Number of Lines Mystery? - c

My goal is to implement a function that calculates the number of lines in a file. And empty file is considered to have no lines. If the last line of the given file is not empty, it should be counted as a line despite not ending with a newline character.
I've come up with the following code:
int linecount(const char *filename)
{
FILE *f = fopen(filename, "r");
if(!f)
return -1;
int lines = 0;
int c = 0;
int n = 0;
while((c = fgetc(f)) != EOF){
if(c == '\n')
lines++;
n++;
}
if(n==0)
return 0; //return 0 if the file is empty
if(c!='\n' && !isspace(c))
lines++; //count the last line if it's not empty
fclose(f);
return lines;
}
However, even after playing with it for over an hour I can't figure out why its return value lines is one too large in some cases...

You were close, here how you could do it:
int linecount(const char *filename) {
FILE *f = fopen(filename, "r");
if (!f)
return -1;
int lines = 0;
int c = 0;
int n = 0;
int read_line = 0;
while ((c = fgetc(f)) != EOF) {
read_line = 1;
if (c == '\n') {
lines++;
read_line = 0;
}
n++;
}
if (n == 0)
return 0; //return 0 if the file is empty
if(read_line)
lines++;
fclose(f);
return lines;
}
The idea is that we want to know if we started reading a line AND if we met a newline, at end of this line. So, we use another variable, called read_line and we use it as a flag.
We set it to 1 (true) if we just started reading a line and we set it to 0 (false) if we just met a newline (end of the line).
Now, if we have something like:
1[newline]
2[newline]
3
we will be OK, since we need to check if read_line after we read the file. Is so, we have to increment our line counter by one.
This is also OK:
1[newline]
2[newline]
3[newline]
since we saw three newlines and the read_line is 0 after we read the file.
Same goes for this case:
1[newline]
2[newline]
3[newline]
[nothing here]
since our flag is going to be equal to 0 after reading the file, since the 3rd newline should set it to 0 and we never actually enter the 4th line in our loop, since there is nothing to read.
With your previous implementation, as stated in the comments, this line:
if(c!='\n' && !isspace(c))
would be executed with c being equal to EOF.
Or you could just use fgets() and you are done. Check the example:
#include <stdio.h>
#include <string.h>
#define bufSize 1024
int main(int argc, char *argv[])
{
FILE *fp;
char buf[bufSize];
if ((fp = fopen("test.txt", "rb")) == NULL)
{ /* Open source file. */
perror("fopen source-file");
return 1;
}
int lines = 0;
while (fgets(buf, sizeof(buf), fp) != NULL)
{ /* While we don't reach the end of source. */
/* Read characters from source file to fill buffer. */
/* fgets will stop when it finds a newline. */
lines++;
}
printf("lines = %d\n", lines);
fclose(fp);
return 0;
}

Modify sample
int linecount(const char *filename)
{
FILE *f = fopen(filename, "r");
if(!f)
return -1;
int lines = 0;
int c = 0;
int flag = 1;
while((c = fgetc(f)) != EOF){
if(flag = (c == '\n'))
lines++;
}
if(!flag)
lines++; //count the last line if it's not empty
fclose(f);
return lines;
}

A simple solution can be
int linecount(const char *filename)
{
FILE *stream;
char *line = NULL;
size_t len = 0;
ssize_t read;
int numOfLines = 0;
stream = fopen(filename, "r");
if (stream == NULL)
exit(EXIT_FAILURE);
while ((read = getline(&line, &len, stream)) != -1) {
numOfLines++;
}
free(line);
fclose(stream);
return numOfLines;
}

