email to dataset

Dockerfile

@@ -6,7 +6,7 @@ COPY app /app
 # App deps only (no LaTeX)
 RUN python -m pip install --upgrade pip && \
     python -m pip install --no-cache-dir \
-      fastapi "uvicorn[standard]" pydantic python-dotenv google-genai
+      fastapi "uvicorn[standard]" pydantic python-dotenv google-genai huggingface-hub
 
 ENV PORT=7860
 EXPOSE 7860

app/main.py

@@ -1,4 +1,5 @@
-import os, re, uuid, subprocess, sys, time, traceback, threading
+import json, os, re, uuid, subprocess, sys, time, traceback, threading
+from io import BytesIO
 from collections import deque
 from pathlib import Path
 from typing import Optional, Tuple
@@ -7,6 +8,8 @@ from fastapi import FastAPI, HTTPException, Response
 from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel, validator
 
+from huggingface_hub import HfApi, create_repo, CommitOperationAdd
+
 # Optional .env for local testing
 from dotenv import load_dotenv
 load_dotenv()
@@ -34,8 +37,23 @@ app.add_middleware(
 
 
 RUNS = Path("runs"); RUNS.mkdir(parents=True, exist_ok=True)
-DATA_DIR = Path("data"); DATA_DIR.mkdir(parents=True, exist_ok=True)
-EMAILS_FILE = DATA_DIR / "emails.txt"
+
+HF_DATASET_ID = os.getenv("HF_DATASET_ID", "MathFrames/email-log")
+HF_TOKEN = os.getenv("HF_TOKEN", "")
+
+hf_api = HfApi(token=HF_TOKEN) if HF_TOKEN else None
+if hf_api:
+    try:
+        create_repo(
+            HF_DATASET_ID,
+            repo_type="dataset",
+            private=True,
+            exist_ok=True,
+            token=HF_TOKEN,
+        )
+    except Exception:
+        # Ignore startup race/permission errors; individual writes will surface issues.
+        pass
 
 # ---------------- simple 10 RPM rate limiter ----------------
 class RateLimiter:
@@ -394,7 +412,7 @@ class EmailIn(BaseModel):
 
     @validator("email")
     def validate_email(cls, value: str) -> str:
-        cleaned = value.strip()
+        cleaned = value.strip().lower()
         if not cleaned:
             raise ValueError("Email cannot be empty")
         if not re.match(r"^[^@\s]+@[^@\s]+\.[^@\s]+$", cleaned):
@@ -429,11 +447,29 @@ def generate_and_render(inp: PromptIn):
 
 @app.post("/store-email")
 def store_email(email: EmailIn):
-    """Append the provided email address to a server-side text file."""
+    """Store the provided email address in the configured Hugging Face dataset."""
+    if not hf_api or not HF_TOKEN:
+        raise HTTPException(500, "Email logging is not configured")
+
     sanitized_email = email.sanitized
+    timestamp = int(time.time())
+    key = f"emails/{int(time.time() * 1000)}-{uuid.uuid4().hex}.json"
+    payload = {"email": sanitized_email, "ts": timestamp}
+
     try:
-        with EMAILS_FILE.open("a", encoding="utf-8") as fh:
-            fh.write(f"{sanitized_email}\n")
-    except Exception:
+        hf_api.create_commit(
+            repo_id=HF_DATASET_ID,
+            repo_type="dataset",
+            operations=[
+                CommitOperationAdd(
+                    path_in_repo=key,
+                    path_or_fileobj=BytesIO(json.dumps(payload).encode("utf-8")),
+                )
+            ],
+            commit_message=f"Log email: {sanitized_email}",
+            token=HF_TOKEN,
+        )
+    except Exception as exc:
+        print("Failed to log email to Hugging Face:", exc, file=sys.stderr)
         raise HTTPException(500, "Failed to save email address")
-    return {"stored": True}
+    return {"stored": True, "path": key}