk
+import re
+
 from ansible.errors import AnsibleError
 from ansible.plugins.action import ActionBase
 from ansible.utils.unicode import to_str
 
+
 class ActionModule(ActionBase):
 
     TRANSFERS_FILES = False
 
-    def run(self, tmp=None, task_vars=None):
+    def _mutually_exclusive(self):
+        dir_arguments = [
+            self.source_dir, self.files_matching, self.ignore_files,
+            self.depth
+        ]
+        if self.source_file and None not in dir_arguments:
+            err_msg = (
+                "Can not include {0} with file argument"
+                .format(", ".join(self.VALID_DIR_ARGUMENTS))
+            )
+            raise AnsibleError(err_msg)
+
+        elif self.source_dir and self.source_file:
+            err_msg = (
+                "Need to pass either file or dir"
+            )
+            raise AnsibleError(err_msg)
+
+    def _set_dir_defaults(self):
+        if not self.depth:
+            self.depth = 0
+
+        if self.files_matching:
+            self.matcher = re.compile(r'{0}'.format(self.files_matching))
+        else:
+            self.matcher = None
+
+        if not self.ignore_files:
+            self.ignore_files = list()
+
+        if isinstance(self.ignore_files, str):
+            self.ignore_files = self.ignore_files.split()
 
-        varname = self._task.args.get('name')
-        source = self._task.args.get('file')
-        if not source:
-            source = self._task.args.get('_raw_params')
-            if source is None:
-                raise AnsibleError("No filename was found for the included vars. " + \
-                                   "Use `- include_vars: <filename>` or the `file:` option " + \
-                                   "to specify the vars filename.", self._task._ds)
+        elif isinstance(self.ignore_files, dict):
+            return {
+                'failed': True,
+                'message': '{0} must be a list'.format(self.ignore_files)
+            }
 
-        if task_vars is None:
+    def _set_args(self):
+        """ Set instance variables based on the arguments that were passed
+        """
+        self.VALID_DIR_ARGUMENTS = [
+            'dir', 'depth', 'files_matching', 'ignore_files'
+        ]
+        self.VALID_FILE_ARGUMENTS = ['file', '_raw_params']
+        self.GLOBAL_FILE_ARGUMENTS = ['name']
+
+        self.VALID_ARGUMENTS = (
+            self.VALID_DIR_ARGUMENTS + self.VALID_FILE_ARGUMENTS +
+            self.GLOBAL_FILE_ARGUMENTS
+        )
+        for arg in self._task.args:
+            if arg not in self.VALID_ARGUMENTS:
+                err_msg = '{0} is not a valid option in debug'.format(arg)
+                raise AnsibleError(err_msg)
+
+        self.return_results_as_name = self._task.args.get('name', None)
+        self.source_dir = self._task.args.get('dir', None)
+        self.source_file = self._task.args.get('file', None)
+        if not self.source_dir and not self.source_file:
+            self.source_file = self._task.args.get('_raw_params')
+
+        self.depth = self._task.args.get('depth', None)
+        self.files_matching = self._task.args.get('files_matching', None)
+        self.ignore_files = self._task.args.get('ignore_files', None)
+
+        self._mutually_exclusive()
+
+    def run(self, tmp=None, task_vars=None):
+        """ Load yml files recursively from a directory.
+        """
+        self.VALID_FILE_EXTENSIONS = ['yaml', 'yml', '.json']
+        if not task_vars:
             task_vars = dict()
 
+        self.show_content = True
+        self._set_args()
+
+        results = dict()
+        if self.source_dir:
+            self._set_dir_defaults()
+            self._set_root_dir()
+            if path.exists(self.source_dir):
+                for root_dir, filenames in self._traverse_dir_depth():
+                    failed, err_msg, updated_results = (
+                        self._load_files_in_dir(root_dir, filenames)
+                    )
+                    if not failed:
+                        results.update(updated_results)
+                    else:
+                        break
+            else:
+                failed = True
+                err_msg = (
+                    '{0} directory does not exist'.format(self.source_dir)
+                )
+        else:
+            try:
+                self.source_file = self._find_needle('vars', self.source_file)
+                failed, err_msg, updated_results = (
+                    self._load_files(self.source_file)
+                )
+                if not failed:
+                    results.update(updated_results)
+
+            except AnsibleError as e:
+                err_msg = to_str(e)
+                raise AnsibleError(err_msg)
+
+        if self.return_results_as_name:
+            scope = dict()
+            scope[self.return_results_as_name] = results
+            results = scope
+
         result = super(ActionModule, self).run(tmp, task_vars)
 
