diff --git a/routes/email/email_routes.py b/routes/email/email_routes.py
index 0af2d0a11..541a7189f 100644
--- a/routes/email/email_routes.py
+++ b/routes/email/email_routes.py
@@ -25,6 +25,7 @@ import json
import re
import html
import io
+import tempfile
import zipfile
from urllib.parse import parse_qs, quote, unquote, urlparse
from html.parser import HTMLParser as _HTMLParser
@@ -642,6 +643,11 @@ def _uid_from_fetch_meta(meta_b: bytes) -> str:
return m.group(1).decode() if m else ""
+def _size_from_fetch_meta(meta_b: bytes) -> int | None:
+ m = re.search(rb"\bRFC822\.SIZE\s+(\d+)\b", meta_b, re.IGNORECASE)
+ return int(m.group(1)) if m else None
+
+
def _parse_list_unsubscribe_header(value: str | None) -> list[dict]:
"""Parse RFC List-Unsubscribe entries into safe reviewable actions.
@@ -4212,15 +4218,16 @@ def setup_email_routes():
date_span = selected_dates[0] if len(selected_dates) == 1 else f"{selected_dates[0]} - {selected_dates[-1]}"
export_folder = f"emails-{category}" + (f" ({date_span})" if date_span else "")
- zip_buf = io.BytesIO()
+ # 512 MiB safety cap prevents unbounded memory growth when exporting
+ # many large messages.
+ max_total_bytes = 512 * 1024 * 1024
+ max_message_bytes = 64 * 1024 * 1024
+ spool_threshold_bytes = 32 * 1024 * 1024
used_names: set[str] = set()
exported = 0
skipped = 0
total_bytes = 0
- # 512 MiB safety cap prevents unbounded memory growth when exporting
- # many large messages.
- max_total_bytes = 512 * 1024 * 1024
- single_export: tuple[str, bytes] | None = None
+ validated_accounts: set[str] = set()
def _email_filename(row: dict, uid: str) -> str:
date = str(row.get("date") or "").strip()
@@ -4281,120 +4288,249 @@ def setup_email_routes():
logger.debug("Skipping malformed fixture email attachment uid=%s", uid, exc_info=True)
return msg.as_bytes(policy=email_policy.SMTP)
- seen_items: set[tuple[str | None, str, str]] = set()
- remote_groups: dict[tuple[str | None, str], dict[str, tuple[str, dict]]] = {}
- with zipfile.ZipFile(zip_buf, "w", compression=zipfile.ZIP_DEFLATED) as zf:
- for row in messages:
- if not isinstance(row, dict):
- skipped += 1
- continue
- uid = str(row.get("uid") or "").strip()
- if not uid:
- skipped += 1
- continue
- folder = str(row.get("folder") or "INBOX").strip() or "INBOX"
- account_id = str(row.get("account_id") or row.get("account") or "").strip() or None
- if account_id:
- _assert_owns_account(account_id, owner)
+ # Optimization: single message exports return raw .eml directly without ZIP compression.
+ if len(messages) == 1:
+ row = messages[0]
+ if not isinstance(row, dict):
+ raise HTTPException(status_code=404, detail="No selected emails could be exported")
+ uid = str(row.get("uid") or "").strip()
+ if not uid:
+ raise HTTPException(status_code=404, detail="No selected emails could be exported")
+ folder = str(row.get("folder") or "INBOX").strip() or "INBOX"
+ account_id = str(row.get("account_id") or row.get("account") or "").strip() or None
+ if account_id:
+ _assert_owns_account(account_id, owner)
- item_key = (account_id, folder, uid)
- if item_key in seen_items:
- skipped += 1
- continue
- seen_items.add(item_key)
+ filename = _email_filename(row, uid)
+ message_bytes: bytes | None = None
- fixture = _fixture_email_read(uid, folder, owner)
- if fixture is not None and not fixture.get("error"):
- filename = _email_filename(row, uid)
- unique_name = filename
- suffix = 2
- while unique_name.lower() in used_names:
- unique_name = f"{Path(filename).stem}-{suffix}.eml"
- suffix += 1
- used_names.add(unique_name.lower())
- message_bytes = _fixture_message_bytes(fixture, uid, folder)
- total_bytes += len(message_bytes)
- if total_bytes > max_total_bytes:
- raise HTTPException(status_code=413, detail="Export too large")
- zf.writestr(f"{export_folder}/{unique_name}", message_bytes)
- single_export = (unique_name, message_bytes)
- exported += 1
- continue
- if not re.fullmatch(r"\d+", uid):
- skipped += 1
- continue
- group = remote_groups.setdefault((account_id, folder), {})
- group.setdefault(uid, (_email_filename(row, uid), row))
-
- # Fetch original RFC822 data in batches and store it byte-for-byte.
