TkTextGetIndex is Erroneous

See the following example:

wish8.6 will produce the result "tTes", because it is interpreting the special index specifier end as a mark name. But this is wrong. The documentation says:

The base for an index must have one of the following forms:

line.char
@x,y
end
mark
tag.first
tag.last
pathName
imageName

If the base could match more than one of the above forms, such as a mark and imageName both having the same value, then the form earlier in the above list takes precedence. …

This means that the precedence of special index specifier end is higher than the precedence of a mark named "end", so the example must produce "Test", and the revised version is doing this. Furthermore the latter is the expected result when reading the script above. The consequence is that end is a reserved identifier and cannot be used as a mark name.

The same applies to the following example:

The right result is "1.0", because the form line.char has a higher precedence than any mark, but wish8.6 gives the wrong result "1.1".