From a277b375264d9ef56044d7a2bd2f883d163b1a1a Mon Sep 17 00:00:00 2001
From: "Edgar P. Burkhart" <git@edgarpierre.fr>
Date: Sun, 5 Jan 2025 18:42:58 +0100
Subject: [PATCH] Refactor invoice handling to remove unused formset and
 improve metadata extraction from PDF files Fix #46

---
 nummi/transaction/forms.py  |  4 ----
 nummi/transaction/models.py | 32 ++++++++++++++++++--------------
 nummi/transaction/views.py  |  5 +----
 3 files changed, 19 insertions(+), 22 deletions(-)

diff --git a/nummi/transaction/forms.py b/nummi/transaction/forms.py
index 8eec37a..6498581 100644
--- a/nummi/transaction/forms.py
+++ b/nummi/transaction/forms.py
@@ -3,7 +3,6 @@ import json
 from account.forms import AccountSelect
 from category.forms import CategorySelect
 from django import forms
-from django.forms import formset_factory
 from django.utils.translation import gettext_lazy as _
 from main.forms import DatalistInput, NummiFileInput, NummiForm
 from statement.forms import StatementSelect
@@ -146,9 +145,6 @@ class MultipleInvoicesForm(forms.Form):
     invoices = MultipleFileField()
 
 
-InvoicesFormSet = formset_factory(MultipleInvoicesForm)
-
-
 class DateInput(forms.DateInput):
     input_type = "date"
 
diff --git a/nummi/transaction/models.py b/nummi/transaction/models.py
index f50e4fa..06f90d9 100644
--- a/nummi/transaction/models.py
+++ b/nummi/transaction/models.py
@@ -108,23 +108,27 @@ class Invoice(NummiModel):
             if _prever.file and _prever.file != self.file:
                 Path(_prever.file.path).unlink(missing_ok=True)
 
-        reader = PdfReader(self.file)
+        if self.file:
+            reader = PdfReader(self.file)
 
-        self.metadata = " ".join(
-            (
-                m
-                for m in (
-                    reader.metadata.title,
-                    reader.metadata.author,
-                    reader.metadata.subject,
+            if reader.metadata:
+                self.metadata = " ".join(
+                    (
+                        m.replace("\x00", "").strip()
+                        for m in (
+                            reader.metadata.title,
+                            reader.metadata.author,
+                            reader.metadata.subject,
+                        )
+                        if m
+                    )
                 )
-                if m
-            )
-        )
 
-        _tags = pdf_outline_to_str(reader.outline)
-        _tags += " ".join((page.extract_text() for page in reader.pages))
-        self.tags = " ".join((tag for tag in _tags.split() if len(tag) >= 3))
+            _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)
 
diff --git a/nummi/transaction/views.py b/nummi/transaction/views.py
index 85a086c..de81a69 100644
--- a/nummi/transaction/views.py
+++ b/nummi/transaction/views.py
@@ -70,7 +70,6 @@ class MultipleInvoiceCreateView(FormView):
             self.request.user.transaction_set, pk=self.kwargs["transaction"]
         )
 
-        invoices = []
         for file in form.cleaned_data["invoices"]:
             invoice = Invoice(
                 transaction=transaction,
@@ -80,6 +79,7 @@ class MultipleInvoiceCreateView(FormView):
             )
             try:
                 invoice.full_clean()
+                invoice.save()
             except ValidationError as err:
                 for msg in err.messages:
                     messages.error(
@@ -91,10 +91,7 @@ class MultipleInvoiceCreateView(FormView):
                             err=msg,
                         ),
                     )
-            else:
-                invoices.append(invoice)
 
-        Invoice.objects.bulk_create(invoices)
         return super().form_valid(form)
 
     def get_success_url(self):