GP32 Its All Upside Down And Back To Front?!?!

[Pirotic]

Hiya!

Working on adding the loading and saving into Gigas, finally figured out how to save and load which i was chuffed about, but now i notice that anything which was in array seems to be jumbled, as in it was written back to front.

for example, the map was stored like this

map[x][y] (x and y are both the same number, always)

but when i load it from the routine, i have to switch the x and y around as it seems to of loaded it backwards or something, this isnt a problem as both X and Y are always the same so i can just flip it around like so...

//Array to hold the current map layer in
unsigned short temp_map[map_settings.MAPW][map_settings.MAPH];

//i think maps have to be even shaped for this to work

//Load the map from the SMC
char filename[8];
int i;
for (i = 0; i < LAYERS; i++ ) {
sprintf((char*)filename, "M%dL%d.DAT", map_number, i);
GpFileOpen(filename, OPEN_R, &h_file );
        GpFileRead(h_file, temp_map, sizeof(temp_map),NULL);
GpFileClose( h_file );

  int x,y;
  for(x = 0; x < map_settings.MAPW; x++)
  {
  for(y = 0; y < map_settings.MAPH; y++)
  {
  map[x][y] = temp_map[y][x];
  }
  }

}

That all works, because like i said - both y and x go upto the same value.

However, im having problems loading arrays where the two values dont match, for example,

//Now load the entitys
//Emulate the .DAT file for entitys
short temp_entitys[100][16]; //was 100 16

//Load the entity information from SMC into that dat file
sprintf((char*)filename, "M%E.DAT", map_number);
GpFileOpen(filename, OPEN_R, &h_file );
        GpFileRead(h_file, temp_entitys, sizeof(temp_entitys),NULL);
GpFileClose( h_file );

//Apply the settings to the entitys
for(i = 0; i < (ENTITY_COUNT - 1); i++)
{

  //Load the info
   ENTITYS.x = temp_entitys[0];
   ENTITYS.y = temp_entitys[1];
   ENTITYS.enabled = temp_entitys[2];
   ENTITYS.ai = temp_entitys[3];
   ENTITYS.animation = temp_entitys[4];
   ENTITYS.attack_animation = temp_entitys[5];
   ENTITYS.state = temp_entitys[6];
   ENTITYS.team = temp_entitys[7];
   ENTITYS.STAT_BASE_HP = temp_entitys[8];
   ENTITYS.STAT_BASE_MP = temp_entitys[9];
   ENTITYS.STAT_BASE_ATK = temp_entitys[10];
   ENTITYS.STAT_BASE_DEF = temp_entitys[11];
   ENTITYS.STAT_EXP = temp_entitys[12];
   ENTITYS.STAT_EXP_GRAPH = temp_entitys[13];
   ENTITYS.speed = temp_entitys[14];
   ENTITYS.STAT_BASE_ATTACK_RECHARGE = temp_entitys[15];

} //end of entity loop

The problem here is, the array i wrote into the file was a short[100][16]. but because each [] doesn't store the same amount, i cannot just reverse it to put everything back in its correct place, any suggestions on how i can do this?

 

[jcbnetwork]

I think you should show us your SAVE function as well, maybe thats the one causing the problem.

 

[Pirotic]

	//save the entitys
	GpFileCreate("M1E.DAT", NOT_IF_EXIST, &h_file );
	GpFileOpen("M1E.DAT", OPEN_W, &h_file );
	GpFileWrite(h_file,map1_entitys,sizeof(map1_entitys));
	GpFileClose( h_file );

the Save function is nice and simple, the problem is the data always seems to get written/read back to front, as it has two arrays [100][16],

if it was [100][100] i could just do [x][y] = [y][x] but as both array has a different length i cannot figure out how to flip it back to its normal position i figure this must be a common problem and therefore has already been solved by somebody, somewhere.

 

[rcx21000]

Load into something like temp_arr[1600], then use a for loop to put it into the array.

short temp_arr[1600];
int x, y, onarr = 0;
/* File loading code (into temp_arr) */
for(x = 1; x < 16; x++) for(y = 1; y < 100; y++) {temp_entities[x-1][y-1] = temp_arr[onarr]; onarr++;}

If that still reverses, just change temp_entities[x-1][y-1] to temp_entities[y-1][x-1]. Theres prolly a better way, but that works

 

[Pirotic]

no luck with that either, to begin with i just tried to load it like that - but it just came out all '255's'

so i decided to make it convert the data into the one [1600] array then save it, then when it loads, to load back that [1600] array and peice of back together like you mentioned.

it generates a 200kb file for some reason - where as saving the arrays before only used 4kb had a little play and tried to fix it, but couldn't really see any problem other than your x and y's are back to front

it goes [100][16], but i fixed that and still no avail, even tried loading it backwards, so like:

data[x][y] = big_data_string[(1600 - 1) - onarr]

nothing seems to work, which is annoying as it should be such a simple task - i'll talk to you more when your on MSN mate and see if we can fix it.

 

[rcx21000]

Another idea: Someone told me that sizeof() didn't work well on multidementional arrays. What about trying: sizeof(short) * 1600 ?

 

[Pirotic]

Thanks to rcx21000 the problem is now solved, cheers everyone.