Improved email scripts for Codex to use

scripts/email_fetch.py

@@ -2,6 +2,7 @@
 import email
 import imaplib
 import os
+import re
 from email.header import decode_header
 from pathlib import Path
 
@@ -38,6 +39,46 @@ def _write_index(index_path, values):
         f.write(line + "\n")
 
 
+def _decoded_text_parts(msg):
+    parts = []
+    if msg.is_multipart():
+        for part in msg.walk():
+            if part.get_content_type() != "text/plain":
+                continue
+            payload = part.get_payload(decode=True)
+            if payload is None:
+                continue
+            charset = part.get_content_charset() or "utf-8"
+            parts.append(payload.decode(charset, errors="replace"))
+    else:
+        payload = msg.get_payload(decode=True)
+        if payload is not None:
+            charset = msg.get_content_charset() or "utf-8"
+            parts.append(payload.decode(charset, errors="replace"))
+    return "\n".join(parts)
+
+
+def _extract_original_headers(text):
+    if not text:
+        return "", "", ""
+    # Find forwarded blocks and keep the deepest match.
+    pattern = re.compile(
+        r"Forwarded message.*?From:\s*(?P<from>.+?)\r?\n"
+        r"(?:To:\s*.+?\r?\n)?"
+        r"Date:\s*(?P<date>.+?)\r?\n"
+        r"Subject:\s*(?P<subject>.+?)\r?\n",
+        re.IGNORECASE | re.DOTALL,
+    )
+    matches = list(pattern.finditer(text))
+    if not matches:
+        return "", "", ""
+    last = matches[-1]
+    original_from = " ".join(last.group("from").split())
+    original_date = " ".join(last.group("date").split())
+    original_subject = " ".join(last.group("subject").split())
+    return original_from, original_subject, original_date
+
+
 def fetch():
     if not IMAP_PASS:
         raise SystemExit("Set ELPHEL_IMAP_PASS")
@@ -59,6 +100,8 @@ def fetch():
             from_ = _decode(msg.get("From"))
             to_ = _decode(msg.get("To"))
             cc_ = _decode(msg.get("Cc"))
+            text = _decoded_text_parts(msg)
+            original_from, original_subject, original_date = _extract_original_headers(text)
             date_dir = OUT_DIR / (date[:10].replace(" ", "_") if date else "unknown-date")
             date_dir.mkdir(parents=True, exist_ok=True)
             seq = f"{idx:04d}"
@@ -67,10 +110,20 @@ def fetch():
             out_path.write_bytes(raw)
             meta_path = date_dir / f"{seq}_meta.txt"
             meta_path.write_text(
-                f"From: {from_}\nTo: {to_}\nCc: {cc_}\nDate: {date}\nSubject: {subject}\n",
+                f"From: {from_}\n"
+                f"To: {to_}\n"
+                f"Cc: {cc_}\n"
+                f"Date: {date}\n"
+                f"Subject: {subject}\n"
+                f"Original-From: {original_from}\n"
+                f"Original-Date: {original_date}\n"
+                f"Original-Subject: {original_subject}\n",
                 encoding="utf-8",
             )
-            _write_index(OUT_DIR / "index.csv", [date, from_, to_, cc_, subject, str(out_path)])
+            _write_index(
+                OUT_DIR / "index.csv",
+                [date, from_, to_, cc_, subject, original_from, original_date, original_subject, str(out_path)],
+            )
 
 
 if __name__ == "__main__":