- for (account_id, folder), rows_by_uid in remote_groups.items():
- rows = list(rows_by_uid.items())
- remaining_rows = len(rows)
+ fixture = _fixture_email_read(uid, folder, owner)
+ if fixture is not None and not fixture.get("error"):
+ message_bytes = _fixture_message_bytes(fixture, uid, folder)
+ elif re.fullmatch(r"\d+", uid):
try:
with _imap(account_id, owner=owner) as conn:
conn.select(_q(folder), readonly=True)
- for batch_start in range(0, len(rows), 20):
- batch = rows[batch_start:batch_start + 20]
- wanted = {uid: filename for uid, (filename, _row) in batch}
- uid_set = ",".join(wanted).encode()
- status, msg_data = _imap_uid_fetch(conn, uid_set, "(UID RFC822)")
- if status != "OK" or not msg_data:
- skipped += len(batch)
- remaining_rows -= len(batch)
- continue
- found: set[str] = set()
- for meta, raw_message in _group_uid_fetch_records(msg_data):
- uid = _uid_from_fetch_meta(meta)
- if uid not in wanted or not raw_message:
- continue
- filename = wanted[uid]
- unique_name = filename
- suffix = 2
- while unique_name.lower() in used_names:
- unique_name = f"{Path(filename).stem}-{suffix}.eml"
- suffix += 1
- used_names.add(unique_name.lower())
- message_bytes = bytes(raw_message)
- total_bytes += len(message_bytes)
- if total_bytes > max_total_bytes:
- raise HTTPException(status_code=413, detail="Export too large")
- zf.writestr(f"{export_folder}/{unique_name}", message_bytes)
- single_export = (unique_name, message_bytes)
- found.add(uid)
- exported += 1
- skipped += len(batch) - len(found)
- remaining_rows -= len(batch)
+ uid_b = uid.encode()
+ try:
+ st_sz, sz_data = _imap_uid_fetch(conn, uid_b, "(UID RFC822.SIZE)")
+ if st_sz == "OK" and sz_data:
+ for item in sz_data:
+ meta = item[0] if isinstance(item, tuple) else item
+ if isinstance(meta, bytes):
+ sz = _size_from_fetch_meta(meta)
+ if sz is not None and (sz > max_message_bytes or sz > max_total_bytes):
+ raise HTTPException(status_code=413, detail="Export too large")
+ except HTTPException:
+ raise
+ except Exception as sz_err:
+ logger.debug("Single message size preflight error: %s", sz_err)
+
+ st, msg_data = _imap_uid_fetch(conn, uid_b, "(UID RFC822)")
+ if st == "OK" and msg_data:
+ for meta, raw in _group_uid_fetch_records(msg_data):
+ if _uid_from_fetch_meta(meta) == uid and raw:
+ message_bytes = bytes(raw)
+ break
except HTTPException:
raise
except Exception as exc:
- skipped += remaining_rows
- logger.warning("Bulk email export skipped mailbox folder=%s account=%s count=%s: %s", folder, account_id, remaining_rows, exc)
+ logger.warning("Single email export failed for uid=%s folder=%s: %s", uid, folder, exc)
+
+ if not message_bytes:
+ raise HTTPException(status_code=404, detail="No selected emails could be exported")
+
+ if len(message_bytes) > max_message_bytes or len(message_bytes) > max_total_bytes:
+ raise HTTPException(status_code=413, detail="Export too large")
