Hi @anitagraham,

They should mostly work. Perhaps you are missing some libraries?

For avif, heic, heif save to buffer you need heifsave_buffer, you can test for it like this:

$ vips heifsave_buffer
save image in HEIF format
usage:
   heifsave_buffer in [--option-name option-value ...]
where:
   in           - Image to save, input VipsImage
outputs:
   buffer       - Buffer to save to, output VipsBlob
optional arguments:
   Q            - Q factor, input gint
			default: 50
...

In 8.11, GIF save is done via magicksave_buffer. 8.12 has a new, dedicated GIF saver and should be out soon.

BMP is also via magicksave_buffer. Perhaps you have built libvips without imagemagick?

WEBP should work, perhaps you are missing the webp libraries?

JP2K should also work. Again, you need to configure with OpenJPEG.

I would check the summary table at the end of libvips configure. It sounds like a lot of things are not there.

I meant to say, libvips from 8.10 has a new low-level IO system based on source and target. These are generic read and write classes which can be connected to any source and destination.

The _buffer classes are tiny wrappers over source and target, so they ought to exist for every operation, with only a few exceptions.

The docs have some info on these classes in Ruby:

http://libvips.github.io/ruby-vips/Vips/Target.html

I made you a small test prog:

#!/usr/bin/ruby

require 'vips'

x = Vips::Image.new_from_file ARGV[0]

%w(dz hdr jpc jpt jp2 j2c j2k webp).each do |format|
  begin
    buf = x.write_to_buffer ".#{format}"
    puts "#{format}: written #{buf.length} bytes"
  rescue Vips::Error
    puts "#{format}: buffer write not supported"
  end
end

# imagemagick-based savers need to be told the for format to write twice :(
%w(bmp gif).each do |format|
  begin
    buf = x.write_to_buffer ".#{format}", format: format
    puts "#{format}: written #{buf.length} bytes"
  rescue Vips::Error
    puts "#{format}: buffer write not supported"
  end
end

# libheif buffer savers need to be told the format too
buf = x.write_to_buffer ".heif", compression: "hevc"
puts "heic: written #{buf.length} bytes"

buf = x.write_to_buffer ".heif", compression: "av1"
puts "avif: written #{buf.length} bytes"

I see:

$ ./anita.rb ~/pics/k2.jpg
dz: written 467759 bytes
hdr: written 9132971 bytes
jpc: written 274315 bytes
jpt: written 274315 bytes
jp2: written 274315 bytes
j2c: written 274315 bytes
j2k: written 274315 bytes
webp: written 122962 bytes
bmp: written 8913034 bytes
gif: written 1968473 bytes
heic: written 209233 bytes
avif: written 254465 bytes

The extra parameter for imagemagick is very unfortunate. That's with 8.11, 8.12 will fix the GIF one at least.

We ought to extend the target API a bit and fix all these stupid special cases.

Yes - running the test script from my home folder gets the expected results. I'll have to see what's happening in the test environment. Thanks.

Thanks John. Tests all passing now.

I appreciate the effort you put in - and have been putting in for a long time.
🥇

8.12 will have a fixed-up buffer save system. With this code:

%w(bmp gif heif avif ppm pfm pgm pbm).each do |format|
  buf = x.write_to_buffer ".#{format}"
  puts "#{format}: written #{buf.length} bytes"
end

I now see:

$ ./anita.rb ~/pics/k2.jpg
bmp: written 8913034 bytes
gif: written 1653970 bytes
heif: written 209233 bytes
avif: written 254465 bytes
ppm: written 8908859 bytes
pfm: written 35635258 bytes
pgm: written 2969659 bytes
pbm: written 372791 bytes

Thanks for suggesting this!