Merge pull request #384 from brightsparklabs/feature/APPCLI-112-valid…

…ate-config-files

Feature/appcli 112 validate config files

CHANGELOG.md

@@ -14,6 +14,7 @@ The changelog is applicable from version `1.0.0` onwards.
 
 - APPCLI-133: Add fix for Git repository ownership issue to the AppCli Dockerfile.
 - APED-67: Add support for running `NullOrchestrator` apps on Windows OS.
+- APPCLI-112: Autovalidate config file when a `.schema.json` file is provided.
 - APPCLI-115: Enable automerge for dependabot PRs.
 
 ### Fixed

README.md

@@ -228,6 +228,32 @@ def main():
   - `resources/templates/baseline` - for templates which the end user **is not** expected to modify.
   - `resources/templates/configurable` - for templates which the end user is expected to modify.
 
+#### Schema validation
+
+Configuration files will be automatically validated against provided schema files whenever
+`configure apply` is run.
+Validation is done with [jsonschema](https://json-schema.org/) and is only available for `yaml/yml`
+and `json/jsn` files.
+The JSON schema file must match the name of the file to validate with a suffix of `.schema.json.`.
+It must be placed in the same directory as the file to validate,
+The `settings.yml`, `stack_settings.yml` file, and any files in the `resource/templates` or
+`resources/overrides` directory can be validated.
+
+```yaml
+# resources/templates/configurable/my-config.yml
+foobar: 5
+```
+
+```json
+# resources/templates/configurable/my-config.yml.schema.json
+{
+    "type": "object",
+    "properties" : {
+        "foobar" : {"type": "number"}
+    }
+}
+```
+
 #### Application context files
 
 Template files are templated with Jinja2. The 'data' passed into the templating engine

appcli/commands/configure_cli.py

@@ -74,6 +74,12 @@ def init(ctx):
             logger.debug("Running pre-configure init hook")
             hooks.pre_configure_init(ctx)
 
+            # Validate the configuration schema.
+            logger.debug("Validating configuration files")
+            ConfigurationManager(
+                cli_context, self.cli_configuration
+            ).validate_configuration()
+
             # Initialise configuration directory
             logger.debug("Initialising configuration directory")
             ConfigurationManager(

appcli/configuration_manager.py

@@ -2,7 +2,7 @@
 # # -*- coding: utf-8 -*-
 
 """
-Manages configuration.
+Manages the configuration of the application.
 ________________________________________________________________________________
 
 Created by brightSPARK Labs
@@ -19,20 +19,22 @@
 from pathlib import Path
 from typing import Iterable
 
+import jsonschema
+import yaml
+
 # vendor libraries
 from jinja2 import StrictUndefined, Template
 
-from appcli.crypto import crypto
-
 # local libraries
+from appcli.crypto import crypto
 from appcli.crypto.crypto import decrypt_values_in_file
 from appcli.functions import error_and_exit, print_header
 from appcli.git_repositories.git_repositories import (
     ConfigurationGitRepository,
     GeneratedConfigurationGitRepository,
 )
 from appcli.logger import logger
-from appcli.models.cli_context import CliContext
+from appcli.models.cli_context import SCHEMA_SUFFIX, CliContext
 from appcli.models.configuration import Configuration
 from appcli.variables_manager import VariablesManager
 
@@ -43,6 +45,14 @@
 METADATA_FILE_NAME = "metadata-configure-apply.json"
 """ Name of the file holding metadata from running a configure (relative to the generated configuration directory) """
 
+FILETYPE_LOADERS = {
+    ".json": json.load,
+    ".jsn": json.load,
+    ".yaml": yaml.safe_load,
+    ".yml": yaml.safe_load,
+}
+""" The supported filetypes for the validator. """
+
 # ------------------------------------------------------------------------------
 # PUBLIC CLASSES
 # ------------------------------------------------------------------------------
@@ -59,6 +69,39 @@ def __init__(self, cli_context: CliContext, configuration: Configuration):
         self.cli_configuration: Configuration = configuration
         self.variables_manager: VariablesManager = cli_context.get_variables_manager()
 
