Merge branch 'dev' into dev

.github/workflows/ci.yml

@@ -22,6 +22,13 @@ env:
 
 jobs:
   pre-commit:
+    # We want to run on external PRs, but not on our own internal PRs as they'll be run
+    # by the push to the branch. Without this if check, checks are duplicated since
+    # internal PRs match both the push and pull_request events.
+    if:
+      github.event_name == 'push' || github.event.pull_request.head.repo.full_name !=
+      github.repository
+
     name: Linting Checks
     runs-on: ubuntu-22.04
     steps:

src-ui/src/styles.scss

@@ -254,6 +254,10 @@ a.btn-link:focus-visible,
     min-height: calc(1.5em + 0.75rem + 5px);
     line-height: 1.5;
 
+    .ng-select-container {
+      border-radius: $border-radius;
+    }
+
     .ng-select-container .ng-value-container .ng-input {
       top: 7px;
     }
@@ -311,6 +315,7 @@ textarea,
   color: var(--bs-body-color);
   background-color: var(--bs-body-bg);
   border-color: var(--bs-border-color);
+  font-size: 0.875rem;
 
   &:focus {
     background-color: var(--pngx-bg-darker);

src/documents/management/commands/document_consumer.py

@@ -32,7 +32,7 @@ except ImportError:  # pragma: nocover
 logger = logging.getLogger("paperless.management.consumer")
 
 
-def _tags_from_path(filepath) -> set[Tag]:
+def _tags_from_path(filepath) -> list[int]:
     """
     Walk up the directory tree from filepath to CONSUMPTION_DIR
     and get or create Tag IDs for every directory.
@@ -47,7 +47,7 @@ def _tags_from_path(filepath) -> set[Tag]:
             Tag.objects.get_or_create(name__iexact=part, defaults={"name": part})[0].pk,
         )
 
-    return tag_ids
+    return list(tag_ids)
 
 
 def _is_ignored(filepath: str) -> bool:

src/paperless_mail/mail.py

@@ -8,6 +8,7 @@ import traceback
 from datetime import date
 from datetime import timedelta
 from fnmatch import fnmatch
+from typing import Optional
 from typing import Union
 
 import magic
@@ -21,6 +22,7 @@ from django.utils.timezone import is_naive
 from django.utils.timezone import make_aware
 from imap_tools import AND
 from imap_tools import NOT
+from imap_tools import MailAttachment
 from imap_tools import MailBox
 from imap_tools import MailboxFolderSelectError
 from imap_tools import MailBoxUnencrypted
@@ -422,14 +424,19 @@ class MailAccountHandler(LoggingMixin):
 
     logging_name = "paperless_mail"
 
-    def _correspondent_from_name(self, name):
+    def _correspondent_from_name(self, name: str) -> Optional[Correspondent]:
         try:
             return Correspondent.objects.get_or_create(name=name)[0]
         except DatabaseError as e:
             self.log.error(f"Error while retrieving correspondent {name}: {e}")
             return None
 
-    def _get_title(self, message, att, rule):
+    def _get_title(
+        self,
+        message: MailMessage,
+        att: MailAttachment,
+        rule: MailRule,
+    ) -> Optional[str]:
         if rule.assign_title_from == MailRule.TitleSource.FROM_SUBJECT:
             return message.subject
 
@@ -444,7 +451,11 @@ class MailAccountHandler(LoggingMixin):
                 "Unknown title selector.",
             )  # pragma: nocover
 
-    def _get_correspondent(self, message: MailMessage, rule):
+    def _get_correspondent(
+        self,
+        message: MailMessage,
+        rule: MailRule,
+    ) -> Optional[Correspondent]:
         c_from = rule.assign_correspondent_from
 
         if c_from == MailRule.CorrespondentSource.FROM_NOTHING:
@@ -606,7 +617,6 @@ class MailAccountHandler(LoggingMixin):
             f"{len(message.attachments)} attachment(s)",
         )
 
-        correspondent = self._get_correspondent(message, rule)
         tag_ids = [tag.id for tag in rule.assign_tags.all()]
         doc_type = rule.assign_document_type
 
@@ -617,7 +627,6 @@ class MailAccountHandler(LoggingMixin):
             processed_elements += self._process_eml(
                 message,
                 rule,
-                correspondent,
                 tag_ids,
                 doc_type,
             )
@@ -629,7 +638,6 @@ class MailAccountHandler(LoggingMixin):
             processed_elements += self._process_attachments(
                 message,
                 rule,
-                correspondent,
                 tag_ids,
                 doc_type,
             )
@@ -640,7 +648,6 @@ class MailAccountHandler(LoggingMixin):
         self,
         message: MailMessage,
         rule: MailRule,
-        correspondent,
         tag_ids,
         doc_type,
     ):
@@ -669,6 +676,8 @@ class MailAccountHandler(LoggingMixin):
                 # as this is system dependent otherwise
                 continue
 
+            correspondent = self._get_correspondent(message, rule)
+
             title = self._get_title(message, att, rule)
 
             # don't trust the content type of the attachment. Could be
@@ -750,7 +759,6 @@ class MailAccountHandler(LoggingMixin):
         self,
         message: MailMessage,
         rule: MailRule,
-        correspondent,
         tag_ids,
         doc_type,
     ):
@@ -781,6 +789,8 @@ class MailAccountHandler(LoggingMixin):
 
             f.write(message.obj.as_bytes())
 
+        correspondent = self._get_correspondent(message, rule)
+
         self.log.info(
             f"Rule {rule}: "
             f"Consuming eml from mail "