EasyCSV/src/easycsv.c

825 lines
20 KiB
C

#include "easycsv.h"
/* Flags */
typedef enum {
EASYCSV_NONE,
EASYCSV_STRING,
EASYCSV_INT,
EASYCSV_UNKNOWN
} _EASYCSV_TYPE;
typedef enum {
/* Generic errors */
EASYCSV_NOERROR = 0, // no error
EASYCSV_NULLCSV, // easycsv* is NULL
EASYCSV_NULLPTR, // generic pointer is NULL
EASYCSV_EMPTYSTRING, // empty string
EASYCSV_OVERMAXROW, // int exceeds row limit
EASYCSV_OVERMAXCOL, // int exceeds col limit
EASYCSV_ZEROROW,
EASYCSV_ZEROCOL,
/* File input/output errors */
EASYCSV_UNKNOWNMODE, // unknown file IO mode
EASYCSV_OPENFAIL, // fail to open
EASYCSV_REOPENFAIL, // fail to reopen
EASYCSV_EMPTYCSV, // csv is empty or does not exist
EASYCSV_UNWRITABLE, // csv mode not at EASYCSV_W
EASYCSV_UNREADABLE, // csv mode not at EASYCSV_R
EASYCSV_UPDATEFAIL, // update file fail
EASYCSV_FILEPTRFAIL, // move file pointer fail
/* Non-existant elements */
EASYCSV_ROWNOTEXIST, // row does not exist
EASYCSV_COLNOTEXIST, // column does not exist
/* Subroutine failure */
EASYCSV_PUSHCOLFAIL, // push column value fail
EASYCSV_COLNUMFAIL // column number retrieval fail
} EASYCSV_ERROR;
typedef struct _easycsv {
FILE *file; // original CSV file
FILE *temp; // temporary CSV file for writing
char *fp;
char *tmpfp;
unsigned int rows;
unsigned int cols;
EASYCSV_ERRORMSG error;
#ifdef EASYCSV_DEBUG
clock_t start;
#endif
} _easycsv;
/* PRIVATE FUNCTIONS */
/* Constructor of private members */
static _easycsv *_easycsv_priv_init(const char*, const EASYCSV_MODE, const EASYCSV_ERRORMSG);
/* Destructor of private members */
static void _easycsv_priv_free(_easycsv*);
/* Verifies mode of file */
// static int _easycsv_checkmode(struct easycsv*, const char);
/* Copies data from temp FILE to file FILE */
static int _easycsv_update(easycsv*);
/* Rewind easycsv, checks for readability */
static int _easycsv_rewind(_easycsv*, const EASYCSV_MODE);
/* Returns string of a specific row */
static char *_easycsv_getrow(const easycsv*, const unsigned int);
/* Return column number of a named column */
static int _easycsv_getcolumn(const easycsv*, const char*);
/* Verifies the type of the value (eg: string or int) */
static int _easycsv_checktype(const easycsv*, const int, const int);
/* Verifies if there is a value or not */
static int _easycsv_checkifvalue(const easycsv*, const int, const int);
/* */
// static char *_easycsv_getvalueinrow(const int, const char*);
/* Calculate rows */
static int _easycsv_rows(_easycsv*, const EASYCSV_MODE);
/* Calculate columns*/
static int _easycsv_columns(_easycsv*, const EASYCSV_MODE);
static void _easycsv_printerror(const _easycsv*, const EASYCSV_ERROR);
/* Print error from _easycsv struct in stderr */
static void _easycsv_geterror(const _easycsv*, const EASYCSV_ERROR, FILE*);
/* Check if easycsv* and const char* are NULL */
static int _easycsv_checkcsvandstring_one(const easycsv*, const char*);
/* Check if easycsv* and two const char* are NULL*/
static int _easycsv_checkcsvandstring_two(const easycsv*, const char*, const char*);
