From 40242ba51b761e036749eada3d41fe0db7fa3533 Mon Sep 17 00:00:00 2001
From: Alex Pyrgiotis <alex.p@freedom.press>
Date: Thu, 19 Dec 2024 18:55:36 +0200
Subject: [PATCH] WIP: Work with HWPX files

---
 Dockerfile                             |  3 ++-
 dangerzone/conversion/common.py        | 17 ++++-------------
 dangerzone/conversion/doc_to_pixels.py |  2 ++
 3 files changed, 8 insertions(+), 14 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index d1760f9..4c4e412 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -30,7 +30,8 @@ RUN mkdir /libreoffice_ext && cd libreoffice_ext \
     && H2ORESTART_FILENAME=h2orestart.oxt \
     && wget https://github.com/ebandal/H2Orestart/releases/download/$H2ORESTART_VERSION/$H2ORESTART_FILENAME \
     && echo "$H2ORESTART_CHECKSUM  $H2ORESTART_FILENAME" | sha256sum -c \
-    && install -dm777 "/usr/lib/libreoffice/share/extensions/"
+    && install -dm777 "/usr/lib/libreoffice/share/extensions/" \
+    && rm /root/.wget-hsts
 
 RUN mkdir -p /opt/dangerzone/dangerzone && \
   touch /opt/dangerzone/dangerzone/__init__.py && \
diff --git a/dangerzone/conversion/common.py b/dangerzone/conversion/common.py
index 2ff10ec..9bbfe8b 100644
--- a/dangerzone/conversion/common.py
+++ b/dangerzone/conversion/common.py
@@ -8,15 +8,6 @@ DEFAULT_DPI = 150  # Pixels per inch
 INT_BYTES = 2
 
 
-class CommandError(RuntimeError):
-
-    def __init__(self, msg, stdout, stderr):
-        self.stdout = stdout
-        self.stderr = stderr
-        msg += f"\n====\nCommand output:\n{stdout}\n=====\nCommand stderr:\n{stderr}\n======"
-        super().__init__(msg)
-
-
 def running_on_qubes() -> bool:
     # https://www.qubes-os.org/faq/#what-is-the-canonical-way-to-detect-qubes-vm
     return os.path.exists("/usr/share/qubes/marker-vm")
@@ -105,7 +96,7 @@ class DangerzoneConverter:
         Run a command using asyncio.subprocess, consume its standard streams, and return its
         output in bytes.
 
-        :raises CommandError: if the process returns a non-zero exit status
+        :raises RuntimeError: if the process returns a non-zero exit status
         """
         # Start the provided command, and return a handle. The command will run in the
         # background.
@@ -134,15 +125,15 @@ class DangerzoneConverter:
         # Wait until the command has finished. Then, verify that the command
         # has completed successfully. In any other case, raise an exception.
         ret = await proc.wait()
+        if ret != 0:
+            raise RuntimeError(error_message)
+
 
         # Wait until the tasks that consume the command's standard streams have exited as
         # well, and return their output.
         stdout = await stdout_task
         stderr = await stderr_task
 
-        if ret != 0:
-            raise CommandError(error_message, stdout, stderr)
-
         return (stdout, stderr)
 
     @abstractmethod
diff --git a/dangerzone/conversion/doc_to_pixels.py b/dangerzone/conversion/doc_to_pixels.py
index 6737607..a7a7776 100644
--- a/dangerzone/conversion/doc_to_pixels.py
+++ b/dangerzone/conversion/doc_to_pixels.py
@@ -129,6 +129,8 @@ class DocumentToPixels(DangerzoneConverter):
             # At least .odt, .docx, .odg, .odp, .ods, and .pptx
             "application/zip": {
                 "type": "libreoffice",
+                # NOTE: Older `file` command cannot detect hwpx files properly.
+                "libreoffice_ext": "h2orestart.oxt",
             },
             # At least .doc, .docx, .odg, .odp, .odt, .pdf, .ppt, .pptx, .xls, and .xlsx
             "application/octet-stream": {