Closed sparr closed 1 year ago
I think it was my intention to deprecate the old syntax, so I've now completely removed the residual code!
The intended way to do this moving forward is with the backquote syntax 8`5
.
Could we get a mention in the docs, perhaps in a section dedicated to deprecated functionality, so that people will know how to update old code? This was nontrivial for me because I didn't know what the syntax I was seeing was supposed to do. I had to read the source and/or browse old versions.
On Sun, Feb 19, 2023, 23:58 Lorenzi @.***> wrote:
I think it was my intention to deprecate the old syntax, so I've now completely removed the residual code!
— Reply to this email directly, view it on GitHub https://github.com/hlorenzi/customasm/issues/158#issuecomment-1436157966, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAA5TU7KZZF6D22E3QC2EDTWYK6S5ANCNFSM6AAAAAAS3QIWLY . You are receiving this because you authored the thread.Message ID: @.***>
With v0.10 the literal
a'b
represented the constantb
output with a width ofa
bits, equivalent tob`a
. In 0.11.7, this syntax produces an error 'invalid digits'.The lowest level code for calculating widths and parsing
'
characters hasn't changed.'
is still a valid character inis_number_mid
and still recognized as the separator inparse_width
. I wasn't able to track down the change in behavior.Test case:
v0.10
outputs
0x05
v0.11.7
error
invalid digits
for the8'5
excerpt