Merge pull request #1357 from digitalfabrik/bugfix/pdf-export

Fix PDF export

.gitignore

@@ -32,9 +32,6 @@ integreat_cms/locale/de/LC_MESSAGES/django.mo
 # Compressed JS & CSS Files
 integreat_cms/static/dist/
 
-# File cache
-integreat_cms/cache/
-
 # Django stuff:
 *.log
 *.pot
@@ -60,6 +57,9 @@ webpack-stats.json
 # Media Library
 integreat_cms/media/
 
+# PDF files
+integreat_cms/pdf/
+
 # XLIFF files folder
 integreat_cms/xliff/upload
 integreat_cms/xliff/download

CHANGELOG.md

@@ -6,6 +6,8 @@ UNRELEASED
 * [ [#1328](https://github.com/digitalfabrik/integreat-cms/issues/1328) ] Fix missing entries in broken link checker
 * [ [#1289](https://github.com/digitalfabrik/integreat-cms/issues/1289) ] Prevent submitting feedback for a non-existent imprint
 * [ [#1359](https://github.com/digitalfabrik/integreat-cms/issues/1359) ] Cascade delete imprint feedback when imprint is deleted
+* [ [#1350](https://github.com/digitalfabrik/integreat-cms/issues/1350) ] Fix font support of PDF export
+* [ [#1349](https://github.com/digitalfabrik/integreat-cms/issues/1349) ] Fix network error when downloading PDF files
 
 
 2022.4.0

example-configs/integreat-cms.ini

@@ -52,6 +52,8 @@ ALLOWED_HOSTS =
 STATIC_ROOT = /var/www/integreat-cms/static
 # The directory for media files [optional, defaults to "media" in the application directory]
 MEDIA_ROOT = /var/www/integreat-cms/media
+# The directory for PDF files [optional, defaults to "pdf" in the application directory]
+PDF_ROOT = /var/www/integreat-cms/pdf
 # The directory for xliff files [optional, defaults to "xliff" in the application directory]
 XLIFF_ROOT = /var/www/integreat-cms/xliff
 # Enable the possibility to upload legacy file formats [optional, defaults to False]
@@ -74,8 +76,6 @@ DB_PORT = <port>
 REDIS_CACHE = True
 # Set this if you want to connect to redis via socket [optional, defaults to None]
 REDIS_UNIX_SOCKET = /var/run/redis/redis-server.sock
-# The location of the file-based cache (e.g. for PDF files) [optional, defaults to "cache" in the application directory]
-FILE_CACHE = /var/cache/integreat-cms
 
 [email]
 # Sender email [optional, defaults to "[email protected]"]

integreat_cms/api/v3/pdf_export.py

@@ -33,8 +33,8 @@ def pdf_export(request, region_slug, language_slug):
 
     :raises ~django.http.Http404: HTTP status 404 if the requested page translation cannot be found.
 
-    :return: The requested pages as PDF document (inline)
-    :rtype: ~django.http.HttpResponse
+    :return: The redirect to the generated PDF document
+    :rtype: ~django.http.HttpResponseRedirect
     """
     region = request.region
     # Request unrestricted queryset because pdf generator performs further operations (e.g. aggregation) on the queryset

integreat_cms/cms/templates/pages/page_pdf.html

@@ -48,7 +48,7 @@
             <table>
                 <tr>
                     <td id="first-footer">
-                        <pdf:pagenumber>
+                        <pdf:pagenumber/>
                     </td>
                     <td id="second-footer">
                         {{ region.full_name }}
@@ -69,9 +69,9 @@ <h1 id="title_page">{{ language.table_of_contents }}</h1>
             <!-- according to xhtml2pdf documentation all custom tags like <pdf:toc/>, <pdf:nextpage>, etc.
                 should be wrapped inside block tags like <div> to avoid problems -->
             <div class="toc">
-                <pdf:toc />
+                <pdf:toc/>
             </div>
-            <div><pdf:nextpage></div>
+            <div><pdf:nextpage/></div>
         {% endif %}
 
         {% for page, info in annotated_pages %}

integreat_cms/cms/utils/pdf_utils.py

@@ -4,9 +4,11 @@
 
 from django.conf import settings
 from django.contrib.staticfiles import finders
-from django.core.cache import caches
 from django.db.models import Min
+from django.core.files.base import ContentFile
+from django.core.files.storage import FileSystemStorage
 from django.http import HttpResponse
+from django.shortcuts import redirect
 from django.template.loader import get_template
 from django.utils.translation import ugettext as _
 from django.views.decorators.cache import never_cache
@@ -18,6 +20,8 @@
 
 logger = logging.getLogger(__name__)
 
