Repository navigation
A few cleanups - #3
Conversation
ImageBase was used in ImageInTerminal to recursively downsample the image into a reasonable size. But for a pure encoding/decoding package, we don't need such heavy dependency. ColorTypes is sufficient since we only use the types for dispatch purposes.
Codecov Report
@@ Coverage Diff @@
## master #3 +/- ##
==========================================
- Coverage 98.09% 95.16% -2.94%
==========================================
Files 3 4 +1
Lines 105 124 +19
==========================================
+ Hits 103 118 +15
- Misses 2 6 +4
Continue to review full report at Codecov.
|
|
|
I'll continue or follow up your devs the next day, so let me merge this first. |
This commit introduces a few changes: - the 256 encoding algorithm is revisited to give a visually more consistant result for decoding. This changes the numerical results and is a breaking change. - 256 decoding is added - the TermColorDepth type becomes a functor -- both an encoder and a decoder.
also git ignore .vscode folder
2 participants
This commit picks the result from JuliaImages/ImageInTerminal.jl#62
(Not yet done) In the spirit of LazyModules, we might also want to make the final API depends on basic Base types:
Curiously, @t-bltg I noticed that you have explicitly changed a few
IOBufferusages toPipeBufferin many places, what's the benefit ofPipeBufferoverIOBuffer?