Related

How to convert a text file from DOS format to UNIX format

I am trying to make a program in C, that reads a text file and replace \r\n with \n to the same file converting the line ending from DOS to UNIX. I use fgetc and treat the file as a binary file. Thanks in advance.
#include <stdio.h>
int main()
{
FILE *fptr = fopen("textfile.txt", "rb+");
if (fptr == NULL)
{
printf("erro ficheiro \n");
return 0;
}
while((ch = fgetc(fptr)) != EOF) {
if(ch == '\r') {
fprintf(fptr,"%c", '\n');
} else {
fprintf(fptr,"%c", ch);
}
}
fclose(fptr);
}
If we assume the file uses a single byte character set, we just need to ignore all the '\r' characters when converting a text file form DOS to UNIX.
We also assume that the size of the file is less than the highest unsigned integer.
The reason we do these assumptions, is to keep the example short.
Be aware that the example below overwrites the original file, as you asked. Normally you shouldn't do this, as you can lose the contents of the original file, if an error occurs.
#include <stdio.h>
#include <stdlib.h>
#include <sys/stat.h>
// Return a negative number on failure and 0 on success.
int main()
{
const char* filename = "textfile.txt";
// Get the file size. We assume the filesize is not bigger than UINT_MAX.
struct stat info;
if (stat(filename, &info) != 0)
return -1;
size_t filesize = (size_t)info.st_size;
// Allocate memory for reading the file
char* content = (char*)malloc(filesize);
if (content == NULL)
return -2;
// Open the file for reading
FILE* fptr = fopen(filename, "rb");
if (fptr == NULL)
return -3;
// Read the file and close it - we assume the filesize is not bigger than UINT_MAX.
size_t count = fread(content, filesize, 1, fptr);
fclose(fptr);
if (count != 1)
return -4;
// Remove all '\r' characters
size_t newsize = 0;
for (long i = 0; i < filesize; ++i) {
char ch = content[i];
if (ch != '\r') {
content[newsize] = ch;
++newsize;
}
}
// Test if we found any
if (newsize != filesize) {
// Open the file for writing and truncate it.
FILE* fptr = fopen(filename, "wb");
if (fptr == NULL)
return -5;
// Write the new output to the file. Note that if an error occurs,
// then we will lose the original contents of the file.
if (newsize > 0)
count = fwrite(content, newsize, 1, fptr);
fclose(fptr);
if (newsize > 0 && count != 1)
return -6;
}
// For a console application, we don't need to free the memory allocated
// with malloc(), but normally we should free it.
// Success
return 0;
} // main()
To only remove '\r' followed by '\n' replace the loop with this loop:
// Remove all '\r' characters followed by a '\n' character
size_t newsize = 0;
for (long i = 0; i < filesize; ++i) {
char ch = content[i];
char ch2 = (i < filesize - 1) ? content[i + 1] : 0;
if (ch == '\r' && ch2 == '\n') {
ch = '\n';
++i;
}
content[newsize++] = ch;
}

How to read a file with numbers on each line in C?

I'm trying to read a file that contains 10 numbers then adding them to an array so I can sort them later on but I'm having trouble reading them in. Not sure why this isn't working for me, can someone explain what is wrong? There's only a number on each lines.
10.05
11.01
9.03
double nums[10] = {0};
int count;
if ((fptr = fopen("filename", "r")) == NULL){
printf("Error opening file.\n");
}
while ((c = getc(fptr)) != EOF){
if (c != '\n'){
nums[count] = (double)c;
count = count + 1;
}
}
fclose(fptr);
What is wrong:
You are storing only one character.
You are updating count each times on non-newline characters while updating should be on newline characters.
count is used without being initialized.
Casting to double is not for this usage.
Possible fix:
int c;
FILE* fptr;
char line[1024]; // add line buffer and size tracking
int lineCount = 0;
double nums[10] = {0};
int count = 0; // initialize count
if ((fptr = fopen("filename", "r")) == NULL){
printf("Error opening file.\n");
} else { // avoid using NULL to read file
while ((c = getc(fptr)) != EOF){
if (c == '\n'){ // update nums on newline character
line[lineCount] = '\0'; // don't forget to terminate the string
nums[count] = atof(line); // atof() from stdlib.h is useful to convert string to number
count = count + 1;
lineCount = 0; // start to read next line
} else { // read line contents
line[lineCount] = (char)c;
lineCount = lineCount + 1;
}
}
fclose(fptr);
}
Here I go
#include <stdio.h>
#include <stdlib.h>
int main()
{
double values[10];
int count;
FILE *f = fopen("filename", "r");
if (f == NULL)| {
fprintf(stderr, "Some error message");
return EXIT_FAILURE; // We cannot go any further - file is dead
}
// This is basic - you could overcome error problems
// When able to read (including white space) we carry on until the array is full
// This is an area for improvement - error checking etc.
for (count = 0; count < 10 && fscanf(f, " %lf", &values[count]) != 1; count ++);
fclose(f);
return EXIT_SUCCESS;
}

Compare each line from two different files and print the lines that are different in C

