Add a function to do in-memory conversion #59
Conversation
|
Thank you for your contribution. I don't mind a rewrite of the cli at all. As per the other issue, do you think it's also needed to add |
|
I’m not sure the right one should be |
|
Looking at the docs https://doc.rust-lang.org/beta/reference/linkage.html again I think adding |
This lets us decouple file reading/writing from the actual conversion.
Instead the input_path and output_path have to be passed to convert_image_to_svg() manually.
linkmauve
force-pushed
the
in-memory-conversion
branch
from
78f283b
to
1361c50
Compare
|
Thanks for the I’ve pushed a new version, rebased on top of master, with the Python binding fixed to use the new API (the Python API stays identical). |
This lets the user convert an image from memory, without encoding it to PNG, writing it to a file, then reopening this file and decoding it using the image crate. It also allows the user to not write a SVG directly to a file.
linkmauve
force-pushed
the
in-memory-conversion
branch
from
1361c50
to
75e4b90
Compare
|
Thank you for the update. Allow me a bit of time to go through this ) |
|
I published on https://crates.io/crates/vtracer/0.6.3 |
This PR is best viewed commit per commit.
The goal is to make vtracer more usable as a library, in order to do so I have moved out path handling and clap parsing from the
Configstruct. This makes it much better than going through a PNG file, especially when the application already has pixels in memory.A future commit can make clap and image optional for building the library (in the latter case the
convert_image_to_svg()function won’t be available).This is a public API change, because
convert_image_to_svg()now takes the input and output paths as arguments, whereas before they were passed in theConfigstruct, so a minor version bump will have to follow for the next release.