Merge pull request #232 from kingthorin/sorting

sort tech files - add sorting and duplication checks

Author: enthec-opensource

.github/workflows/scripts/order_validator.py

@@ -0,0 +1,64 @@
+import json
+import pathlib
+from typing import Final, Any
+
+
+class DuplicateKeyException(Exception):
+    def __init__(self, msg: str):
+        super().__init__(msg)
+
+
+class OrderingException(Exception):
+    def __init__(self, msg: str):
+        super().__init__(msg)
+
+
+class InvalidStructureException(Exception):
+    def __init__(self, msg: str):
+        super().__init__(msg)
+
+
+class OrderValidator:
+    def __init__(self):
+        self._SOURCE_DIR: Final[str] = "src"
+        self._TECH_DIR: Final[str] = "technologies"
+        self._FULL_TECH_DIR: Final[pathlib.Path] = pathlib.Path(self._SOURCE_DIR).joinpath(self._TECH_DIR)
+
+    def validate(self) -> None:
+        if not self._FULL_TECH_DIR.is_dir():
+            raise InvalidStructureException(f"{self._FULL_TECH_DIR} is not a valid directory")
+        for tech_file in sorted(self._FULL_TECH_DIR.iterdir()):
+            if not tech_file.name.endswith(".json"):
+                continue
+            self._check_ordering(tech_file)
+
+    def _check_ordering(self, tech_file: pathlib.Path) -> None:
+        with tech_file.open("r", encoding="utf8") as f:
+            data: dict = json.load(f, object_pairs_hook=self._duplicate_key_validator)
+        if not isinstance(data, dict):
+            raise InvalidStructureException(f"{tech_file.name} root must be an object")
+        keys: list[str] = list(data.keys())
+        sorted_keys: list[str] = sorted(keys, key=lambda x: x.lower())
+        if keys != sorted_keys:
+            mismatches: list[str] = [
+                f"Found '{actual}' expected '{expected}'"
+                for actual, expected in zip(keys, sorted_keys)
+                if actual != expected
+            ]
+            raise OrderingException(
+                f"{tech_file.name} entries are not ordered alphabetically by keys:\n  " +
+                "\n  ".join(mismatches)
+            )
+
+    @classmethod
+    def _duplicate_key_validator(cls, pairs: list[tuple[str, Any]]) -> dict[str, Any]:
+        result: dict[str, Any] = {}
+        for key, value in pairs:
+            if key in result:
+                raise DuplicateKeyException(f"Duplicate key found: '{key}'")
+            result[key] = value
+        return result
+
+
+if __name__ == '__main__':
+    OrderValidator().validate()

.github/workflows/validate.yml

@@ -18,6 +18,24 @@ jobs:
       - name: run structure validator
         run: python3 .github/workflows/scripts/structure_validator.py
 
+  validate_order_duplication:
+    runs-on: ubuntu-22.04
+    needs: validate_structure
+    strategy:
+      matrix:
+        python-version: [ "3.12" ]
+    steps:
+      - name: checkout repository
+        uses: actions/checkout@v4
+
+      - name: set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: run order validator
+        run: python3 .github/workflows/scripts/order_validator.py src/technologies
+
   validate_schema:
     runs-on: ubuntu-24.04
     needs: validate_structure

scripts/fix_order.py

@@ -0,0 +1,41 @@
+import json
+import pathlib
+from typing import Final, Any
+
+
+class DuplicateKeyException(Exception):
+    def __init__(self, msg: str):
+        super().__init__(msg)
+
+
+class OrderFixer:
+    def __init__(self):
+        self._SOURCE_DIR: Final[str] = "src"
+        self._TECH_DIR: Final[str] = "technologies"
+        self._FULL_TECH_DIR: Final[pathlib.Path] = pathlib.Path(self._SOURCE_DIR).joinpath(self._TECH_DIR)
+
+    def fix(self) -> None:
+        for tech_file in sorted(self._FULL_TECH_DIR.iterdir()):
+            if not tech_file.name.endswith(".json"):
+                continue
+            with tech_file.open("r", encoding="utf8") as f:
+                data: dict = json.load(f, object_pairs_hook=self._duplicate_key_validator)
+            sorted_data: dict[str, Any] = {
+                key: data[key] for key in sorted(data.keys(), key=lambda x: x.lower())
+            }
+            with tech_file.open("w", encoding="utf8") as f:
+                json.dump(sorted_data, f, indent=2, ensure_ascii=False)
+                f.write("\n")
+
+    @classmethod
+    def _duplicate_key_validator(cls, pairs: list[tuple[str, Any]]) -> dict[str, Any]:
+        result: dict[str, Any] = {}
+        for key, value in pairs:
+            if key in result:
+                raise DuplicateKeyException(f"Duplicate key found: '{key}'")
+            result[key] = value
+        return result
+
+
+if __name__ == '__main__':
+    OrderFixer().fix()

src/technologies/_.json

@@ -292,6 +292,27 @@
     ],
     "website": "https://www.5centscdn.net"
   },
+  "6sense": {
+    "cats": [
+      32,
+      76
+    ],
+    "description": "6sense is a B2B predictive intelligence platform for marketing and sales.",
+    "headers": {
+      "Content-Security-Policy": "\\.6sc\\.co/"
+    },
+    "icon": "6sense.svg",
+    "pricing": [
+      "poa",
+      "high",
+      "recurring"
+    ],
+    "saas": true,
+    "scriptSrc": [
+      "\\.6sc\\.co/"
+    ],
+    "website": "https://6sense.com"
+  },
   "6Valley eCommerce CMS": {
     "cats": [
       1,
@@ -310,26 +331,18 @@
     ],
     "website": "https://6valley.app"
   },
-  "6sense": {
+  "7moor": {
     "cats": [
       32,
-      76
+      53
     ],
-    "description": "6sense is a B2B predictive intelligence platform for marketing and sales.",
-    "headers": {
-      "Content-Security-Policy": "\\.6sc\\.co/"
+    "description": "7moor is an integrated customer service marketing solution that combines communication, customer support, and marketing tools.",
+    "icon": "7moor.svg",
+    "js": {
+      "moor7Source": ""
     },
-    "icon": "6sense.svg",
-    "pricing": [
-      "poa",
-      "high",
-      "recurring"
-    ],
     "saas": true,
-    "scriptSrc": [
-      "\\.6sc\\.co/"
-    ],
-    "website": "https://6sense.com"
+    "website": "https://www.7moor.com"
   },
   "7Shifts": {
     "cats": [
@@ -348,19 +361,6 @@
     "saas": true,
     "website": "https://www.7shifts.com"
   },
-  "7moor": {
-    "cats": [
-      32,
-      53
-    ],
-    "description": "7moor is an integrated customer service marketing solution that combines communication, customer support, and marketing tools.",
-    "icon": "7moor.svg",
-    "js": {
-      "moor7Source": ""
-    },
-    "saas": true,
-    "website": "https://www.7moor.com"
-  },
   "8base": {
     "cats": [
       3,
@@ -467,4 +467,4 @@
     ],
     "website": "https://hyperscript.org"
   }
-}
+}