- if exported <= 0:
- raise HTTPException(status_code=404, detail="No selected emails could be exported")
- if len(messages) == 1 and single_export:
- filename, message_bytes = single_export
return StreamingResponse(
io.BytesIO(message_bytes),
media_type="message/rfc822",
headers={
"Content-Disposition": f"attachment; filename=\"email.eml\"; filename*=UTF-8''{quote(filename, safe='')}",
"X-Exported-Emails": "1",
+ "X-Skipped-Emails": "0",
+ },
+ )
+
+ seen_items: set[tuple[str | None, str, str]] = set()
+ remote_groups: dict[tuple[str | None, str], dict[str, tuple[str, dict]]] = {}
+ spooled_file = tempfile.SpooledTemporaryFile(max_size=spool_threshold_bytes, mode="w+b")
+ response_started = False
+ try:
+ with zipfile.ZipFile(spooled_file, "w", compression=zipfile.ZIP_DEFLATED) as zf:
+ for row in messages:
+ if not isinstance(row, dict):
+ skipped += 1
+ continue
+ uid = str(row.get("uid") or "").strip()
+ if not uid:
+ skipped += 1
+ continue
+ folder = str(row.get("folder") or "INBOX").strip() or "INBOX"
+ account_id = str(row.get("account_id") or row.get("account") or "").strip() or None
+ if account_id and account_id not in validated_accounts:
+ _assert_owns_account(account_id, owner)
+ validated_accounts.add(account_id)
+
+ item_key = (account_id, folder, uid)
+ if item_key in seen_items:
+ skipped += 1
+ continue
+ seen_items.add(item_key)
+
+ fixture = _fixture_email_read(uid, folder, owner)
+ if fixture is not None and not fixture.get("error"):
+ filename = _email_filename(row, uid)
+ unique_name = filename
+ suffix = 2
+ while unique_name.lower() in used_names:
+ unique_name = f"{Path(filename).stem}-{suffix}.eml"
+ suffix += 1
+ used_names.add(unique_name.lower())
+ message_bytes = _fixture_message_bytes(fixture, uid, folder)
+ if len(message_bytes) > max_message_bytes:
+ raise HTTPException(status_code=413, detail="Export too large")
+ total_bytes += len(message_bytes)
+ if total_bytes > max_total_bytes:
+ raise HTTPException(status_code=413, detail="Export too large")
+ try:
+ zf.writestr(f"{export_folder}/{unique_name}", message_bytes)
+ except Exception as zip_err:
+ logger.error("ZIP entry write failed for fixture: %s", zip_err)
+ raise HTTPException(status_code=500, detail="Failed to create export archive")
+ exported += 1
+ continue
+ if not re.fullmatch(r"\d+", uid):
+ skipped += 1
+ continue
+ group = remote_groups.setdefault((account_id, folder), {})
+ group.setdefault(uid, (_email_filename(row, uid), row))
+
+ # Fetch original RFC822 data in batches and store it byte-for-byte.
+ for (account_id, folder), rows_by_uid in remote_groups.items():
+ rows = list(rows_by_uid.items())
+ remaining_rows = len(rows)
+ try:
+ with _imap(account_id, owner=owner) as conn:
+ conn.select(_q(folder), readonly=True)
+
+ # Size preflight query to catch oversized exports early and bound batch sizes.
