Converting character array to Floating Point with 2 Decimal Places - c

Question

I have defined a character array like this.
char pkgamount[20]="";
This can be populated by values like 120, -120, -120.23 etc (Up to 2 decimal places)
I currently convert the char array to integer like this.
int amt2;
sscanf(pkgamount, "%d", &amt2);
What is the proper way to convert the character array that can containing negative decimals to a float value with only 2 decimal places?
UPDATE: #1
I read values from a database (VARCHAR) to the character array.These value will have a maximum of 2 decimal places.The value can be 0/With 2 decimal places/with no decimal places at all. I think integer data type cannot have decimal places. I need to add/subtract these values.So I need to convert the character array to float.
UPDATE #2
I tried using float and I keep getting some random output.Please the extracted code below
float bamt1,bamt2,amt1,amt2;
float balance1,balance2;
sscanf(stbamount, "%f", &amt1);
sscanf(pkgamount, "%f", &amt2);
//amoun1=32 amount2=23 any decimal or non decimal values..using API
//Using API on device to read this..skipping the code for that
sscanf(amount1, "%f", &bamt1);
sscanf(amount2, "%f", &bamt2);
balance1=amt1-bamt1;
balance2=amt2-bamt2;
// buffers used for conversion and printing
char buf1[sizeof(int)*3+2]="";
snprintf(buf1, sizeof buf1, "%f", balance1);
char buf2[sizeof(int)*3+2]="";
char sbal[20]="";
char pbal[20]="";
strcat(sbal,"STB Balance: ");
strcat(pbal,"PKG Balance: ");
strcat(sbal,buf1);
strcat(pbal,buf2);
//prn_write_text is API Call to run thermal printer on device
prn_write_text(sbal,strlen(sbal),1);
prn_write_text(pbal,strlen(pbal),1);
snprintf(buf2, sizeof buf2, "%f", balance2);
UPDATE #3
This is the API Call to use the thermal printer to write some text. I need to convert the floating point value to a string and cut it, so that it includes only 2 decimal values.

Answer 1

Using floats to store the numbers..
float amtf;
sscanf(pkgamount, "%f", &amtf);
To print the numbers to 2 decimal places
printf (".2%f",amtf);
You can use floats to store the numbers as long as you are willing to tolerate errors of possibly 0.01 during addition or subtraction. It makes the code easier.
However, if you absolutely have to be accurate down to the last 0.00 then you should use integers * 100 to store the data, as suggested by M Oehem.

Answer 2

Just loop through your characters
enum parseFloatErrorCodes {
SUCCESS, DOUBLE_DOT,PARSE_ERROR, INTEGER_OVERFLOW,
TOO_MANY_DIGITS_AFTER_DECIMAL_POINT
};
int parse_float(char* inp, float *x) {
char *p;
int ival, dot;
for (ival = 0, dot = 0, p = inp; *p; ++p) {
int overflow_check = ival;
if (('0' <= *p) && (*p <= '9')) {
ival = ival * 10 + (int)(*p - '0');
if (dot > 0) ++dot;
} else if (*p == '.') {
if (dot > 0) { return DOUBLE_DOT; }
dot = 1;
} else { return PARSE_ERROR; }
if (overflow_check > ival) { return INTEGER_OVERFLOW; }
}
*x = ival;
if (dot == 0) *x *= 100.0;
else if (dot == 1) *x *= 100.0;
else if (dot == 2) *x *= 10.0;
else if (dot > 3) { return TOO_MANY_DIGITS_AFTER_DECIMAL_POINT; }
*x /= 100.0;
return SUCCESS;
}
See the gist for test code that works.

Answer 3

UPDATE #2 writes outside the the character array. --> UB
Assume sizeof(int)== 4 or 8.
// buf1[14] or buf1[26]
char buf1[sizeof(int)*3+2]="";
// the string length of buf1 is at least 8 and likely 10+
snprintf(buf1, sizeof buf1, "%f", balance1);
char sbal[20]="";
strcat(sbal,"STB Balance: "); // 14 char used
// attempts to write at least 14+8 = 22 char, sbal too small.
strcat(sbal,buf1);
Perhaps OP wants
#include <float.h>
// - Digits in FLT_MAX .xx \0
#define FLT_SIZE_2_DP (1 + (FLT_MAX_10_EXP + 1) + (1 + 2) + 1
char sbal[sizeof("STB Balance: ") + FLT_SIZE_2_DP];
int len = snprintf(sbal, sizeof sbal, "STB Balance: %.2f", balance1);
if (len >= 0 && len < sizeof sbal) {
prn_write_text(sbal,strlen(sbal),1);
else {
prn_write_text("Error",5,1);
}
What is the proper way to convert the character array that can containing negative decimals to a float value with only 2 decimal places?
When code has a special input require, useful to create a helper function to meet the need. The below reads 2 integers and then combines them into a float.
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
// EOF: end-of-file or IO error
// 0: no conversion
// 1: success
int read_float_2dp(FILE *istream, float *f) {
intmax_t ipart;
char fpart_str[3];
int count = fscanf(istream, "%jd.%2[0-9]", &ipart, fpart_str);
if (count <= 0) return count;
int fpart = (count == 2) ? atoi(fpart_str) : 0;
if (ipart < 0) fpart = -fpart;
*f = ipart + fpart / 100.0f;
return 1;
}

Answer 4

char pkgamount[10],pkgamount1[10];
char something[][3] = {"1.3","53","123","569"};
int amt2;
float amt1;
sscanf(pkgamount,"%2d",&amt2);
sscanf(pkgamount1,"%2.2f",&am1);
the %2d will only scan upto decimal values, same with the float 2 before decimal point and 2 after...Now the problem is i dont know exactly what yu are expecting, thats the reason i defined the something[][3] 2D array, cause thats how yu store values which means something[0][3] something[1][3] something[2][3] so on and so forth that in memory means that array something in place 0 can hold 3 chars, on place 1 can hold 3.....Now its upto yu how exactly yu use a loop to take in values... I can give yu an example..
char something[10][3];
int i;
for (1..10)
{
scanf("%2s",&something[i]);
}
printf("\nSomething array");
for (i..10)
{
printf("\n%2s",something[i]);
}