JSON Schema validation and URL validation for PRs

Author: adamziel

.github/workflows/ships-blueprint-json.yml .github/workflows/has-valid-blueprint-json.yml

@@ -1,4 +1,4 @@
-name: Ships a valid Blueprint.json file?
+name: Has a valid Blueprint.json file?
 
 on:
     pull_request:
@@ -15,6 +15,10 @@ jobs:
               run: |
                   git fetch origin trunk
 
+            - name: Install pre-requisites
+              run: |
+                  pip install -r requirements.txt
+
             - name: Run validate_pr.py
               run: |
                   python validate_pr.py

blueprints/custom-post/blueprint.json

@@ -4,7 +4,10 @@
 		"title": "Custom Post Type: Books",
 		"description": "Blueprint that added a custom post type to playground",
 		"author": "bph",
-		"categories": ["Content", "CPT"]
+		"categories": [
+			"Content",
+			"CPT"
+		]
 	},
 	"landingPage": "/wp-admin/",
 	"steps": [
@@ -28,4 +31,4 @@
 			"pluginPath": "books/books.php"
 		}
 	]
-}
+}

requirements.txt

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

validate_pr.py

@@ -1,9 +1,13 @@
 import os
 import re
+from jsonschema import Draft7Validator, validate, ValidationError
 import json
 import sys
+import requests
 
 def validate_blueprints():
+    errors = []
+
     # Get the list of directories touched in the current branch
     touched_dirs = get_touched_directories()
 
@@ -14,17 +18,71 @@ def validate_blueprints():
 
         # Check if blueprint.json file exists
         if not os.path.exists(blueprint_json_path):
-            print(f"Error: {dir}/{blueprint_json_path} file does not exist in this PR.")
-            sys.exit(1)
+            errors.append(f"Error: {dir}/{blueprint_json_path} file does not exist in this PR.")
+            continue
 
         # Read and validate the JSON file
         try:
             with open(blueprint_json_path, 'r') as f:
                 blueprint_json = json.load(f)
         except json.JSONDecodeError as e:
-            print(f"Error: Invalid JSON in {blueprint_json_path}: {str(e)}")
-            sys.exit(1)
+            errors.append(f"Error: Invalid JSON in {blueprint_json_path}: {str(e)}")
+            continue 
+
+        # Validate the Blueprint against the JSON schema
+        schema_url = 'https://playground.wordpress.net/blueprint-schema.json'
+        schema = json.loads(requests.get(schema_url).text)
+        
+        try:
+            validate(instance=blueprint_json, schema=schema, cls=Draft7Validator)
+        except ValidationError as e:
+            error_path = " > ".join(e.absolute_path)
+            at_error = f"at {error_path}" if error_path else "at root" 
+            errors.append(
+                f"Error: {dir}/{blueprint_json_path} does not match the JSON schema.\n"
+                f"{str(e.message)} {at_error}.\n"
+            )
+            continue
+
+        # Recursively find all urls in the blueprint.json file
+        urls = find_urls(blueprint_json)
 
+        # Check if the URLs all point to raw.githubusercontent.com/adamziel/blueprints/{CURRENT BRANCH}
+        urls_valid = True
+        current_branch = os.popen('git rev-parse --abbrev-ref HEAD').read().strip()
+        for url in urls:
+            if not url.startswith('https://') and not url.startswith('http://'):
+                continue
+            if not url.startswith(f'https://raw.githubusercontent.com/adamziel/blueprints/{current_branch}/'):
+                urls_valid = False
+                errors.append(
+                    f"Error: {dir}/{blueprint_json_path} contains a URL that is not allowed: \n* {url}\n"
+                    f"Since the current branch is {current_branch}, the URL should start with \n* https://raw.githubusercontent.com/adamziel/blueprints/{current_branch}/\n"
+                    "In general, all URLs in the blueprint.json file must start with https://raw.githubusercontent.com/adamziel/blueprints/{CURRENT BRANCH}/"
+                )
+
+        if not urls_valid:
+            continue
+
+    if len(errors):
+        for error in errors:
+            print(error)
+        sys.exit(1)
+    
+
+def find_urls(obj):
+    urls = []
+    if isinstance(obj, dict):
+        for key, value in obj.items():
+            if key == 'url':
+                urls.append(value)
+            urls.extend(find_urls(value))
+    elif isinstance(obj, list):
+        for item in obj:
+            urls.extend(find_urls(item))
+    return urls
+
+        
 
 def get_touched_directories():
     # Run git diff command to get the list of directories touched in the current branch
@@ -37,5 +95,6 @@ def get_touched_directories():
         touched_dirs.add(dir_path)
 
     return touched_dirs
+
 # Call the function to validate blueprints
 validate_blueprints()