Getting alignement (row padding) from View

[sdebionne]

Getting back to binding IPP and other libraries to GIL, here is an example:

    gray8_image_t in(dim, dim);
    generate_pixels(view(in), [i = 0]() mutable -> std::uint8_t { return ++i; });
    gray8_image_t out(dim, dim);

    IppiSize srcRoi = { dim, dim };

    ippiMirror_8u_C1R(
        boost::gil::interleaved_view_get_raw_data(const_view(in)), (int) in.width(), // <- in.width() should be nb_bytes_per_row !
        boost::gil::interleaved_view_get_raw_data(view(out)), (int) out.width(),
        srcRoi,
        ippAxsVertical);

I wonder if there is way with a View model to get the number of bytes per row (depending on the alignement of the image)? I guess the underlying step_iterator has this information?

[sdebionne]

It looks like the answer is const_view(in).pixels().row_size()!

Digging into the code, I found out that image does not align rows but just the beginning of the image, is that correct?

[mloskot]

@sdebionne TBH, I can't remember. AFAIR, row_size used to be named row_bytes and the original release notes are still available at https://stlab.adobe.com/gil/gil2.1_changes.pdf - not very explanatory though.