Move data files

.gitignore

@@ -1,5 +1,5 @@
 .env
-data/
+fetched_data/
 valid_data/
 
 # Fasttext model

fetch_files.sh

@@ -5,8 +5,8 @@ pv commits.json |
     while read url; do
         # Strip the url prefix to get the path
         path=$(echo "$url" | cut -d/ -f4-)
-        if ! [ -f "data/$path" ]; then
-            curl "$url" --silent --create-dirs -o "data/$path"
+        if ! [ -f "fetched_data/$path" ]; then
+            curl "$url" --silent --create-dirs -o "fetched_data/$path"
             sleep 1
         fi
     done

train_split.py

@@ -31,7 +31,7 @@ def files_list(licenses):
 
 def write_schemas(filename, schema_list, schema_data):
     sys.stderr.write(f"Writing {filename}…\n")
-    with gzip.open(filename, "wt") as f:
+    with gzip.open(Path("data") / filename, "wt") as f:
         for schema in tqdm.tqdm(list(schema_list)):
             filename = str(os.path.join(*Path(schema).parts[1:]))
 

validate_schemas.py

@@ -36,5 +36,5 @@ if __name__ == "__main__":
     # Increase the recursion limit to handle large schemas
     sys.setrecursionlimit(10000)
 
-    data_path = Path("data")
+    data_path = Path("fetched_data")
     process_map(process_file, list(data_path.rglob("*.json")), chunksize=10)