Add support for converting docx to pdf using headless libreoffice

dangerzone/tasks.py

@@ -95,6 +95,12 @@ class ConvertToPixels(TaskBase):
         ]
         output = self.execute_podman(args)
 
+        # Did we hit an error?
+        for line in output.split("\n"):
+            if "conversion failed" in line or "The document format is not supported" in line:
+                self.task_failed.emit(output)
+                return
+
         # How many pages was that?
         num_pages = None
         for line in output.split("\n"):

share/container/Containerfile

@@ -1,7 +1,7 @@
 FROM ubuntu:18.04
 
 RUN apt-get update && \
-    apt-get install -y file poppler-utils imagemagick ghostscript tesseract-ocr
+    apt-get install -y file poppler-utils imagemagick ghostscript tesseract-ocr libreoffice
 
 # TODO: when we support OCR in other languages, we need tesseract-ocr-all
 

share/container/document-to-pixels

@@ -1,11 +1,33 @@
 #!/bin/bash
 
-IMG_DEPTH=8
+die() {
+    echo "$1" >&2
+    exit 1
+}
+
+# Detect the mime type
+MIME_TYPE=$(file -b --mime-type /tmp/input_file)
+echo "Documet MIME type is $MIME_TYPE"
+echo
+
+# .docx
+if [ $MIME_TYPE = "application/pdf" ]; then
+    ORIGINAL_PDF=/tmp/input_file
+elif [ $MIME_TYPE = "application/vnd.openxmlformats-officedocument.wordprocessingml.document" ]; then
+    echo "Converting document to PDF"
+    CONVERT_MSGS=$(libreoffice --headless --convert-to pdf:writer_pdf_Export --outdir /tmp /tmp/input_file 2>&1)
+    if [ $? -ne 0 ]; then
+        die "Page $PAGE conversion failed: $CONVERT_MSGS"
+    fi
+    ORIGINAL_PDF=/tmp/input_file.pdf
+else
+    die "The document format is not supported"
+fi
 
 echo "Separating document into pages"
+pdfseparate $ORIGINAL_PDF /tmp/page-%d.pdf
 
-pdfseparate /tmp/input_file /tmp/page-%d.pdf
+IMG_DEPTH=8
-
 NUM_PAGES=$(find /tmp/page-*.pdf |wc -l)
 echo "Document has $NUM_PAGES pages"
 echo