/* Verifies if the string is not NULL or empty, returns 0 on success and -1 on failure */
// static int _easycsv_checkstring(const char*);
/* Verifies if easycsv is not NULL or unallocated */
// static int _easycsv_checkeasycsv(const struct easycsv*);
/* Verifies if int is */
// static int _easycsv_checkunsigned(const int);
/*
* ====================
* FUNCTION DEFINITIONS
* ====================
*/
/* PUBLIC FUNCTIONS */
easycsv*
easycsv_init(const char *fp, const EASYCSV_MODE mode)
{
return easycsv_init_errormsg(fp, mode, EASYCSV_ERRORSTDERR);
}
easycsv*
easycsv_init_errormsg(const char *fp, const EASYCSV_MODE mode, const EASYCSV_ERRORMSG error)
{
#ifdef EASYCSV_DEBUG
csv->start = clock();
#endif
easycsv *csv = malloc(sizeof *csv);
csv->mode = mode;
csv->_priv = _easycsv_priv_init(fp, mode, error);
if (csv->_priv == NULL) {
easycsv_free(csv);
return NULL;
}
return csv;
}
void
easycsv_free(easycsv *csv)
{
if (csv == NULL) return;
_easycsv_priv_free(csv->_priv);
free(csv);
}
_easycsv*
_easycsv_priv_init(const char *fp, const EASYCSV_MODE mode, const EASYCSV_ERRORMSG error)
{
_easycsv *_priv = malloc(sizeof *_priv);
memset(_priv, 0, sizeof(*_priv)); // set all members to NULL
_priv->error = error;
/* Open file according to mode */
switch (mode) {
case EASYCSV_R:
case EASYCSV_W: {
_priv->file = fopen(fp, "r");
break;
}
default: {
_easycsv_printerror(_priv, EASYCSV_UNKNOWNMODE);
_easycsv_priv_free(_priv);
return NULL;
}
if (_priv->file == NULL)
{
_easycsv_printerror(_priv, EASYCSV_OPENFAIL);
_easycsv_priv_free(_priv);
return NULL;
}
size_t stfp = strlen(fp);
/* Allocate memory for char* */
_priv->fp = malloc(stfp + 1); // + 1 for null
strncpy(_priv->fp, fp, stfp);
/* Calculate rows and cols if file exists */
if (access(_priv->fp, F_OK) == 0) {
_priv->rows = _easycsv_rows(_priv, mode);
_priv->cols = _easycsv_columns(_priv, mode);
}
else {
_priv->rows = 0;
_priv->cols = 0;
}
if (mode == EASYCSV_W) {
/* csv->tmpfp = malloc(16 + stfp + 1); */
/* strncpy(csv->tmpfp, prefix, 16); */
/* strncat(csv->tmpfp, fp, stfp); */
do {
/* Write to temporary file */
unsigned int i = 1;
char buffer[21] = "/tmp/easycsv-"; // 13 char, 3 for digits, 4 for .csv, 1 for NULL
char *pbuffer = &buffer[13]; // don't free, is used for digit con
if (i < 100) {
strncpy(pbuffer, "0", 1);
pbuffer++;
}
if (i < 10) {
strncpy(pbuffer, "0", 1);
pbuffer++;
}
sprintf(pbuffer, "%i", i);
strncpy(++pbuffer, ".csv", 4);
buffer[20] = '\0';
if (access(buffer, F_OK) < 0) {
i++;
}
else {
_priv->tmpfp = malloc(21);
strncpy(_priv->tmpfp, buffer, 21);
break;
}
} while (1);
_priv->temp = fopen(_priv->tmpfp, "w");
if (_priv->temp == NULL) {
_easycsv_printerror(_priv, EASYCSV_OPENFAIL);
_easycsv_priv_free(_priv);
return NULL;
}
/*
#ifdef EASYCSV_DEBUG
printf("[%i] easycsv_debug: temp file %s opened\n", clock() - csv->start, csv->tmpfp);
#endif*/
/*
if (freopen(csv->fp, csv->mode, csv->file) == NULL) {
fprintf(stderr, "easycsv: failed to set temporary file %s to %s mode\n", csv->tmpfp, csv->mode);
return NULL;
}
if (freopen(csv->tmpfp, csv->mode, csv->temp) == NULL) {
fprintf(stderr, "easycsv: failed to set temporary file %s to %s mode\n", csv->tmpfp, csv->mode);
return NULL;
}
*/
}
return _priv;
}
void
_easycsv_priv_free(_easycsv *_priv)
{
if (_priv == NULL) return;
if (_priv->file != NULL) fclose(_priv->file);
if (_priv->temp != NULL) fclose(_priv->temp);
if (_priv->fp != NULL) free(_priv->fp);
if (_priv->tmpfp != NULL) free(_priv->tmpfp);
free(_priv);
}
int
easycsv_pushcolumn(easycsv *csv, const char *col)
{
if (_easycsv_checkcsvandstring_one(csv, col) < 0) return -1;
if (csv->mode == EASYCSV_R) {
_easycsv_printerror(csv->_priv, EASYCSV_UNWRITABLE);
return -1;
}
if ((csv->mode == EASYCSV_W) && (access(csv->_priv->fp, F_OK) == 0)) {
/* If the file doesn't exist, just put the col in the file */
if (fputs(col, csv->_priv->file) < 0) {
_easycsv_printerror(csv->_priv, EASYCSV_PUSHCOLFAIL);
return -1;
}
return 0;
}
char *str = _easycsv_getrow(csv, 1);
size_t ststr = strlen(str);
size_t stcol = strlen(col);
realloc(str, ststr + stcol + 2); // 1 for null and 1 for comma
strncat(str + ststr, ",", 1);
strncat(str + ststr + 1, col, stcol);
/*
#ifdef EASYCSV_DEBUG
printf("[%i] easycsv_debug: push column str: %s", clock() - csv->_priv->start, str);
#endif
*/
/* Put str in temp file */
if (fputs(str, csv->_priv->temp) < 0) {
_easycsv_printerror(csv->_priv, EASYCSV_PUSHCOLFAIL);
return -1;
}
free(str);
/* Copy every row following the first into temp */
for (int i = 2; i <= csv->_priv->rows; i++) {
char *row = _easycsv_getrow(csv, i);
fputs(row, csv->_priv->temp);
free(row);
}
/* Update the file */
if (_easycsv_update(csv) < 0) {
_easycsv_printerror(csv->_priv, EASYCSV_UPDATEFAIL);
return -1;
}
return 0;
}
int
easycsv_pushcolumnvalue(const easycsv *csv, const char *col, const char *val)
{
if (_easycsv_checkcsvandstring_two(csv, col, val) < 0) return -1;
int colnum = _easycsv_getcolumn(csv, col);
if (colnum < 0) {
_easycsv_printerror(csv->_priv, EASYCSV_COLNUMFAIL);
return -1;
}
/* Find a free cell under col within csv->_priv->cols limits, if there is none, generate a new
row */
unsigned int row;
for (row = 2; row <= csv->_priv->cols; row++) {
if (easycsv_readvalue(csv, colnum, row) != NULL) break;
}
/* All rows are filled, generate new row */
if (row == csv->_priv->cols) {
row++;
/* Set file pointer to end */
if (fseek(csv->_priv->file, 0L, SEEK_END) < 0) {
_easycsv_printerror(csv->_priv, EASYCSV_FILEPTRFAIL);
return -1;
}
char *rowstr = malloc(csv->_priv->cols + strlen(val) + 2); // 1 extra for '\n'
strncpy(rowstr, "\n", 1);
for (unsigned int i = 1; i < colnum; i++) {
strncpy(rowstr, ",", 1);
}
strncpy(rowstr, val, strlen(val));
unsigned int colsleft = csv->_priv->cols - colnum;
for (unsigned i = 1; i < colsleft; i++) {
strncpy(rowstr, ",", 1);
}
strncpy(rowstr, "/0", 1);
csv->_priv->rows++;
}
/* Fill in rows before */
for (unsigned int i = 1; i < row; i++) {
char *str = _easycsv_getrow(csv, i);
if (fputs(str, csv->_priv->temp) == EOF) {
fprintf(stderr, "easycsv: failed to write \"%s\" in temporary file %s for %s\n", str, csv->_priv->tmpfp, csv->_priv->fp);
return -1;
}
free(str);
}
return 0;
}
int
easycsv_insertcolumnvalue(const easycsv *csv, const char *col, const unsigned int row, const char *val, const EASYCSV_VALMODE valmode)
{
if (_easycsv_checkcsvandstring_two(csv, col, val) < 0) return -1;
int colnum = _easycsv_getcolumn(csv, col);
if (colnum < 0) {
_easycsv_printerror(csv->_priv, EASYCSV_COLNUMFAIL);
return -1;
}
char *rowstr = _easycsv_getrow(csv, row);
char *pch = rowstr;
for (unsigned int i = 1; i < colnum; i++) {
pch = strchr(pch, ',');
pch++;
}
size_t st = strcspn(pch, ',');
/* Empty cell */
if (st == 0) {
}
return 0;
}
int
easycsv_printrows(const easycsv *csv)
{
return csv->_priv->rows;
}
int
easycsv_printcolumns(const easycsv *csv)
{
return csv->_priv->cols;
}
int
_easycsv_rows(_easycsv *_priv, const EASYCSV_MODE mode)
{
// no need to check _priv for NULL
if (_easycsv_rewind(_priv, mode) < 0) return -1;
int rows = 1;
char c;
/* Go through each character in the file and count the number of \n
in it */
while ((c = fgetc(_priv->file)) != EOF) {
if (c == '\n') rows++;
}
return rows;
}
int
_easycsv_columns(_easycsv *_priv, const EASYCSV_MODE mode)
{
// no need to check _priv for NULL
/* Prepare it for reading */
if (_easycsv_rewind(_priv, mode) < 0) return -1;
/*
1. check if empty file
2. check if only one column -> 0 commas
3. if >1 column, n commas = n+1 columns
*/
unsigned int col = 1;
char c;
while ((c = fgetc(_priv->file)) != '\n') {
if (c == ',') col++;
}
return col;
}
const char*
easycsv_readcolumnvalue(const easycsv *csv, const char *col, const unsigned int row)
{
if (_easycsv_checkcsvandstring_one(csv, col) < 0) return NULL;
if (row > csv->_priv->rows) {
_easycsv_printerror(csv->_priv, EASYCSV_OVERMAXROW);
return NULL;
}
int i = _easycsv_getcolumn(csv, col);
if (i < 1) {
_easycsv_printerror(csv->_priv, EASYCSV_COLNUMFAIL);
return NULL;
}
return easycsv_readvalue(csv, i, row);
}
const char*
easycsv_readvalue(const easycsv *csv, const unsigned int col, const unsigned int row)
{
if (row == 0) {
_easycsv_printerror(csv->_priv, EASYCSV_ZEROROW);
return NULL;
}
if (col == 0) {
_easycsv_printerror(csv->_priv, EASYCSV_ZEROCOL);
return NULL;
}
char *str = _easycsv_getrow(csv, row);
if (str == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_ROWNOTEXIST);
return NULL;
}
/*
#ifdef EASYCSV_DEBUG
printf("[%i] rv_col: %i\nstr: %s\n", clock() - csv->_priv->start, col, str);
#endif
*/
/* Return if first col */
if (col == 1) {
size_t st = strcspn(str, ",");
if (st == 0) {
free(str);
return NULL;
}
char *pch = malloc(st);
strncpy(pch, str, st);
free(str);
return pch;
}
/* Get first occurance of comma in str,
the first value is ommited but not the comma */
char *pch = strpbrk(str, ",");
/* Repeat until desired col is found */
for (int i = 2; i < col; i++) {
pch = strpbrk(pch + 1, ",");
}
/* Get span from start of string to first occurence of comma */
size_t st = strcspn(pch + 1, ",");
/* If 0, no string exists! */
if (st == 0) {
free(str);
return NULL;
}
char *val = malloc(st + 1);
strncpy(val, pch + 1, st);
strncat(val, "\0", 1);
free(str);
return val;
}
/* PRIVATE FUNCTIONS */
/*
