Hello. Thank you for your recent PR approval. Today, I suggest that the value of least_score of GetNearNameColors be set to 0.5.

if least_score is 0.05, it means the colors suggested have more than 5 percent similarity to the target. Before, mplcolors -s greem shows below suggestions. I think the result is not natural, and that there are too many colors. For example, red is far away from greem.

Now, set its value 0.5 and mplcolors -s greem shows This result is natural. If least_score is bigger or small than 0.5, The suggestion colors are too many or a few; If 0.4: If 0.6:

Thus, least_score = 0.5 is better than least_score = 0.05 and may be best value. Another idea about this value is that user can change this value from some setting file or, more simply, from command line argument. In either case, I think that the default value of least_score should be set 0.5.

Also, this PR includes deleting print(suggestions) in mplcolors.py at line 258. suggestions is dict object. Therefore simple print is not user friendly. I guessed this print was for debugging, so I deleted it. If the result shows not only color names but also color codes, I think it is better to introduce another option argument and implement another function.

Thanks.

Hello. Your project mplcolors is really good and useful! And here is pull request about implementation to search color names. The features added in PR are below.

1. search color name (not color bar names...)

Search by whether the target name is included or not.

2. suggest color name if no color is found

If no color name is found in above method, suggest color names based on difflib package (since python 3.2).

If there is something wrong, please tell me it!