1 cfa37a7b 2004-04-10 devnull .TH JOIN 1
  2 cfa37a7b 2004-04-10 devnull .CT 1 files
  3 cfa37a7b 2004-04-10 devnull .SH NAME
  4 cfa37a7b 2004-04-10 devnull join \- relational database operator
  5 cfa37a7b 2004-04-10 devnull .SH SYNOPSIS
  6 cfa37a7b 2004-04-10 devnull .B join
  7 cfa37a7b 2004-04-10 devnull [
  8 cfa37a7b 2004-04-10 devnull .I options
  9 cfa37a7b 2004-04-10 devnull ]
 10 cfa37a7b 2004-04-10 devnull .I file1 file2
 11 cfa37a7b 2004-04-10 devnull .SH DESCRIPTION
 12 cfa37a7b 2004-04-10 devnull .I Join
 13 cfa37a7b 2004-04-10 devnull forms, on the standard output,
 14 cfa37a7b 2004-04-10 devnull a join
 15 cfa37a7b 2004-04-10 devnull of the two relations specified by the lines of
 16 cfa37a7b 2004-04-10 devnull .I file1
 17 cfa37a7b 2004-04-10 devnull and
 18 cfa37a7b 2004-04-10 devnull .IR file2 .
 19 cfa37a7b 2004-04-10 devnull If one of the file names is
 20 cfa37a7b 2004-04-10 devnull .LR - ,
 21 cfa37a7b 2004-04-10 devnull the standard input is used.
 22 cfa37a7b 2004-04-10 devnull .PP
 23 cfa37a7b 2004-04-10 devnull .I File1
 24 cfa37a7b 2004-04-10 devnull and
 25 cfa37a7b 2004-04-10 devnull .I file2
 26 cfa37a7b 2004-04-10 devnull must be sorted in increasing
 27 cfa37a7b 2004-04-10 devnull .SM ASCII 
 28 cfa37a7b 2004-04-10 devnull collating
 29 cfa37a7b 2004-04-10 devnull sequence on the fields
 30 cfa37a7b 2004-04-10 devnull on which they are to be joined,
 31 cfa37a7b 2004-04-10 devnull normally the first in each line.
 32 cfa37a7b 2004-04-10 devnull .PP
 33 cfa37a7b 2004-04-10 devnull There is one line in the output
 34 cfa37a7b 2004-04-10 devnull for each pair of lines in
 35 cfa37a7b 2004-04-10 devnull .I file1
 36 cfa37a7b 2004-04-10 devnull and
 37 cfa37a7b 2004-04-10 devnull .I file2
 38 cfa37a7b 2004-04-10 devnull that have identical join fields.
 39 cfa37a7b 2004-04-10 devnull The output line normally consists of the common field,
 40 cfa37a7b 2004-04-10 devnull then the rest of the line from
 41 cfa37a7b 2004-04-10 devnull .IR file1 ,
 42 cfa37a7b 2004-04-10 devnull then the rest of the line from
 43 cfa37a7b 2004-04-10 devnull .IR file2 .
 44 cfa37a7b 2004-04-10 devnull .PP
 45 cfa37a7b 2004-04-10 devnull Input fields are normally separated spaces or tabs;
 46 cfa37a7b 2004-04-10 devnull output fields by space.
 47 cfa37a7b 2004-04-10 devnull In this case, multiple separators count as one, and
 48 cfa37a7b 2004-04-10 devnull leading separators are discarded.
 49 cfa37a7b 2004-04-10 devnull .PP
 50 cfa37a7b 2004-04-10 devnull The following options are recognized, with POSIX syntax.
 51 cfa37a7b 2004-04-10 devnull .TP
 52 cfa37a7b 2004-04-10 devnull .BI -a " n
 53 cfa37a7b 2004-04-10 devnull In addition to the normal output,
 54 cfa37a7b 2004-04-10 devnull produce a line for each unpairable line in file
 55 cfa37a7b 2004-04-10 devnull .IR n ,
 56 cfa37a7b 2004-04-10 devnull where
 57 cfa37a7b 2004-04-10 devnull .I n
 58 cfa37a7b 2004-04-10 devnull is 1 or 2.