static int _easycsv_checkmode(struct easycsv *csv, const char mode)
{
if (strchr(csv->mode, mode) == NULL)
{
return -1;
}
}
*/
int
_easycsv_update(easycsv *csv)
{
if (csv == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_NULLCSV);
return -1;
}
/* Set temp file to read binary */
if (freopen(csv->_priv->tmpfp, "rb", csv->_priv->temp) == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_REOPENFAIL);
easycsv_free(csv);
return -1;
}
/* Set file to write binary */
if (freopen(csv->_priv->fp, "wb", csv->_priv->file) == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_REOPENFAIL);
easycsv_free(csv);
return -1;
}
char buf[BUFSIZ];
size_t size;
while (size = fread(buf, 1, BUFSIZ, csv->_priv->temp)) {
fwrite(buf, 1, size, csv->_priv->file);
}
/* Set temp file back to write */
if (freopen(csv->_priv->tmpfp, "w", csv->_priv->temp) == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_REOPENFAIL);
easycsv_free(csv);
return -1;
}
/* Set file back to read */
if (freopen(csv->_priv->fp, "r", csv->_priv->file) == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_REOPENFAIL);
easycsv_free(csv);
return -1;
}
return 0;
}
char*
_easycsv_getrow(const easycsv *csv, const unsigned int row)
{
if (csv == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_NULLCSV);
return NULL;
}
if (row > csv->_priv->rows) {
_easycsv_printerror(csv->_priv, EASYCSV_OVERMAXROW);
return NULL;
}
if (row == 0) {
_easycsv_printerror(csv->_priv, EASYCSV_ZEROROW);
return NULL;
}
/* Allocate memory */
char *str = malloc(BUFSIZ);
/* Set file pointer to start */
rewind(csv->_priv->file);
for (int i = 1; i < row; i++)
{
/* skip until row is reached */
fscanf(csv->_priv->file, "%*[^\n]\n", NULL);
}
/* Grab the row and store it in str */
fscanf(csv->_priv->file, "%s\n", str);
// printf("row: %s\n", str);
return str;
}
int
_easycsv_rewind(_easycsv *_priv, const EASYCSV_MODE mode)
{
/* Check if empty file */
if (fscanf(_priv->file, "\n") == EOF) {
_easycsv_printerror(_priv, EASYCSV_EMPTYCSV);
return -1;
}
/* Check if file is readable */
if (mode != EASYCSV_R) {
_easycsv_printerror(_priv, EASYCSV_UNREADABLE);
return -1;
}
/* Set file pointer to the start */
rewind(_priv->file);
return 0;
}
int
_easycsv_getcolumn(const easycsv *csv, const char *col)
{
if (_easycsv_checkcsvandstring_one(csv, col) < 0) return -1;
/* Grab str of row 1 */
char *firstrow = _easycsv_getrow(csv, 1);
if (firstrow == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_ROWNOTEXIST);
return -1;
}
unsigned int commas = 0;
// printf("FIRST COLUMN: %s\n", firstrow);
/* Find first occurance of col in firstrow */
char *str = strstr(firstrow, col);
if (str == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_COLNOTEXIST);
return -1;
}
/* Count numbers of commas following str */
char *c = strpbrk(str, ",");
while (c != NULL) {
commas++;
c = strpbrk(c + 1, ",");
}
/* no need to free c as it is already NULL at this point */
// free((char*) str); apparently invalid pointer
// printf("ROW: %i\nCOL: %i\n", row, csv->_priv->cols - commas);
/*
#ifdef EASYCSV_DEBUG
printf("[%i] rcv_commas: %i\n", clock() - csv->_priv->start, commas);
