Merge pull request #114 from MPEGGroup/pr_check_fixes

PR Check Fixes

conformance-search/src/types/json.ts

@@ -4,7 +4,9 @@ export type Coverage = {
         percentage: number;
         boxes: string[];
         missing_extensions: string[];
-        paths: string[];
+        paths: {
+            [path: string]: string[];
+        };
     };
     lists: {
         boxes: {

data/schemas/gpac-extension.schema.json

@@ -31,13 +31,18 @@
         },
         "@Specification": { "type": "string" },
         "@Container": { "type": "string" },
+        "@data": {
+          "type": "string",
+          "description": "dump of the box in hexadecimal format",
+          "pattern": "^0x[0-9a-fA-F]+$"
+        },
         "children": {
           "type": "array",
           "items": { "$ref": "#/$defs/box" }
         }
       },
       "additionalProperties": true,
-      "required": ["@Type"]
+      "required": ["@Type", "@data"]
     }
   },
   "type": "object",

src/construct/coverage.py

@@ -94,17 +94,29 @@ def main():
         "percentage": len(files["not_found"]) / len(files["path_file_map"]),
         "boxes": list(set(p.split(".")[-1] for p in files["not_found"])),
         "missing_extensions": list(missing_extensions),
-        "paths": files["not_found"],
+        "paths": list(files["not_found"].keys()),
     }
 
     # FIXME: All the logs here should be errors, except for info
-    for upath in NOT_FOUND["paths"]:
+    for upath, in_files in files["not_found"].items():
         # Easy to access variables
         container_path = upath.split(".")[:-1]
         box_fourcc = upath.split(".")[-1]
         known_box = box_fourcc in dictionary["fourccs"]
 
         if not known_box:
+            # Check if this was in under consideration files
+            if any(["under_consideration" in f for f in in_files]):
+                extra = ""
+                if box_fourcc in get_mp4ra_boxes():
+                    extra = " It exists in MP4RA though."
+
+                logger.error(
+                    f"Box {box_fourcc} was found in under consideration files but it is not in our database."
+                    + extra
+                )
+                continue
+
             if box_fourcc not in get_mp4ra_boxes():
                 logger.info(f"Box {box_fourcc} is not in standard features or MP4RA")
             else:

src/construct/files.py

@@ -145,7 +145,7 @@ def main():
         logger.warning(f"Found {len(ignored)} ignored files.")
 
     file_metadata = {}
-    not_found = set()
+    not_found = {}
 
     for file in files:
         # metadata is the one without the _gpac.json
@@ -187,7 +187,9 @@ def main():
 
         for path, variants in paths_contained.items():
             if path not in path_file_map:
-                not_found.add(path)
+                if path not in not_found:
+                    not_found[path] = set()
+                not_found[path].add(key_name)
                 continue
 
             for variant in variants:
@@ -247,7 +249,7 @@ def main():
     with open("output/files.json", "w", encoding="utf-8") as f:
         json.dump(
             {
-                "not_found": list(not_found),
+                "not_found": {path: sorted(files) for path, files in not_found.items()},
                 "path_file_map": path_file_map,
                 "feature_file_map": feature_file_map,
                 "file_metadata": file_metadata,

src/tests/test_files.py

@@ -313,9 +313,11 @@ def test_gpac_ext_consistency(check):
         with open(gpac_ext, "r", encoding="utf-8") as f:
             gpac_ext_dict = json.load(f)
 
-        # Test if locations are the same
-        gt_locations = [ub["location"] for ub in unknown_boxes]
-        ref_locations = [ub["location"] for ub in gpac_ext_dict["extensions"]]
+        # Test if boxes are the same
+        gt_locations = [(ub["location"], ub["box"]["@data"]) for ub in unknown_boxes]
+        ref_locations = [
+            (ub["location"], ub["box"]["@data"]) for ub in gpac_ext_dict["extensions"]
+        ]
 
         # Reference must match exactly
         with check: