First-record scanning let trailing text be parsed as number

[PaulWessel]

The gmtio_examine_current_record failed to distinguish text from numbers.
In an example, the data record

0 2 -JX8cd/8c

ended up giving the (harmless but annoying) warning

-JX8c/8 not a valid number and may not be decoded properly

because gmt_scanf_arg was fooled. This tightens the check.

[joa-quim]

I'm still getting a

psrose [WARNING]: 12c/8 not a valid number and may not be decoded properly.

[PaulWessel]

Ok, which test?

[joa-quim]

It's a Julia one.

julia> data=[20 5.4 5.4 2.4 1.2; 40 2.2 2.2 0.8 0.7; 60 1.4 1.4 0.7 0.7; 80 1.1 1.1 0.6 0.6; 100 1.2 1.2 0.7 0.7; 120 2.6 2.2 1.2 0.7; 140 8.9 7.6 4.5 0.9; 160 10.6 9.3 5.4 1.1; 180 8.2 6.2 4.2 1.1; 200 4.9 4.1 2.5 1.5; 220 4 3.7 2.2 1.5; 240 3 3 1.7 1.5; 260 2.2 2.2 1.3 1.2; 280 2.1 2.1 1.4 1.3;; 300 2.5 2.5 1.4 1.2; 320 5.5 5.3 2.5 1.2; 340 17.3 15 8.8 1.4; 360 25 14.2 7.5 1.3];

julia> rose(data, yx=[], A=20, R="0/25/0/360", B="xa10g10 ya10g10 +t\"Sector Diagram\"", W=1, G="orange", F=1, D=1, S=4, Vd=1)
        psrose  -R0/25/0/360 -JX12c/8c -Bxa10g10 -Bya10g10 -B+t"Sector -BDiagram" -: -A20 -D1 -F1 -S4 -Gorange -W1 -P -K > C:\TMP\GMTjl_tmp.ps
psrose [WARNING]: 12c/8 not a valid number and may not be decoded properly.

BTW, this should be a INFO too, no?

julia> G = surface(rand(100,3) * 150, R="0/150/0/150", I=1, Ll=-100, upper=100);
surface [WARNING]: Your upper value is < than max data value.

[joa-quim]

and this

julia> gmtsimplify([0.0 0; 1.1 1.1; 2 2.2; 3.3 3], T="3k", Vd=1);
        gmtsimplify  -T3k
gmtsimplify [WARNING]: gmtsimplify only implemented using Flat-Earth calculations.

[PaulWessel]

OK, please fix - going on hike.

[PaulWessel]

Odd to give differernt height and width for a square (circle) plot. I will look. psrose was the last module to actually tolerate -J. It had its own -Sradius forever. I probably am taking some short-cut and did not anticipate your crazy -J argument.

[joa-quim]

That's automatic generated code. Must add a check to make it square by default for roses.

[PaulWessel]

OK, now that argument will cause an error on the psrose side (same for psternary)

[joa-quim]

Why forbid? If user want to deform just let him do that. And what if a subplot with rectangular subplots want to plot a rose? A warning is fine here, but erroring is is extreme.

[joa-quim]

Even if I use -JX12c/0, the complain is the same

julia> rose(data, yx=[], A=20, R="0/25/0/360", B="xa10g10 ya10g10 +t\"Sector Diagram\"", W=1, G="orange", F=1, D=1, S=4, Vd=1, show=1)
        psrose  -R0/25/0/360 -JX12c/0 -Bxa10g10 -Bya10g10 -B+t"Sector -BDiagram" -: -A20 -D1 -F1 -S4 -Gorange -W1 -P -K > C:\TMP\GMTjl_tmp.ps
psrose [WARNING]: 12c/0 not a valid number and may not be decoded properly.

[PaulWessel]

Please update.

[joa-quim]

It gives psrose -R0/25/0/360 -JX12c/8c -Bxa10g10 -Bya10g10 -B+t"Sector -BDiagram" -: -A20 -D1 -F1 -S4 -Gorange -W1 -P -K > C:\TMP\GMTjl_tmp.ps psrose [ERROR]: Option -J: Must specify -JX<diameter>[<unit>] From: Paul Wessel <[email protected]> Sent: Sunday, February 2, 2020 8:03 PM To: GenericMappingTools/gmt <[email protected]> Cc: Joaquim Manuel Freire Luís <[email protected]>; State change <[email protected]> Subject: Re: [GenericMappingTools/gmt] First-record scanning let trailing text be parsed as number (#2613) Please update. — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub<#2613?email_source=notifications&email_token=AAEDF2PMZUQFL4AXBIOHV2DRA4RFRA5CNFSM4KORQJZKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKR7RPI#issuecomment-581171389>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAEDF2NZ76OM5QUSGGZYI2LRA4RFRANCNFSM4KORQJZA>.

[PaulWessel]

Even if we allowed -JXa/b we are drawing the rose per polar coordinates using a radial scale. It makes absolutely no sense to give -JXa/b to prose and psternary. As for subplots, we are supposed to give a -JX? or -Jx? as the projection. Not sure if we have a test for this with rose and ternaries so that would be a good test.