Upload source/upload_data.py with huggingface_hub

source/upload_data.py

@@ -0,0 +1,113 @@
+import os, time, glob, datetime
+from huggingface_hub import HfApi
+
+TOKEN = "HF_TOKEN_REDACTED"
+REPO = "OpenTransformer/web-crawl-2026"
+STALE_SEC = 300  # 5 min no writes = done
+
+api = HfApi(token=TOKEN)
+
+def log(msg):
+    ts = datetime.datetime.utcnow().isoformat()
+    print(f"{ts} {msg}", flush=True)
+
+def find_ready_files():
+    ready = []
+    now = time.time()
+    for d in ["/workspace/staging", "/workspace/scraped_data_go", "/workspace/scraped_data_rust", "/workspace/scraped_data"]:
+        for f in glob.glob(os.path.join(d, "*.gz")):
+            age = now - os.path.getmtime(f)
+            sz = os.path.getsize(f)
+            if age > STALE_SEC and sz > 1024*1024:  # stale and >1MB
+                ready.append((f, sz, age))
+                log(f"  Ready: {f} ({sz/(1024*1024):.0f}MB, {age/3600:.1f}h old)")
+    return ready
+
+def upload_file(filepath, size):
+    ts = datetime.datetime.utcnow().strftime("%Y%m%d_%H%M%S")
+    basename = os.path.basename(filepath)
+    remote = f"crawl/combined/{basename.replace(chr(46)+chr(106),chr(95)+ts+chr(46)+chr(106))}"
+    log(f"Uploading {basename} ({size/(1024*1024):.0f}MB) -> {remote}")
+    try:
+        api.upload_file(
+            path_or_fileobj=filepath,
+            path_in_repo=remote,
+            repo_id=REPO,
+            repo_type="dataset",
+            commit_message=f"Crawl data: {basename} ({size/(1024*1024):.0f}MB)"
+        )
+        log(f"Uploaded! Removing {filepath}")
+        os.remove(filepath)
+        return True
+    except Exception as e:
+        log(f"Upload failed: {e}")
+        return False
+
+def combine_and_upload(files):
+    ts = datetime.datetime.utcnow().strftime("%Y%m%d_%H%M%S")
+    combined = f"/workspace/crawl_batch_{ts}.jsonl.gz"
+    total = sum(s for _, s, _ in files)
+    log(f"Combining {len(files)} files ({total/(1024*1024):.0f}MB)")
+    with open(combined, "wb") as out:
+        for f, _, _ in files:
+            with open(f, "rb") as inp:
+                while True:
+                    chunk = inp.read(8*1024*1024)
+                    if not chunk:
+                        break
+                    out.write(chunk)
+    remote = f"crawl/combined/crawl_batch_{ts}.jsonl.gz"
+    final = os.path.getsize(combined)
+    log(f"Uploading combined {final/(1024*1024):.0f}MB -> {remote}")
+    try:
+        api.upload_file(
+            path_or_fileobj=combined,
+            path_in_repo=remote,
+            repo_id=REPO,
+            repo_type="dataset",
+            commit_message=f"Crawl batch {ts} ({final/(1024*1024):.0f}MB, {len(files)} files)"
+        )
+        log(f"Uploaded! Cleaning up...")
+        for f, _, _ in files:
+            os.remove(f)
+        os.remove(combined)
+        return True
+    except Exception as e:
+        log(f"Upload failed: {e}")
+        if os.path.exists(combined):
+            os.remove(combined)
+        return False
+
+def main():
+    log("Upload daemon v2 starting")
+    while True:
+        log("Scanning...")
+        ready = find_ready_files()
+        if not ready:
+            log("No files ready, sleeping 30min")
+            time.sleep(1800)
+            continue
+        
+        # If any single file >= 100MB, upload individually
+        big = [(f, s, a) for f, s, a in ready if s >= 100*1024*1024]
+        small = [(f, s, a) for f, s, a in ready if s < 100*1024*1024]
+        
+        for f, s, a in big:
+            upload_file(f, s)
+        
+        # Combine small files if total >= 100MB, or if any are >12h old (avoid data loss)
+        if small:
+            total_small = sum(s for _, s, _ in small)
+            max_age = max(a for _, _, a in small)
+            if total_small >= 100*1024*1024 or max_age > 43200:  # 12 hours
+                if len(small) == 1:
+                    upload_file(small[0][0], small[0][1])
+                else:
+                    combine_and_upload(small)
+            else:
+                log(f"Small files total {total_small/(1024*1024):.0f}MB, newest {max_age/3600:.1f}h old, waiting")
+        
+        time.sleep(1800)
+
+if __name__ == "__main__":
+    main()