Add insensitive search for documents

src/documents/admin.py

@@ -13,12 +13,11 @@
 from django.utils.safestring import mark_safe
 from djangoql.admin import DjangoQLSearchMixin
 
-from documents.actions import (
- add_tag_to_selected,
- remove_correspondent_from_selected,
- remove_tag_from_selected,
- set_correspondent_on_selected
-)
+from documents.actions import (add_tag_to_selected,
+ remove_correspondent_from_selected,
+ remove_tag_from_selected,
+ set_correspondent_on_selected)
+from paperless.utils import slugify
 
 from .models import Correspondent, Document, Log, Tag
 
@@ -173,7 +172,12 @@ class Media:
  "all": ("paperless.css",)
  }
 
- search_fields = ("correspondent__name", "title", "content", "tags__name")
+ search_fields = (
+ "correspondent__name",
+ "searchable_title",
+ "searchable_content",
+ "tags__name",
+ )
  readonly_fields = ("added", "file_type", "storage_type",)
  list_display = ("title", "created", "added", "thumbnail", "correspondent",
  "tags_")
@@ -338,6 +342,10 @@ def _html_tag(kind, inside=None, **kwargs):
 
  return format_html("<{} {}/>", kind, attributes)
 
+ def get_search_results(self, request, queryset, search_term):
+ search_term = slugify(search_term)
+ return super().get_search_results(request, queryset, search_term)
+
 
 class LogAdmin(CommonAdmin):
 

src/documents/migrations/0023_document_searchable_content.py

@@ -0,0 +1,44 @@
+# Generated by Django 2.0.13 on 2019-05-30 14:50
+import unicodedata
+
+from django.db import migrations, models
+
+from paperless.utils import slugify as slugifyOCR
+
+
+class Migration(migrations.Migration):
+ dependencies = [("documents", "0022_auto_20181007_1420")]
+
+ reversible = True
+
+ def casefold_forwards(apps, schema_editor):
+ Document = apps.get_model("documents", "Document")
+ for doc in Document.objects.all():
+ if doc.title is not None:
+ doc.searchable_title = slugifyOCR(doc.title)
+ if doc.content is not None:
+ doc.searchable_content = slugifyOCR(doc.content)
+ doc.save()
+
+ def casefold_backwards(apps, schema_editor):
+ pass
+
+ def database_backwards(self, app_label, schema_editor, from_state, to_state):
+ migrations.RemoveField(model_name="document", name="searchable_content"),
+ migrations.RemoveField(model_name="document", name="searchable_title"),
+
+ operations = [
+ migrations.AddField(
+ model_name="document",
+ name="searchable_content",
+ field=models.TextField(blank=True, db_index=True, editable=False),
+ ),
+ migrations.AddField(
+ model_name="document",
+ name="searchable_title",
+ field=models.CharField(
+ max_length=128, blank=True, db_index=True, editable=False
+ ),
+ ),
+ migrations.RunPython(casefold_forwards, casefold_backwards),
+ ]

src/documents/models.py

@@ -14,6 +14,8 @@
 from django.utils.text import slugify
 from fuzzywuzzy import fuzz
 
+from paperless.utils import slugify as slugifyOCR
+
 from .managers import LogManager
 
 try:
@@ -221,6 +223,19 @@ class Document(models.Model):
  "primarily used for searching."
  )
 
+ searchable_content = models.TextField(
+ db_index=True,
+ blank=True,
+ editable=False,
+ )
+
+ searchable_title = models.CharField(
+ max_length=128,
+ blank=True,
+ db_index=True,
+ editable=False,
+ )
+
  file_type = models.CharField(
  max_length=4,
  editable=False,
@@ -266,6 +281,13 @@ def __str__(self):
  return "{}: {}".format(created, self.correspondent or self.title)
  return str(created)
 
+ def save(self, *args, **kwargs):
+ if self.title is not None:
+ self.searchable_title = slugifyOCR(self.title)
+ if self.content is not None:
+ self.searchable_content = slugifyOCR(self.content)
+ return super().save(*args, **kwargs)
+
  @property
  def source_path(self):
 

src/documents/tests/test_document_model.py

@@ -21,3 +21,28 @@ def test_file_deletion(self):
  mock_unlink.assert_any_call(file_path)
  mock_unlink.assert_any_call(thumb_path)
  self.assertEqual(mock_unlink.call_count, 2)
+
+ def test_searchable_title_and_content(self):
+ document = Document.objects.create(
+ title="Title",
+ content="Content",
+ checksum="azerty1"
+ )
+ self.assertEqual(document.title, "Title")
+ self.assertEqual(document.content, "Content")
+ self.assertEqual(document.searchable_title, "title")
+ self.assertEqual(document.searchable_content, "content")
+
+ document = Document.objects.create(
+ title="Zürich Weiß",
+ content="Telefónica ééé aaa",
+ checksum="azerty2"
+ )
+ self.assertEqual(document.searchable_title, "zurich weiss")
+ self.assertEqual(document.searchable_content, "telefonica eee aaa")
+
+ document = Document.objects.create(checksum="azerty3")
+ self.assertEqual(document.title, '')
+ self.assertEqual(document.content, '')
+ self.assertEqual(document.searchable_title, '')
+ self.assertEqual(document.searchable_content, '')

src/paperless/utils.py

@@ -0,0 +1,9 @@
+import unicodedata
+
+
+def slugify(content):
+ return (
+ unicodedata.normalize("NFKD", content.casefold())
+ .encode("ASCII", "ignore")
+ .decode("utf-8")
+ )