Error with haishoku backend on wallpapers with limited colors

[serin113]

Sample image that triggers the error: https://i.imgur.com/1gKSbJ0.png

[I] image: Using image Untitled.png.
[I] colors: Generating a colorscheme.
[I] colors: Using haishoku backend.
Traceback (most recent call last):
  File "/usr/bin/wal", line 11, in <module>
    load_entry_point('pywal==3.3.0', 'console_scripts', 'wal')()
  File "/usr/lib/python3.7/site-packages/pywal/__main__.py", line 208, in main
    parse_args(parser)
  File "/usr/lib/python3.7/site-packages/pywal/__main__.py", line 164, in parse_args
    sat=args.saturate)
  File "/usr/lib/python3.7/site-packages/pywal/colors.py", line 144, in get
    colors = getattr(backend, "get")(img, light)
  File "/usr/lib/python3.7/site-packages/pywal/backends/haishoku.py", line 37, in get
    return adjust(cols, light)
  File "/usr/lib/python3.7/site-packages/pywal/backends/haishoku.py", line 31, in adjust
    return colors.generic_adjust(raw_colors, light)
  File "/usr/lib/python3.7/site-packages/pywal/colors.py", line 71, in generic_adjust
    colors[15] = colors[7]
IndexError: list assignment index out of range

[jakobkukla]

Same issue!

[Neightro]

If it helps, I've found another image that triggers this bug.

[PRESFIL]

This bug starts in this place (in current code base).

I tried to see why (with REPL):

from haishoku.haishoku import Haishoku
palette = Haishoku.getPalette("https://i.imgur.com/1gKSbJ0.png")
print(len(palette))
# 8

palette = Haishoku.getPalette("https://user-images.githubusercontent.com/23202923/128759243-fecd7209-fed4-44fb-8019-13d086579260.jpg")
print(len(palette))
# 7

For some reason, the length returned by haishoku may be 7, not 8.

A chain of calls led me to:

1. Haishoku.getPalette
2. -- Haishoku.getColorsMean

[khughitt]

Good find. Going a little deeper, in the call to getColorsMean, by the time the weighted mean colors are computed and the nested for loops have finished executing, the resulting colors_mean has only 7 entries:

# haishoku/haishoku/haishoku.py

  # get the weighted mean of all colors
  colors_mean = []
  for i in range(3):
      for j in range(3):
          for k in range(3):
              grouped_image_color = grouped_image_colors[i][j][k]
              if 0 != len(grouped_image_color):
                  color_mean = alg.get_weighted_mean(grouped_image_color)
                  colors_mean.append(color_mean)