#endif
*/
free(firstrow);
return csv->_priv->cols - commas;
}
/*
int _easycsv_checkifvalue(struct easycsv *csv, const int col, const int row)
{
const char *rowstr = _easycsv_getrow(csv, row);
return 0;
}
char *_easycsv_getvalueinrow(const int col, const char *row)
{
char *pbrk = strchr(row, ',');
for (unsigned int i = 1; i < col; i++) {
}
}
*/
void
_easycsv_printerror(const _easycsv *_priv, const EASYCSV_ERROR error)
{
switch (_priv->error) {
case EASYCSV_ERROROFF: return;
case EASYCSV_ERRORSTDOUT: _easycsv_geterror(_priv, error, stdout); break;
case EASYCSV_ERRORSTDERR:
default: _easycsv_geterror(_priv, error, stderr); return;
}
}
void
_easycsv_geterror(const _easycsv *_priv, const EASYCSV_ERROR error, FILE *fs)
{
#ifdef EASYCSV_DEBUG
fprintf(stderr, "[%i] ", clock() - _priv->start);
#endif
fputs("easycsv: ", fs);
switch (error) {
case EASYCSV_NOERROR: fputs("no error", fs); break;
case EASYCSV_NULLCSV: fputs("easycsv pointer is NULL", fs); break;
case EASYCSV_NULLPTR: fputs("pointer is NULL", fs); break;
case EASYCSV_EMPTYSTRING: fputs("string is empty", fs); break;
case EASYCSV_OVERMAXROW: fprintf(fs, "int exceeds row limit %i", _priv->rows); break;
case EASYCSV_OVERMAXCOL: fprintf(fs, "int exceeds column limit %i", _priv->cols); break;
case EASYCSV_ZEROROW: fputs("parameterised row number is zero", fs); break;
case EASYCSV_ZEROCOL: fputs("parameterised column number is zero", fs); break;
case EASYCSV_UNKNOWNMODE: fputs("unknown file IO mode", fs); break;
case EASYCSV_OPENFAIL: fputs("failed to open file", fs); break;
case EASYCSV_REOPENFAIL: fputs("failed to reopen file", fs); break;
case EASYCSV_EMPTYCSV: fputs("CSV file is empty", fs); break;
case EASYCSV_UNWRITABLE: fputs("CSV file is not in a writable mode", fs); break;
case EASYCSV_UNREADABLE: fputs("CSV file is not in a readable mode", fs); break;
case EASYCSV_UPDATEFAIL: fputs("CSV file has failed to update", fs); break;
case EASYCSV_FILEPTRFAIL: fputs("failed to move FILE pointer", fs); break;
case EASYCSV_ROWNOTEXIST: fputs("given row does not exist", fs); break;
case EASYCSV_COLNOTEXIST: fputs("given column does not exist", fs); break;
case EASYCSV_PUSHCOLFAIL: fputs("failed to push value under column", fs); break;
case EASYCSV_COLNUMFAIL: fputs("failed to determine the column number of a value in the first row", fs); break;
default: fputs("unknown error, easycsv* is possibly NULL", fs); break;
}
if (_priv->fp != NULL) {
fprintf(fs, " in %s", _priv->fp);
}
fputc('\n', fs);
}
int
_easycsv_checkcsvandstring_one(const easycsv *csv, const char *one)
{
if (csv == NULL) {
_easycsv_printerror(csv->_priv, EASYCSV_NULLCSV);
return -1;
}
if ((one == NULL) || (one[0] == '\0')) {
_easycsv_printerror(csv->_priv, EASYCSV_EMPTYSTRING);
return -1;
}
return 0;
}
int
_easycsv_checkcsvandstringtwo(const easycsv *csv, const char *one, const char *two)
{
if (_easycsv_checkcsvandstring_one(csv, one) < 0) return -1;
if ((two == NULL) || (two[0] == '\0')) {
_easycsv_printerror(csv->_priv, EASYCSV_EMPTYSTRING);
return -1;
}
return 0;
}