Softpanorama (slightly skeptical) Open Source Software Educational Society

May the source be with you, but remember the KISS principle ;-)

Unix sort command

The Unix sort command sorts ASCII files. It is not always wise to use it as options are obscure and specifying delimiters and sorting keys can be a nightmare. In a sense Unix sort can serve as a good example of programming of utilities in early 70th of the last century :-).

For small sets a good alternative is Perl sort  function which is more modern and more flexible.  I would say that in case of complex keys and for small files  Perl-based solution is almost always superior. But for large files (for example logs) Unix sort command might be the only tool that is able to handle jobs.

Unix sort can sort pretty large files. I successfully sorted proxy log files with the size over 10G using  Solaris sort on a pretty old V210 with 80G 10RPM drives and single 1.34GHz CPU.  

Lines need not to have the same length or the number of fields to be sorted successfully. All you need is the presence of the key.  

The input can be from files or the standard input.  In case of files Unix sort can accept as input one or several files. In the latter case all input files are merged. As for output, there is always a single file to be written -- sorted sequence of all input records:

The output can be a file or the standard output.

Data can be sorted in several ways:

• interpreting sorting keys alphabetically or numerically (-n option)
• in ascending or descending order  (-r -- sort in reverse option)

By default fields in each record (line of input) are delimited by blanks, but you can specify a different delimiter using option -t. You can also sort character columns (see below).

Selection of sorting keys is similar to selection of fields in cut and is extremely obscure. We have a separate page addressing this issue in depth (see UNIX sort keys definition for more details).  Be careful and always test your sorting keys on small sample before sorting a large file. As a rule of thump you can assume that no specification works from the first time. So testing it on a small sample is of paramount importance. 

The most common mistake is to forget to use -n option for sorting numeric fields. Also specifying delimiter (option -t) with an unquoted character after it can be a source of problems; it's better to use single quotes around the character that you plan to use as a delimiter. for example -t ':'

Here is a standard example of usage of the sort utility, sorting /etc/passwd file (user database) by UID (the third colon-separated field in the passwd file structure):

sort -t ':' +2 /etc/passwd  # incorrect result, the field is numeric
sort -n -t ':' +2 /etc/passwd # order of the numbers is now correct

sort -t ':' -k 3,3n /etc/passwd
sort -t ':' +2 -3n /etc/passwd
sort -n -t ':' -k 3,3 /etc/passwd

See Sorting key definitions and  Examples for more details.  Generally you will be surprised how often the result is not what you want due to the obscurity of the definitions

By default sort sorts the file in ascending order using the entire line as a sorting key.  Please note that a lot of WEB resources interpret this sort utility behavior incorrectly (most often they state that by default sorting is performed on the first key).

The three most important options of Unix soft are -n (numeric keys sorting), +n (sorting using n-th field, counting from zero) and -r (sort in reverse). For example:

Comparisons are based on one or more sort keys extracted from each line of input. Again, please remember that by default, there is one sort key, the entire input line.

Lines are ordered according to the collating sequence of the current locale. By changing locale you can change the behavior of the sort.

In Solaris there are two variants of sort: System V version and BSD version. Both have identical options:

• /usr/bin/sort [ -cmu ] [ -o output ] [ -T directory ] [ -y [ kmem ]] [ -z recsz ] [ -dfiMnr ] [ - b ] [ t char ] [ -k keydef ] [ +pos1 [ -pos2 ]] [ file...]
  
• /usr/xpg4/bin/sort [ -cmu ] [ -o output ] [ -T directory ] [ -y [ kmem ]] [ -z recsz ] [ - dfiMnr ] [ -b ] [ -t char ] [ -k keydef ] [ +pos1 [ -pos2 ]] [ file...]

