unified document matching, legacy and automatching work alongside now

src/documents/signals/handlers.py

@@ -9,7 +9,7 @@ from django.contrib.contenttypes.models import ContentType
 from django.utils import timezone
 
 from documents.classifier import DocumentClassifier
-from .. import index
+from .. import index, matching
 from ..models import Document, Tag
 
 
@@ -17,35 +17,107 @@ def logger(message, group):
     logging.getLogger(__name__).debug(message, extra={"group": group})
 
 
-#TODO: global? really?
-classifier = DocumentClassifier()
-
-
 def index_document(sender, document=None, logging_group=None, **kwargs):
     index.add_document_to_index(sender, instance=document)
 
 
-def classify_document(sender, document=None, logging_group=None, **kwargs):
-    global classifier
-    try:
-        classifier.reload()
-        classifier.classify_document(
-            document,
-            classify_correspondent=True,
-            classify_tags=True,
-            classify_document_type=True
-        )
-    except FileNotFoundError:
-        logging.getLogger(__name__).fatal(
-            "Cannot classify document, classifier model file was not found."
-        )
-
-
 def add_inbox_tags(sender, document=None, logging_group=None, **kwargs):
     inbox_tags = Tag.objects.filter(is_inbox_tag=True)
     document.tags.add(*inbox_tags)
 
 
+def set_correspondent(sender, document=None, logging_group=None, classifier=None, replace=False, use_first=True, **kwargs):
+    if document.correspondent and not replace:
+        return
+
+    potential_correspondents = matching.match_correspondents(document.content, classifier)
+
+    potential_count = len(potential_correspondents)
+    if potential_correspondents:
+        selected = potential_correspondents[0]
+    else:
+        selected = None
+    if potential_count > 1:
+        if use_first:
+            message = "Detected {} potential correspondents, so we've opted for {}"
+            logger(
+                message.format(potential_count, selected),
+                logging_group
+            )
+        else:
+            message = "Detected {} potential correspondents, not assigning any correspondent"
+            logger(
+                message.format(potential_count),
+                logging_group
+            )
+            return
+
+    logger(
+        'Assigning correspondent "{}" to "{}" '.format(selected, document),
+        logging_group
+    )
+
+    document.correspondent = selected
+    document.save(update_fields=("correspondent",))
+
+
+def set_document_type(sender, document=None, logging_group=None, classifier=None, replace=False, use_first=True, **kwargs):
+    if document.document_type and not replace:
+        return
+
+    potential_document_type = matching.match_document_types(document.content, classifier)
+
+    potential_count = len(potential_document_type)
+    if potential_document_type:
+        selected = potential_document_type[0]
+    else:
+        selected = None
+
+    if potential_count > 1:
+        if use_first:
+            message = "Detected {} potential document types, so we've opted for {}"
+            logger(
+                message.format(potential_count, selected),
+                logging_group
+            )
+        else:
+            message = "Detected {} potential document types, not assigning any document type"
+            logger(
+                message.format(potential_count),
+                logging_group
+            )
+            return
+
+    logger(
+        'Assigning document type "{}" to "{}" '.format(selected, document),
+        logging_group
+    )
+
+    document.document_type = selected
+    document.save(update_fields=("document_type",))
+
+
+def set_tags(sender, document=None, logging_group=None, classifier=None, replace=False, **kwargs):
+    if replace:
+        document.tags.clear()
+        current_tags = set([])
+    else:
+        current_tags = set(document.tags.all())
+
+    relevant_tags = set(matching.match_tags(document.content, classifier)) - current_tags
+
+    if not relevant_tags:
+        return
+
+    message = 'Tagging "{}" with "{}"'
+    logger(
+        message.format(document, ", ".join([t.slug for t in relevant_tags])),
+        logging_group
+    )
+
+    document.tags.add(*relevant_tags)
+
+
 def run_pre_consume_script(sender, filename, **kwargs):
 
     if not settings.PRE_CONSUME_SCRIPT: