如何 fscanf 字数未知的字符串？

Question

I have a text file formatted like this:

code subject_name ects

Where code and ects are integers, and subject_name is a string that can be multiple words long and sometimes contains a number.

I've tried fscanf(f, "%d %s %d", &code, subject_name, &ects); which doesn't work because there are spaces in the string.

"%d %[^\\n] %d" also won't work because the string swallows up ects .

What would be the correct way to do this?

Answer 1

Step 1: read a line

#define LINE_MAX_EXPECTED_SIZE 100
char buf[LINE_MAX_EXPECTED_SIZE + 2];// Let code read lines that are too long

if (fgets(buf, sizeof buf, f)) {
  buf[strcspn(buf, "\n\r")] = '\0';  // lop off potential end-of-line
  if (strlen(buf) >= LINE_MAX_EXPECTED_SIZE || buf[0] == '\0') {
    fprintf(stderr, "Line too long/short. <%s>\n", buf); 
    exit(EXIT_FAILURE);
  }
  ...

OK, now we have the line read and saved as a string .

Step 2: Since subject_name and ects can be a number, let code look for ects first since it is one and only one number.

  // Start at end
  char *end = strlen(buf) - 1;
  if (!isdigit((unsigned char) *end)) {
    fprintf(stderr, "No number at end. <%s>\n", buf); 
    exit(EXIT_FAILURE);
  }
  while (end > buf && isdigit((unsigned char) * --end)) {
    ;
  }
  if (end > buf && (*end == '-' || *end == '+')) {
    end--;
  }
  ects = atoi(end + 1); // or better strtol()
  end[1] = '\0'; // lop off ects

Now buf has, hopefully, code and subject_name which can be parsed with user code, sscanf() , strtol() , etc. Leave that for OP.