The sort command can (and should) be used in pipes or have its output redirected as desired. Here are some practically important examples that illustrates using of this utility (for more examples please look into our sort examples collection page): 

1. Sort the file and display the sorted file, one page at a time. (If you prefer the standard command "more", you can use this instead of "less". "less" is an enhanced version of "more" - for example, it allows you to move backwards and forwards in the file; to exit "less" , use "q".)

   sort file | less
   

2. Read and sorts infile and display the result on standard output, which has been redirected to outfile:
   
   sort infile > outfile # sorts infile and writes the results to outfile
   
3. Write sorted output directly to  outfile.
   
   sort -o outfile infile  # same as in 1, but using an option -o
   
4. Read and sort infile  "in place" writing to the same file
   
   sort -o infile infile  # sort file "in place"
   
5. Sort the file and pipe it to uniq command with the number of identical keys counter printed (-c option in uniq)
   
   sort infile | uniq -c
   
6. Pipe the output of the who command to the input of the sort command:
   
   who | sort
   
7. Classic "number of instances" analysis of log files:
   
   cat messages | awk '{"select the keywords"}' | sort | uniq -c | sort -nr

   In simple cases cut can be used instead of AWK. For example the following example couts distinc visitors from HTTP logs (assuming this is the first field in the logs):

   cat http.log | cut -d " " -f 1 | sort | uniq -c | sort -nr
   

8. Sort the file, then prepend line numbers to each line (not many Unix adminitrators know that cat can be used to number lines):

   sort -n file | cat -n
   
   This can be useful if you want to count the number of lines in which the first entry is in a given range: simply subtract the line numbers corresponding to the beginning and end of the range.
   

As I mentioned about by default the sort command uses entire lines as a key. It compares the characters starting with the first, until non-matching character or the end of the shortest line. Leading blanks (spaces and tabs) are considered valid characters to compare. Thus a line beginning with a space precedes a line beginning with the letter A. If you do not want this effect you need to delete leading spaces beforehand.  

Multiple sort keys may be used on the same command line. If two lines have the same value in the same field, sort uses the next set of sort keys to resolve the equal comparison. For example,

     sort +4 -5 +1 -2 infile

means to sort based on field 5 (+4 -5). If two lines have the same value in field 5, sort those two lines based on field 2 (+1 -2).

Beside sorting Unix sort is useful for merging files (option -m).  It can also checked whether the file is sorted or not (option -c). It can also suppress duplicates (option -u):

•  -c     Check whether the given files are  already  sorted: if  they are not all sorted, print an error message and exit with a status of 1.
•  -m     Merge the given files by sorting them as  a  group.  Each  input  file  should  already  be individually sorted.  It always works to sort instead of  merge; merging  is  provided  because it is faster, in the  case where it works.
• -u      Suppress all but one occurrence of matching keys.

In case Unix sort does not produce the required results you might want to look into Perl built-in function. If it is too slow more memory can be specified on invocation.

The most important options

The following list describes the options and their arguments that may be used to control how sort functions.

• - Forces sort to read from the standard input. Useful for reading from pipes and files simultaneously.
• -c Verifies that the input is sorted according to the other options specified on the command line. If the input is sorted correctly then no output is provided. If the input is not sorted then sort informs you of the situation. The message resembles this.

• sort: disorder: This line not in sorted order.
  

• -m Merges the sorted input. sort assumes the input is already sorted. sort normally merges input as it sorts. This option informs sort that the input is already sorted, thus sort runs much faster.
• -o output Sends the output to file output instead of the standard output. The output file may be the same name as one of the input files.
• -u Suppress all but one occurrence of matching keys. Normally, the entire line is the key. If field or character keys are specified, then the suppressing is done based on the keys.
• -y kmem Use kmem kilobytes of main memory to initially start the sorting. If more memory is needed, sort automatically requests it from the operating system. The amount of memory allocated for the sort impacts the speed of the sort significantly. If no kmem is specified, sort starts with the default amount of memory (usually 32K). The maximum (usually 1 Megabyte) amount of memory may be allocated if needed. If 0 is specified for kmem, the minimum (usually 16K) amount of memory is allocated.
• -z recsz Specifies the record size used to store each line. Normally the recsz is set to the longest line read during the sort phase. If the -c or -m options are specified, the sort phase is not performed and thus the record size defaults to a system size. If this default size is not large enough, sort may abort during the merge phase. To alleviate this problem you can specify a recsz that will allow the merge phase to run without aborting.

