Parsing code for GPS NMEA string - c

Question

i am trying to parse the incoming GPGGA NMEA GPS string using Arduino uno and below code.
What i am trying to do is that i am using only GPGGA NMEA string to get the values of Latitude, longitude and altitude.In my below code, i had put certain checks to check if incoming string is GPGGA or not, and then store the further string in a array which can be further parsed suing strtok function and all the 3 GPS coordinates can be easily find out.
But i am unable to figure out how to store only GPGGA string and not the further string.I am using a for loop but it isn't working.
I am not trying to use any library.I had came across certain existing codes like this.
Here is the GPGGA string information link
i am trying to have following functionlity
i) Check if incoming string is GPGGA
ii) If yes, then store the following string upto EOL or upto * (followed by checksum for the array) in a array, array length is variable(i am unable to find out solution for this)
iii) Then parse the stored array(this is done, i tried this with a different array)
#include <SoftwareSerial.h>
SoftwareSerial mySerial(10,11); // 10 RX / 11 TX
void setup()
{
Serial.begin(9600);
mySerial.begin(9600);
}
void loop()
{
uint8_t x;
char gpsdata[65];
if((mySerial.available()))
{
char c = mySerial.read();
if(c == '$')
{char c1 = mySerial.read();
if(c1 == 'G')
{char c2 = mySerial.read();
if(c2 == 'P')
{char c3 = mySerial.read();
if(c3 == 'G')
{char c4 = mySerial.read();
if(c4 == 'G')
{char c5 = mySerial.read();
if(c5 == 'A')
{for(x=0;x<65;x++)
{
gpsdata[x]=mySerial.read();
while (gpsdata[x] == '\r' || gpsdata[x] == '\n')
{
break;
}
}
}
else{
Serial.println("Not a GPGGA string");
}
}
}
}
}
}
}
Serial.println(gpsdata);
}
Edit 1:
Considering Joachim Pileborg, editing the for loop in the code.
I am adding a pic to show the undefined output of the code.
Input for the code:
$GPGGA,092750.000,5321.6802,N,00630.3372,W,1,8,1.03,61.7,M,55.2,M,,*76
$GPGSA,A,3,10,07,05,02,29,04,08,13,,,,,1.72,1.03,1.38*0A
$GPGSV,3,1,11,10,63,137,17,07,61,098,15,05,59,290,20,08,54,157,30*70
$GPGSV,3,2,11,02,39,223,19,13,28,070,17,26,23,252,,04,14,186,14*79
$GPGSV,3,3,11,29,09,301,24,16,09,020,,36,,,*76
$GPRMC,092750.000,A,5321.6802,N,00630.3372,W,0.02,31.66,280511,,,A*43
$GPGGA,092751.000,5321.6802,N,00630.3371,W,1,8,1.03,61.7,M,55.3,M,,*75
$GPGSA,A,3,10,07,05,02,29,04,08,13,,,,,1.72,1.03,1.38*0A
$GPGSV,3,1,11,10,63,137,17,07,61,098,15,05,59,290,20,08,54,157,30*70
$GPGSV,3,2,11,02,39,223,16,13,28,070,17,26,23,252,,04,14,186,15*77
$GPGSV,3,3,11,29,09,301,24,16,09,020,,36,,,*76
$GPRMC,092751.000,A,5321.6802,N,00630.3371,W,0.06,31.66,280511,,,A*45

Answer 1

After a quick check of the linked article on the NMEA 0183 protocol, this jumped out at me:
<CR><LF> ends the message.
This means, that instead of just read indiscriminately from the serial port, you should be looking for that sequence. If found, you should terminate the string, and break out of the loop.
Also, you might want to zero-initialize the data string to begin with, to easily see if there actually is any data in it to print (using e.g. strlen).

Answer 2

You could use some functions from the C library libnmea. Theres functions to split a sentence into values by comma and then parse them.