+pdf_storage = FileSystemStorage(location=settings.PDF_ROOT, base_url=settings.PDF_URL)
+
 
 @never_cache
 # pylint: disable=too-many-locals
@@ -36,8 +40,8 @@ def generate_pdf(region, language_slug, pages):
     :param pages: at least on page to render as PDF document
     :type pages: ~treebeard.ns_tree.NS_NodeQuerySet
 
-    :return: PDF document wrapped in a HtmlResponse
-    :rtype: ~django.http.HttpResponse
+    :return: Redirection to PDF document
+    :rtype: ~django.http.HttpResponseRedirect
     """
     # first all necessary data for hashing are collected, starting at region slug
     # region last_updated field taking into account, to keep track of maybe edited region icons
@@ -55,12 +59,7 @@ def generate_pdf(region, language_slug, pages):
     # finally combine all list entries to a single hash key
     pdf_key_string = "_".join(map(str, pdf_key_list))
     # compute the hash value based on the hash key
-    pdf_hash = hashlib.sha256(bytes(pdf_key_string, "utf-8")).hexdigest()
-    cache = caches["pdf"]
-    cached_response = cache.get(pdf_hash, "has_expired")
-    if cached_response != "has_expired":
-        # if django cache already contains a response object
-        return cached_response
+    pdf_hash = hashlib.sha256(bytes(pdf_key_string, "utf-8")).hexdigest()[:10]
     amount_pages = pages.count()
     if amount_pages == 0:
         return HttpResponse(
@@ -81,36 +80,39 @@ def generate_pdf(region, language_slug, pages):
             # In any other case, take the region name
             title = region.name
     language = Language.objects.get(slug=language_slug)
-    filename = f"Integreat - {language.translated_name} - {title}.pdf"
-    # Convert queryset to annotated list which can be rendered better
-    annotated_pages = Page.get_annotated_list_qs(pages)
-    context = {
-        "right_to_left": language.text_direction == text_directions.RIGHT_TO_LEFT,
-        "region": region,
-        "annotated_pages": annotated_pages,
-        "language": language,
-        "amount_pages": amount_pages,
-        "prevent_italics": ["ar", "fa"],
-    }
-    response = HttpResponse(content_type="application/pdf")
-    response["Content-Disposition"] = f'filename="{filename}"'
-    html = get_template("pages/page_pdf.html").render(context)
-    pisa_status = pisa.CreatePDF(
-        html, dest=response, link_callback=link_callback, encoding="UTF-8"
-    )
-    # pylint: disable=no-member
-    if pisa_status.err:
-        logger.error(
-            "The following PDF could not be rendered: %r, %r, %r",
-            region,
-            language,
-            pages,
-        )
-        return HttpResponse(
-            _("The PDF could not be successfully generated."), status=500
-        )
-    cache.set(pdf_hash, response, 60 * 60 * 24)
-    return response
+    filename = f"{pdf_hash}/Integreat - {language.translated_name} - {title}.pdf"
+    # Only generate new pdf if not already exists
+    if not pdf_storage.exists(filename):
+        # Convert queryset to annotated list which can be rendered better
+        annotated_pages = Page.get_annotated_list_qs(pages)
+        context = {
+            "right_to_left": language.text_direction == text_directions.RIGHT_TO_LEFT,
+            "region": region,
+            "annotated_pages": annotated_pages,
+            "language": language,
+            "amount_pages": amount_pages,
+            "prevent_italics": ["ar", "fa"],
+        }
+        html = get_template("pages/page_pdf.html").render(context)
+        # Save empty file
+        pdf_storage.save(filename, ContentFile(""))
+        # Write PDF content into file
+        with pdf_storage.open(filename, "w+b") as pdf_file:
+            pisa_status = pisa.CreatePDF(
+                html, dest=pdf_file, link_callback=link_callback, encoding="UTF-8"
+            )
+        # pylint: disable=no-member
+        if pisa_status.err:
+            logger.error(
+                "The following PDF could not be rendered: %r, %r, %r",
+                region,
+                language,
+                pages,
+            )
+            return HttpResponse(
+                _("The PDF could not be successfully generated."), status=500
+            )
+    return redirect(pdf_storage.url(filename))
 
 
 # pylint: disable=unused-argument

integreat_cms/cms/views/pages/page_bulk_actions.py

@@ -50,18 +50,12 @@ def post(self, request, *args, **kwargs):
         :return: The redirect
         :rtype: ~django.http.HttpResponseRedirect
         """
-        # Generate PDF document wrapped in a HtmlResponse object
-        response = generate_pdf(
+        # Generate PDF document and redirect to it
+        return generate_pdf(
             request.region,
             kwargs.get("language_slug"),
             self.get_queryset(),
         )
-        if response.status_code == 200:
-            # Offer PDF document for download
-            response["Content-Disposition"] = (
-                response["Content-Disposition"] + "; attachment"
-            )
-        return response
 
