Fix double filtering in psevents when using -q option

[Esteban82]

Assisted by Qwen 3.7 Plus

Bug:

psevents applies the -q filter, and writes the selected data to temporary files. Afterwards, it calls internal modules to plot the symbols/labels from these temporary files. Because the -q option remained active, the internal modules applied the exact same filter a second time on the already-filtered temporary files.

Solution:

Implemented a local patch to temporarily suspend the -q option before calling the internal modules, and restore it immediately after.

Alternative solution:

I also considered modifying the global gmt_disable_bghio_opts() and gmt_reenable_bghio_opts() functions in gmt_io.c to include -q. However, this will affect dozens of other modules, and I'm not sure if the -q option should also be disabled for them. But maybe it should.

Test with:

I use -qi0:30: to plot a data point every 30.

gmt math -T0/360/1 T SIND = sin_curve.txt
cat << 'EOF' > main.sh
gmt begin
	gmt basemap -R0/360/-1.2/1.6 -JX22c/11.5c -X1c -Y1c -Bxa30g30f -By
	gmt events sin_curve.txt -i0,1,0 -T${MOVIE_FRAME} -Sc0.2c -Gblack -qi0:30:
gmt end
EOF
gmt movie main.sh -Chd -Tsin_curve.txt -Ntest -Ml,png -Zs

Closes #8460.

[joa-quim]

I'll be very extra cautious in using other models than Anthropic's (even Sonnet, though Sonnet 5 is a good surprise but previous one was moderately bad). GLM, for example, is ver acclaimed but my experience with it was bad. Have no experience with OpenAI top models though.
This fix seams simple so it should not harm.

[Esteban82]

I agree. I will ask Sonnet to review the patch before merging, just to get an additional set of eyes on it.

[Esteban82]

Based on Claude's suggestion:

• I initialized the values to false or 0.
• I added the {} when the settings are recovered so the if applies to all three statements instead of only the first one.