+    def validate_configuration(self):
+        """Validates all settings files that have associated schema files."""
+
+        # Define all the config directories and files.
+        settings_schema: Path = self.cli_context.get_app_configuration_file_schema()
+        stack_settings_schema: Path = (
+            self.cli_context.get_stack_configuration_file_schema()
+        )
+        overrides_dir: Path = self.cli_context.get_baseline_template_overrides_dir()
+        templates_dir: Path = self.cli_context.get_configurable_templates_dir()
+
+        # Parse the directories to get the schema files.
+        schema_files = []
+        if settings_schema.is_file():
+            schema_files.append(settings_schema)
+        if stack_settings_schema.is_file():
+            schema_files.append(stack_settings_schema)
+        schema_files.extend(overrides_dir.glob(f"**/*{SCHEMA_SUFFIX}"))
+        schema_files.extend(templates_dir.glob(f"**/*{SCHEMA_SUFFIX}"))
+
+        for schema_file in schema_files:
+            # Take out the `schema` suffix to get the original config file.
+            # NOTE: As we only need to remove part of the suffix, it is easier to convert to string
+            # and just remove part of the substring.
+            config_file: Path = Path(str(schema_file).removesuffix(SCHEMA_SUFFIX))
+            if not config_file.exists():
+                logger.warning(f"Found {schema_file} but no matching config file.")
+                continue
+
+            # Load and validate the config/schema.
+            logger.debug(f"Found schema for {config_file}. Validating...")
+            self.__validate_schema(config_file, schema_file)
+
     def initialise_configuration(self):
         """Initialises the configuration repository"""
 
@@ -206,6 +249,37 @@ def set_variable(self, variable: str, value: any):
     def get_stack_variable(self, variable: str):
         return self.variables_manager.get_stack_variable(variable)
 
+    def __validate_schema(self, config_file: Path, schema_file: Path) -> None:
+        """Attempt to validate a config file against a provided schema file.
+        The function will try and determine the file content based off the extension.
+        It will throw if the extension is unknown or the contents do not match the expected format.
+
+        Args:
+            config_file (Path): Path to the config file to validate.
+            schema_file (Path): Path to the schema file.
+        """
+        # Check we actually have a loader for the filetype.
+        try:
+            loader = FILETYPE_LOADERS[config_file.suffix]
+        except KeyError:
+            error_and_exit(
+                f"The `{config_file}` file does not have an associated loader function."
+                f"Check that the suffix is one of the supported types: {[k for k in FILETYPE_LOADERS.keys()]}"
+            )
+
+        # Load the files.
+        data = loader(open(config_file, "r"))
+        schema = json.load(open(schema_file, "r"))
+
+        # Validate the config.
+        try:
+            jsonschema.validate(instance=data, schema=schema)
+            logger.debug(
+                f"The configuration file `{config_file}` matched the provided schema."
+            )
+        except jsonschema.exceptions.ValidationError as e:
+            error_and_exit(f"Validation of `{config_file}` failed at:\n{e}")
+
     def __create_new_configuration_branch_and_files(self):
         app_version: str = self.cli_context.app_version
         app_version_branch: str = self.config_repo.generate_branch_name(app_version)
@@ -256,11 +330,11 @@ def __seed_configuration_dir(self):
         os.makedirs(target_app_configuration_file.parent, exist_ok=True)
         shutil.copy2(seed_app_configuration_file, target_app_configuration_file)
 
+        # Copy in the stack configuration file.
         stack_configuration_file = self.cli_configuration.stack_configuration_file
         target_stack_configuration_file = (
             self.cli_context.get_stack_configuration_file()
         )
-        # Copy in the stack configuration file
         if stack_configuration_file.is_file():
             shutil.copy2(stack_configuration_file, target_stack_configuration_file)
 

appcli/models/cli_context.py

@@ -1,6 +1,14 @@
 #!/usr/bin/env python3
 # # -*- coding: utf-8 -*-
 
+"""
+Outlines the CLI context.
+________________________________________________________________________________
+
+Created by brightSPARK Labs
+www.brightsparklabs.com
+"""
+
 # standard libraries
 from pathlib import Path
 from typing import Dict, Iterable, NamedTuple, Tuple
@@ -13,6 +21,18 @@
 from appcli.logger import logger
 from appcli.variables_manager import VariablesManager
 
