How do I correctly sort strings in C?

Question

Here is my code:

#include <stdio.h>
#include <string.h>
#include <errno.h>

int cmp(const void *a, const void *b) {
    const char **ia = (const char **)a;
    const char **ib = (const char **)b;
    return strcmp(*ia, *ib);
}

void print_array(char **array, size_t len) {
    size_t i;
    for(i=0; i<len; i++) {
        printf("%s, ", array[i]);
    }
    putchar('\n');
}

int main(int argc, char *argv[]) {

        char *strings[] = { "z1.doc", "z100.doc",  "z2.doc", "z3.doc", "z20.doc"};
        size_t strings_len = sizeof(strings) / sizeof(char *);
        print_array(strings, strings_len);
        qsort(strings, strings_len, sizeof(char *), cmp);
        print_array(strings, strings_len);

        system("PAUSE");
        return 1;
}

the actual output is

z1.doc, z100.doc, z2.doc, z20.doc, z3.doc

and I want it to be

z1.doc, z2.doc, z3.doc, z20.doc, z100.doc

What am doing wrong?

Yikes! So many stars! const char *ia = a; /* ... */ return strcmp(ia, ib); — pmg
– pmg, Commented Nov 5, 2009 at 0:22
Came across this at mail-archive.com/[email protected]/…: Explorer uses the StrCmpLogicalW() function in shlwapi.dll to sort filenames into "logical" order. The exact sort order is not guaranteed and may change with each Windows release or service pack! — Bob Kaufman
– Bob Kaufman, Commented Nov 5, 2009 at 0:26
@pmg: That's incorrect. Doing it as you suggest will surely result in segmentation fault. The alternative variant would be const char *ia = *(const char **) a;...; return strcmp(ia, ib);, but absolutely not what you suggest. — AnT stands with Russia
– AnT stands with Russia, Commented Nov 5, 2009 at 0:28
@pmg: No, it doesn't. You simply failed to test it properly. You are re-interpreting pointer values as C-strings and comparing the resulting "garbage" strings. If it didn't crash for you, it is simply because you got lucky. Run it and take a look at your "sorted" array for a laugh. Then think about what you done. Again, your version makes no sense whatsoever. — AnT stands with Russia
– AnT stands with Russia, Commented Nov 5, 2009 at 0:41

progrmr · Accepted Answer · 2009-11-05 00:10:02Z

4

The actual output is correct, the string "z100.doc" is less than "z2.doc". The strcmp compares character by character and when it gets to the '1' that is less than '2' and it stops there, so z100 < z2.

If you name the files z001.doc, z002.doc, z003.doc, z020.doc, z100.doc it will sort the way you want.

answered Nov 5, 2009 at 0:10

progrmr

77.5k16 gold badges116 silver badges148 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Chris Lutz Over a year ago

-1 Renaming the files is not the answer to this question. The OP wants to do work for the user, not make the user do (unnecessary) work to get his/her files displayed properly. Anyone who upvotes this doesn't understand the concept of "interface usability" - no one other than us programmers expects lists to be sorted in the order that the standard strcmp() sorts in.

Chris Lutz Over a year ago

The question was "Why is the sorted output I get not the sorted output I expect, and how can I fix that?"

Dan · Accepted Answer · 2009-11-05 00:11:42Z

3

Change your comparator to say:

return (atoi(*ib + 1) - atoi(*ia + 1));

answered Nov 5, 2009 at 0:11

Dan

541 bronze badge

2 Comments

progrmr Over a year ago

that will only work if there is only 1 number in the filename and all the characters are identical. it won't work properly for "z1aa.doc", "z1zz.doc", "z1bb.doc". it wont' work for "z10_10.doc" and "z10_5.doc", etc.

netdur Over a year ago

return (atoi(*ia + 1) - atoi(*ib + 1));

Cœur · Accepted Answer · 2018-10-21 03:41:29Z

1

You want a natural sorting algorithm as opposed to the ASCIIbetical sorting algorithm provided by default. See Jeff's blog entry for a long rant about this, and some nice links to implementations of this natural algorithm. Just as a warning, the correct implementation (as opposed to the hacky answer you've accepted) is quite complicated, so don't try to implement it yourself, take someone else's (hopefully public domain or liberally licensed) implementation and use that instead.

edited Oct 21, 2018 at 3:41

Cœur

39k25 gold badges207 silver badges282 bronze badges

answered Nov 5, 2009 at 2:43

Chris Lutz

75.9k16 gold badges132 silver badges184 bronze badges

Collectives™ on Stack Overflow

How do I correctly sort strings in C?

3 Answers 3

2 Comments

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related