fix youtube in production and add more recursion limit

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.mp4 filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.xlsx filter=lfs diff=lfs merge=lfs -text

app.py

@@ -16,15 +16,6 @@ import re
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-import requests, logging, os
-TEST_URL = "https://www.youtube.com/watch?v=L1vXCYZAYYM"
-def test_youtube():
-    try:
-        r = requests.head(TEST_URL, timeout=8, allow_redirects=True)
-        logging.warning("YT status %s %s", r.status_code, r.url)
-    except Exception as e:
-        logging.error("YT check failed: %s", e)
-
 def run_and_submit_all( profile: gr.OAuthProfile | None):
     """
     Fetches all questions, runs the BasicAgent on them, submits all answers,
@@ -104,7 +95,12 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
                         for chunk in r.iter_content(chunk_size=8192):
                             if chunk:                             # evita keep-alives vacíos
                                 f.write(chunk)
-        
+
+            #use youtube from data/ directory as scrapping doesn't work in spaces
+            if os.exists(f"data/{task_id}.mp4"):
+                path_filename = f"data/{task_id}.mp4"
+
+            if path_filename is not None:
                 gcp_uri_file = upload_file_to_bucket(path_filename)
                 
             #submitted_answer = agent(question_text)
@@ -113,7 +109,7 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
                                "check_final_answer": False,
                                "path_filename": path_filename,
                                "gcp_path": gcp_uri_file},
-                              {"recursion_limit": 20})
+                              {"recursion_limit": 30})
             
             submitted_answer = result['final_answer']
             answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
@@ -207,7 +203,6 @@ with gr.Blocks() as demo:
     )
 
 if __name__ == "__main__":
-    test_youtube()
     print("\n" + "-"*30 + " App Starting " + "-"*30)
     # Check for SPACE_HOST and SPACE_ID at startup for information
     space_host_startup = os.getenv("SPACE_HOST")

data/1f975693-876d-457b-a649-393859e79bf3.mp3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:200f767e732b49efef5c05d128903ee4d2c34e66fdce7f5593ac123b2e637673
+size 280868

data/7bd855d8-463d-4ed5-93ca-5fe35145f733.xlsx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:196b64ea8f72841cb0c4fc972ca82a28e74de926c402bb043305e12e05e012b7
+size 5285

data/99c9cc74-fdc8-46c6-8f8d-3ce2d3bfeea3.mp3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b218c951c1f888f0bbe6f46c080f57afc7c9348fffc7ba4da35749ff1e2ac40f
+size 179304

data/9d191bce-651d-4746-be2d-7ef8ecadb9c2.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbd1ad3d2a2da43173c26d14bef4591b585fa1d541d6622c37d785e826d52a6b
+size 1247103

data/a1e91b78-d3d8-4675-bb8d-62741b4b68a6.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c52beb9924f94949983450a758f9fe8bd3e5c099c09eb2595934e02b31a4aae
+size 8141554

data/f918266a-b3e0-4914-865d-4faa564f1aef.py

@@ -0,0 +1,35 @@
+from random import randint
+import time
+
+class UhOh(Exception):
+    pass
+
+class Hmm:
+    def __init__(self):
+        self.value = randint(-100, 100)
+
+    def Yeah(self):
+        if self.value == 0:
+            return True
+        else:
+            raise UhOh()
+
+def Okay():
+    while True:
+        yield Hmm()
+
+def keep_trying(go, first_try=True):
+    maybe = next(go)
+    try:
+        if maybe.Yeah():
+            return maybe.value
+    except UhOh:
+        if first_try:
+            print("Working...")
+            print("Please wait patiently...")
+        time.sleep(0.1)
+        return keep_trying(go, first_try=False)
+
+if __name__ == "__main__":
+    go = Okay()
+    print(f"{keep_trying(go)}")

graph.py

@@ -79,7 +79,11 @@ def attach_data(state: TaskState) -> dict:
                 response = f"Path of the attached file: {path_filename}"
                 
         elif filename_type == 'youtube':
-            _, gcp_path = download_youtube_video(router_decision.youtube_url, "video")
+            if state.get('gcp_path'):
+                gcp_path = state["gcp_path"]
+            else:
+                _, gcp_path = download_youtube_video(router_decision.youtube_url, "video")
+                
             response = f"video GCP uri: {gcp_path}"
 
         elif filename_type == 'audio':