marcovde / warframemarketrelicocr Goto Github PK

This project uses: https://hub.docker.com/r/otiai10/ocrserver as an OCR server, it's based on gosseract, using the OCR server removes the need to install it for dev purposes. The last iteration might include adding gosseract to the final release, meanwhile this can be run on a server instead, it does not require a lot of compute overhead.

To check if someone already has an item: https://github.com/WFCD/warframe-items this is to allow user to know what items they already have to help with selection.

The value of items is from: https://warframe.market/ , use https://docs.google.com/document/d/1121cjBNN4BeZdMBGil6Qbuqse-sWpEXPpitQH5fb_Fo/edit#heading=h.irwashnbboeo which has an open api, please stay under 3 requests per second average, 4 once-off requests should be all right.

Probably going to use an OpenGL borderless implementation: https://github.com/go-gl/gl Might also look into whether fyne-io has support for it at the time: fyne-io/fyne#231 , then can use it for the user interface as well.

It is intended to use: https://github.com/kbinani/screenshot This screenshots the main display, further refinement can be done to isolate the image within percentage bounds. Also capturing the window rather than the entire screen.

Possible improvement is to ask user which screen to capture: https://github.com/kbinani/screenshot/blob/master/example/main.go since it seems one cannot specify the process to capture directly as this would be not be platform agnostic. Best option would be to use the Windows API for it.

Tests still have to be built, currently there is a sample file for testing within the assets folder.

Please use the same filename and append _test for test files.