+ uid_set_all = ",".join(uid for uid, _ in rows).encode()
+ size_by_uid: dict[str, int] = {}
+ try:
+ st_sz, sz_data = _imap_uid_fetch(conn, uid_set_all, "(UID RFC822.SIZE)")
+ if st_sz == "OK" and sz_data:
+ for item in sz_data:
+ meta = item[0] if isinstance(item, tuple) else item
+ if isinstance(meta, bytes):
+ u = _uid_from_fetch_meta(meta)
+ s = _size_from_fetch_meta(meta)
+ if u and s is not None:
+ size_by_uid[u] = s
+ except Exception as sz_err:
+ logger.debug("IMAP size preflight query failed or unsupported: %s", sz_err)
+
+ if size_by_uid:
+ for u, s in size_by_uid.items():
+ if s > max_message_bytes:
+ raise HTTPException(status_code=413, detail="Export too large")
+ if total_bytes + sum(size_by_uid.values()) > max_total_bytes:
+ raise HTTPException(status_code=413, detail="Export too large")
+
+ # Form batches with at most 20 messages and capped at 10 MiB payload.
+ batches: list[list[tuple[str, tuple[str, dict]]]] = []
+ cur_batch: list[tuple[str, tuple[str, dict]]] = []
+ cur_batch_bytes = 0
+ max_batch_bytes = 10 * 1024 * 1024
+ for item in rows:
+ uid_val = item[0]
+ sz = size_by_uid.get(uid_val, 0)
+ if cur_batch and (
+ len(cur_batch) >= 20 or
+ (sz > 0 and cur_batch_bytes + sz > max_batch_bytes)
+ ):
+ batches.append(cur_batch)
+ cur_batch = []
+ cur_batch_bytes = 0
+ cur_batch.append(item)
+ cur_batch_bytes += sz
+ if cur_batch:
+ batches.append(cur_batch)
+
+ for batch in batches:
+ wanted = {uid: filename for uid, (filename, _row) in batch}
+ uid_set = ",".join(wanted).encode()
+ status, msg_data = _imap_uid_fetch(conn, uid_set, "(UID RFC822)")
+ if status != "OK" or not msg_data:
+ skipped += len(batch)
+ remaining_rows -= len(batch)
+ continue
+ found: set[str] = set()
+ for meta, raw_message in _group_uid_fetch_records(msg_data):
+ uid = _uid_from_fetch_meta(meta)
+ if uid not in wanted or not raw_message:
+ continue
+ filename = wanted[uid]
+ unique_name = filename
+ suffix = 2
+ while unique_name.lower() in used_names:
+ unique_name = f"{Path(filename).stem}-{suffix}.eml"
+ suffix += 1
+ used_names.add(unique_name.lower())
+ message_bytes = bytes(raw_message)
+ if len(message_bytes) > max_message_bytes:
+ raise HTTPException(status_code=413, detail="Export too large")
+ total_bytes += len(message_bytes)
+ if total_bytes > max_total_bytes:
+ raise HTTPException(status_code=413, detail="Export too large")
+ try:
+ zf.writestr(f"{export_folder}/{unique_name}", message_bytes)
+ except Exception as zip_err:
+ logger.error("ZIP entry write failed for remote message: %s", zip_err)
+ raise HTTPException(status_code=500, detail="Failed to create export archive")
+ found.add(uid)
+ exported += 1
+ remaining_rows -= 1
+ missing = len(batch) - len(found)
+ skipped += len(batch) - len(found)
+ remaining_rows -= missing
+ except HTTPException:
+ raise
+ except Exception as exc:
+ skipped += remaining_rows
+ logger.warning("Bulk email export skipped mailbox folder=%s account=%s count=%s: %s", folder, account_id, remaining_rows, exc)
+
+ if exported <= 0:
+ raise HTTPException(status_code=404, detail="No selected emails could be exported")
+
+ spooled_file.seek(0)
+ zip_name = _safe_attachment_zip_name(f"{export_folder}.zip", "selected-emails.zip")
+ logger.info("Bulk email export complete owner=%s selected=%s exported=%s skipped=%s", owner, len(messages), exported, skipped)
