Should __get_item__() be reverted since it is confusing as to what the i-th entry of a Cigar is?

[yfarjoun]

There was a lively discussion here #155 and elsewhere about the merits of having a __len__ method in Cigar since there are meny possible interpretations of length.

in the same vein, one could argue that there are many interpretations of cigar[i]:

• the I-th operator
• the i-th cigar-element
• the i-th character in the string representation
• the offset in the read of the i-th base
• the offset in the reference of the i-th base

Since we concluded that having len is a "footgun" and may cause more problems than it's worth, this issue hopes to embody the discussion for and against removing __get_item__ i.e. reverting #154

[yfarjoun]

I agree that if one can do cigar[i] it can be understood that len(cigar)>i so that imposes a relationship between whatever it is we are counting with i when we use it in cigar[i] and the meaning of len(cigar), i.e. the meaning of __len__().

If we accept vetoing __len__() due to possible confusion with other length-adjacent concepts, it seems reasonable to me to not have __get_item__ thereby forcing the user to access elements explicitly, and also helping the reader understand the code.