diff --git a/examples/run_with_formats.py b/examples/run_with_formats.py index 382a5eae..aa915578 100644 --- a/examples/run_with_formats.py +++ b/examples/run_with_formats.py @@ -25,6 +25,7 @@ input_paths = [ Path("tests/data/lorem_ipsum.docx"), Path("tests/data/powerpoint_sample.pptx"), Path("tests/data/2206.01062.pdf"), + # Path("tests/data/2305.03393v1-pg9-img.png"), ] input = DocumentConversionInput.from_paths(input_paths) @@ -35,6 +36,7 @@ input = DocumentConversionInput.from_paths(input_paths) doc_converter = DocumentConverter( # all of the below is optional, has internal defaults. formats=[ InputFormat.PDF, + # InputFormat.IMAGE, InputFormat.DOCX, ], # whitelist formats, other files are ignored. format_options={ diff --git a/tests/data/2305.03393v1-pg9-img.png b/tests/data/2305.03393v1-pg9-img.png new file mode 100644 index 00000000..a936b772 Binary files /dev/null and b/tests/data/2305.03393v1-pg9-img.png differ