Answer 3

Offering this as a suggestion in support of what you are doing...
Would it not be useful to replace all of the nested if()s in your loop with something like:
EDIT added global string to copy myString into once captured
char globalString[100];//declare a global sufficiently large to hold you results
void loop()
{
int chars = mySerial.available();
int i;
char *myString;
if (chars>0)
{
myString = calloc(chars+1, sizeof(char));
for(i=0;i<chars;i++)
{
myString[i] = mySerial.read();
//test for EOF
if((myString[i] == '\n') ||(myString[i] == '\r'))
{
//pick this...
myString[i]=0;//strip carriage - return line feed(or skip)
//OR pick this... (one or the other. i.e.,I do not know the requirements for your string)
if(i<chars)
{
myString[i+1] = mySerial.read() //get remaining '\r' or '\n'
myString[i+2]=0;//add null term if necessary
}
break;
}
}
if(strstr(myString, "GPGGA") == NULL)
{
Serial.println("Not a GPGGA string");
//EDIT
strcpy(globalString, "");//if failed, do not want globalString populated
}
else
{ //EDIT
strcpy(globalString, myString);
}
}
//free(myString) //somewhere when you are done with it
}
Now, the return value from mySerial.available() tells you exactly how many bytes to read, you can read the entire buffer, and test for validity all in one.

Answer 4

I have a project that will need to pull the same information out of the same sentence.
I got this out of a log file
import serial
import time
ser = serial.Serial(1)
ser.read(1)
read_val = ("nothing")
gpsfile="gpscord.dat"
l=0
megabuffer=''
def buffThis(s):
global megabuffer
megabuffer +=s
def buffLines():
global megabuffer
megalist=megabuffer.splitlines()
megabuffer=megalist.pop()
return megalist
def readcom():
ser.write("ati")
time.sleep(3)
read_val = ser.read(size=500)
lines=read_val.split('\n')
for l in lines:
if l.startswith("$GPGGA"):
if l[:len(l)-3].endswith("*"):
outfile=open('gps.dat','w')
outfile.write(l.rstrip())
outfile.close()
readcom()
while 1==1:
readcom()
answer=raw_input('not looping , CTRL+C to abort')
The result is this:
gps.dat
$GPGGA,225714.656,5021.0474,N,00412.4420,W,0,00,50.0,0.0,M,18.0,M,0.0,0000*5B

Answer 5

Using "malloc" every single time you read a string is an enormous amount of computational overhead. (And didn't see the corresponding free() function call. Without that, you never get that memory back until program termination or system runs out of memory.) Just pick the size of the longest string you will ever need, add 10 to it, and declare that your string array size. Set once and done.
There are several C functions for getting substrings out of a string, strtok() using the coma is probably the least overhead.
You are on an embedded microcontroller. Keep it small, keep overhead down. :)