-        try:
-            source = self._find_needle('vars', source)
-        except AnsibleError as e:
-            result['failed'] = True
-            result['message'] = to_str(e)
-            return result
+        if failed:
+            result['failed'] = failed
+            result['message'] = err_msg
+
+        result['ansible_facts'] = results
+        result['_ansible_no_log'] = not self.show_content
+
+        return result
+
+    def _set_root_dir(self):
+        if self._task._role:
+            if self.source_dir.split('/')[0] == 'vars':
+                path_to_use = (
+                    path.join(self._task._role._role_path, self.source_dir)
+                )
+                if path.exists(path_to_use):
+                    self.source_dir = path_to_use
+            else:
+                path_to_use = (
+                    path.join(
+                        self._task._role._role_path, 'vars', self.source_dir
+                    )
+                )
+                self.source_dir = path_to_use
+        else:
+            current_dir = (
+                "/".join(self._task._ds._data_source.split('/')[:-1])
+            )
+            self.source_dir = path.join(current_dir, self.source_dir)
+
+    def _traverse_dir_depth(self):
+        """ Recursively iterate over a directory and sort the files in
+            alphabetical order. Do not iterate pass the set depth.
+            The default depth is unlimited.
+        """
+        current_depth = 0
+        sorted_walk = list(walk(self.source_dir))
+        sorted_walk.sort(key=lambda x: x[0])
+        for current_root, current_dir, current_files in sorted_walk:
+            current_depth += 1
+            if current_depth <= self.depth or self.depth == 0:
+                current_files.sort()
+                yield (current_root, current_files)
+            else:
+                break
+
+    def _ignore_file(self, filename):
+        """ Return True if a file matches the list of ignore_files.
+        Args:
+            filename (str): The filename that is being matched against.
+
+        Returns:
+            Boolean
+        """
+        for file_type in self.ignore_files:
+            try:
+                if re.search(r'{0}$'.format(file_type), filename):
+                    return True
+            except Exception:
+                err_msg = 'Invalid regular expression: {0}'.format(file_type)
+                raise AnsibleError(err_msg)
+        return False
 
-        (data, show_content) = self._loader._get_file_contents(source)
+    def _is_valid_file_ext(self, source_file):
+        """ Verify if source file has a valid extension
+        Args:
+            source_file (str): The full path of source file or source file.
+
+        Returns:
+            Bool
+        """
+        success = False
+        file_ext = source_file.split('.')
+        if len(file_ext) >= 1:
+            if file_ext[-1] in self.VALID_FILE_EXTENSIONS:
+                success = True
+                return success
+        return success
+
+    def _load_files(self, filename):
+        """ Loads a file and converts the output into a valid Python dict.
+        Args:
+            filename (str): The source file.
+
+        Returns:
+            Tuple (bool, str, dict)
+        """
+        results = dict()
+        failed = False
+        err_msg = ''
+        if not self._is_valid_file_ext(filename):
+            failed = True
+            err_msg = (
+                '{0} does not have a valid extension: {1}'
+                .format(filename, ', '.join(self.VALID_FILE_EXTENSIONS))
+            )
+            return failed, err_msg, results
+
+        data, show_content = self._loader._get_file_contents(filename)
+        self.show_content = show_content
         data = self._loader.load(data, show_content)
-        if data is None:
-            data = {}
+        if not data:
+            data = dict()
         if not isinstance(data, dict):
-            result['failed'] = True
-            result['message'] = "%s must be stored as a dictionary/hash" % source
+            failed = True
+            err_msg = (
+                '{0} must be stored as a dictionary/hash'
+                .format(filename)
+            )
         else:
-            if varname:
-                scope = {}
-                scope[varname] = data
-                data = scope
-            result['ansible_facts'] = data
-            result['_ansible_no_log'] = not show_content
+            results.update(data)
+        return failed, err_msg, results
 