Ordering Options

• -d Specifies dictionary sort. Only blanks, digits, and alphabetic characters are significant in the comparison.
• -f Fold lowercase letters to uppercase. Ignores the difference between upper and lowercase ASCII characters.
• -i Ignore characters outside the ASCII range of 040 (octal) and 0176 (octal). Only alphabetic characters, blanks, digits, and punctuation are used for comparison (printable characters). Control characters are ignored. This is only valid for nonnumeric sorts.
• -M Compare fields as months. The first three nonblank characters are folded (see -i option) to uppercase and compared. Thus January is compared as JAN. JAN precedes FEB, and fields not containing months precede JAN. The -b option is placed in effect automatically.
• -n Sorts the input numerically. The comparison is based on numerical value instead of alphabetic value. The number field used for comparison can contain optional blanks, optional minus signs, optional decimal point, and zero or more digits. The -b option is placed in effect automatically. Exponential numbers are not sorted correctly.
• -r Reverse the order of the output.

Old-style Sort Key Options

Old News ;-)

[Jul 14, 2007] My SysAd Blog -- UNIX Sort Files by Their Filesizes

Here's a convenient way of finding those space hogs in your home directory (can be any directory). For me, those large files are usually a result of mkfile event (testing purposes) and can be promptly deleted. Here's an example of its use.

#cd /export/home/esofthub
#ls -l | sort +4n | awk '{print $5 "\t" $9}'

Find recursively (a little awkward)
#ls -lR | sort +4n | awk '{print $5 "\t" $9}' | more

[Jul 14, 2007] Learn Unix The sort command

ps -ef | sort

This command pipeline sorts the output of the "ps -ef" command. Because no arguments are supplied to the sort command, the output is sorted in alphabetic order by the first column of the ps -ef output (i.e., the output is sorted alphabetically by username).

ls -al | sort +4n

This command performs a numeric sort on the fifth column of the "ls -al" output. This results in a file listing where the files are listed in ascending order, from smallest in size to largest in size.

ls -al | sort +4n | more

The same command as the previous, except the output is piped into the more command. This is useful when the output will not all fit on one screen.

ls -al | sort +4nr

This command reverses the order of the numeric sort, so files are listed in descending order of size, with the largest file listed first, and the smallest file listed last.

[Feb 15, 2007] UNIX Disk Usage Simplifying Analysis with sort

The output of du has been very informative, but it's difficult to scan a listing to ascertain the four or five largest directories, particularly as more and more directories and files are included in the output. The good news is that the Unix sort utility is just the tool we need to sidestep this problem.

# du -s * | sort -nr
13984  Lynx
10464  IBM
3092   Gator
412    bin
196    DEMO
84     etcpasswd
76     CBO_MAIL
48     elance
36     CraigsList
16     Exchange
4      gettermsheet.sh
4      getstocks.sh
4      getmodemdriver.sh
4      buckaroo
4      browse.sh
4      badjoke.rot13
4      badjoke
0      gif.gif

One final concept and we're ready to move along. If you want to only see the five largest files or directories in a specific directory, all that you'd need to do is pipe the command sequence to head:

# du -s * | sort -nr | head -5
13984  Lynx
10464  IBM
3092   Gator
412    bin
196    DEMO

[Feb 14, 2007] UNIX Power Tools, 3rd Edition Examples!

