revisit "or" condition when fill or True is evaluated

[jbfaden]

I believe that True.or(fill) is fill right now, and this should be reconsidered. Sadie and I had a very strange bug where we couldn't figure out how some points were getting through, and it was because of this.

showBug= getParam('showBug',True,'Show the bug',[True,False])

dens= getDataSet('vap+cdaweb:ds=PSP_SWP_SPI_SF00_L3_MOM&filter=PSP&id=DENS&timerange=2023-03-13+through+2023-03-16')
ds=getDataSet('vap+cdaweb:ds=PSP_SWP_SPI_SF00_L3_MOM&filter=PSP&id=VEL_RTN_SUN&timerange=2023-03-13+through+2023-03-16')
ds = magnitude(ds)
ds = synchronize(dens,ds)

r= where( within( xtags(ds), '2023-03-13 13:30:30 to 13:35:00' ) )
ds= ds[r]
dens= dens[r]

if showBug:
    vals = where(dens.lt(10.0).or(invalid(dens)))
    ds = removeValues(ds,vals)
else:
    vals = where(dens.lt(10.0))
    ds = removeValues(ds,vals)
    vals= where(invalid(dens))
    ds = removeValues(ds,vals)

reset()
plot(0,ds,ytitle='ds',renderType='scatter>symbolSize=10')
plot(1,dens,ytitle='dens',renderType='scatter>symbolSize=10')

[jbfaden]

SetScriptDescription('This seems to contradict')

ds1= zeros(4)
ds1.putProperty(QDataSet.FILL_VALUE,0)

ds2= dataset([10,20,30,40])

print ds1[0],ds2[0]

r=where( invalid(ds1).or(ds2.gt(100) ) )
print r

r=where( ds2.gt(100).or(invalid(ds1)) )
print r

[jbfaden]

This code confirms that or(fill,one) is fill:

fill= dataset([Double.NaN])
one= dataset([1])

print fill[0], one[0], fill.or(one)[0], where( fill.or(one) )

which prints:

fill 1.0  fill dataset[0] (dimensionless)

[jbfaden]

I noticed the greaterOf function returns fill where one of its arguments contains fill. I started coding this and quickly realized to solve this correctly (where performance is not compromised) is not trivial.