g.region: Fix the flat flag

[HuidaeCho]

This PR prints projection= and zone= in the same one line for the flat -f flag.

Before:

$ g.region -pgf
projection=3
zone=0
n=37.0625 s=31.375 w=-109.0625 e=-103 nsres=0.03125 ewres=0.03125 rows=182 cols=194 cells=35308 

After:

$ g.region -pgf
projection=3 zone=0 n=37.0625 s=31.375 w=-109.0625 e=-103 nsres=0.03125 ewres=0.03125 rows=182 cols=194 cells=35308 

[HuidaeCho]

Maybe later for a new release in a separate PR, it would be better to imply -g when -f is used. Currently, both -g and -f are needed and -f without -g has no effect (surprise!). Just using -f should replace -gf IMO, UNLESS -f flattening is also used for other formats.

Or it just needs a better description because other flags are also accumulative as well?

[petrasovaa]

There are print flags that add information (-l, -e, -c, -n, -3, -b). I don't think all of them are printed in flat format on 1 line with this PR yet.

I agree the -f flag should stand on its own in the same way you don't have to use -g flag together with -p flag. Ultimately we could introduce some kind of format method to include json format and get rid of all these flags?

There are some flags that don't seem to be behaving consistently, e.g. -t (no key printed), -w (always key=value format) and maybe some others.

[wenzeslaus]

+1 for making flat really one line. This strangeness was there for a while and it prevents use in shell scripting without addition of a grep filter.

A test for -gf and -f would be nice, e.g., testing that -f really yields one line only which was apparently missing. There is already general/g.region/testsuite/ so it should be relatively easy to add a new test.

[wenzeslaus]

There are print flags that add information (-l, -e, -c, -n, -3, -b). I don't think all of them are printed in flat format on 1 line with this PR yet.

I think that falls into the general revision of all output formats below.

I agree the -f flag should stand on its own in the same way you don't have to use -g flag together with -p flag.

I also expect that. -g usually means "key=value" line by line in other tools and there are no modifications to the format.

Ultimately we could introduce some kind of format method to include json format and get rid of all these flags?

That would be ideal.

[HuidaeCho]

There are print flags that add information (-l, -e, -c, -n, -3, -b). I don't think all of them are printed in flat format on 1 line with this PR yet.

Done.

[HuidaeCho]

There are print flags that add information (-l, -e, -c, -n, -3, -b). I don't think all of them are printed in flat format on 1 line with this PR yet.

I think that falls into the general revision of all output formats below.

I agree the -f flag should stand on its own in the same way you don't have to use -g flag together with -p flag.

I also expect that. -g usually means "key=value" line by line in other tools and there are no modifications to the format.

Ultimately we could introduce some kind of format method to include json format and get rid of all these flags?

That would be ideal.

Agreed, but I think that's for another PR.

[HuidaeCho]

There are some flags that don't seem to be behaving consistently, e.g. -t (no key printed), -w (always key=value format) and maybe some others.

Yep, GMT doesn't have a key. bbox= (already used with -w...) might be good? Or when -p prints the same info already, both -t and -w should be exclusive to -f? I think both -t and -w are meant to be standalone.

[petrasovaa]

I tested it and added a test for the f flag.

I tested the standalone -f flag, here is the code for it, but probably you are right this should be a separate PR.

diff --git a/general/g.region/main.c b/general/g.region/main.c
index 06c33a2e8c..33ad06f383 100644
--- a/general/g.region/main.c
+++ b/general/g.region/main.c
@@ -361,7 +361,7 @@ int main(int argc, char *argv[])
     G_option_required(
         flag.dflt, flag.savedefault, flag.print, flag.lprint, flag.eprint,
         flag.center, flag.gmt_style, flag.wms_style, flag.dist_res, flag.nangle,
-        flag.z, flag.bbox, flag.gprint, flag.res_set, flag.noupdate,
+        flag.z, flag.bbox, flag.gprint, flag.flprint, flag.res_set, flag.noupdate,
         parm.region, parm.raster, parm.raster3d, parm.vect, parm.north,
         parm.south, parm.east, parm.west, parm.top, parm.bottom, parm.rows,
         parm.cols, parm.res, parm.res3, parm.nsres, parm.ewres, parm.tbres,
@@ -371,7 +371,6 @@ int main(int argc, char *argv[])
                       flag.eprint, flag.center, flag.gmt_style, flag.wms_style,
                       flag.dist_res, flag.nangle, flag.z, flag.bbox,
                       flag.gprint, parm.save, NULL);
-    G_option_requires(flag.flprint, flag.gprint, NULL);
 
     if (G_parser(argc, argv))
         exit(EXIT_FAILURE);
@@ -383,7 +382,7 @@ int main(int argc, char *argv[])
     if (flag.print->answer)
         print_flag |= PRINT_REG;
 
-    if (flag.gprint->answer)
+    if (flag.gprint->answer || flat_flag)
         print_flag |= PRINT_SH;
 
     if (flag.lprint->answer)

[echoix]

I read the changes and they make sense for what it says it does ;)

[echoix]

Created issues to not forget about the other points, so this PR should be good to merge, it was already reviewed and approved, and I took another look at it.