Update app.py

app.py

@@ -2,6 +2,10 @@ import gradio as gr
 from internetarchive import search_items, get_item
 import requests
 import time
+import subprocess
+import json
+import re
+from bs4 import BeautifulSoup
 
 # --- VirusTotal helper functions ---
 def scan_url_vt(url, api_key):
@@ -22,9 +26,35 @@ def scan_url_vt(url, api_key):
             stats = attr.get("stats", {})
             return stats.get("malicious", 0) == 0
 
+# --- FFprobe metadata extraction ---
+def extract_ffprobe_metadata(url_or_path):
+    cmd = [
+        "ffprobe", "-v", "error", "-print_format", "json",
+        "-show_format", "-show_streams",
+        url_or_path
+    ]
+    out = subprocess.check_output(cmd)
+    return json.loads(out)
+
+# --- Scrape basic page metadata (title + og: tags) ---
+def fetch_page_metadata(url):
+    try:
+        resp = requests.get(url, timeout=5)
+        resp.raise_for_status()
+        html = resp.text
+        soup = BeautifulSoup(html, "html.parser")
+        meta = {"url": url, "title": soup.title.string if soup.title else None}
+        # grab OpenGraph tags
+        for tag in soup.find_all("meta"):
+            prop = tag.get("property") or tag.get("name")
+            if prop and prop.startswith(("og:", "twitter:")):
+                meta[prop] = tag.get("content")
+        return meta
+    except Exception as e:
+        return {"url": url, "error": str(e)}
+
 # --- Core search & scan logic ---
 def fetch_clean_videos(keywords, api_key, scan_enabled):
-    # build IA query from comma-separated keywords
     query = " OR ".join([f"{kw.strip().replace(' ', '+')}" for kw in keywords.split(",")])
     ia_query = f"mediatype:(movies) AND ({query})"
     results = list(search_items(ia_query))[:50]
@@ -35,7 +65,7 @@ def fetch_clean_videos(keywords, api_key, scan_enabled):
         item = get_item(identifier)
         for f in item.files:
             fmt = f.get("format", "").lower()
-            if fmt.startswith(('mpeg','mp4','avi','mov','webm','m4v')):
+            if fmt.startswith(("mpeg","mp4","avi","mov","webm","m4v")):
                 url = f"https://archive.org/download/{identifier}/{f['name']}"
                 if scan_enabled and api_key:
                     try:
@@ -50,63 +80,79 @@ def fetch_clean_videos(keywords, api_key, scan_enabled):
 
 # --- Gradio UI setup ---
 with gr.Blocks() as demo:
-    gr.Markdown("# 📼 IA Drone‑Strike Video Browser with VT Scan + Raw Metadata")
+    gr.Markdown("# 📼 IA Drone‑Strike Explorer  \nEnable VT scan, FFprobe & Origin Tracing")
     with gr.Row():
-        kw_input     = gr.Textbox(label="Search keywords (comma-separated)", value="drone strike, military uav, kamikaze drone")
-        key_input    = gr.Textbox(label="VirusTotal API Key", type="password")
-    scan_toggle     = gr.Checkbox(label="Enable VirusTotal scan", value=True)
-    run_btn         = gr.Button("Search & Scan")
+        kw_input     = gr.Textbox(label="Search keywords", value="drone strike, military uav")
+        vt_key_input = gr.Textbox(label="VirusTotal API Key", type="password")
+    scan_toggle    = gr.Checkbox(label="Enable VT scan", value=True)
+    ffprobe_toggle = gr.Checkbox(label="Enable FFprobe metadata", value=False)
+    run_btn        = gr.Button("Search & Scan")
 
-    url_dropdown    = gr.Dropdown(label="Clean Video URLs", choices=[], interactive=True, allow_custom_value=True)
-    video_player    = gr.Video(label="Video Player")
-    metadata_json   = gr.JSON(label="► Raw IA Metadata")
+    url_dropdown   = gr.Dropdown(label="Clean Video URLs", choices=[], interactive=True)
+    video_player   = gr.Video(label="Video Player")
+    ia_meta_json   = gr.JSON(label="► Raw IA Metadata")
+    ffprobe_json   = gr.JSON(label="► FFprobe Metadata")
+    origins_json   = gr.JSON(label="► Source‑Origin Metadata")
 
     def search_and_populate(keywords, api_key, scan_enabled):
         urls = fetch_clean_videos(keywords, api_key, scan_enabled)
         return gr.update(choices=urls, value=urls[0] if urls else None)
 
-    def update_player_and_metadata(selected_url):
-        # extract IA identifier from URL: ".../download/{identifier}/{filename}"
+    def update_all(selected_url, ff_on, api_key):
+        # no selection guard
+        if not selected_url:
+            return None, {}, {}, []
+
+        # 1) IA metadata + file list
         parts = selected_url.split("/")
         identifier = parts[4] if len(parts) > 4 else None
-
-        raw = {
-            "identifier": identifier,
-            "metadata": {},
-            "files": []
-        }
-
+        raw_ia = {"identifier": identifier, "metadata": {}, "files": []}
         if identifier:
             try:
                 item = get_item(identifier)
-                # full descriptive metadata
-                raw["metadata"] = item.metadata
-                # per-file metadata (name, format, size, md5, + any extra tags)
-                raw["files"] = [
+                raw_ia["metadata"] = item.metadata
+                raw_ia["files"] = [
                     {
                         "name": f.get("name"),
                         "format": f.get("format"),
                         "size": f.get("size"),
                         "md5": f.get("md5"),
-                        **{k: v for k, v in f.items() if k not in ("name", "format", "size", "md5")}
+                        **{k: v for k,v in f.items() if k not in ("name","format","size","md5")}
                     }
                     for f in item.files
                 ]
             except Exception:
-                raw["error"] = "could not fetch IA metadata"
+                raw_ia["error"] = "could not fetch IA metadata"
+
+        # 2) FFprobe metadata if toggled
+        ff_md = {}
+        if ff_on:
+            try:
+                ff_md = extract_ffprobe_metadata(selected_url)
+            except Exception as e:
+                ff_md = {"error": str(e)}
+
+        # 3) Origin tracing: scrape each URL in description
+        origins = []
+        desc = raw_ia["metadata"].get("description", "")
+        urls_found = re.findall(r'https?://[^\s"<]+', desc)
+        for url in urls_found:
+            meta = fetch_page_metadata(url)
+            origins.append(meta)
+            # stop at first “real” origin (you can remove this break to collect all)
+            break
 
-        return gr.update(value=selected_url), raw
+        return selected_url, raw_ia, ff_md, origins
 
-    # wire up the interactions
     run_btn.click(
         fn=search_and_populate,
-        inputs=[kw_input, key_input, scan_toggle],
+        inputs=[kw_input, vt_key_input, scan_toggle],
         outputs=[url_dropdown]
     )
     url_dropdown.change(
-        fn=update_player_and_metadata,
-        inputs=[url_dropdown],
-        outputs=[video_player, metadata_json]
+        fn=update_all,
+        inputs=[url_dropdown, ffprobe_toggle, vt_key_input],
+        outputs=[video_player, ia_meta_json, ffprobe_json, origins_json]
     )
 
 if __name__ == "__main__":