The ! command (pronounced "bang") creates a temporary file to be used with a program that requires a filename in its command line. This is useful with shells that don't support process substitution. For example, to diff two files after sorting them, you might do:

diff `! sort file1` `! sort file2`

commer

commer is a shell script that uses comm to compare two sorted files; it processes comm's output to make it easier to read. (See article 11.9.)
[Overview] [List]

lensort

lensort sorts lines from shortest to longest. (See article 22.7.)
[Overview] [List]

namesort

The namesort program sorts a list of names by the last name. (See article 22.8.) See also namesort.pl.
[Overview] [List]

namesort.pl

The namesort.pl script uses the Perl module Lingua::EN::NameParse to sort a list of names by the last name. (See article 22.8.) See also namesort.
[Overview] [List]

[Feb 11, 2007] Learn Unix The sort command

This command pipeline sorts the output of the "ps -ef" command. Because no arguments are supplied to the sort command, the output is sorted in alphabetic order by the first column of the ps -ef output (i.e., the output is sorted alphabetically by username).

ls -al | sort +4n

This command performs a numeric sort on the fifth column of the "ls -al" output. This results in a file listing where the files are listed in ascending order, from smallest in size to largest in size.

ls -al | sort +4n | more

The same command as the previous, except the output is piped into the more command. This is useful when the output will not all fit on one screen.

ls -al | sort +4nr

This command reverses the order of the numeric sort, so files are listed in descending order of size, with the largest file listed first, and the smallest file listed last.

[Feb 10, 2007] using -t option with unix sort

Daniel Malaby dan at malaby.com
Thu Jul 14 08:02:47 GMT 2005

• Previous message: DHCP assigned unregistered IP address
• Next message: using -t option with unix sort ?
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hi All,

I am trying to sort a tab delimited file with sort. The problem I am
having is with the -t option. I do not know how to pass a tab.

Things I have tried:

sort -t \t
sort -t '\t'
sort -t "\t"
sort -t 0x09
sort -t '0x09'
sort -t "0x09"
sort -t ^I
sort -t '^I'
sort -t "^I"

Any suggestions would be much appreciated.

Thanks

Daniel Malaby dan at malaby.com 
Thu Jul 14 17:48:05 GMT 2005 

Nelis Lamprecht wrote:
> On 7/14/05, Nelis Lamprecht <nlamprecht at gmail.com> wrote:
> 
>>On 7/14/05, Daniel Malaby <dan at malaby.com> wrote:
>>
>>>Hi All,
>>>
>>>I am trying to sort a tab delimited file with sort. The problem I am
>>>having is with the -t option. I do not know how to pass a tab.
>>
>><snip>
>>
>>>sort -t \t
>>
>></snip>
>>
>>>Any suggestions would be much appreciated.
>>
>>remove the space between -t and \t and it should work
> 
> 
> actually scratch that, it works either way. can you give a sample of the data ?
> 
> Regards,
> Nelis

The sample data has 9 fields, I am trying to sort on the fifth field, 
here is what I have tried.

sort -t\t +4 -5 -o test.txt sample.txt

I did try removing the space and it did not work, I have also tried 
removing the -5. I think the spaces in the third field are confusing sort.

BTW this is being done on a PC running FBSD 4.11 prerelease #1

Thanks for your help and suggestions.

