Download the Judgment PDF from the PRIVATE_ASSET_BUCKET

In an earlier PR we added the ability to download the Judgment PDF from the
public asset bucket. This is incorrect - editors shoud download the PDF from the
*private* asset bucket, so they can check it before publication.

We have refactored this to work more in line with downloading the docx -
rather than getting the PDF from the bucket and returning it in an HTTPResponse,
we can generate a presigned URL and download the PDF in the same way we download
the docx. This simplifies the views and urls considerably.

ds_caselaw_editor_ui/templates/includes/judgment_text_toolbar.html

@@ -18,8 +18,8 @@
           {% if context.docx_url %}
             <a href="{{ context.docx_url }}">{% translate "judgment.download_docx" %}</a>
           {% endif %}
-          {% if context.has_pdf %}
-            <a href="{% url 'detail_pdf' %}?judgment_uri={{context.judgment_uri}}">{% translate "judgment.download_pdf" %}</a>
+          {% if context.pdf_url %}
+            <a href="{{ context.pdf_url }}">{% translate "judgment.download_pdf" %}</a>
           {% endif %}
           {% if context.is_failure %}
             <a class="judgment-toolbar__delete" href="{% url 'delete' %}?judgment_uri={{ context.judgment_uri }}">{% translate "judgment.delete" %}</a>

ds_caselaw_editor_ui/templates/judgment/edit.html

@@ -2,9 +2,11 @@
 {% block content %}
 {% load i18n %}
    <aside class="judgment-info-component">
-    <a href="{{ context.docx_url }}">{% translate "judgment.download_docx" %}</a>
-    {% if context.has_pdf %}
-      <a href="{% url 'detail_pdf' %}?judgment_uri={{context.judgment_uri}}">{% translate "judgment.download_pdf" %}</a>
+    {% if context.docx_url %}
+      <a href="{{ context.docx_url }}">{% translate "judgment.download_docx" %}</a>
+    {% endif %}
+    {% if context.pdf_url %}
+      <a href="{{ context.pdf_url }}">{% translate "judgment.download_pdf" %}</a>
     {% endif %}
   </aside>
   <div class="edit-component">

judgments/tests.py

@@ -1,25 +1,12 @@
-import os
 import re
-from unittest import mock, skip
-from unittest.mock import patch
+from unittest import skip
 
 from django.test import TestCase
 
 from judgments import converters, views
 from judgments.models import Judgment
 
 
-def mocked_requests_head(*args):
-    class MockResponse:
-        def __init__(self, status_code):
-            self.status_code = status_code
-
-    if "ewca/civ/2004/632" in args[0]:
-        return MockResponse(200)
-
-    return MockResponse(404)
-
-
 class TestJudgment(TestCase):
     @skip
     def test_valid_content(self):
@@ -35,24 +22,6 @@ def test_404_response(self):
         self.assertIn("Judgment was not found", decoded_response)
         self.assertEqual(response.status_code, 404)
 
-    @patch.dict(
-        os.environ,
-        {"PUBLIC_ASSET_BUCKET": "public-asset-bucket", "S3_REGION": "eu-west-2"},
-        clear=True,
-    )
-    @mock.patch("requests.head", side_effect=mocked_requests_head)
-    def test_has_pdf_true(self, mock_head):
-        self.assertEqual(views.has_pdf("ewca/civ/2004/632"), True)
-
-    @patch.dict(
-        os.environ,
-        {"PUBLIC_ASSET_BUCKET": "public-asset-bucket", "S3_REGION": "eu-west-2"},
-        clear=True,
-    )
-    @mock.patch("requests.head", side_effect=mocked_requests_head)
-    def test_has_pdf_false(self, mock_head):
-        self.assertEqual(views.has_pdf("ewca/civ/2004/XXX"), False)
-
 
 class TestJudgmentModel(TestCase):
     def test_can_parse_judgment(self):

judgments/urls.py

@@ -5,7 +5,6 @@
 urlpatterns = [
     path("edit", views.EditJudgmentView.as_view(), name="edit"),
     path("detail", views.detail, name="detail"),
-    path("detail_pdf", views.detail_pdf, name="detail_pdf"),
     path("results", views.results, name="results"),
     path("delete", views.delete, name="delete"),
     path("", views.index, name="home"),

judgments/views.py

@@ -7,7 +7,6 @@
 import botocore.client
 import caselawclient.xml_tools as xml_tools
 import environ
-import requests
 from botocore.exceptions import ClientError
 from caselawclient.Client import (
     RESULTS_PER_PAGE,
@@ -17,7 +16,6 @@
 )
 from caselawclient.xml_tools import JudgmentMissingMetadataError
 from django.http import Http404, HttpResponse
-from django.shortcuts import redirect
 from django.template import loader
 from django.utils.translation import gettext
 from django.views.generic import View
@@ -54,7 +52,7 @@ def get_metadata(self, uri: str, judgment: ET.Element) -> dict:
             )
             meta["judgment_date"] = xml_tools.get_judgment_date_value(judgment)
             meta["docx_url"] = generate_docx_url(uri)
-            meta["has_pdf"] = has_pdf(uri)
+            meta["pdf_url"] = generate_pdf_url(uri)
             meta["previous_versions"] = self.get_versions(uri)
         except JudgmentMissingMetadataError:
             meta[
@@ -147,7 +145,7 @@ def detail(request):
         context["judgment"] = judgment
         context["page_title"] = metadata_name
         context["docx_url"] = generate_docx_url(judgment_uri)
-        context["has_pdf"] = has_pdf(judgment_uri)
+        context["pdf_url"] = generate_pdf_url(judgment_uri)
 
         if version_uri:
             context["version"] = re.search(r"([\d])-([\d]+)", version_uri).group(1)
@@ -157,21 +155,6 @@ def detail(request):
     return HttpResponse(template.render({"context": context}, request))
 
 
-def detail_pdf(request):
-    params = request.GET
-    judgment_uri = params.get("judgment_uri")
-    pdf_path = f'{judgment_uri}/{judgment_uri.replace("/", "_")}.pdf'
-    pdf_uri = f'https://{env("PUBLIC_ASSET_BUCKET")}.s3.{env("S3_REGION")}.amazonaws.com/{pdf_path}'
-    return redirect(pdf_uri)
-
-
-def has_pdf(judgment_uri):
-    pdf_path = f'{judgment_uri}/{judgment_uri.replace("/", "_")}.pdf'
-    pdf_uri = f'https://{env("PUBLIC_ASSET_BUCKET")}.s3.{env("S3_REGION")}.amazonaws.com/{pdf_path}'
-    response = requests.head(pdf_uri)
-    return response.status_code == 200
-
-
 def delete(request):
     judgment_uri = request.GET.get("judgment_uri", None)
     context = {
@@ -403,3 +386,17 @@ def generate_docx_url(uri: str):
     return client.generate_presigned_url(
         "get_object", Params={"Bucket": bucket, "Key": key}
     )
+
+
+def generate_pdf_url(uri: str):
+    bucket = env("PRIVATE_ASSET_BUCKET", None)
+    if not bucket:
+        return ""
+
+    client = create_s3_client()
+
+    key = f'{uri}/{uri.replace("/", "_")}.pdf'
+
+    return client.generate_presigned_url(
+        "get_object", Params={"Bucket": bucket, "Key": key}
+    )