fix: improve dataset validation and update script paths in documentation

Author: tae898

bindings/python/examples/download_data.py

@@ -724,19 +724,43 @@ def download_stackoverflow(size="small"):
               'large' (~10 GB subset), 'xlarge' (~50 GB subset), or
               'full' (~323 GB)
     """
+    required_xml_files = (
+        "Posts.xml",
+        "Users.xml",
+        "Comments.xml",
+        "Tags.xml",
+        "Badges.xml",
+        "PostLinks.xml",
+        "PostHistory.xml",
+        "Votes.xml",
+    )
+
+    def has_required_xml_files(dataset_dir: Path) -> bool:
+        return dataset_dir.exists() and all(
+            (dataset_dir / filename).exists() for filename in required_xml_files
+        )
+
     # Create data directory
     data_dir = Path(__file__).parent / "data"
     data_dir.mkdir(exist_ok=True)
 
     if size == "tiny":
         source_dir = data_dir / "stackoverflow-small"
-        if not source_dir.exists():
+        if not has_required_xml_files(source_dir):
+            print(
+                "[INFO] stackoverflow-small is missing required XML files; "
+                "downloading a fresh source dataset"
+            )
             download_stackoverflow(size="small")
         return create_stackoverflow_tiny(source_dir=source_dir)
 
     if size == "large":
         source_dir = data_dir / "stackoverflow-full"
-        if not source_dir.exists():
+        if not has_required_xml_files(source_dir):
+            print(
+                "[INFO] stackoverflow-full is missing required XML files; "
+                "downloading a fresh source dataset"
+            )
             download_stackoverflow(size="full")
         return create_stackoverflow_large(
             source_dir=source_dir,
@@ -747,7 +771,11 @@ def download_stackoverflow(size="small"):
 
     if size == "xlarge":
         source_dir = data_dir / "stackoverflow-full"
-        if not source_dir.exists():
+        if not has_required_xml_files(source_dir):
+            print(
+                "[INFO] stackoverflow-full is missing required XML files; "
+                "downloading a fresh source dataset"
+            )
             download_stackoverflow(size="full")
         return create_stackoverflow_large(
             source_dir=source_dir,

bindings/python/scripts/fix_markdown.py

@@ -254,7 +254,7 @@ def main() -> int:
     )
     parser.add_argument(
         "--docs",
-        default="/mnt/ssd2/repos/arcadedb-embedded-python/bindings/python/docs",
+        default="./docs",
         help="Path to docs directory",
     )
     args = parser.parse_args()

bindings/python/scripts/profile-python/README.md

@@ -31,7 +31,7 @@ Required fixes are done. Any further work is optional tuning.
 From `bindings/python`:
 
 ```bash
-/mnt/ssd2/repos/arcadedb-embedded-python/.venv/bin/python scripts/profile-python/profile_bindings.py --preset full --runs 3 --records 5000 --person-count 2000 --vector-records 1500 --vector-k 10 --query-runs 100 --heap-size 4g
+python scripts/profile-python/profile_bindings.py --preset full --runs 3 --records 5000 --person-count 2000 --vector-records 1500 --vector-k 10 --query-runs 100 --heap-size 4g
 ```
 
 ## Read The Report

bindings/python/scripts/profile-python/results.md

@@ -8,8 +8,10 @@ This directory keeps one current benchmark artifact and one short summary.
 
 Run used to produce it:
 
+from `bindings/python`:
+
 ```bash
-cd /mnt/ssd2/repos/arcadedb-embedded-python/bindings/python && /mnt/ssd2/repos/arcadedb-embedded-python/.venv/bin/python scripts/profile-python/profile_bindings.py --preset full --runs 3 --records 5000 --person-count 2000 --vector-records 1500 --vector-k 10 --query-runs 100 --heap-size 4g
+python scripts/profile-python/profile_bindings.py --preset full --runs 3 --records 5000 --person-count 2000 --vector-records 1500 --vector-k 10 --query-runs 100 --heap-size 4g
 ```
 
 ## Current Numbers