The content is disordered when converting a PDF file.

[tosone]

Here, should be fixed. https://github.com/sajari/docconv/blob/master/pdf_ocr.go#L75C2-L92C3

We should also set a concurrency number because many images will make app OOM.