[SVCS-528] [Blocked] Improve gdoc rendering - WB Part

tests/providers/googledrive/test_metadata.py

@@ -44,6 +44,7 @@ def test_file_metadata_drive(self, basepath, root_provider_fixtures):
         assert parsed.materialized_path == str(path)
         assert parsed.is_google_doc is False
         assert parsed.export_name == item['title']
+        assert parsed.alt_export_name == 'PART_1420130849837.pdf'
 
     def test_file_metadata_drive_slashes(self, basepath, root_provider_fixtures):
         item = root_provider_fixtures['file_forward_slash']
@@ -66,6 +67,7 @@ def test_file_metadata_drive_slashes(self, basepath, root_provider_fixtures):
         assert parsed.materialized_path == str(path)
         assert parsed.is_google_doc is False
         assert parsed.export_name == item['title']
+        assert parsed.alt_export_name == 'PART_1420130849837.pdf'
 
     def test_file_metadata_docs(self, basepath, root_provider_fixtures):
         item = root_provider_fixtures['docs_file_metadata']
@@ -80,6 +82,7 @@ def test_file_metadata_docs(self, basepath, root_provider_fixtures):
         }
         assert parsed.is_google_doc is True
         assert parsed.export_name == item['title'] + '.docx'
+        assert parsed.alt_export_name == 'version-test.pdf'
 
     def test_folder_metadata(self, root_provider_fixtures):
         item = root_provider_fixtures['folder_metadata']

tests/providers/googledrive/test_provider.py

@@ -588,6 +588,7 @@ class TestDownload:
         ds.DRIVE_IGNORE_VERSION)
 
     GDOC_EXPORT_MIME_TYPE = 'application/vnd.openxmlformats-officedocument.wordprocessingml.document'
+    GDOC_ALT_EXPORT_MIME_TYPE = 'application/pdf'
 
     @pytest.mark.asyncio
     @pytest.mark.aiohttpretty
@@ -619,6 +620,36 @@ async def test_download_editable_gdoc_no_revision(self, provider, sharing_fixtur
         assert aiohttpretty.has_call(method='GET', uri=revisions_url)
         assert aiohttpretty.has_call(method='GET', uri=download_file_url)
 
+    @pytest.mark.asyncio
+    @pytest.mark.aiohttpretty
+    async def test_download_editable_gdoc_as_mfr(self, provider, sharing_fixtures):
+        metadata_body = sharing_fixtures['editable_gdoc']['metadata']
+        path = GoogleDrivePath(
+            '/sharing/editable_gdoc',
+            _ids=['1', '2', metadata_body['id']]
+        )
+
+        metadata_query = provider._build_query(path.identifier)
+        metadata_url = provider.build_url('files', path.identifier)
+        aiohttpretty.register_json_uri('GET', metadata_url, body=metadata_body)
+
+        revisions_body = sharing_fixtures['editable_gdoc']['revisions']
+        revisions_url = provider.build_url('files', metadata_body['id'], 'revisions')
+        aiohttpretty.register_json_uri('GET', revisions_url, body=revisions_body)
+
+        file_content = b'we love you conrad'
+        download_file_url = metadata_body['exportLinks'][self.GDOC_ALT_EXPORT_MIME_TYPE]
+        aiohttpretty.register_uri('GET', download_file_url, body=file_content, auto_length=True)
+
+        result = await provider.download(path, mfr='true')
+        assert result.name == 'editable_gdoc.pdf'
+
+        content = await result.read()
+        assert content == file_content
+        assert aiohttpretty.has_call(method='GET', uri=metadata_url)
+        assert aiohttpretty.has_call(method='GET', uri=revisions_url)
+        assert aiohttpretty.has_call(method='GET', uri=download_file_url)
+
     @pytest.mark.asyncio
     @pytest.mark.aiohttpretty
     async def test_download_editable_gdoc_good_revision(self, provider, sharing_fixtures):
@@ -1577,6 +1608,19 @@ async def test_intra_copy_file(self, provider, root_provider_fixtures):
 
 class TestOperationsOrMisc:
 
+    def test_misc_utils(self):
+        metadata = {
+            'mimeType': 'application/vnd.google-apps.drawing',
+            'exportLinks': {
+                'image/jpeg': 'badurl.osf.899'
+            }
+        }
+        ext = drive_utils.get_alt_download_extension(metadata)
+        link = drive_utils.get_alt_export_link(metadata)
+
+        assert ext == '.jpg'
+        assert link == 'badurl.osf.899'
+
     @pytest.mark.asyncio
     @pytest.mark.aiohttpretty
     async def test_can_duplicate_names(self, provider):

waterbutler/providers/googledrive/metadata.py

@@ -136,6 +136,14 @@ def export_name(self):
             title += ext
         return title
 