Answer 6

#include <stdio.h>
#include <string.h>
#define GNSS_HEADER_LENGTH 5
#define GNSS_PACKET_START '$'
#define GNSS_TOKEN_SEPARATOR ','
#define bool int
#define FALSE 0
#define TRUE 1
//To trim a string contains \r\n
void str_trim(char *str){
while(*str){
if(*str == '\r' || *str == '\n'){
*str = '\0';
}
str++;
}
}
/**
* To parse GNSS data by header and the index separated by comma
*
* $GPGSV,1,1,03,23,39,328,30,18,39,008,27,15,33,035,33,1*5A
* $GNRMC,170412.000,V,,,,,,,240322,,,N,V*2D
* $GNGGA,170412.000,,,,,0,0,,,M,,M,,*57
*
* #data_ptr the pointer points to gps data
* #header the header for parsing GPGSV
* #repeat_index the header may repeat for many lines
* so the header index is for identifying repeated header
* #token_index is the index of the parsing data separated by ","
* the start is 1
* #result to store the result of the parser input
*
* #result bool - parsed successfully
**/
bool parse_gnss_token(char *data_ptr, char *header, int repeat_index, int token_index, char *result) {
bool gnss_parsed_result = FALSE; // To check GNSS data parsing is success
bool on_header = FALSE;
// For header
int header_repeat_counter = 0;
int header_char_index = 0; // each char in header index
// For counting comma
int counted_token_index = 0;
// To hold the result character index
bool data_found = FALSE;
char *result_start = result;
char header_found[10];
while (*data_ptr) {
// 1. Packet start
if (*data_ptr == GNSS_PACKET_START) {
on_header = TRUE;
header_char_index = 0; // to index each character in header
data_found = FALSE; // is data part found
data_ptr++;
}
// 2. For header parsing
if (on_header) {
if (*data_ptr == GNSS_TOKEN_SEPARATOR || header_char_index >= GNSS_HEADER_LENGTH) {
on_header = FALSE;
} else {
header_found[header_char_index] = *data_ptr;
if (header_char_index == GNSS_HEADER_LENGTH - 1) { // Now Header found
header_found[header_char_index + 1] = '\0';
on_header = FALSE;
if (!strcmp(header, header_found)) {
// Some headers may repeat - to identify it set the repeat index
if (header_repeat_counter == repeat_index) {
//printf("Header: %s\r\n", header_found );
data_found = TRUE;
}
header_repeat_counter++;
}
}
header_char_index++;
}
}
// 3. data found
if (data_found) {
// To get the index data separated by comma
if (counted_token_index == token_index && *data_ptr != GNSS_TOKEN_SEPARATOR) {
// the data to parse
*result++ = *data_ptr;
gnss_parsed_result = TRUE;
}
if (*data_ptr == GNSS_TOKEN_SEPARATOR) { // if ,
counted_token_index++; // The comma counter for index
}
// Break if the counted_token_index(token_counter) greater than token_index(search_token)
if (counted_token_index > token_index) {
break;
}
}
// Appending \0 to the end
*result = '\0';
// To trim the data if ends with \r or \n
str_trim(result_start);
// Input data
data_ptr++;
}
return gnss_parsed_result;
}
int main()
{
char res[100];
char *nem = "\
$GNRMC,080817.000,A,0852.089246,N,07636.289920,E,0.00,139.61,270322,,,A,V*04\r\n\\r\n\
$GNGGA,080817.000,0852.089246,N,07636.289920,E,1,5,1.41,11.246,M,-93.835,M,,*5E\r\n\
$GNVTG,139.61,T,,M,0.00,N,0.00,K,A*2F\r\n\
$GNGSA,A,3,30,19,17,14,13,,,,,,,,1.72,1.41,0.98,1*0A\r\n\
$GNGSA,A,3,,,,,,,,,,,,,1.72,1.41,0.98,3*02\r\n\
$GNGSA,A,3,,,,,,,,,,,,,1.72,1.41,0.98,6*07\r\n\
$GPGSV,3,1,12,06,64,177,,30,60,138,15,19,51,322,18,17,42,356,27,1*68\r\n\
$GPGSV,3,2,12,14,36,033,17,07,34,142,17,13,32,267,17,02,21,208,,1*6C\r\n\
$GPGSV,3,3,12,15,05,286,,01,05,037,,03,03,083,,20,02,208,,1*6B\r\n\
$GAGSV,1,1,00,7*73\r\n\
$GIGSV,1,1,00,1*7D\r\n\
$GNGLL,0852.089246,N,07636.289920,E,080817.000,A,A*43\r\n\
$PQTMANTENNASTATUS,1,0,1*4F\r\n";
printf("Parsing GNRMC\r\n");
printf("===============\r\n");
for(int i=1;i<=16;i++){
parse_gnss_token(nem, "GNRMC", 0, i, res);
printf("Index: %d, Result: %s\r\n", i, res);
}
printf("Parsing GNVTG (First Parameter)\r\n");
printf("================================");
// GNVTG - Header, 0 - Repeat Index(if header is repeating), 1 - Value Index,
parse_gnss_token(nem, "GNVTG", 0, 1, res);
printf("\r\nGNVTG: %s\r\n", res);
return 0;
}