fix: coderabbit

Author: Integer-Ctrl

README.md

@@ -67,8 +67,8 @@ Commands to download the [DBpedia Knowledge Graphs](#dbpedia-knowledge-graphs) g
 
 To download BUSL 1.1 licensed datasets, you need to register and get an access token.
 
-1. If you do not have a DBpedia Account yet (Forum/Databus), please register at https://account.dbpedia.org
-2. Log in at https://account.dbpedia.org and create your token.
+1. If you do not have a DBpedia Account yet (Forum/Databus), please register at [https://account.dbpedia.org](https://account.dbpedia.org)
+2. Log in at [https://account.dbpedia.org](https://account.dbpedia.org) and create your token.
 3. Save the token to a file, e.g. `vault-token.dat`.
 
 ### DBpedia Knowledge Graphs
@@ -182,7 +182,7 @@ Options:
   --databus TEXT      Databus URL (if not given, inferred from databusuri,
                       e.g. https://databus.dbpedia.org/sparql)
   --vault-token TEXT  Path to Vault refresh token file
-  --databus-key TEXT  Databus API key to donwload from protected databus
+  --databus-key TEXT  Databus API key to download from protected databus
   --authurl TEXT      Keycloak token endpoint URL  [default:
                       https://auth.dbpedia.org/realms/dbpedia/protocol/openid-
                       connect/token]
@@ -191,7 +191,7 @@ Options:
   --help              Show this message and exit.
 ```
 
-### Examples of using the download command
+#### Examples of using the download command
 
 **Download File**: download of a single file
 ```bash
@@ -397,7 +397,7 @@ docker run --rm -v $(pwd):/data dbpedia/databus-python-client deploy \
   ./data_folder
 ```
 
-<a id="module-delete"></a>
+<a id="cli-delete"></a>
 ### Delete
 
 With the delete command you can delete collections, groups, artifacts, and versions from the Databus. Deleting files is not supported via API.
@@ -437,11 +437,9 @@ To authenticate the delete request, you need to provide an API key with `--datab
 
 If you want to perform a dry run without actual deletion, use the `--dry-run` option. This will show you what would be deleted without making any changes.
 
-As securety measure, the delete command will prompt you for confirmation before proceeding with the deletion. If you want to skip this prompt, you can use the `--force` option.
+As security measure, the delete command will prompt you for confirmation before proceeding with the deletion. If you want to skip this prompt, you can use the `--force` option.
 
-**Example of using the delete command**
-
-### Examples of using the download command
+#### Examples of using the delete command
 
 **Delete Version**: delete a specific version
 ```bash

databusclient/api/delete.py

@@ -65,7 +65,7 @@ def _delete_resource(databusURI: str, databus_key: str, dry_run: bool = False, f
         print(f"[DRY RUN] Would delete: {databusURI}")
         return
 
-    response = requests.delete(databusURI, headers=headers)
+    response = requests.delete(databusURI, headers=headers, timeout=30)
 
     if response.status_code in (200, 204):
         print(f"Successfully deleted: {databusURI}")
@@ -112,11 +112,10 @@ def _delete_artifact(databusURI: str, databus_key: str, dry_run: bool = False, f
         raise ValueError("No versions found in artifact JSON-LD")
 
     # Delete all versions
-    _delete_list(version_uris, databus_key, dry_run=dry_run)
+    _delete_list(version_uris, databus_key, dry_run=dry_run, force=force)
 
     # Finally, delete the artifact itself
-    _delete_resource(databusURI, databus_key, dry_run=dry_run)
-
+    _delete_resource(databusURI, databus_key, dry_run=dry_run, force=force)
 
 def _delete_group(databusURI: str, databus_key: str, dry_run: bool = False, force: bool = False):
     """
@@ -167,7 +166,7 @@ def delete(databusURIs: List[str], databus_key: str, dry_run: bool, force: bool)
     """
 
     for databusURI in databusURIs:
-        host, account, group, artifact, version, file = get_databus_id_parts_from_uri(databusURI)
+        _host, _account, group, artifact, version, file = get_databus_id_parts_from_uri(databusURI)
 
         if group == "collections" and artifact is not None:
             print(f"Deleting collection: {databusURI}")
@@ -185,4 +184,4 @@ def delete(databusURIs: List[str], databus_key: str, dry_run: bool, force: bool)
             print(f"Deleting group and all its artifacts and versions: {databusURI}")
             _delete_group(databusURI, databus_key, dry_run=dry_run, force=force)
         else:
-            print(f"Deleting ${databusURI} is not supported.")
+            print(f"Deleting {databusURI} is not supported.")

databusclient/api/utils.py

@@ -17,7 +17,7 @@ def get_databus_id_parts_from_uri(uri: str) -> Tuple[Optional[str], Optional[str
     parts += [None] * (6 - len(parts))  # pad with None if less than 6 parts
     return tuple(parts[:6])  # return only the first 6 parts
 
-def get_json_ld_from_databus(uri: str, databus_key: str = None) -> str:
+def get_json_ld_from_databus(uri: str, databus_key: str | None = None) -> str:
     """
     Retrieve JSON-LD representation of a databus resource.
 
