fix: mark_seen searches UNSEEN only; patterns run over every text before the digit fallback

T2-12: retrieve_otp re-served an already-consumed code inside max_age, and
fell back to an ALL search on every empty poll. with mark_seen=True the
query now includes UNSEEN; the ALL fallback is gone. mark_seen=False is
unchanged.

T2-13: the default catch-all \b(\d{4,8})\b pattern ran on the subject
before the body (picking a year from a date) and bypassed lengths=. drop
it; run the patterns over subject + bodies, then the lengths fallback over
the same texts.

Signed-off-by: disqualifier <dev@disqualifier.me>

src/aiomail/extract.py

@@ -15,12 +15,13 @@ log = logging.getLogger(__name__)
 MatchSpec = Union[str, Pattern, Callable[[Optional[str]], bool], None]
 
 # patterns tried (in order) against the subject then each body part; the first
-# capturing group is the code, or the whole match if there are no groups
+# capturing group is the code, or the whole match if there are no groups. no bare
+# digit catch-all here - that is the `lengths` fallback, which runs only after every
+# text missed every pattern
 DEFAULT_PATTERNS: Sequence[Union[str, Pattern]] = (
     r"(\d{4,8})\s+is\s+your",
     r"your.{0,20}?code(?:\s+is)?[:\s]+(\d{4,8})",
     r"code(?:\s+is)?[:\s]+(\d{4,8})",
-    r"\b(\d{4,8})\b",
 )
 
 # default digit-run lengths to treat as a standalone code when no pattern hits
@@ -79,15 +80,12 @@ def _iter_text(message: email.message.Message) -> Iterator[str]:
         yield text
 
 
-def _scan(text: str, patterns: list[Pattern], lengths: set[int]) -> Optional[str]:
-    """run patterns then a standalone digit-run fallback over one block of text"""
+def _match(text: str, patterns: list[Pattern]) -> Optional[str]:
+    """first pattern hit in one block of text"""
     for pat in patterns:
         m = pat.search(text)
         if m:
             return m.group(1) if m.groups() else m.group(0)
-    for run in re.findall(r"\d+", text):
-        if len(run) in lengths:
-            return run
     return None
 
 
@@ -100,20 +98,21 @@ def extract_code(
     """extract a one-time code from a message, subject first then body parts
 
     `patterns` are regexes tried in order (first capturing group wins, else the whole
-    match); if none hit, a standalone digit run whose length is in `lengths` is returned.
+    match) over the subject and every body part; only if none hit anywhere is a
+    standalone digit run whose length is in `lengths` returned, from the same texts.
     """
     compiled = _compile(patterns)
     length_set = set(lengths)
+    texts = [decode_header_value(message.get("Subject", "") or ""), *_iter_text(message)]
 
-    subject = decode_header_value(message.get("Subject", "") or "")
-    hit = _scan(subject, compiled, length_set)
-    if hit:
-        return hit
-
-    for text in _iter_text(message):
-        hit = _scan(text, compiled, length_set)
+    for text in texts:
+        hit = _match(text, compiled)
         if hit:
             return hit
+    for text in texts:
+        for run in re.findall(r"\d+", text):
+            if len(run) in length_set:
+                return run
     return None
 
 

src/aiomail/retrieve.py

@@ -22,9 +22,12 @@ log = logging.getLogger(__name__)
 DEFAULT_FOLDERS: Sequence[str] = ("INBOX", "Junk", "Spam", "Archive", "All Mail")
 
 
-def _server_query(sender: MatchSpec, subject: MatchSpec, match_field: str = "from") -> str:
-    """build a narrowing IMAP query from plain-string specs, falling back to ALL for regex/callable specs"""
-    parts: List[str] = []
+def _server_query(sender: MatchSpec, subject: MatchSpec, match_field: str = "from", unseen: bool = False) -> str:
+    """build a narrowing IMAP query from plain-string specs (ALL when nothing narrows)
+
+    `unseen` adds UNSEEN so a code already consumed (marked seen) is never served again
+    """
+    parts: List[str] = ["UNSEEN"] if unseen else []
     if isinstance(sender, str):
         if match_field == "to":
             parts.append(f'OR TO "{sender}" FROM "{sender}"')
@@ -73,20 +76,20 @@ async def retrieve_otp(
     sender/subject accept a substring, a compiled regex, or a callable.
     `match_field="to"` matches the recipient address and additionally accepts a
     forwarded match on From. set `max_age=None` to disable the freshness check.
+    with `mark_seen=True` (default) only UNSEEN messages are searched, so a code is
+    returned once; `mark_seen=False` searches seen messages too.
     """
     folders = list(folders) if folders is not None else list(DEFAULT_FOLDERS)
     lengths = list(lengths)
     sender_ok = as_predicate(sender)
     subject_ok = as_predicate(subject)
-    query = _server_query(sender, subject, match_field)
+    query = _server_query(sender, subject, match_field, unseen=mark_seen)
 
     for attempt in range(retries + 1):
         for folder in folders:
             if not await client.select(folder):
                 continue
             ids = await client.search(query)
-            if not ids and query != "ALL":
-                ids = await client.search("ALL")
             if not ids:
                 continue