https://docs.python.org/3/reference/lexical_analysis.html#indentation reads Point 1: "Tabs are replaced (from left to right) by one to eight spaces such that the total number of characters up to and including the replacement is a multiple of eight" and in the next paragraph Point 2: "Indentation is rejected as inconsistent if a source file mixes tabs and spaces in a way that makes the meaning dependent on the worth of a tab in spaces" In point 1, each tab has definitely a unique space counterpart, in point 2, tabs may have different space counterpart, which one is reliable ? The documentation should state that Point 1 concerns cPython, or at least indicate that the 8 may depend on the implementation, which then gives sense to point 2.
To be more precise, consider code def f(x): \tx=0 # 7 spaces + one tab return x # 8 spaces In cpython, both indentation levels are 8 and no indentation error is reported (this is the case where both tab size and alt tab size are equal) If instead of 8 the tab would count for 6 spaces, then we would have 12 and 8 as indentation level, resulting in a mismatch and an indentation error being reported, according to the documentation. This is inconsistent. Then either the documentation is faulty or cpython is. Actually, cpython accepts a mix of space and tabs only when tabs are in 8, 16, 24... positions.
Reading from the source code (Parser/tokenizer.c from line 1364 as in Python 3.10), I derive these as the actual indentation rules: – Tab makes the indentation amount the next multiple of 8. – Among lines with the same indentation amount in the same parent line, the total number of tabs and spaces must match, in other words, the tabs' total padded width must be equal. – If it's an indent, the total number of spaces and tabs must exceed that of the parent line.