@@ -31,7 +31,7 @@ def get_json_ld_from_databus(uri: str, databus_key: str = None) -> str:
     headers = {"Accept": "application/ld+json"}
     if databus_key is not None:
         headers["X-API-KEY"] = databus_key
-    response = requests.get(uri, headers=headers)
+    response = requests.get(uri, headers=headers, timeout=30)
     response.raise_for_status()
 
     return response.text

databusclient/cli.py

@@ -96,7 +96,7 @@ def deploy(version_id, title, abstract, description, license_url, apikey,
 @click.option("--localdir", help="Local databus folder (if not given, databus folder structure is created in current working directory)")
 @click.option("--databus", help="Databus URL (if not given, inferred from databusuri, e.g. https://databus.dbpedia.org/sparql)")
 @click.option("--vault-token", help="Path to Vault refresh token file")
-@click.option("--databus-key", help="Databus API key to donwload from protected databus")
+@click.option("--databus-key", help="Databus API key to download from protected databus")
 @click.option("--authurl", default="https://auth.dbpedia.org/realms/dbpedia/protocol/openid-connect/token", show_default=True, help="Keycloak token endpoint URL")
 @click.option("--clientid", default="vault-token-exchange", show_default=True, help="Client ID for token exchange")
 def download(databusuris: List[str], localdir, databus, vault_token, databus_key, authurl, clientid):

databusclient/client.py

@@ -568,8 +568,7 @@ def __download_file__(url, filename, vault_token_file=None, databus_key=None, au
 
     # TODO: could be a problem of github raw / openflaas
     if total_size_in_bytes != 0 and progress_bar.n != total_size_in_bytes:
-        # raise IOError("Downloaded size does not match Content-Length header")
-        print(f"Warning: Downloaded size does not match Content-Length header:\nExpected {total_size_in_bytes}, got {progress_bar.n}")
+        raise IOError("Downloaded size does not match Content-Length header")
 
 
 def __get_vault_access__(download_url: str,
@@ -622,13 +621,14 @@ def __get_vault_access__(download_url: str,
     return vault_token
 
 
-def __query_sparql__(endpoint_url, query) -> dict:
+def __query_sparql__(endpoint_url, query, databus_key=None) -> dict:
     """
     Query a SPARQL endpoint and return results in JSON format.
 
     Parameters:
     - endpoint_url: the URL of the SPARQL endpoint
     - query: the SPARQL query string
+    - databus_key: Optional API key for authentication
 
     Returns:
     - Dictionary containing the query results
@@ -637,12 +637,14 @@ def __query_sparql__(endpoint_url, query) -> dict:
     sparql.method = 'POST'
     sparql.setQuery(query)
     sparql.setReturnFormat(JSON)
+    if databus_key is not None:
+        sparql.setCustomHttpHeaders({"X-API-KEY": databus_key})
     results = sparql.query().convert()
     return results
 
 
-def __handle_databus_file_query__(endpoint_url, query) -> List[str]:
-    result_dict = __query_sparql__(endpoint_url, query)
+def __handle_databus_file_query__(endpoint_url, query, databus_key=None) -> List[str]:
+    result_dict = __query_sparql__(endpoint_url, query, databus_key=databus_key)
     for binding in result_dict['results']['bindings']:
         if len(binding.keys()) > 1:
             print("Error multiple bindings in query response")
@@ -760,6 +762,7 @@ def download(
     endpoint: the databus endpoint URL
     databusURIs: identifiers to access databus registered datasets
     token: Path to Vault refresh token file
+    databus_key: Databus API key for protected downloads
     auth_url: Keycloak token endpoint URL
     client_id: Client ID for token exchange
     """
@@ -823,5 +826,5 @@ def download(
             print("QUERY {}", databusURI.replace("\n", " "))
             if endpoint is None:  # endpoint is required for queries (--databus)
                 raise ValueError("No endpoint given for query")
-            res = __handle_databus_file_query__(endpoint, databusURI)
+            res = __handle_databus_file_query__(endpoint, databusURI, databus_key=databus_key)
             __download_list__(res, localDir, vault_token_file=token, databus_key=databus_key, auth_url=auth_url, client_id=client_id)