Supposing that I have two files like this:
file1.txt
john
is
the new
guy
file2.txt
man
the old
is
rick
cat
dog
I'd like to compare first line from file1 with all the lines from file2 and verify if it exist. If not, go two the second line from file1 and compare it with all the lines from file2.. and so on until eof is reached by file1.
The output that I expect is:
john
the new
guy
How I thought this should be done:
read file1 and file2
create a function which returns the line number of each of them
take the first line from file1 and compare it to all the lines from file2
do this until all the lines from file1 are wasted
Now, I don't know what I'm doing wrong, but I don't get the result that I expect:
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
int countlines(char *filename)
{
int ch = 0, lines = 0;
FILE *fp = fopen(filename, "r");
if (fp == NULL)
return 0;
do {
ch = fgetc(fp);
if (ch == '\n')
lines++;
} while (ch != EOF);
if (ch != '\n' && lines != 0)
lines++;
fclose(fp);
return lines;
}
int main(int argc, char *argv[])
{
FILE *template_file = fopen(argv[1], "r");
FILE *data_file = fopen(argv[2], "r");
char buffer_line_template_file[100];
char buffer_line_data_file[100];
if (argc != 3)
{
perror("You didn't insert all the arguments!\n\n");
exit(EXIT_FAILURE);
}
if (template_file == NULL || data_file == NULL)
{
perror("Error while opening the file!\n\n");
exit(EXIT_FAILURE);
}
int counter = 0;
for (int i = 0; i < countlines(argv[1]); i++)
{
fgets(buffer_line_template_file, 100, template_file);
for (int j = 0; j < countlines(argv[2]); j++)
{
fgets(buffer_line_data_file, 100, data_file);
if (strcmp(buffer_line_template_file, buffer_line_data_file) != 0)
{
counter++;
printf("%d", counter);
}
}
}
printf("\n\n");
return 0;
}
Could someone please point me into the right direction ? For testing purposes I created a counter at the end which was a part of a small debug. There should be the print() function
As per #chux answer I got the following simplified code:
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
int main(int argc, char *argv[])
{
FILE *template_file = fopen(argv[1], "r");
FILE *data_file = fopen(argv[2], "r");
char buffer_line_template_file[100];
char buffer_line_data_file[100];
if (argc != 3)
{
perror("You didn't insert all the arguments!\n\n");
exit(EXIT_FAILURE);
}
if (template_file == NULL || data_file == NULL)
{
perror("Error while opening the file!\n\n");
exit(EXIT_FAILURE);
}
while(fgets(buffer_line_template_file, 100, template_file))
{
buffer_line_template_file[strcspn(buffer_line_template_file, "\n")] = '\0';
rewind(data_file);
while (fgets(buffer_line_data_file, 100, data_file))
{
buffer_line_data_file[strcspn(buffer_line_data_file, "\n")] = '\0';
if (strcmp(buffer_line_template_file, buffer_line_data_file) != 0)
{
printf("%s\n", buffer_line_template_file);
}
}
}
printf("\n\n");
return 0;
}
The above code is giving me the following output, which is not what is expected:
john
john
john
john
john
john
is
is
is
is
is
the new
the new
the new
the new
the new
the new
guy
guy
guy
guy
guy
guy
Problems with OP's code
Imprecise definition of line.
Excessive recalculation
Fuzzy determination of the number of lines in a file.
Unlike string, which has a precise definition in C, reading a line is not so well defined. The primary specificity issue: does a line contain the trailing '\n'. If the first answer is Yes, then does the last text in a file after a '\n' constitute a line? (Excessively long lines are another issue, but let us not deal with that today.)
Thus possibly some lines end with '\n' and others do not, fooling strcmp("dog", "dog\n").
The easiest solution is to read a line until either 1) a '\n' is encountered, 2) EOF occurs or 3) line buffer is full. Then after getting a line, lop off the potential trailing '\n'.
Now all lines code subsequently works with have no '\n'.
fgets(buffer_line_template_file, 100, template_file);
buffer_line_template_file[strcspn(buffer_line_template_file, "\n")] = '\0';
OP's loop is incredible wasteful. Consider a file with 1000 lines. Code will loop, calling 1000 times countlines() (each countlines() call reads 1000 lines) times when one countlines() call would suffice.
// for (int j = 0; j < countlines(argv[2]); j++)
int j_limit = countlines(argv[2]);