+
+ def _iter_spooled(f, chunk_size=64 * 1024):
+ try:
+ while True:
+ chunk = f.read(chunk_size)
+ if not chunk:
+ break
+ yield chunk
+ finally:
+ f.close()
+
+ response_started = True
+ return StreamingResponse(
+ _iter_spooled(spooled_file),
+ media_type="application/zip",
+ headers={
+ "Content-Disposition": f'attachment; filename="{zip_name}"',
+ "X-Exported-Emails": str(exported),
"X-Skipped-Emails": str(skipped),
},
)
- zip_buf.seek(0)
- zip_name = _safe_attachment_zip_name(f"{export_folder}.zip", "selected-emails.zip")
- logger.info("Bulk email export complete owner=%s selected=%s exported=%s skipped=%s", owner, len(messages), exported, skipped)
- return StreamingResponse(
- zip_buf,
- media_type="application/zip",
- headers={
- "Content-Disposition": f'attachment; filename="{zip_name}"',
- "X-Exported-Emails": str(exported),
- "X-Skipped-Emails": str(skipped),
- },
- )
+ finally:
+ if not response_started:
+ spooled_file.close()
@router.get("/inline-image/{uid}")
async def inline_image(
diff --git a/static/js/emailLibrary/index.js b/static/js/emailLibrary/index.js
index 543fdfd55..cc08f2d0c 100644
--- a/static/js/emailLibrary/index.js
+++ b/static/js/emailLibrary/index.js
@@ -1132,32 +1132,70 @@ async function _printSelectedEmails(messages) {
while (nextIndex < messages.length) {
const index = nextIndex++;
const message = messages[index];
- const account = message.account_id ? `&account_id=${encodeURIComponent(message.account_id)}` : '';
- const url = `${API_BASE}/api/email/read/${encodeURIComponent(message.uid)}?folder=${encodeURIComponent(message.folder || 'INBOX')}${account}&mark_seen=false&full=true`;
- const response = await fetch(url, { credentials: 'same-origin' });
- if (!response.ok) throw new Error(`Could not load "${message.subject || 'email'}" (HTTP ${response.status})`);
- const data = await response.json();
- if (data?.error) throw new Error(data.error);
- loaded[index] = { message, data };
+ try {
+ const account = message.account_id ? `&account_id=${encodeURIComponent(message.account_id)}` : '';
+ const url = `${API_BASE}/api/email/read/${encodeURIComponent(message.uid)}?folder=${encodeURIComponent(message.folder || 'INBOX')}${account}&mark_seen=false&full=true`;
+ const response = await fetch(url, { credentials: 'same-origin' });
+ if (!response.ok) throw new Error(`HTTP ${response.status}`);
+ const data = await response.json();
+ if (data?.error) throw new Error(data.error);
+ loaded[index] = { ok: true, message, data };
+ } catch (err) {
+ loaded[index] = { ok: false, message, error: err?.message || 'Failed to load' };
+ }
}
};
await Promise.all(Array.from({ length: Math.min(4, messages.length) }, readWorker));
+ const successful = loaded.filter(item => item && item.ok);
+ const omitted = loaded.filter(item => !item || !item.ok);
+
+ if (successful.length === 0) {
+ throw new Error('None of the selected emails could be loaded for printing');
+ }
+
// Build sanitized HTML for the print frame -- scripts and event handlers
// cannot execute inside the sandboxed iframe.
const articles = [];
- for (const { message, data } of loaded) {
+ if (omitted.length > 0) {
+ const count = omitted.length;
+ articles.push(
+ `Note: ${count} selected email${count === 1 ? ' was' : 's were'} unavailable and omitted from this print.
