parsing a text data as...

[AngKar]

Hi,

I've a data file as:

AB01 \t G01550 \t 4.0153- \t N001b \n
AB03 \t G01560 \t 4.0173+ \t N001w \n
.
.
.

I want to store only the numbers associated with G. I mean, I want to store 1550, 1560 ...
And also I want to store the associated + or -.

I'll be obliged if anyone could suggest some means to parse in this manner?

My algo may be :
After the 1st tab I'll reject G and store that in an id_array
And I'll also store the character before the 3rd tab in the sign_array...

Do I sound right? Also how can I implement this?

Thanks,
Angkar

[dwks]

Read in a buffer with fgets() and parse it (possibly with sscanf()) at will.

[AngKar]

Thanks a ton.
Now I've got G01550, G01560 how can I rip the G off and store the rest in an int array? Any functions?

[dwks]

You can get rid of the first character just by using &array[1] or memmove().

I would suggest sscanf() again.

[AngKar]

Thanks...

[King Mir]

Why use sscanf? Just use fscanf.

[dwks]

That's what I usually suggest, but someone usually corrects me and says "fgets()+sscanf() is better" . . .

I guess you can use fscanf() unless you need the buffer later.

[Salem]

> Why use sscanf? Just use fscanf.
Because with very little effort, fgets() + sscanf() allows you to recover from badly formatted files.

Using fscanf(), if the format of the file breaks down, you end up with a big problem of how to recover.