-------------- next part --------------
E002	19085	GENERAL DYNAMICS	5031802	E-GL/VX/B/R1.0	SFT CD, GL VXWORKS BOREALIS R1.0	06/30/05	     1	$995.00	$995.00
E016	19096	TGA INGENIERIA Y ELECTRONICS S	5881-2	E-AD600729C501	ARGUS PMC,2 DVI 16MB PERCHAN USB A/V	12/01/05	    30	$2,312.00	$69,360.00
E016	19096	TGA INGENIERIA Y ELECTRONICS S	5881-2	E-DDX/SO/R4.0	SFT CD, DDX SOL 2.6-9 BOREALIS R4.0	12/01/05	    30	$74.00	$2,220.00
E016	19096	TGA INGENIERIA Y ELECTRONICS S	5881-2	E-VIN/SO/R1.0	SFT CD, VID CAP SOL 2.6-9 BOREALIS R1.0	12/01/05	    30	$74.00	$2,220.00
E021	19093	GANYMED COMPUTER GMBH	7103879	E-AD90073913011	GARNET PMC RIO8 C2, REAR I/O 16MB	07/19/05	     2	$1,848.00	$3,696.00
E024	19080	DRS LAUREL TECHNOLOGIES	94358	E-AC7007121115A	ECLIPSE3 PMC, VGA 16MB Q70	08/18/05	     1	$846.00	$846.00
E024	19080	DRS LAUREL TECHNOLOGIES	94358	E-AC7007121115A	ECLIPSE3 PMC, VGA 16MB Q70	10/19/05	    19	$846.00	$16,074.00
E024	19080	DRS LAUREL TECHNOLOGIES	94358	E-AC7007121115A	ECLIPSE3 PMC, VGA 16MB Q70	09/20/05	     2	$846.00	$1,692.00
E024	19080	DRS LAUREL TECHNOLOGIES	94358	E-AC7007121115A	ECLIPSE3 PMC, VGA 16MB Q70	11/17/05	     7	$846.00	$5,922.00

[Jan 30, 2007] lf131, UNIX Basics GNU file utilities

Let´s assume that we want to sort /etc/passwd using the geco field. To achieve this, we will use sort, the unix sorting tool

$ sort -t: +4 /etc/passwd
murie:x:500:500:Manuel Muriel Cordero:/home/murie:/bin/bash
practica:x:501:501:Usuario de practicas para Ksh:/home/practica:/bin/ksh
wizard:x:502:502:Wizard para nethack:/home/wizard:/bin/bash
root:x:0:0:root:/root:/bin/bash

It is very easy to see that the file has been sorted, but using the ASCII table order. If we don´t want to make a difference among capital letter, we can use:

$ sort -t: +4f  /etc/passwd
murie:x:500:500:Manuel Muriel Cordero:/home/murie:/bin/bash
root:x:0:0:root:/root:/bin/bash
practica:x:501:501:Usuario de practicas para Ksh:/home/practica:/bin/ksh
wizard:x:502:502:Wizard para nethack:/home/wizard:/bin/bash

-t is the option to select the field separator. +4 stands for the number of field to jump before ordering the lines, and f means to sort regardless of upper and lowercase.

A much more complicated sort can be achieved. For example, we can sort using the shell in a first step then sort using the geco:

$ sort -t: +6r +4f /etc/passwd
practica:x:501:501:Usuario de practicas para Ksh:/home/practica:/bin/ksh
murie:x:500:500:Manuel Muriel Cordero:/home/murie:/bin/bash
root:x:0:0:root:/root:/bin/bash
wizard:x:502:502:Wizard para nethack:/home/wizard:/bin/bash

You have a file with some people you lend money and the amount of money you gave them. Take ´deudas.txt´ as an example:

Son Goku:23450
Son Gohan:4570
Picolo:356700
Ranma 1/2:700

If you want to know the first one to ´visit´, you need a sorted list.
Just type

$ sort +1 deudas
Ranma 1/2:700
Son Gohan:4570
Son Goku:23450
Picolo:356700

which is not the desired result because the number of fields is not the same across the file. The solution is the ´n´ option:

$ sort +1n deudas
Picolo:356700
Son Goku:23450
Son Gohan:4570
Ranma 1/2:700

Basic options for sort are
+n.m jumps over the first n fields and the next m characters before begin the sort
-n.m stops the sorting when arriving to the m-th character of the n-th field

The following are modification parameters:

-b jumps over leading whitespaces
-d dictionary sort (just using letters, numbers and whitespace)
-f ignores case distinction
-n sort numerically
-r reverse order

[Jan 22, 2007]  Oracle UNIX pipe command

For example, suppose we want to list the distinct file owners in a directory. To do this, we must perform three discrete tasks:

1. We must list all files in the directory (ls –al)
2. We must parse this output and extract the file owner from the fourth column of the output. (awk ‘{ print $3 }’)
3. We must then take the list of file owners and remove duplicate entries (sort –u)

Using the pipe command, we can tie these three functions together into a single UNIX command, piping the output from one command as sending it as input to the next UNIX command:

root> ls -al|awk '{ print $3 }'|sort -u

marion
oracle
root

[Jul 3, 2005] anagram -- an interesting use of sort (sorting letters)

% awk '{ print NF " " $0}' < out | sort -n | tail

[Sep 16 2004] loganalysis 2004-09 Re [logs] Faster unix 'sort' replacement

From: Ed Schmollinger (schmolli@private)
Date: Thu Sep 16 2004 - 09:14:32 PDT

• Previous message: Marcus J. Ranum: "Re: [logs] Faster unix 'sort' replacement?"
• In reply to: Mike Blomgren: "[logs] Faster unix 'sort' replacement?"
• Next in thread: Mike Blomgren: "RE: [logs] Faster unix 'sort' replacement?"
• Reply: Mike Blomgren: "RE: [logs] Faster unix 'sort' replacement?"
• Reply: cadams@private: "Re: [logs] Faster unix 'sort' replacement?"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

---

On Thu, Sep 16, 2004 at 12:33:12AM +0200, Mike Blomgren wrote:
> I'm having trouble with 'sort' taking alot of cpu-time on a Solaris machine,
> and I'm wondering if anyone knows of a replacement for the gnu 'sort'
> command, which is faster and will compile on Solaris and preferably Linux
> too?
> 
> I'm using sort in the standard 'cat <file> | awk '{"compute..."}' | sort |
> uniq -c | sort -n -r' type analysis.

You can get rid of the multiple sorts/uniq thing by doing it all at
once:

--- CUT HERE ---
#!/usr/bin/perl -wT

use strict;

my %msg = ();

while (<>) { chomp; $msg{$_} = $msg{$_} ? $msg{$_} + 1 : 1; }

for(sort { $msg{$a} <=> $msg{$b} } keys %msg) { print "$msg{$_}\t$_\n"; }
--- CUT HERE ---

I've found that for my datasets, the awk/sed stage is what constitues
the bulk of the bottleneck.  You may want to look at optimizing that
part as well.

-- 
Ed Schmollinger - schmolli@private

[Sep 20, 2004] Faster unix 'sort' replacement

Russell Fulton r.fulton at auckland.ac.nz
Mon Sep 20 13:13:59 MDT 2004

• Previous message: [logs] Faster unix 'sort' replacement?
• Next message: [logs] Faster unix 'sort' replacement?
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

---

On Fri, 2004-09-17 at 05:59, Mike Blomgren wrote:
> Thanks for the tip -  I'll have to try that one with perl doing the sort
> instead of gnu sort. I have been somewhat reluctant to use perl since I find
> it has a severe performance impact in some cases - but that may be related
> to my regexp's and not the sorting. For a fact though, I do know that using
> associative arrays is a good way to consume memory in a hurry. And thus
> causing the os to start swapping memory to disk, which is not very
> beneficial for speed, to say the least... 

If you are short of memory sort may be swapping stuff out to disk and
hence your performance problems.  It depends on the implementations but
some sorts are smart enough to work out how much memory is really
available and then do sort & merges with in this. This is much better
than sorts that simply assume that virtual memory is endless and cause
the OS to thrash madly but is much slower than doing the whole thing in
memory.

This will not show up as OS level swapping though, just as lots of disk
activity during the sort.

-- 
Russell Fulton, Information Security Officer, The University of Auckland
New Zealand

Recommended Links

Internal

• Softpanorama Unix Sort Examples

• Unix Sort Command Sorting Fields (Keys) Definitions

• Sorting Algorithms

External

• sort (Unix) - Wikipedia, the free encyclopedia

• docs.sun.com man pages section 1 User Commands Sun man page

• Linux and UNIX sort command help
• UNIX man pages - sort ()
• A UNIX Survival Guide for DOS Users
• eXtropia Tutorials Introduction to UNIX for Web Technicians The sort Utility
• Unix commands (math.uiuc.edu/~hildeb)
• Native Win32 ports of some GNU utilities
• UNIX Disk Usage Simplifying Analysis with sort

GNU sort command options

       Write sorted concatenation of all FILE(s) to standard output.

       Mandatory  arguments  to  long  options are mandatory for short options
       too.  Ordering options:

       -b, --ignore-leading-blanks
      ignore leading blanks

       -d, --dictionary-order
      consider only blanks and alphanumeric characters

       -f, --ignore-case
      fold lower case to upper case characters

       -g, --general-numeric-sort
      compare according to general numerical value

       -i, --ignore-nonprinting
      consider only printable characters

       -M, --month-sort
      compare (unknown) < `JAN' < ... < `DEC'

       -n, --numeric-sort
      compare according to string numerical value

       -r, --reverse
      reverse the result of comparisons

       Other options:

       -c, --check
      check whether input is sorted; do not sort

       -k, --key=POS1[,POS2]
      start a key at POS1, end it at POS2 (origin 1)

       -m, --merge
      merge already sorted files; do not sort

       -o, --output=FILE   write result to FILE instead of standard output

       -s, --stable    stabilize sort by disabling last-resort comparison

       -S, --buffer-size=SIZE    use SIZE for main memory buffer

       -t, --field-separator=SEP    use SEP instead of non-blank to blank transition

       -T, --temporary-directory=DIR    use DIR for temporaries, not $TMPDIR or /tmp;  multiple  options
      specify multiple directories

       -u, --unique   with  -c, check for strict ordering; without -c, output only the
      first of an equal run

       -z, --zero-terminated    end lines with 0 byte, not newline

       --help display this help and exit

       --version     output version information and exit

       POS is F[.C][OPTS], where F is the field number and  C the  character
       position  in  the  field.   OPTS  is one or more single-letter ordering
       options, which override global ordering options for that  key. If  no
       key is given, use the entire line as the key.

       SIZE  may be followed by the following multiplicative suffixes: % 1% of
       memory, b 1, K 1024 (default), and so on for M, G, T, P, E, Z, Y.

       With no FILE, or when FILE is -, read standard input.

       *** WARNING *** The locale specified by the  environment  affects  sort
       order.  Set LC_ALL=C to get the traditional sort order that uses native
       byte values.

Solaris sort command options

/usr/bin/sort

-c

Check that the single input file is ordered as specified by the arguments and the collating sequence of the current locale. The exit code is set and no output is produced unless the file is out of sort.

-c

Same as /usr/bin/sort except no output is produced under any circumstances.

-m

Merge only. The input files are assumed to be already sorted.

-u

Unique: suppress all but one in each set of lines having equal keys. If used with the -c option, check that there are no lines with duplicate keys in addition to checking that the input file is sorted.

-o output

Specify the name of an output file to be used instead of the standard output. This file can be the same as one of the input files.

-T directory

The directory argument is the name of a directory in which to place temporary files.

-y kmem

The amount of main memory initially used by sort. If this option is omitted, sort begins using a system default memory size, and continues to use more space as needed. If kmem is present, sort will start using that number of Kbytes of memory, unless the administrative minimum or maximum is exceeded, in which case the corresponding extremum will be used. Thus, -y 0 is guaranteed to start with minimum memory. -y with no kmem argument starts with maximum memory.

-z recsz

(obsolete). This option was used to prevent abnormal termination when lines longer than the system-dependent default buffer size are encountered. Because sort automatically allocates buffers large enough to hold the longest line, this option has no effect.

-d

``Dictionary'' order: only letters, digits, and blanks (spaces and tabs) are significant in comparisons.

-f

Fold lower-case letters into upper case.

-i

Ignore non-printable characters.