for (int j = 0; j < j_limit; j++)
There really is no need to count the line anyways, just continue until EOF (fgets() returns NULL). So no need to fix its fuzzy definition. (fuzzy-ness concerns same issues as #1)
int counter = 0;
for (fgets(buffer_line_template_file, 100, template_file)) {
buffer_line_template_file[strcspn(buffer_line_template_file, "\n")] = '\0';
rewind(data_file);
while ((fgets(buffer_line_data_file, 100, data_file)) {
buffer_line_data_file[strcspn(buffer_line_data_file, "\n")] = '\0';
if (strcmp(buffer_line_template_file, buffer_line_data_file) != 0) {
counter++;
printf("%d", counter);
}
}
}
Other simplifications possible - for another day.
FWIW, following counts lines of text allowing the last line in the file to optionally end with a '\n'.
unsigned long long FileLineCount(FILE *istream) {
unsigned long long LineCount = 0;
rewind(istream);
int previous = '\n';
int ch;
while ((ch = fgetc(inf)) != EOF) {
if (previous == '\n') LineCount++;
previous = ch;
}
return LineCount;
}
Note that this function may get a different result that fgets() calls. Consider a file of one line of 150 characters. fgets(..., 100,...) will report 2 lines. FileLineCount() reports 1.
[Edit] Updated code to conform to OP functionality.
int found = 0;
while (fgets(buffer_line_data_file, 100, data_file))
{
buffer_line_data_file[strcspn(buffer_line_data_file, "\n")] = '\0';
if (strcmp(buffer_line_template_file, buffer_line_data_file) == 0)
{
found = 1;
break;
}
}
if (!found) printf("%s\n", buffer_line_template_file);
This program prints the diff of two files file1.txt and file2.txt.
#include<stdio.h>
#include <stdlib.h>
#include <memory.h>
int main() {
FILE *fp1, *fp2;
int ch1, ch2;
char fname1[40], fname2[40];
char *line = NULL;
size_t len = 0;
ssize_t read;
char *line2 = NULL;
size_t len2 = 0;
ssize_t read2;
fp1 = fopen("file1.txt", "r");
fp2 = fopen("file2.txt", "r");
if (fp1 == NULL) {
printf("Cannot open %s for reading ", fname1);
exit(1);
} else if (fp2 == NULL) {
printf("Cannot open %s for reading ", fname2);
exit(1);
} else {
while ((read = getline(&line, &len, fp1)) != -1 && (read2 = getline(&line2, &len2, fp2)) != -1) {
if (!strcmp(line, line2)) {
printf("Retrieved diff on line %zu :\n", read);
printf("%s", line);
}
}
if (ch1 == ch2)
printf("Files are identical \n");
else if (ch1 != ch2)
printf("Files are Not identical \n");
fclose(fp1);
fclose(fp2);
}
return (0);
}
You already have a very good answer (and always will from chux), but here is a slightly different approach to the problem. It uses automatic storage to reading file2 into an array of strings and then compares each line in file1 against every line in file2 to determine whether it is unique. You can easily convert the code to dynamically allocate memory, but for sake of complexity that was omitted:
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
enum { MAXC = 256, MAXL = 512 };
void file1infile2 (FILE *fp2, FILE *fp1, size_t *n2, size_t *n1);
int main (int argc, char **argv) {
FILE *fp1 = fopen (argc > 1 ? argv[1] : "file1.txt", "r");
FILE *fp2 = fopen (argc > 2 ? argv[2] : "file2.txt", "r");
size_t n1 = 0, n2 = 0;
if (!fp1 || !fp2) {
fprintf (stderr, "error: file open failed.\n");
return 1;
}
printf ("\nunique words in file1, not in file 2.\n\n");
file1infile2 (fp2, fp1, &n2, &n1);
printf ("\nanalyzed %zu lines in file1 against %zu lines in file2.\n\n",
n1, n2);
return 0;
}
void file1infile2 (FILE *fp2, FILE *fp1, size_t *n2, size_t *n1)
{
char buf[MAXC] = "";
char f2buf[MAXL][MAXC] = { "" };
size_t i;
*n1 = *n2 = 0;
while (*n2 < MAXL && fgets (buf, MAXC, fp2)) {
char *np = 0;
if (!(np = strchr (buf, '\n'))) {
fprintf (stderr, "error: line exceeds MAXC chars.\n");
exit (EXIT_FAILURE);
}
*np = 0;
strcpy (f2buf[(*n2)++], buf);
}
while (*n1 < MAXL && fgets (buf, MAXC, fp1)) {
char *np = 0;
if (!(np = strchr (buf, '\n'))) {
fprintf (stderr, "error: line exceeds MAXC chars.\n");
exit (EXIT_FAILURE);
}
*np = 0, (*n1)++;
for (i = 0; i < *n2; i++)
if (!(strcmp (f2buf[i], buf)))
goto matched;
printf (" %s\n", buf);
matched:;
}
}
Look over the code and let me know if you have any questions.
Example Use/Output
$ ./bin/f1inf2 dat/f1 dat/f2
unique words in file1, not in file 2.
john
the new
guy
analyzed 4 lines in file1 against 6 lines in file2.

