From 2abb943d3210e3be269801e4c8f3928e02192d7f Mon Sep 17 00:00:00 2001
From: Ricardo Garcia Silva <ricardo.garcia.silva@gmail.com>
Date: Fri, 22 Mar 2024 21:47:39 +0000
Subject: [PATCH] 1600 allow providing default value in config (#1604)

---
 docs/source/configuration.rst | 20 ++++++++++++++++++
 pygeoapi/util.py              | 40 ++++++++++++++++++++++++-----------
 tests/test_util.py            | 36 +++++++++++++++++++++++++++++++
 3 files changed, 84 insertions(+), 12 deletions(-)

diff --git a/docs/source/configuration.rst b/docs/source/configuration.rst
index b9422d8..b11cfdd 100644
--- a/docs/source/configuration.rst
+++ b/docs/source/configuration.rst
@@ -412,6 +412,26 @@ Below is an example of how to integrate system environment variables in pygeoapi
            host: ${MY_HOST}
            port: ${MY_PORT}
 
+Multiple environment variables are supported as follows:
+
+.. code-block:: yaml
+
+   data: ${MY_HOST}:${MY_PORT}
+
+It is also possible to define a default value for a variable in case it does not exist in
+the environment using a syntax like: ``value: ${ENV_VAR:-the default}``
+
+.. code-block:: yaml
+
+   server:
+       bind:
+           host: ${MY_HOST:-localhost}
+           port: ${MY_PORT:-5000}
+   metadata:
+       identification:
+           title:
+               en: This is pygeoapi host ${MY_HOST} and port ${MY_PORT:-5000}, nice to meet you!
+
 
 Hierarchical collections
 ------------------------
diff --git a/pygeoapi/util.py b/pygeoapi/util.py
index 280362d..2ccbeb2 100644
--- a/pygeoapi/util.py
+++ b/pygeoapi/util.py
@@ -163,23 +163,39 @@ def yaml_load(fh: IO) -> dict:
     :returns: `dict` representation of YAML
     """
 
-    # support environment variables in config
-    # https://stackoverflow.com/a/55301129
-    path_matcher = re.compile(r'.*\$\{([^}^{]+)\}.*')
+    # # support environment variables in config
+    # # https://stackoverflow.com/a/55301129
 
-    def path_constructor(loader, node):
-        env_var = path_matcher.match(node.value).group(1)
-        if env_var not in os.environ:
-            msg = f'Undefined environment variable {env_var} in config'
-            raise EnvironmentError(msg)
-        return get_typed_value(os.path.expandvars(node.value))
+    env_matcher = re.compile(
+        r'.*?\$\{(?P<varname>\w+)(:-(?P<default>[^}]+))?\}')
+
+    def env_constructor(loader, node):
+        result = ""
+        current_index = 0
+        raw_value = node.value
+        for match_obj in env_matcher.finditer(raw_value):
+            groups = match_obj.groupdict()
+            varname_start = match_obj.span('varname')[0]
+            result += raw_value[current_index:(varname_start-2)]
+            if (var_value := os.getenv(groups['varname'])) is not None:
+                result += var_value
+            elif (default_value := groups.get('default')) is not None:
+                result += default_value
+            else:
+                raise EnvironmentError(
+                    f'Could not find the {groups["varname"]!r} environment '
+                    f'variable'
+                )
+            current_index = match_obj.end()
+        else:
+            result += raw_value[current_index:]
+        return get_typed_value(result)
 
     class EnvVarLoader(yaml.SafeLoader):
         pass
 
-    EnvVarLoader.add_implicit_resolver('!path', path_matcher, None)
-    EnvVarLoader.add_constructor('!path', path_constructor)
-
+    EnvVarLoader.add_implicit_resolver('!env', env_matcher, None)
+    EnvVarLoader.add_constructor('!env', env_constructor)
     return yaml.load(fh, Loader=EnvVarLoader)
 
 
diff --git a/tests/test_util.py b/tests/test_util.py
index 98097f5..262cfcd 100644
--- a/tests/test_util.py
+++ b/tests/test_util.py
@@ -31,6 +31,8 @@ from datetime import datetime, date, time
 from decimal import Decimal
 from contextlib import nullcontext as does_not_raise
 from copy import deepcopy
+from io import StringIO
+from unittest import mock
 
 import pytest
 from pyproj.exceptions import CRSError
@@ -77,6 +79,40 @@ def test_yaml_load(config):
             util.yaml_load(fh)
 
 
+@pytest.mark.parametrize('env,input_config,expected', [
+    pytest.param({}, 'foo: something', {'foo': 'something'}, id='no-env-expansion'),  # noqa E501
+    pytest.param({'FOO': 'this'}, 'foo: ${FOO}', {'foo': 'this'}),  # noqa E501
+    pytest.param({'FOO': 'this'}, 'foo: the value is ${FOO}', {'foo': 'the value is this'}, id='no-need-for-yaml-tag'),  # noqa E501
+    pytest.param({}, 'foo: ${FOO:-some default}', {'foo': 'some default'}),  # noqa E501
+    pytest.param({'FOO': 'this', 'BAR': 'that'}, 'composite: ${FOO}:${BAR}', {'composite': 'this:that'}),  # noqa E501
+    pytest.param({}, 'composite: ${FOO:-default-foo}:${BAR:-default-bar}', {'composite': 'default-foo:default-bar'}),  # noqa E501
+    pytest.param(
+        {
+            'HOST': 'fake-host',
+            'USER': 'fake',
+            'PASSWORD': 'fake-pass',
+            'DB': 'fake-db'
+        },
+        'connection: postgres://${USER}:${PASSWORD}@${HOST}:${PORT:-5432}/${DB}',  # noqa E501
+        {
+            'connection': 'postgres://fake:fake-pass@fake-host:5432/fake-db'
+        },
+        id='multiple-no-need-yaml-tag'
+    ),
+])
+def test_yaml_load_with_env_variables(
+        env: dict[str, str], input_config: str, expected):
+
+    def mock_get_env(env_var_name):
+        result = env.get(env_var_name)
+        return result
+
+    with mock.patch('pygeoapi.util.os') as mock_os:
+        mock_os.getenv.side_effect = mock_get_env
+        loaded_config = util.yaml_load(StringIO(input_config))
+        assert loaded_config == expected
+
+
 def test_str2bool():
     assert not util.str2bool(False)
     assert not util.str2bool('0')