pdf: make thumbnail url vary with PDF form file content (#78879)

passerelle/apps/pdf/models.py

@@ -15,6 +15,8 @@
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 import base64
+import functools
+import hashlib
 import os
 import subprocess
 import tempfile
@@ -139,6 +141,14 @@ class Resource(BaseResource):
     class Meta:
         verbose_name = _('PDF')
 
+    @functools.cached_property
+    def pdf_hash(self):
+        if self.fill_form_file:
+            with self.fill_form_file.open() as fd:
+                return hashlib.md5(fd.read()).hexdigest()
+        else:
+            return '0'
+
     @classmethod
     def get_manager_form_class(cls, **kwargs):
         kwargs['exclude'] = tuple(kwargs.get('exclude') or ()) + ('fields_mapping',)

passerelle/apps/pdf/templates/pdf/fields_mapping_edit.html

@@ -46,7 +46,7 @@
             <map name="map-page-{{ page_number }}">
               {{ image_map|safe }}
             </map>
-            <img src="{% url "pdf-page-thumbnail" connector="pdf" slug=object.slug page_number=page_number %}" usemap="#map-page-{{ page_number }}">
+            <img src="{% url "pdf-page-thumbnail" connector="pdf" slug=object.slug page_number=page_number %}?hash={{ object.pdf_hash }}" usemap="#map-page-{{ page_number }}">
           </div>
         </div>
         <div class="pdf-fields-mapping-edit-form--fields">

tests/test_pdf.py

@@ -34,6 +34,7 @@ from tests.utils import generic_endpoint_url, setup_access_rights
 with open(os.path.join(os.path.dirname(__file__), 'data', 'minimal.pdf'), 'rb') as fd:
     pdf_content = fd.read()
     pdf_b64content = base64.b64encode(pdf_content).decode()
+
 with open(os.path.join(os.path.dirname(__file__), 'data', 'pdf-form.pdf'), 'rb') as fd:
     acroform_content = fd.read()
     acroform_b64content = base64.b64encode(acroform_content).decode()
@@ -325,3 +326,28 @@ def test_add_or_edit(app, admin_user, pdf, cerfa_content):
     pdf.fill_form_file.save('form.pdf', ContentFile(cerfa_content))
     resp = app.get('/manage/pdf/test/edit')
     assert b'fields_mapping' not in resp.content
+
+
+def test_thumbnail_url_contains_hash(app, admin_user, pdf, cerfa_content):
+    pdf.fill_form_file.save('form.pdf', ContentFile(cerfa_content))
+    app = login(app)
+    resp = app.get('/pdf/test/')
+    resp = resp.click('Fill form: Edit fields mapping')
+    img_tags = resp.pyquery('img')
+    urls = [img_tag.attrib['src'] for img_tag in img_tags]
+    assert len(urls) == 5  # cerfa_content has 5 pages
+    assert all('?hash=' in url for url in urls)
+    assert not any(url.endswith('?hash=0') for url in urls)
+
+    first_url = urls[0]
+
+    pdf.fill_form_file.save('form.pdf', ContentFile(acroform_content))
+
+    resp = app.get('/pdf/test/')
+    resp = resp.click('Fill form: Edit fields mapping')
+    img_tags = resp.pyquery('img')
+    urls = [img_tag.attrib['src'] for img_tag in img_tags]
+    assert len(urls) == 1  # acroform_content has 1 page
+    assert all('?hash=' in url for url in urls)
+    assert not any(url.endswith('?hash=0') for url in urls)
+    assert urls[0] != first_url