`
+ );
+ }
+
+ for (const { message, data } of successful) {
const subject = String(data.subject || message.subject || '(no subject)');
const from = data.from_name && data.from_address
? `${data.from_name} <${data.from_address}>`
: data.from_name || data.from_address;
const safeBody = new DOMParser().parseFromString(_safeRenderEmailBody(data), 'text/html');
- safeBody.querySelectorAll('img').forEach(img => {
- const placeholder = safeBody.createElement('span');
- placeholder.textContent = img.alt ? `[Image: ${img.alt}]` : '[Inline image omitted]';
- img.replaceWith(placeholder);
+
+ // Replace interactive image placeholder components with clean print-safe notices
+ safeBody.querySelectorAll('.email-inline-image-placeholder').forEach(ph => {
+ const title = ph.querySelector('.email-inline-image-title')?.textContent?.trim();
+ const ariaLabel = ph.getAttribute('aria-label')?.trim();
+ const label = title || ariaLabel || '';
+ const isGeneric = !label || /^inline image$|^remote image$|^image unavailable$|^image$/i.test(label);
+ const span = safeBody.createElement('span');
+ span.className = 'email-export-image-notice';
+ span.textContent = isGeneric ? '[Image omitted]' : `[Image: ${label}]`;
+ ph.replaceWith(span);
});
+
+ // Replace any remaining raw ![]()
tags
+ safeBody.querySelectorAll('img').forEach(img => {
+ const alt = (img.getAttribute('alt') || img.getAttribute('title') || '').trim();
+ const span = safeBody.createElement('span');
+ span.className = 'email-export-image-notice';
+ span.textContent = alt ? `[Image: ${alt}]` : '[Image omitted]';
+ img.replaceWith(span);
+ });
+
+ // Strip interactive controls (buttons, inputs, actions) from the print document
+ safeBody.querySelectorAll('button, input, select, textarea, .email-inline-image-actions, .email-inline-image-btn').forEach(el => el.remove());
+
safeBody.querySelectorAll('details').forEach(detail => { detail.open = true; });
const bodyHtml = safeBody.body.innerHTML;
@@ -1192,11 +1230,14 @@ async function _printSelectedEmails(messages) {
'.email-export-pdf-body a{color:#135f9b;text-decoration:underline}',
'.email-export-pdf-body blockquote{margin-left:12px;padding-left:10px;border-left:2px solid #ccc;color:#555}',
'.email-export-pdf-body pre{white-space:pre-wrap}',
+ '.email-inline-image-placeholder,.email-inline-image-actions,.email-inline-image-btn{display:none !important}',
+ '.email-export-image-notice{display:inline-block;color:#666;font-style:italic;font-size:11px;margin:4px 0}',
+ '.email-export-pdf-omitted-notice{margin:0 0 16px;padding:8px 12px;background:#f8f9fa;border:1px solid #e2e8f0;border-radius:4px;font-size:11px;color:#64748b}',
'@page{margin:10mm}',
].join('\n');
- const safeBaseName = messages.length === 1
- ? String(loaded[0]?.data?.subject || messages[0]?.subject || 'email').replace(/[^\p{L}\p{N} ._-]+/gu, '_').trim().slice(0, 100) || 'email'
+ const safeBaseName = successful.length === 1
+ ? String(successful[0]?.data?.subject || successful[0]?.message?.subject || 'email').replace(/[^\p{L}\p{N} ._-]+/gu, '_').trim().slice(0, 100) || 'email'
: 'selected-emails';
// Use a sandboxed same-origin iframe following the established document.js
@@ -1209,6 +1250,17 @@ async function _printSelectedEmails(messages) {
frame.setAttribute('sandbox', 'allow-same-origin allow-modals');
frame.style.cssText = 'position:fixed;left:-10000px;top:0;width:800px;height:600px;border:0;';
document.body.appendChild(frame);
+
+ let cleanedUp = false;
+ const cleanupFrame = () => {
+ if (cleanedUp) return;
+ cleanedUp = true;
+ try {
+ frame.remove();
+ } catch (_) {}
+ window.removeEventListener('afterprint', cleanupFrame);
+ };
+
try {
const printDoc = frame.contentDocument;
printDoc.open();
@@ -1223,11 +1275,17 @@ async function _printSelectedEmails(messages) {
container.innerHTML = articles.join('');
printDoc.body.appendChild(container);
await printDoc.fonts.ready;
- frame.contentWindow.addEventListener('afterprint', () => frame.remove(), { once: true });
+ frame.contentWindow?.addEventListener('afterprint', cleanupFrame, { once: true });
+ window.addEventListener('afterprint', cleanupFrame, { once: true });
+ setTimeout(cleanupFrame, 60000);
frame.contentWindow.focus();
frame.contentWindow.print();
+ if (omitted.length > 0) {
+ const msg = `Printed ${successful.length} email${successful.length === 1 ? '' : 's'}; ${omitted.length} unavailable`;
+ showToast(msg);
+ }
} catch (error) {
- frame.remove();
+ cleanupFrame();
throw error;
}
}
diff --git a/tests/test_email_library_bulk_actions.py b/tests/test_email_library_bulk_actions.py
index 6840b5916..1d8464270 100644
--- a/tests/test_email_library_bulk_actions.py
+++ b/tests/test_email_library_bulk_actions.py
@@ -411,3 +411,161 @@ def test_messages_download_bulk_enforces_account_ownership_guard(monkeypatch):
assert res.status_code == 404
assert "Account not found" in res.text
+
+
+def test_messages_download_bulk_mid_batch_failure_accounting(monkeypatch):
+ """When an exception occurs mid-batch after partial progress, already-exported
+ records must not be double-counted as skipped."""