+# ------------------------------------------------------------------------------
+# CONSTANTS
+# ------------------------------------------------------------------------------
+
+SCHEMA_SUFFIX = ".schema.json"
+""" The suffix for the validation schema files. """
+
+
+# ------------------------------------------------------------------------------
+# PUBLIC CLASSES
+# ------------------------------------------------------------------------------
+
 
 class CliContext(NamedTuple):
     """Shared context from a run of the CLI."""
@@ -131,6 +151,14 @@ def get_app_configuration_file(self) -> Path:
         """
         return self.configuration_dir.joinpath("settings.yml")
 
+    def get_app_configuration_file_schema(self) -> Path:
+        """Get the location of the configuration schema file
+
+        Returns:
+            Path: location of the configuration schema file
+        """
+        return self.configuration_dir.joinpath("settings.yml", SCHEMA_SUFFIX)
+
     def get_stack_configuration_file(self) -> Path:
         """Get the location of the configuration file
 
@@ -139,6 +167,14 @@ def get_stack_configuration_file(self) -> Path:
         """
         return self.configuration_dir.joinpath("stack-settings.yml")
 
+    def get_stack_configuration_file_schema(self) -> Path:
+        """Get the location of the configuration schema file
+
+        Returns:
+            Path: location of the configuration schema file
+        """
+        return self.configuration_dir.joinpath("stack-settings.yml", SCHEMA_SUFFIX)
+
     def get_baseline_template_overrides_dir(self) -> Path:
         """Get the directory of the configuration template overrides
 

appcli/models/configuration.py

@@ -1,6 +1,14 @@
 #!/usr/bin/env python3
 # # -*- coding: utf-8 -*-
 
+"""
+Configuration for building the CLI.
+________________________________________________________________________________
+
+Created by brightSPARK Labs
+www.brightsparklabs.com
+"""
+
 # standard libraries
 import inspect
 import os

setup.py

@@ -64,6 +64,7 @@ def get_version():
         "dataclasses-json==0.5.7",
         "deepdiff==6.5.0",
         "GitPython==3.1.35",
+        "jsonschema==4.19.1",
         "jinja2==3.1.2",
         "pre-commit==3.3.3",
         "pycryptodome==3.18.0",

tests/configuration_manager/invalid_resources/settings.yml

@@ -0,0 +1 @@
+---

tests/configuration_manager/invalid_resources/templates/configurable/invalid.yml

@@ -0,0 +1 @@
+foobar: "5"  #   <- This should be int.

tests/configuration_manager/invalid_resources/templates/configurable/invalid.yml.schema.json

@@ -0,0 +1,6 @@
+{
+    "type": "object",
+    "properties" : {
+        "foobar" : {"type": "number"}
+    }
+}

tests/configuration_manager/invalid_resources/templates/stack_settings.yml

@@ -0,0 +1 @@
+---

tests/configuration_manager/resources/templates/configurable/valid_json.json

@@ -0,0 +1,6 @@
+{
+    "alphabet": {
+        "a": 1,
+        "b": 2
+    }
+}

tests/configuration_manager/resources/templates/configurable/valid_json.json.schema.json

@@ -0,0 +1,12 @@
+{
+    "type": "object",
+    "properties" : {
+        "alphabet" : {
+            "type": "object",
+            "properties" : {
+                "a": {"type": "number"},
+                "b": {"type": "number"}
+            }
+        }
+    }
+}

tests/configuration_manager/resources/templates/configurable/valid_yaml.yml

@@ -0,0 +1,3 @@
+alphabet:
+  a: 1
+  b: 2

tests/configuration_manager/resources/templates/configurable/valid_yaml.yml.schema.json

@@ -0,0 +1,12 @@
+{
+    "type": "object",
+    "properties" : {
+        "alphabet" : {
+            "type": "object",
+            "properties" : {
+                "a": {"type": "number"},
+                "b": {"type": "number"}
+            }
+        }
+    }
+}