-        return result
+    def _load_files_in_dir(self, root_dir, var_files):
+        """ Load the found yml files and update/overwrite the dictionary.
+        Args:
+            root_dir (str): The base directory of the list of files that is being passed.
+            var_files: (list): List of files to iterate over and load into a dictionary.
+
+        Returns:
+            Tuple (bool, str, dict)
+        """
+        results = dict()
+        failed = False
+        err_msg = ''
+        for filename in var_files:
+            stop_iter = False
+            # Never include main.yml from a role, as that is the default included by the role
+            if self._task._role:
+                if filename == 'main.yml':
+                    stop_iter = True
+                    continue
+
+            filepath = path.join(root_dir, filename)
+            if self.files_matching:
+                if not self.matcher.search(filename):
+                    stop_iter = True
+
+            if not stop_iter and not failed:
+                if path.exists(filepath) and not self._ignore_file(filename):
+                    failed, err_msg, loaded_data = self._load_files(filepath)
+                    if not failed:
+                        results.update(loaded_data)
+
+        return failed, err_msg, results
diff --git a/test/integration/roles/test_include_vars/defaults/main.yml b/test/integration/roles/test_include_vars/defaults/main.yml
new file mode 100644
index 0000000000..901fb220d2
--- /dev/null
+++ b/test/integration/roles/test_include_vars/defaults/main.yml
@@ -0,0 +1,3 @@
+---
+testing: 1
+base_dir: defaults
diff --git a/test/integration/roles/test_include_vars/tasks/main.yml b/test/integration/roles/test_include_vars/tasks/main.yml
new file mode 100644
index 0000000000..5ef6b3ef05
--- /dev/null
+++ b/test/integration/roles/test_include_vars/tasks/main.yml
@@ -0,0 +1,86 @@
+---
+- name: verify that the default value is indeed 1
+  assert:
+    that:
+      - "testing == 1"
+      - "base_dir == 'defaults'"
+
+- name: include the vars/environments/development/all.yml
+  include_vars:
+    file: environments/development/all.yml
+    
+- name: verify that the default value is indeed 789
+  assert:
+    that:
+      - "testing == 789"
+      - "base_dir == 'environments/development'"
+
+- name: include the vars/environments/development/all.yml and save results in all
+  include_vars:
+    file: environments/development/all.yml
+    name: all
+    
+- name: verify that the values are stored in the all variable
+  assert:
+    that:
+      - "all['testing'] == 789"
+      - "all['base_dir'] == 'environments/development'"
+
+- name: include the all directory in vars
+  include_vars:
+    dir: all
+    depth: 1
+    
+- name: verify that the default value is indeed 123
+  assert:
+    that:
+      - "testing == 123"
+      - "base_dir == 'all'"
+
+- name: include every directory in vars
+  include_vars:
+    dir: vars
+    
+- name: verify that the variable overwrite based on alphabetical order
+  assert:
+    that:
+      - "testing == 456"
+      - "base_dir == 'services'"
+      - "webapp_containers == 10"
+
+- name: include every directory in vars except files matching webapp.yml
+  include_vars:
+    dir: vars
+    ignore_files:
+      - webapp.yml
+    
+- name: verify that the webapp.yml file was not included
+  assert:
+    that:
+      - "testing == 789"
+      - "base_dir == 'environments/development'"
+
+- name: include only files matching webapp.yml
+  include_vars:
+    dir: environments
+    files_matching: webapp.yml
+
+- name: verify that only files matching webapp.yml and in the environments directory get loaded.
+  assert:
+    that:
+      - "testing == 101112"
+      - "base_dir == 'development/services'"
+      - "webapp_containers == 20"
+
+- name: include only files matching webapp.yml and store results in webapp
+  include_vars:
+    dir: environments
+    files_matching: webapp.yml
+    name: webapp
+
+- name: verify that only files matching webapp.yml and in the environments directory get loaded into stored variable webapp.
+  assert:
+    that:
+      - "webapp['testing'] == 101112"
+      - "webapp['base_dir'] == 'development/services'"
+      - "webapp['webapp_containers'] == 20"
diff --git a/test/integration/roles/test_include_vars/vars/all/all.yml b/test/integration/roles/test_include_vars/vars/all/all.yml
new file mode 100644
index 0000000000..14c3e92b8e
--- /dev/null
+++ b/test/integration/roles/test_include_vars/vars/all/all.yml
@@ -0,0 +1,3 @@
+---
+testing: 123
+base_dir: all
diff --git a/test/integration/roles/test_include_vars/vars/environments/development/all.yml b/test/integration/roles/test_include_vars/vars/environments/development/all.yml
new file mode 100644
index 0000000000..9f370de549
--- /dev/null
+++ b/test/integration/roles/test_include_vars/vars/environments/development/all.yml
@@ -0,0 +1,3 @@
+---
+testing: 789
+base_dir: 'environments/development'
diff --git a/test/integration/roles/test_include_vars/vars/environments/development/services/webapp.yml b/test/integration/roles/test_include_vars/vars/environments/development/services/webapp.yml
new file mode 100644
index 0000000000..a0a809c9e5
--- /dev/null
+++ b/test/integration/roles/test_include_vars/vars/environments/development/services/webapp.yml
@@ -0,0 +1,4 @@
+---
+testing: 101112
+base_dir: 'development/services'
+webapp_containers: 20
diff --git a/test/integration/roles/test_include_vars/vars/services/webapp.yml b/test/integration/roles/test_include_vars/vars/services/webapp.yml
new file mode 100644
index 0000000000..f0dcc8b517
--- /dev/null
+++ b/test/integration/roles/test_include_vars/vars/services/webapp.yml
@@ -0,0 +1,4 @@
+---
+testing: 456
+base_dir: services
+webapp_containers: 10
diff --git a/test/integration/test_include_vars.yml b/test/integration/test_include_vars.yml
new file mode 100644
index 0000000000..cb6aa7ec8d
--- /dev/null
+++ b/test/integration/test_include_vars.yml
@@ -0,0 +1,5 @@
+---
+- hosts: 127.0.0.1
+  gather_facts: False
+  roles:
+    - { role: test_include_vars }