+ class MidBatchFailImap:
+ def select(self, folder, readonly=True):
+ return "OK", [b"1"]
+ def uid(self, command, *args):
+ records = []
+ for i in range(1, 5):
+ records.append((f"{i} (UID {i} RFC822 {{{10}}}".encode(), f"Subject: Msg {i}\r\n\r\nBody {i}".encode()))
+ class BadRecord:
+ def __bytes__(self):
+ raise RuntimeError("Corrupt memory buffer")
+ records.append((b"5 (UID 5 RFC822 {10}", BadRecord()))
+ return "OK", records
+
+ from contextlib import contextmanager
+ from fastapi import FastAPI
+ from fastapi.testclient import TestClient
+ from routes.email import email_routes as routes
+
+ @contextmanager
+ def mock_imap(account_id=None, owner=""):
+ yield MidBatchFailImap()
+
+ monkeypatch.setattr(routes, "_start_poller", lambda: None)
+ monkeypatch.setattr(routes, "_POOL_HOOKS", {})
+ monkeypatch.setattr(routes, "_imap", mock_imap)
+
+ app = FastAPI()
+ app.include_router(routes.setup_email_routes())
+ app.dependency_overrides[routes.require_owner] = lambda: "test-user"
+ client = TestClient(app)
+
+ rows = [{"uid": str(i)} for i in range(1, 11)]
+ res = client.post("/api/email/messages-download-bulk", json={"messages": rows})
+ assert res.status_code == 200
+ exp = int(res.headers.get("x-exported-emails"))
+ skp = int(res.headers.get("x-skipped-emails"))
+ assert exp == 4
+ assert skp == 6
+ assert exp + skp == 10
+
+
+def test_messages_download_bulk_size_cap_413_boundary(monkeypatch):
+ """Cumulative export size exceeding max_total_bytes returns HTTP 413."""
+ class LargeMessageImap:
+ def select(self, folder, readonly=True):
+ return "OK", [b"1"]
+ def uid(self, command, *args):
+ raw_uids = args[0] if isinstance(args[0], str) else args[0].decode()
+ if "(UID RFC822.SIZE)" in str(args):
+ records = []
+ for u in raw_uids.split(","):
+ records.append(f"{u} (UID {u} RFC822.SIZE 10485760)".encode())
+ return "OK", records
+ records = []
+ big_chunk = b"X" * (10 * 1024 * 1024)
+ for u in raw_uids.split(","):
+ records.append((f"{u} (UID {u} RFC822".encode(), big_chunk))
+ return "OK", records
+
+ from contextlib import contextmanager
+ from fastapi import FastAPI
+ from fastapi.testclient import TestClient
+ from routes.email import email_routes as routes
+
+ @contextmanager
+ def mock_imap(account_id=None, owner=""):
+ yield LargeMessageImap()
+
+ monkeypatch.setattr(routes, "_start_poller", lambda: None)
+ monkeypatch.setattr(routes, "_POOL_HOOKS", {})
+ monkeypatch.setattr(routes, "_imap", mock_imap)
+
+ app = FastAPI()
+ app.include_router(routes.setup_email_routes())
+ app.dependency_overrides[routes.require_owner] = lambda: "test-user"
+ client = TestClient(app)
+
+ # 52 messages of 10 MB = 520 MB > 512 MiB limit
+ payload_52 = [{"uid": str(i + 1), "subject": f"Big {i}"} for i in range(52)]
+ res = client.post("/api/email/messages-download-bulk", json={"messages": payload_52})
+ assert res.status_code == 413
+ assert res.json()["detail"] == "Export too large"
+
+
+def test_messages_download_bulk_per_message_limit_413(monkeypatch):
+ """A single oversized message (>64 MiB) triggers HTTP 413."""
