btscs(): Spell Counter with Panel Gaps

`btscs()` improperly generates its spell variable when a panel has gaps. It continues counting from the spell variable's previous value, instead of continuing the count _plus_ taking the gap into account.

For a quick toy example:
````r
library(pltesim)

dat <- data.frame(id=c(rep(1,5), rep(2,5)),
year=c(1950, 1951, 1953, 1954, 1955, seq(1950,1954,1)),
event=c(rep(0,10)),
stataBTSCS=c(1,2,4,5,6,1,2,3,4,5)
# ^ if no counter reset desired when gaps present (Stata BTSCS's behavior) -> what I'd expect
)
dat

btscs(dat, "event", "year", "id")
````