Infer filename from remote location URL

When no `file` or `value` is given for test inputs or outputs.

lib/galaxy/tool_util/parser/xml.py

@@ -1,6 +1,7 @@
 import json
 import logging
 import math
+import os
 import re
 import uuid
 from typing import (
@@ -697,7 +698,10 @@ def __parse_test_attributes(output_elem, attrib, parse_elements=False, parse_dis
     attributes["sort"] = string_as_bool(attrib.pop("sort", False))
     attributes["decompress"] = string_as_bool(attrib.pop("decompress", False))
     # `location` may contain an URL to a remote file that will be used to download `file` (if not already present on disk).
-    attributes["location"] = attrib.get("location")
+    location = attrib.get("location")
+    if location and file is None:
+        file = os.path.basename(location)  # If no file specified, try to get filename from URL last component
+    attributes["location"] = location
     try:
         attributes["count"] = int(attrib.pop("count"))
     except KeyError:

lib/galaxy/tool_util/verify/test_data.py

@@ -176,12 +176,11 @@ class RemoteLocationDataResolver(FileDataResolver):
             return
         if not is_url(location):
             raise ValueError(f"Invalid 'location' URL for remote test data provided: {location}")
-        if filename:
-            if self._is_direct_url_paste_upload(filename, location):
-                return  # No pre-download required
-            self._ensure_base_dir_exists()
-            dest_file_path = self.path(filename)
-            download_to_file(location, dest_file_path)
+        if not self._is_valid_filename(filename):
+            raise ValueError(f"Invalid 'filename' provided: '{filename}'")
+        self._ensure_base_dir_exists()
+        dest_file_path = self.path(filename)
+        download_to_file(location, dest_file_path)
 
     def _ensure_base_dir_exists(self):
         if not os.path.exists(self.file_dir):
@@ -201,6 +200,10 @@ class RemoteLocationDataResolver(FileDataResolver):
                 f"Failed to validate test data '{filename}' with [{hash_function}] - expected [{expected_hash_value}] got [{calculated_hash_value}]"
             )
 
-    def _is_direct_url_paste_upload(self, filename: str, location: Optional[str]):
-        """Checks if the test data file is an URL and will be directly url_pasted to Galaxy."""
-        return location and filename == location and is_url(location)
+    def _is_valid_filename(self, filename: str):
+        """
+        Checks that the filename does not contain the following
+        characters: <, >, :, ", /, \\, |, ?, *, or any control characters.
+        """
+        pattern = r"^[^<>:\"/\\|?*\x00-\x1F]+$"
+        return bool(re.match(pattern, filename))

lib/galaxy/tools/test.py

@@ -184,8 +184,8 @@ def _process_raw_inputs(
                 param_extra = raw_input_dict["attributes"]
                 location = param_extra.get("location")
                 if param_value is None and location:
-                    # TODO: I bet there is a better way of signaling that we are going to use the location and not the value...
-                    param_value = location
+                    # If no value is given, we try to get the file name directly from the URL
+                    param_value = os.path.basename(location)
                 if not value.type == "text":
                     param_value = _split_if_str(param_value)
                 if isinstance(value, galaxy.tools.parameters.basic.DataToolParameter):

test/functional/tools/remote_test_data_location.xml

@@ -10,7 +10,8 @@
     </outputs>
     <tests>
         <test>
-            <!-- When only the `location` is defined, it will be directly URL pasted to Galaxy -->
+            <!-- When only the `location` is defined, the name of the input file will be infered from the last component of the URL.
+                In this example, it will be equivalent to `value="hello.txt"`. -->
             <param name="input" location="https://raw.githubusercontent.com/galaxyproject/planemo/7be1bf5b3971a43eaa73f483125bfb8cabf1c440/tests/data/hello.txt"/>
             <output name="output">
                 <assert_contents>
@@ -53,5 +54,11 @@
             <param name="input" value="not_local_input.txt" location="https://raw.githubusercontent.com/galaxyproject/planemo/master/tests/data/not_hello.txt" checksum="sha1$3436387a8a45b00ef11e621e501ba23b52f06101"/>
             <output name="output" file="not_local_output.txt" location="https://raw.githubusercontent.com/galaxyproject/planemo/master/tests/data/not_hello.txt" checksum="sha1$3436387a8a45b00ef11e621e501ba23b52f06101"/>
         </test>
+        <test>
+            <!-- If the output `file` or `value` is not specified, the last component of the location URL will be used
+                as filename. In this example, this will be equivalent to `file="not_hello.txt"`. -->
+            <param name="input" value="not_local_input.txt" location="https://raw.githubusercontent.com/galaxyproject/planemo/master/tests/data/not_hello.txt" checksum="sha1$3436387a8a45b00ef11e621e501ba23b52f06101"/>
+            <output name="output" location="https://raw.githubusercontent.com/galaxyproject/planemo/master/tests/data/not_hello.txt" checksum="sha1$3436387a8a45b00ef11e621e501ba23b52f06101"/>
+        </test>
     </tests>
 </tool>