Add hwp hwpx support

hwp/hwpx has several custom MIME types

.hwp:
 - application/x-hwp
 - application/haansofthwp
 - application/vnd.hancom.hwp

.hwpx:
 - application/haansofthwpx
 - application/vnd.hancom.hwpx,
 - application/hwp+zip

Fixes #243

Dockerfile

@@ -14,7 +14,8 @@ RUN apk --no-cache -U upgrade && \
     poppler-data \
     python3 \
     py3-magic \
-    tesseract-ocr
+    tesseract-ocr \
+    font-noto-cjk
 
 # Download the trained models from the latest GitHub release of Tesseract, and
 # store them under /usr/share/tessdata. This is basically what distro packages

README.md

@@ -36,6 +36,7 @@ Dangerzone can convert these types of document into safe PDFs:
 - ODF Spreadsheet (`.ods`)
 - ODF Presentation (`.odp`)
 - ODF Graphics (`.odg`)
+- Hancom HWP (Hangul Word Processor) (`.hwp`, `.hwpx`)
 - Jpeg (`.jpg`, `.jpeg`)
 - GIF (`.gif`)
 - PNG (`.png`)

dangerzone/conversion/doc_to_pixels.py

@@ -77,6 +77,26 @@ class DocumentToPixels(DangerzoneConverter):
             "application/vnd.oasis.opendocument.text-template": {
                 "type": "libreoffice",
             },
+            # .hwp
+            "application/vnd.hancom.hwp": {
+                "type": "libreoffice",
+            },
+            "application/haansofthwp": {
+                "type": "libreoffice",
+            },
+            "application/x-hwp": {
+                "type": "libreoffice",
+            },
+            # .hwpx
+            "application/vnd.hancom.hwpx": {
+                "type": "libreoffice",
+            },
+            "application/haansofthwpx": {
+                "type": "libreoffice",
+            },
+            "application/hwp+zip": {
+                "type": "libreoffice",
+            },
             # At least .odt, .docx, .odg, .odp, .ods, and .pptx
             "application/zip": {
                 "type": "libreoffice",

dangerzone/gui/main_window.py

@@ -553,7 +553,7 @@ class DocSelectionWidget(QtWidgets.QWidget):
         self.file_dialog.setFileMode(QtWidgets.QFileDialog.ExistingFiles)
         self.file_dialog.setNameFilters(
             [
-                "Documents (*.pdf *.docx *.doc *.docm *.xlsx *.xls *.pptx *.ppt *.odt *.odg *.odp *.ods *.jpg *.jpeg *.gif *.png *.tif *.tiff)"
+                "Documents (*.pdf *.docx *.doc *.docm *.xlsx *.xls *.pptx *.ppt *.odt *.odg *.odp *.ods *.hwp *.hwpx *.jpg *.jpeg *.gif *.png *.tif *.tiff)"
             ]
         )
 

install/pyinstaller/pyinstaller.spec

@@ -143,6 +143,20 @@ if p == "Darwin":
                     "CFBundleTypeName": "ODF Graphics Document",
                     "CFBundleTypeRole": "Viewer",
                 },
+                {
+                    "CFBundleTypeExtensions": ["hwp", "hwpx"],
+                    "CFBundleTypeIconFile": "../macos/document.icns",
+                    "CFBundleTypeMIMETypes": [
+                        "application/vnd.hancom.hwp",
+                        "application/haansofthwp",
+                        "application/x-hwp",
+                        "application/vnd.hancom.hwpx",
+                        "application/haansofthwpx",
+                        "application/hwp+zip",
+                    ],
+                    "CFBundleTypeName": "Hancom Office Document",
+                    "CFBundleTypeRole": "Viewer",
+                },
                 {
                     "CFBundleTypeExtensions": ["jpg", "jpeg"],
                     "CFBundleTypeIconFile": "../macos/document.icns",