Using file pointers correctly

I have a basic question about file pointers. In the code below i have a while loop followed by a for loop. The for loop only will show the line count unless i fopen the file again - is that normal? and if so should i fclose it after the while loop beforehand? There's probably some "rewind" function that i'm unaware of, so my whole approach might be wrong. I realize both the while loop and for can be combined, but this my question is about the fopen, fclose and using the data again from fopen.
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#define MAXLEN 200
enum { MAXLINES = 200 };
char lpath[MAXLINES][BUFSIZ];
int main(int argc, char **argv) {
char c;
int l = 0, n = 0, i = 0, count = 0;
char lines[MAXLINES][BUFSIZ];
FILE *fp = fopen(argv[1], "r");
if (fp == 0) {
fprintf(stderr, "failed to open input.txt\n");
exit(1);
}
while (l < MAXLINES && fgets(lines[l], sizeof(lines[0]), fp)) {
lines[l][strlen(lines[l])-1] = '\0';
puts(lines[l]);
l++;
}
// fp = fopen(argv[1], "r"); // below won't output unless fopen again
for (c = getc(fp); c != EOF; c = getc(fp)) {
if (c == '\n') {
count++;
}
printf(">> line count: %i", count);
fclose(fp);
}
looked also at: Pointer best practice
It's normal, you need to rewind() the file. The problem is that when the for loop start the file has reached the end, so reads will fail and feof() will return non-zero.
Two options
rewind(fp);
Or
fseek(fp, 0L, SEEK_SET);
When you call fopen() again you leak resources because you overwrite the pointer and now you can't fclose() the first fopen()ed file.
You could use rewind() as #iharob suggested, or you could just close and re-open the file:
while (l < MAXLINES && fgets(lines[l], sizeof(lines[0]), fp)) {
lines[l][strlen(lines[l])-1] = '\0';
puts(lines[l]);
l++;
}
fclose(fp);
fp = fopen(argv[1], "r"); // below won't output unless fopen again
for (c = getc(fp); c != EOF; c = getc(fp)) {
if (c == '\n') {
count++;
}
}

How to count blank lines from file in C?

So what I'm trying to do is to count blank lines, which means not only just containing '\n'but space and tab symbols as well. Any help is appreciated! :)
char line[300];
int emptyline = 0;
FILE *fp;
fp = fopen("test.txt", "r");
if(fp == NULL)
{
perror("Error while opening the file. \n");
system("pause");
}
else
{
while (fgets(line, sizeof line, fp))
{
int i = 0;
if (line[i] != '\n' && line[i] != '\t' && line[i] != ' ')
{
i++;
}
emptyline++;
}
printf("\n The number of empty lines is: %d\n", emptyline);
}
fclose(fp);
You should try and get your code right when posting on SO. You are incrementing both i and emptyline but the use el in your call to printf(). And then I don't know what that is supposed to be in your code where it has }ine. Please, at least make an effort.
For starters, you are incrementing emptyline for every line because it is outside of your if statement.
Second, you need to test the entire line to see if it contains any character that is not a whitespace character. Only if that is true should you increment emptyline.
int IsEmptyLine(char *line)
{
while (*line)
{
if (!isspace(*line++))
return 0;
}
return 1;
}
Before getting into the line loop increment the emptyLine counter and if an non whitespace character is encountred decrement the emptyLine counter then break the loop.
#include <stdio.h>
#include <string.h>
int getEmptyLines(const char *fileName)
{
char line[300];
int emptyLine = 0;
FILE *fp = fopen("text.txt", "r");
if (fp == NULL) {
printf("Error: Could not open specified file!\n");
return -1;
}
else {
while(fgets(line, 300, fp)) {
int i = 0;
int len = strlen(line);
emptyLine++;
for (i = 0; i < len; i++) {
if (line[i] != '\n' && line[i] != '\t' && line[i] != ' ') {
emptyLine--;
break;
}
}
}
return emptyLine;
}
}
int main(void)
{
const char fileName[] = "text.txt";
int emptyLines = getEmptyLines(fileName);
if (emptyLines >= 0) {
printf("The number of empty lines is %d", emptyLines);
}
return 0;
}
You are incrementing emptyline on every iteration, so you should wrap it in an else block.
Let's think of this problem logically, and let's use functions to make it clear what is going on.
First, we want to detect lines that only consist of whitespace. So let's create a function to do that.
bool StringIsOnlyWhitespace(const char * line) {
int i;
for (i=0; line[i] != '\0'; ++i)
if (!isspace(line[i]))
return false;
return true;
}
Now that we have a test function, let's build a loop around it.
while (fgets(line, sizeof line, fp)) {
if (StringIsOnlyWhitespace(line))
emptyline++;
}
printf("\n The number of empty lines is: %d\n", emptyline);
Note that fgets() will not return a full line (just part of it) on lines that have at least sizeof(line) characters.

Resources