+    @property
+    def alt_export_name(self):
+        title = self._file_title
+        if self.is_google_doc:
+            ext = utils.get_alt_download_extension(self.raw)
+            title += ext
+        return title
+
     @property
     def _file_title(self):
         return self.raw['title']

waterbutler/providers/googledrive/provider.py

@@ -234,9 +234,16 @@ async def download(self,  # type: ignore
 
         metadata = await self.metadata(path, revision=revision)
 
+        if kwargs.get('mfr', None) and kwargs['mfr'].lower() == 'true':
+            download_url = metadata.raw.get('downloadUrl') or drive_utils.get_alt_export_link(metadata.raw)
+            export_name = metadata.alt_export_name
+        else:
+            download_url = metadata.raw.get('downloadUrl') or drive_utils.get_export_link(metadata.raw)
+            export_name = metadata.export_name
+
         download_resp = await self.make_request(
             'GET',
-            metadata.raw.get('downloadUrl') or drive_utils.get_export_link(metadata.raw),  # type: ignore
+            download_url,
             range=range,
             expects=(200, 206),
             throws=exceptions.DownloadError,
@@ -251,7 +258,7 @@ async def download(self,  # type: ignore
         if download_resp.headers.get('Content-Type'):
             # TODO: Add these properties to base class officially, instead of as one-off
             stream.content_type = download_resp.headers['Content-Type']  # type: ignore
-        stream.name = metadata.export_name  # type: ignore
+        stream.name = export_name  # type: ignore
         return stream
 
     async def upload(self, stream, path: wb_path.WaterButlerPath, *args, **kwargs) \

waterbutler/providers/googledrive/utils.py

@@ -4,6 +4,8 @@
         'ext': '.gdoc',
         'download_ext': '.docx',
         'type': 'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
+        'alt_download_ext': '.pdf',
+        'alt_type': 'application/pdf',
     },
     {
         'mime_type': 'application/vnd.google-apps.drawing',
@@ -24,6 +26,7 @@
         'type': 'application/vnd.openxmlformats-officedocument.presentationml.presentation',
     },
 ]
+
 DOCS_DEFAULT_FORMAT = {
     'ext': '',
     'download_ext': '.pdf',
@@ -37,28 +40,42 @@ def is_docs_file(metadata):
 
 
 def get_mimetype_from_ext(ext):
-    for format in DOCS_FORMATS:
-        if format['ext'] == ext:
-            return format['mime_type']
+    for format_type in DOCS_FORMATS:
+        if format_type.get('ext') == ext:
+            return format_type.get('mime_type')
 
 
 def get_format(metadata):
-    for format in DOCS_FORMATS:
-        if format['mime_type'] == metadata['mimeType']:
-            return format
+    for format_type in DOCS_FORMATS:
+        if format_type.get('mime_type') == metadata.get('mimeType'):
+            return format_type
     return DOCS_DEFAULT_FORMAT
 
 
 def get_extension(metadata):
-    format = get_format(metadata)
-    return format['ext']
+    format_type = get_format(metadata)
+    return format_type.get('ext')
 
 
 def get_download_extension(metadata):
-    format = get_format(metadata)
-    return format['download_ext']
+    format_type = get_format(metadata)
+    return format_type.get('download_ext')
+
+
+def get_alt_download_extension(metadata):
+    format_type = get_format(metadata)
+    return format_type.get('alt_download_ext', None) or format_type.get('download_ext')
+
+
+def get_alt_export_link(metadata):
+    format_type = get_format(metadata)
+    export_links = metadata.get('exportLinks')
+    if format_type.get('alt_type'):
+        return export_links.get(format_type.get('alt_type'))
+    else:
+        return export_links.get(format_type.get('type'))
 
 
 def get_export_link(metadata):
-    format = get_format(metadata)
-    return metadata['exportLinks'][format['type']]
+    format_type = get_format(metadata)
+    return metadata.get('exportLinks').get(format_type.get('type'))

waterbutler/server/api/v1/provider/metadata.py

@@ -71,6 +71,7 @@ async def download_file(self):
             range=request_range,
             accept_url='direct' not in self.request.query_arguments,
             mode=self.get_query_argument('mode', default=None),
+            mfr=self.get_query_argument('mfr', default=None)
         )
 
         if isinstance(stream, str):