TM typedef int16_t WTX_SolRad_t { text "5.1lf"; }
With the missing '%', this causes an assertion failure in gen_e_tcvt(), which is sort of a last ditch option in generate_calibrations(). This should probably have been headed off with an error in classify_conv() or even parse_txtfmt().
Has there ever been a reason to support anything except a %-coded format? In theory, it could be used to append a unit or prepend something, but that would interfere with many other things that assume the text conversion is a number.
TM typedef int16_t WTX_SolRad_t { text "5.1lf"; }
With the missing '%', this causes an assertion failure ingen_e_tcvt()
, which is sort of a last ditch option ingenerate_calibrations()
. This should probably have been headed off with an error inclassify_conv()
or evenparse_txtfmt()
.Has there ever been a reason to support anything except a %-coded format? In theory, it could be used to append a unit or prepend something, but that would interfere with many other things that assume the text conversion is a number.