使用 C 在多个 CSV 文件之间读取、转换和写入数据

Question

I have 3 CSV files. The first of them contains 4 columns:

ID,Section1,Section2,Secion3
1,23,12,7
2,11,26,9
. . . .
. . . .
19,30,22,4
20,5,6,16

The first column is the ID and the other three contain random numbers ranging from 0 to 30.
The next file is a "conversion" file. It shows the corresponding value to each number:

30,45,44,45
29,44,42,43
28,43,42,41
. . . .
. . . .
1,22,21,22
0,20,21,21

The first column is the number to be read and the next columns are the values that will replace those numbers in each Section.
Like, if you read a 30 in Section1 it will be replaced by a 45, and if you find a 29 in Section 2 it will be replaced by a 42 and so on.
If I were to write the converted numbers into a third CSV file with the same 4 column format, how should I do it?
So far I've had no problems with generating the files since they are randomized, except for the "conversion" file, but I don't know how to proceed with the conversion.

#include <stdio.h> 
#include <stdlib.h>
#include <math.h>
#include <time.h>

int main (int argc,char *argv[]){
    int i, num1;
    char numbers[20][20];
    char conversions[20][20];

    //File with the initial numbers
    FILE *registry = fopen("registry.csv","w+");
    if (registry == NULL){
        fputs ("File error",stderr); 
        exit (1);
    }
    //Conversion file
    FILE *conversion = fopen("conversion.csv","r");
    if (conversion == NULL){
        fputs ("File error",stderr); 
        exit (1);
    }
    //File where the resulting values will be written
    FILE *results = fopen("results.csv","a");
    if (results == NULL){
        fputs ("File error",stderr); 
        exit (1);
    }

    //Writing the headers
    fprintf(results,"ID,Results1,Results2,Results3\n");
    fprintf(registry,"ID,Section1,Section2,Section3\n");
    srand(time(0));

    //Generating the random numbers
    for(i=1;i<21;i++){
        sprintf(numbers[i],"%d,%d,%d,%d\n",i,rand()%31,rand()%31,
            rand()%31);
        fputs(numbers[i], registry);
    }

    //This is where I don't know how to proceed

    for(i=1;i<21;i++){
        sprintf(conversions[i],"%d,%d,%d,%d\n",i,
            fscanf(registry,"%d,%*s,%*s,%*s\n",num1)...);


    }

I was trying to do something like what I did to generate the random numbers by saving everything into a buffer and then writing it into the file and I found that the fscanf function can be of great use to skip the parts I don't need to read, but I couldn't figure out how to use it to skip the headers and the ID column and I'm still missing the conversion part.
Conversion file:

30,45,44,45
29,44,43,43
28,43,42,42
27,43,41,40
26,41,40,40
25,40,39,39
24,39,37,39
23,38,37,38
22,38,36,37
21,36,36,37
20,35,35,36
19,34,34,35
18,33,34,35
17,33,33,34
16,32,32,33
15,32,31,32
14,31,30,31
13,30,30,31
12,29,29,30
11,29,28,29
10,28,28,29
9,28,27,27
8,27,26,26
7,27,26,25
6,26,25,25
5,25,24,24
4,25,23,24
3,24,23,23
2,23,22,21
1,21,21,20
0,20,21,20

Answer 1

This answer is scoped to address:
//This is where I don't know how to proceed .

To do what you have described, if it is not necessary to include the first columns in registry, results or conversion for storing, then do not. Although nice for human readability, they complicate the task of reading in, and using arrays. The unused column data and header row requires array indexing gymnastics to get the data to line up correctly into the arrays. Without the header row and numbering column, array indexes alone are sufficient to track all that is needed to make the conversion.
However , if you feel the first column/row is needed for your purposes, then adjust the array indexing of the routines shown below to accommodate.

The following steps (some of which you already do) are what I identified as essential toward making the conversion step as simple as possible...

Steps:

• Create registry - 60 - random numbers (0-30) arranged in 20 rows, 3 data columns

• Create int reg[20][3] = {{0}};

• Read registry into reg array using a combination fgets sscanf routine.

Example:(assuming registry has already been created in your code)

registry = fopen(".\\registry.csv","r");
if (registry)
{
    while(fgets(buf, sizeof(buf), registry))
    {
        sscanf(buf, "%d,%d,%d", &reg[index][0],&reg[index][1],&reg[index][2] );
        index++;
    }
    fclose(registry);
}

conversion table - pre-existing file

• Create int conv[31][3] = {{0}};

• Read conversion table into conv array using a combination fgets sscanf routine (see above example)

Once you have these arrays created properly,
Perform conversion:

• Create int res[20][3] = {{0}};

• Create char buf[80] = {0};

• Open 'results' file: FILE *results = fopen(".\\results.csv","w");
• populate results:

Example:

//File where the resulting values will be written
FILE *results = fopen(".\\results.csv","w");
if (results)
{
    for(i=0;i<sizeof(reg)/sizeof(reg[0]);i++)
    {
        for(j=0;j<sizeof(reg[0])/sizeof(reg[0][0]);j++)
        {
            res[i][j] = conv[reg[i][j]][j];
        //      |                         |  |___reg col   == conv col
        //      |                         |______value reg == conv row
        //      |________________________________row and column of conv
        }
        sprintf(buf, "%d,%d,%d\n", res[i][0],res[i][1],res[i][2]);
        fputs(buf, results);

    }
    fclose(results);    
}

At this point, you can print the results file in a format that suits your needs.

Following is an image showing actual results using suggested approaches above for a randomly generated registry , the values you provided for conversion and resulting results . (each created without column 1 shown in your original code, as I suggested in my comments.) Image is showing indexing from 0 - n for each set of array data: