uniq -d选项源代码分析

本文主要是介绍uniq -d选项源代码分析，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！

csdn上还真有人分析这个命令的源代码。

其实不容易啦。

最后一个

d

d

d

d

其实是用得check_file函数里面的//457行的

writeline函数。

 

[root@localhost src]# gdb ./uniq
GNU gdb (GDB) Red Hat Enterprise Linux 7.6.1-94.el7
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-redhat-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /root/coreutils-8.22/src/uniq...done.
(gdb) set args -d 1.txt
(gdb) b 457
Breakpoint 1 at 0x401ee3: file src/uniq.c, line 457.
(gdb) r
Starting program: /root/coreutils-8.22/src/./uniq -d 1.txt
b
c

Breakpoint 1, check_file (delimiter=10 '\n', outfile=<optimized out>, infile=0x7fffffffe55b "1.txt") at src/uniq.c:457
457           writeline (prevline, false, match_count);
Missing separate debuginfos, use: debuginfo-install glibc-2.17-323.el7_9.x86_64
(gdb) c
Continuing.
d
[Inferior 1 (process 12122) exited normally]
(gdb)

 

[root@localhost src]# gdb ./uniq
GNU gdb (GDB) Red Hat Enterprise Linux 7.6.1-94.el7
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-redhat-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /root/coreutils-8.22/src/uniq...done.
(gdb) set args -c 1.txt
(gdb) b 457
Breakpoint 1 at 0x401ee3: file src/uniq.c, line 457.
(gdb) c
The program is not being run.
(gdb) r
Starting program: /root/coreutils-8.22/src/./uniq -c 1.txt
      1 a
      2 b
      3 c

Breakpoint 1, check_file (delimiter=10 '\n', outfile=<optimized out>, infile=0x7fffffffe55b "1.txt") at src/uniq.c:457
457           writeline (prevline, false, match_count);
Missing separate debuginfos, use: debuginfo-install glibc-2.17-323.el7_9.x86_64
(gdb) c
Continuing.
      4 d
[Inferior 1 (process 12128) exited normally]
(gdb)

 

---

static void
check_file (const char *infile, const char *outfile, char delimiter)
{
  struct linebuffer lb1, lb2;
  struct linebuffer *thisline, *prevline;

  if (! (STREQ (infile, "-") || freopen (infile, "r", stdin)))
    error (EXIT_FAILURE, errno, "%s", infile);
  if (! (STREQ (outfile, "-") || freopen (outfile, "w", stdout)))
    error (EXIT_FAILURE, errno, "%s", outfile);

  fadvise (stdin, FADVISE_SEQUENTIAL);

  thisline = &lb1;
  prevline = &lb2;

  initbuffer (thisline);
  initbuffer (prevline);

  /* The duplication in the following 'if' and 'else' blocks is an
     optimization to distinguish between when we can print input
     lines immediately (1. & 2.) or not.

     1. --group => all input lines are printed.
        checking for unique/duplicated lines is used only for printing
        group separators.

     2. The default case in which none of these options has been specified:
          --count, --repeated,  --all-repeated, --unique
        In the default case, this optimization lets uniq output each different
        line right away, without waiting to see if the next one is different.

     3. All other cases.
  */
  if (output_unique && output_first_repeated && countmode == count_none)
    {
      char *prevfield IF_LINT ( = NULL);
      size_t prevlen IF_LINT ( = 0);
      bool first_group_printed = false;

      while (!feof (stdin))
        {
          char *thisfield;
          size_t thislen;
          bool new_group;

          if (readlinebuffer_delim (thisline, stdin, delimiter) == 0)
            break;

          thisfield = find_field (thisline);
          thislen = thisline->length - 1 - (thisfield - thisline->buffer);

          new_group = (prevline->length == 0
                       || different (thisfield, prevfield, thislen, prevlen));

          if (new_group && grouping != GM_NONE
              && (grouping == GM_PREPEND || grouping == GM_BOTH
                  || (first_group_printed && (grouping == GM_APPEND
                                              || grouping == GM_SEPARATE))))
            putchar (delimiter);

          if (new_group || grouping != GM_NONE)
            {
              fwrite (thisline->buffer, sizeof (char),
                      thisline->length, stdout);

              SWAP_LINES (prevline, thisline);
              prevfield = thisfield;
              prevlen = thislen;
              first_group_printed = true;
            }
        }
      if ((grouping == GM_BOTH || grouping == GM_APPEND) && first_group_printed)
        putchar (delimiter);
    }
  else
    {
      char *prevfield;
      size_t prevlen;
      uintmax_t match_count = 0;
      bool first_delimiter = true;

      if (readlinebuffer_delim (prevline, stdin, delimiter) == 0)
        goto closefiles;
      prevfield = find_field (prevline);
      prevlen = prevline->length - 1 - (prevfield - prevline->buffer);

      while (!feof (stdin))
        {
          bool match;
          char *thisfield;
          size_t thislen;
          if (readlinebuffer_delim (thisline, stdin, delimiter) == 0)
            {
              if (ferror (stdin))
                goto closefiles;
              break;
            }
          thisfield = find_field (thisline);
          thislen = thisline->length - 1 - (thisfield - thisline->buffer);
          match = !different (thisfield, prevfield, thislen, prevlen);
          match_count += match;

          if (match_count == UINTMAX_MAX)
            {
              if (count_occurrences)
                error (EXIT_FAILURE, 0, _("too many repeated lines"));
              match_count--;
            }

          if (delimit_groups != DM_NONE)
            {
              if (!match)
                {
                  if (match_count) /* a previous match */
                    first_delimiter = false; /* Only used when DM_SEPARATE */
                }
              else if (match_count == 1)
                {
                  if ((delimit_groups == DM_PREPEND)
                      || (delimit_groups == DM_SEPARATE
                          && !first_delimiter))
                    putchar (delimiter);
                }
            }

          if (!match || output_later_repeated)
            {
              writeline (prevline, match, match_count);//448行
              SWAP_LINES (prevline, thisline);
              prevfield = thisfield;
              prevlen = thislen;
              if (!match)
                match_count = 0;
            }
        }

      writeline (prevline, false, match_count);//457行

//457行是不属于上面的while循环的。和上面的while循环是并列关系。
    }

 closefiles:
  if (ferror (stdin) || fclose (stdin) != 0)
    error (EXIT_FAILURE, 0, _("error reading %s"), infile);

  /* stdout is handled via the atexit-invoked close_stdout function.  */

  free (lb1.buffer);
  free (lb2.buffer);
}

 

---

最后-c选项都是最大7位的统计次数。

(gdb) s
312         printf ("%7" PRIuMAX " ", linecount + 1);
(gdb) s
printf (__fmt=0x405c77 "%7lu ") at /usr/include/bits/stdio2.h:104
104       return __printf_chk (__USE_FORTIFY_LEVEL - 1, __fmt, __va_arg_pack ());
(gdb) n
writeline (match=<optimized out>, linecount=<optimized out>, line=<optimized out>, line=<optimized out>)
    at src/uniq.c:314
314       fwrite (line->buffer, sizeof (char), line->length, stdout);
(gdb)

这篇关于uniq -d选项源代码分析的文章就介绍到这儿，希望我们推荐的文章对编程师们有所帮助！

---