Compress images for PDF/EPUB

[Daltz333]

Should be a way to do this without running a script.

[calcmogul]

I can only find rtd rst docs on including images, not compressing them. Compression can either be done to the images before they are included or after by ghostscript or something. Ghostscript at least takes the dpi of the output into account. It's worth noting that no matter what you do, the screenshots stored in png format are going to have compression artifacts. svg is preferred.

[Daltz333]

After further review, this issue will inevitably conflict with #129. Compressing the images removes the ability to embed image meta data. There are also issues with being unable to determine which build was completed (HTML, latexpdf, epub), as well as Sphinx embedding fonts into the PDF.

I'll leave this issue open, but not a priority for a bit.

[Daltz333]

We may be able to check metadata of the image to verify if it's labview embeddable or not.

[TheTripleV]

# pip install pypng
import png

VI_CHUNK_TYPE = bytes("niVI", "utf-8")
def is_image_a_vi_snippet(image_path):
    with open(image_path, "rb") as image:
        reader = png.Reader(bytes=image.read())
    for chunk_type, chunk_data in reader.chunks():
        if chunk_type == VI_CHUNK_TYPE:
            return True
    return False

works for identifying if an image is a labview snippet.

From https://github.com/rcpacini/labview_snippet, vi snippets are stored in extra png chunks.