use ruby-vips instead of mini-exiftools for low level image stuff, when possible

[ndushay]

Tony C. believes that ruby-vips gem will provide a bunch of the functionality available to us from miniexif-tool gem. https://github.com/libvips/ruby-vips

See also #53

Look for the code that uses exif and see what can be replaced. If all of it - yay! Make a PR. If some of it, confer with Tony C as to when we should favor one tool over the other.

Tony C. is the PO for this work, and this ticket should get Tony's approval, not Andrew's.

Note that autodetection of mime type could introduce errors that Image-Magick (or libvips?) could barf on, for example.

ALERT:

• this gem is also used by was-robot-suite, so libvips will need to go on was-robot VMs sul-dlss/was_robot_suite/blob/main/lib/was_seed_preassembly/thumbnail_generator_service.rb, for example.
  • see sul-dlss/was_robot_suite/issues/471 and sul-dlss/operations-tasks/issues/3071
• this gem is also in pre-assembly Gemfile, and you should see if it is used there ... I couldn't find it.

[ndushay]

The main places to benefit:

• color channels and other low level color stuff
• height and width

and other low level image technical metadata.

[ndushay]

to get height and width:

you can do something like this.

require 'ruby-vips' # and in Gemfile

        vips_image = Vips::Image.new_from_file(file)
        { height: vips_image.height, width: vips_image.width }

[ndushay]

Actions for this ticket:

• [ ] look for code that is about image processing that uses exif, and replace it with vips code. e.g.
  • lib/assembly-image/image.rb
  • lib/assembly-image/images.rb
  • (search on "exif"; leave mimetype stuff alone; use libvips for other stuff) exif-colorspace, exif-profile ...
• [ ] be sure there is good test code that confirms the correct image results for what is being replaced - talk to Tony C about this
• [ ] if test code is impossible, have Tony C vet the work and indicate in the skipped test how to test it manually
• [ ] ensure that circleci can install libvips with the dependencies needed for this to work in circleci. If not, see https://github.com/sul-dlss/assembly-image/issues/55#issuecomment-1178249472 about what I suggest for the test.

[ndushay]

Note: this is NOT blocked by libvips performance problems for common-accessioning; the assembly-image gem stands alone, and the main branch is already libvips; we revert to non-libvips for consumers via gem version.

Note further: for some changes, testing with consumers (common-accessioning and was-robot-suite for was seed thumbnail creation) may ultimately be warranted (as in on stage and qa ...)

[aaron-collier]

It is unclear to me if libvips is going to be useful in this gem. For instance https://github.com/sul-dlss/assembly-image/blob/main/spec/image_spec.rb#L76 is testing the output of exif.colorspace to be sRGB, and later on line 99 it's looking for Grayscale. This appears to be interpretation in libvips (I could be mistaken here) with outputs srgb (lowercase) and b-w instead of Grayscale.

irb(main):013:0> vimg = Vips::Image.new_from_file '/Users/amcollie/github/sul-dlss/assembly-image/spec/test_data/gray_gray_untagged.tif'
=> #<Image 43x36 uchar, 1 bands, b-w>
irb(main):014:0> vimg.interpretation
=> :"b-w"
irb(main):015:0> vimg = Vips::Image.new_from_file '/Users/amcollie/github/sul-dlss/assembly-image/spec/test_data/color_rgb_srgb.tif'
=> #<Image 43x36 uchar, 3 bands, srgb>
irb(main):016:0> vimg.interpretation
=> :srgb

This is an example where we can map the output possibly, but it's unclear if this is even the correct field to use for it.

I will hold on this until we can discuss between @ndushay and @calavano

[jcoyne]

I believe this is complete.