something is squiffy

**cooper1200** · 05-24-2023

i have the following

Code:

int main()
{
    //declare variables
    char buffer[1024] = {'\0'};
    FILE *pRead = NULL;

    pRead = fopen("names.txt", "r");

    if ( pRead == NULL )
    {
        printf("file cant be opened!!\n");
        return 0;
    }

    //fscanf( pRead, "%s", buffer );

    //fread(buffer, sizeof buffer, sizeof(char), pRead);
    while ( (fscanf(pRead, "\"%[^\"]\",", buffer) != EOF) )
    //while (fread(buffer, sizeof buffer, sizeof(char), pRead) != 1)
    {

        printf("%s\n",buffer);
    }
    fclose( pRead );

    return 0;
}

the only one that did something other than a load of empty lines in a infinite loop was but it would print the text file twice before ending if it was inside the while loop or print the text file in a infinite loop outside the while loop.

if i make the buffer smaller it just throws a seg fault when it tries to read the file.

i even copied the current while loop from the official answer and it just produces empty buffer again in an infinite loop

**Salem** · 05-24-2023

Well what's in your text file?

Because if the first character isn't a ", your fscanf will just cause an infinite loop.

This is how you read a text file.

Code:

while ( fgets(buffer, sizeof(buffer), pRead) ) {
  char quoted[1024];
  if ( sscanf(buffer, "\"%[^\"]\",", quoted) == 1 ) {
    printf("%s\n",quoted);
  } else {
    printf("Buffer doesn't match expected format\n");
  }
}

> i even copied the current while loop from the official answer
You keep saying "official answer" like such a thing exists - it doesn't.

**flp1969** · 05-25-2023

if your libc supports ISO 27431-2:2010 (__STDC_WANT_LIB_EXT2__), then you can use getline() function which will allocate the buffer dynamically.

Most glibc supports this.

**cooper1200** · 05-25-2023

the text file is here https://projecteuler.net/project/res...p042_words.txt it is a csv file with words in quotation marks ie "a","above","below"...
i am twisting myself into knots here. All i am trying to do is read in one word at a time and when it gets to the end of the file quit

im getting frustrated with myself because what ever combination of buffers and arrays unless i have stupidly large arrays it all falls over

**Salem** · 05-25-2023

Code:

$ wc p042_words.txt 
    0     1 16345 p042_words.txt

It's just ONE line of 16K characters.

What are you trying to do?

Print each word?

**cooper1200** · 05-25-2023

yes or the entire thing and then split it myself

**Salem** · 05-25-2023

The trick would seem to be dealing with the lack of a comma after the last word.

So maybe something like

Code:

while ( (fscanf(pRead, "\"%[^\"]\"%*c", buffer) == 1 )

The %*c will eat a comma (or indeed any other character), if there happens to be one.

But it won't cause a scan failure at the end of input because it's missing.

**flp1969** · 05-25-2023

CSV format standard (RFC 4180) says these are valid and are the same thing:

"aaa","bbb","ccc"<crlf>
aaa","bbb","ccc<crlf>

And the " char is done by double quotes:

10,"abc""d","fred""<crlf>

The second field here is abc"d. The third is fred" (the final ", closing the string is missing. "" is the character ").

PS: Every line are always terminated by "\r\n", otherwise this isn't a CSV file. There's an exception: The last line may not be terminated by this final "\r\n".

Spaces are part of the fields and cannot be ignored, the line below:

aaa,bbb ,ccc<crlf>

Have 3 fields: "aaa", "bbb " and "ccc".

Thread: something is squiffy

Thread Tools

Search Thread

Display

something is squiffy

Tags for this Thread