+ class HugeSingleMsgImap:
+ def select(self, folder, readonly=True):
+ return "OK", [b"1"]
+ def uid(self, command, *args):
+ return "OK", [b"1 (UID 101 RFC822.SIZE 73400320)"]
+
+ from contextlib import contextmanager
+ from fastapi import FastAPI
+ from fastapi.testclient import TestClient
+ from routes.email import email_routes as routes
+
+ @contextmanager
+ def mock_imap(account_id=None, owner=""):
+ yield HugeSingleMsgImap()
+
+ monkeypatch.setattr(routes, "_start_poller", lambda: None)
+ monkeypatch.setattr(routes, "_POOL_HOOKS", {})
+ monkeypatch.setattr(routes, "_imap", mock_imap)
+
+ app = FastAPI()
+ app.include_router(routes.setup_email_routes())
+ app.dependency_overrides[routes.require_owner] = lambda: "test-user"
+ client = TestClient(app)
+
+ res = client.post("/api/email/messages-download-bulk", json={
+ "messages": [{"uid": "101", "subject": "Huge Email"}]
+ })
+ assert res.status_code == 413
+ assert res.json()["detail"] == "Export too large"
+
+
+def test_messages_download_bulk_account_ownership_memoization(monkeypatch):
+ """Account ownership validation is memoized per distinct account_id to eliminate N+1 DB queries."""
+ owns_calls = []
+ def counting_assert_owns(account_id, owner):
+ owns_calls.append(account_id)
+
+ raw_rfc822 = b"From: sender@example.com\r\nSubject: Test\r\n\r\nBody"
+ msgs = {str(i): raw_rfc822 for i in range(1, 26)}
+ mailbox = _MockImapConn(msgs)
+ client = _create_test_app(monkeypatch, mailbox=mailbox, assert_owns_fn=counting_assert_owns)
+
+ # 25 messages all referencing the same account_id
+ payload = [{"uid": str(i), "account_id": "acc-shared"} for i in range(1, 26)]
+ res = client.post("/api/email/messages-download-bulk", json={"messages": payload})
+ assert res.status_code == 200
+ assert len(owns_calls) == 1
+
+
+def test_messages_download_bulk_corrupt_zip_fails_closed(monkeypatch):
+ """If ZIP construction fails (e.g. disk error), fail closed with 500 instead of returning corrupt archive."""
+ import zipfile
+ raw_rfc822 = b"From: sender@example.com\r\nSubject: Test\r\n\r\nBody"
+ mailbox = _MockImapConn({"101": raw_rfc822, "102": raw_rfc822})
+
+ orig_writestr = zipfile.ZipFile.writestr
+ def failing_writestr(self, zinfo_or_arcname, data=None, *args, **kwargs):
+ raise OSError("Disk write failed: No space left on device")
+
+ monkeypatch.setattr(zipfile.ZipFile, "writestr", failing_writestr)
+ client = _create_test_app(monkeypatch, mailbox=mailbox)
+
+ res = client.post("/api/email/messages-download-bulk", json={
+ "messages": [{"uid": "101"}, {"uid": "102"}]
+ })
+ assert res.status_code == 500
+ assert "Failed to create export archive" in res.json()["detail"]