-M

Compare as months. The first three non-blank characters of the field are folded to upper case and compared. For example, in English the sorting order is "JAN" < "FEB" < ... < "DEC". Invalid fields compare low to "JAN". The - M option implies the -b option (see below).

-n

Restrict the sort key to an initial numeric string, consisting of optional blank characters, optional minus sign, and zero or more digits with an optional radix character and thousands separators (as defined in the current locale), which will be sorted by arithmetic value. An empty digit string is treated as zero. Leading zeros and signs on zeros do not affect ordering.

-r

Reverse the sense of comparisons.

-b

Ignore leading blank characters when determining the starting and ending positions of a restricted sort key. If the -b option is specified before the first sort key option, it is applied to all sort key options. Otherwise, the -b option can be attached independently to each -k field_start, field_end, or +pos1 or -pos2 option-argument (see below).

-t char

Use char as the field separator character. char is not considered to be part of a field (although it can be included in a sort key). Each occurrence of char is significant (for example, <char><char> delimits an empty field). If - t is not specified, blank characters are used as default field separators; each maximal non-empty sequence of blank characters that follows a nonblank character is a field separator.

-k keydef

The keydef argument is a restricted sort key field definition. The format of this definition is: -k field_start [ type ] [ ,field_end [ type ] ] where:

type

is a modifier from the list of characters bdfiMnr. The b modifier behaves like the -b option, but applies only to the field_start or field_end to which it is attached and characters within a field are counted from the first nonblank character in the field. (This applies separately to first_character and last_character.) The other modifiers behave like the corresponding options, but apply only to the key field to which they are attached. They have this effect if specified with field_start, field_end or both. If any modifier is attached to a field_start or to a field_end, no option applies to either.

The notation:

-k field_start[type][,field_end[type]]

A field comprises a maximal sequence of nonseparating characters and, in the absence of option -t, any preceding field separator.

The field_start portion of the keydef optionargument has the form:

field_number[.first_character]

Fields and characters within fields are numbered starting with 1. field_number and first_character, interpreted as positive decimal integers, specify the first character to be used as part of a sort key. If .first_character is omitted, it refers to the first character of the field.

The field_end portion of the keydef optionargument has the form:

field_number[.last_character]

The field_number is as described above for field_start. last_character, interpreted as a non-negative decimal integer, specifies the last character to be used as part of the sort key. If last_character evaluates to zero or .last_character is omitted, it refers to the last character of the field specified by field_number.

If the -b option or b type modifier is in effect, characters within a field are counted from the first non-blank character in the field. (This applies separately to first_character and last_character.)

[+pos1[-pos2]]
(obsolete). Provide functionality equivalent to the -k keydef option.

pos1 and pos2 each have the form m.n optionally followed by one or more of the flags bdfiMnr. A starting position specified by +m.n is interpreted to mean the n+1st character in the m+1st field. A missing .n means .0, indicating the first character of the m+1st field. If the b flag is in effect n is counted from the first non-blank in the m+1st field; +m.0b refers to the first nonblank character in the m+1st field.

A last position specified by -m.n is interpreted to mean the nth character (including separators) after the last character of the mth field. A missing .n means .0, indicating the last character of the mth field. If the b flag is in effect n is counted from the last leading blank in the m+1st field; -m.1b refers to the first non-blank in the m+1st field.

The fully specified +pos1 - pos2 form with type modifiers T and U:
+w.xT -y.zU

is equivalent to:

undefined

(z==0 & U contains b & -t is present)

-k w+1.x+1T,y.0U

(z==0 otherwise)

-k w+1.x+1T,y+1.zU

(z > 0)

Exit status

• 0 All input files were output successfully, or - c was specified and the input file was correctly sorted.
• 1 Under the -c option, the file was not ordered as specified, or if the -c and -u options were both specified, two input lines were found with equal keys.
• >1 An error occurred.

Etc

anagram -- an interesting use of sort

% awk '{ print NF " " $0}' < out | sort -n | tail

Created May 16, 2006. Last modified: September 15, 2008