Add metadata and tags fields to Statement model; enhance search results to include statements

Close #44

nummi/main/models.py

@@ -26,7 +26,7 @@ class NummiQuerySet(models.QuerySet):
                     ),
                     SearchQuery(search, search_type="websearch"),
                 ),
-                similarity=TrigramSimilarity("name", search),
+                similarity=TrigramSimilarity(self.main_field, search),
             )
             .filter(models.Q(rank__gte=0.1) | models.Q(similarity__gte=0.3))
             .order_by("-rank")

nummi/search/templates/search/search_results.html

@@ -1,6 +1,6 @@
 {% extends "main/base.html" %}
 {% load i18n static %}
-{% load main_extras account_extras transaction_extras %}
+{% load main_extras account_extras transaction_extras statement_extras %}
 {% block title %}
   {% translate "Search" %} – Nummi
 {% endblock title %}
@@ -21,6 +21,12 @@
       <div class="split">{% account_table accounts search=True %}</div>
     </section>
   {% endif %}
+  {% if statements %}
+    <section>
+      <h3>{% translate "Statements" %}</h3>
+      {% statement_table statements %}
+    </section>
+  {% endif %}
   {% if categories %}
     <section>
       <h3>{% translate "Categories" %}</h3>

nummi/search/views.py

@@ -25,6 +25,7 @@ class SearchView(LoginRequiredMixin, TemplateView):
         context["search"] = self.kwargs["search"]
         context["transactions"] = _user.transaction_set.search(self.kwargs["search"])
         context["accounts"] = _user.account_set.search(self.kwargs["search"])
+        context["statements"] = _user.statement_set.search(self.kwargs["search"])
         context["categories"] = _user.category_set.search(self.kwargs["search"])
         context["invoices"] = _user.invoice_set.search(self.kwargs["search"])[:10]
 

nummi/statement/migrations/0004_statement_metadata_statement_tags.py

@@ -0,0 +1,22 @@
+# Generated by Django 4.2.7 on 2025-01-05 17:54
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("statement", "0003_remove_statement_diff_remove_statement_sum"),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name="statement",
+            name="metadata",
+            field=models.TextField(blank=True),
+        ),
+        migrations.AddField(
+            model_name="statement",
+            name="tags",
+            field=models.TextField(blank=True),
+        ),
+    ]

nummi/statement/models.py

@@ -7,7 +7,17 @@ from django.core.validators import FileExtensionValidator
 from django.db import models
 from django.urls import reverse
 from django.utils.translation import gettext_lazy as _
+from main.models import NummiQuerySet
+from main.utils import pdf_outline_to_str
 from media.utils import get_path
+from pypdf import PdfReader
+
+
+class StatementQuerySet(NummiQuerySet):
+    main_field = "metadata"
+    fields = {
+        "tags": "B",
+    }
 
 
 class Statement(AccountModel):
@@ -30,6 +40,10 @@ class Statement(AccountModel):
         blank=True,
         default="",
     )
+    metadata = models.TextField(blank=True)
+    tags = models.TextField(blank=True)
+
+    objects = StatementQuerySet.as_manager()
 
     def __str__(self):
         return _("%(date)s statement") % {"date": self.date}
@@ -40,6 +54,28 @@ class Statement(AccountModel):
             if _prever.file and _prever.file != self.file:
                 Path(_prever.file.path).unlink(missing_ok=True)
 
+        if self.file:
+            reader = PdfReader(self.file)
+
+            if reader.metadata:
+                self.metadata = " ".join(
+                    (
+                        m.replace("\x00", "").strip()
+                        for m in (
+                            reader.metadata.title,
+                            reader.metadata.author,
+                            reader.metadata.subject,
+                        )
+                        if m
+                    )
+                )
+
+            _tags = pdf_outline_to_str(reader.outline)
+            _tags += " ".join(
+                (page.extract_text().replace("\x00", "") for page in reader.pages)
+            )
+            self.tags = " ".join((tag for tag in _tags.split() if len(tag) >= 3))
+
         super().save(*args, **kwargs)
 
     @property