 
 # pylint: disable=too-many-ancestors

integreat_cms/core/middleware/access_control_middleware.py

@@ -17,7 +17,15 @@ class AccessControlMiddleware:
     """
 
     #: The namespaces that are whitelisted and don't require access control
-    whitelist = ["api", "public", "sitemap", "i18n", "media_files", "xliff_files"]
+    whitelist = [
+        "api",
+        "public",
+        "sitemap",
+        "i18n",
+        "media_files",
+        "pdf_files",
+        "xliff_files",
+    ]
 
     def __init__(self, get_response):
         """

integreat_cms/core/settings.py

@@ -643,18 +643,10 @@
 
 #: Configuration for caches (see :setting:`django:CACHES` and :doc:`django:topics/cache`).
 #: Use a ``LocMemCache`` for development and a ``RedisCache`` whenever available.
-#: Additionally, a ``FileBasedCache`` is used for PDF caching.
 CACHES = {
     "default": {
         "BACKEND": "django.core.cache.backends.locmem.LocMemCache",
     },
-    "pdf": {
-        "BACKEND": "django.core.cache.backends.filebased.FileBasedCache",
-        "LOCATION": os.path.join(
-            os.environ.get("INTEGREAT_CMS_FILE_CACHE", os.path.join(BASE_DIR, "cache")),
-            "pdf",
-        ),
-    },
 }
 
 # Use RedisCache when activated
@@ -754,6 +746,17 @@
 ]
 
 
+##############
+# PDF EXPORT #
+##############
+
+#: The directory where PDF files are stored
+PDF_ROOT = os.environ.get("INTEGREAT_CMS_PDF_ROOT", os.path.join(BASE_DIR, "pdf"))
+
+#: The URL path where PDF files are served for download
+PDF_URL = "/pdf/"
+
+
 #######################
 # XLIFF SERIALIZATION #
 #######################

integreat_cms/core/urls.py

@@ -56,6 +56,15 @@
             )
         ),
     ),
+    path(
+        "",
+        include(
+            (
+                static(settings.PDF_URL, document_root=settings.PDF_ROOT),
+                "pdf_files",
+            )
+        ),
+    ),
     path(
         "",
         include(

integreat_cms/locale/de/LC_MESSAGES/django.po

@@ -8,7 +8,7 @@ msgid ""
 msgstr ""
 "Project-Id-Version: 1.0\n"
 "Report-Msgid-Bugs-To: \n"
-"POT-Creation-Date: 2022-04-15 14:48+0000\n"
+"POT-Creation-Date: 2022-04-15 20:11+0000\n"
 "PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
 "Last-Translator: Integreat <[email protected]>\n"
 "Language-Team: Integreat <[email protected]>\n"
@@ -3169,7 +3169,7 @@ msgstr "Feedback"
 
 #: cms/templates/_base.html:181
 #: cms/templates/push_notifications/push_notification_list.html:9
-#: core/settings.py:718
+#: core/settings.py:710
 msgid "News"
 msgstr "Nachrichten"
 
@@ -5711,12 +5711,12 @@ msgstr "Unbekannt"
 msgid "CW"
 msgstr "KW"
 
-#: cms/utils/pdf_utils.py:67
+#: cms/utils/pdf_utils.py:66
 msgid "No valid pages selected for PDF generation."
 msgstr ""
 "Es wurden keine gültigen Seiten zur Erzeugung der PDF-Datei ausgewählt."
 
-#: cms/utils/pdf_utils.py:110
+#: cms/utils/pdf_utils.py:113
 msgid "The PDF could not be successfully generated."
 msgstr "PDF-Datei konnte nicht erfolgreich erzeugt werden."
 
@@ -6404,11 +6404,11 @@ msgstr ""
 msgid "Success: The user {user} cannot publish this page anymore."
 msgstr "Erfolg: Der Nutzer {user} kann diese Seite nicht mehr veröffentlichen"
 
-#: cms/views/pages/page_bulk_actions.py:109
+#: cms/views/pages/page_bulk_actions.py:103
 msgid "XLIFF file for translation to {} successfully created."
 msgstr "XLIFF Datei für die Übersetzung nach {} wurde erfolgreich erstellt."
 
-#: cms/views/pages/page_bulk_actions.py:113
+#: cms/views/pages/page_bulk_actions.py:107
 msgid "If the download does not start automatically, please click {}here{}."
 msgstr ""
 "Wenn der Download nicht automatisch startet, klicken Sie bitte {}hier{}."