prog.c

#define _GNU_SOURCE
#include <ctype.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <unistd.h>

#define CHUNK 4096

int main(int argc, char *argv[]) {
  int bufsize = CHUNK;
  int empty = bufsize;
  char *data = malloc(bufsize);
  int size;
  while (1) {
    char *ptr = data + bufsize - empty;
    int result = read(0, ptr, empty);
    if (result == empty) {
      data = realloc(data, bufsize * 2);
      if (data == NULL) {
	fprintf(stderr, "not enough space\n");
	exit(-1);
      }
      empty = bufsize;
      bufsize *= 2;
    } else {
      size = bufsize - empty + result;
      break;
    }
  }

  char *first = malloc(size);
  char *second = malloc(size);
  char *first_ptr = first, *second_ptr = second;
  char *end = data + size - 1;
  if (*end != '\n')
    *++end = '\n';
  *++end = 'x';

  if (first == NULL || second == NULL) {
    fprintf(stderr, "out of space\n");
    exit(-1);
  }

  int count = 0;

  fprintf(stderr, "Got %d bytes\n", size);

  char *ptr = data;
  while (ptr < end) {
    do
      *first_ptr++ = *ptr++;
    while (isalnum(*ptr));
    *first_ptr++ = ' ';
    while (!isalnum(*ptr))
      ptr++;
    while (isalnum(*ptr))
      *second_ptr++ = *ptr++;
    *second_ptr++ = '\n';
    while (!isalnum(*ptr))
      ptr++;
    count++;
  }

  free(data);

  char out_buf[2][CHUNK];
  char *out_ptr = *out_buf;

  char *head = first;
  int line_number = 0;
  for (line_number = 0; line_number < count; line_number++) {
    //fprintf(stderr, "Line %d/%d\n", line_number+1, count);
    int first_len = (char *) rawmemchr(head, ' ') - head;
    char *tail = second;
    int i;
    for (i = 0; i < count; i++) {
      char *eol = rawmemchr(tail, '\n');
      // fprintf(stderr, "%.*s%.*s", first_len, head, eol-tail+1, tail);
      memcpy(out_ptr, head, first_len);
      out_ptr += first_len;
      memcpy(out_ptr, tail, eol-tail+1);
      out_ptr += eol-tail+1;
      if (out_ptr - *out_buf >= CHUNK) {
	write(1, *out_buf, sizeof(*out_buf));
	memcpy(*out_buf, out_buf[1], out_ptr - *out_buf - CHUNK);
	out_ptr -= CHUNK;
      }
      tail = eol+1;
    }

    //fprintf(stderr, "done\n");

    head = rawmemchr(head, ' ') + 1;
  }

  write(1, *out_buf, out_ptr - *out_buf);
}