charwidth of U+2a1d (join)

[stevengj]

I noticed in JuliaLang/julia#23820 that utf8proc returns a different character width for the join symbol ⨝ (U+2a1d) vs the bowtie symbol ⋈ (U+22c8), even though they seem to use an identical glyph in many (but not all!) fonts:

julia> charwidth('⋈') # bowtie U+22c8
1

julia> charwidth('⨝') # join   U+2A1D
2

I'm not sure why this happened — maybe GNU Unifont uses a different width for the two symbols? (cc @jiahao)

The EastAsianWidth-9.0.0.txt file for UAX#11 lists both of these as "neutral":

22C0..22FF;N     # Sm    [64] N-ARY LOGICAL AND..Z NOTATION BAG MEMBERSHIP
2A00..2AFF;N     # Sm   [256] N-ARY CIRCLED DOT OPERATOR..N-ARY WHITE VERTICAL BAR

If we treated neutral width as narrow (as suggested in #83), then both symbols would have width 1.