Parse frames-interp output by whitespace

[grahambates]

The current implementation uses character offsets from the header row, but this can break if long values push rows out of alignment e.g.

   LOC   CFA      r2    r3    r4    r13   r14   ra
00001d7c r15+4    u     u     u     u     u     c-4
00001d80 r13+8    u     u     u     c-8   u     c-4
00001d84 r13+8    c-1048 c-1044 c-1040 c-8   c-1036 c-4

Splitting on whitespace and using element indexes from the header row instead handles this case.

[grahambates]

Fixes #35