 59 cfa37a7b 2004-04-10 devnull .TP
 60 cfa37a7b 2004-04-10 devnull .BI -v " n
 61 cfa37a7b 2004-04-10 devnull Like
 62 cfa37a7b 2004-04-10 devnull .BR -a ,
 63 cfa37a7b 2004-04-10 devnull omitting output for paired lines.
 64 cfa37a7b 2004-04-10 devnull .TP
 65 cfa37a7b 2004-04-10 devnull .BI -e " s
 66 cfa37a7b 2004-04-10 devnull Replace empty output fields by string
 67 cfa37a7b 2004-04-10 devnull .IR s .
 68 cfa37a7b 2004-04-10 devnull .TP
 69 cfa37a7b 2004-04-10 devnull .BI -1 " m
 70 cfa37a7b 2004-04-10 devnull .br
 71 cfa37a7b 2004-04-10 devnull .ns
 72 cfa37a7b 2004-04-10 devnull .TP
 73 cfa37a7b 2004-04-10 devnull .BI -2 " m
 74 cfa37a7b 2004-04-10 devnull Join on the
 75 cfa37a7b 2004-04-10 devnull .IR m th
 76 cfa37a7b 2004-04-10 devnull field of
 77 cfa37a7b 2004-04-10 devnull .I file1
 78 cfa37a7b 2004-04-10 devnull or
 79 cfa37a7b 2004-04-10 devnull .IR file2 .
 80 cfa37a7b 2004-04-10 devnull .TP
 81 cfa37a7b 2004-04-10 devnull .BI -j "n m"
 82 cfa37a7b 2004-04-10 devnull Archaic equivalent for
 83 cfa37a7b 2004-04-10 devnull .BI - n " m"\f1.
 84 cfa37a7b 2004-04-10 devnull .TP
 85 cfa37a7b 2004-04-10 devnull .BI -o fields
 86 cfa37a7b 2004-04-10 devnull Each output line comprises the designated fields.
 87 cfa37a7b 2004-04-10 devnull The comma-separated field designators are either
 88 cfa37a7b 2004-04-10 devnull .BR 0 ,
 89 cfa37a7b 2004-04-10 devnull meaning the join field, or have the form
 90 cfa37a7b 2004-04-10 devnull .IR n . m ,
 91 cfa37a7b 2004-04-10 devnull where
 92 cfa37a7b 2004-04-10 devnull .I n
 93 cfa37a7b 2004-04-10 devnull is a file number and
 94 cfa37a7b 2004-04-10 devnull .I m
 95 cfa37a7b 2004-04-10 devnull is a field number.
 96 cfa37a7b 2004-04-10 devnull Archaic usage allows separate arguments for field designators.
 97 cfa37a7b 2004-04-10 devnull .PP
 98 cfa37a7b 2004-04-10 devnull .TP
 99 cfa37a7b 2004-04-10 devnull .BI -t c
100 cfa37a7b 2004-04-10 devnull Use character
101 cfa37a7b 2004-04-10 devnull .I c
102 cfa37a7b 2004-04-10 devnull as the only separator (tab character) on input and output.
103 cfa37a7b 2004-04-10 devnull Every appearance of
104 cfa37a7b 2004-04-10 devnull .I c
105 cfa37a7b 2004-04-10 devnull in a line is significant.
106 cfa37a7b 2004-04-10 devnull .SH EXAMPLES
107 cfa37a7b 2004-04-10 devnull .TP
108 cfa37a7b 2004-04-10 devnull .L
109 058b0118 2005-01-03 devnull sort /etc/passwd | join -t: -1 1 -a 1 -e "" - bdays
110 cfa37a7b 2004-04-10 devnull Add birthdays to the
111 058b0118 2005-01-03 devnull .B /etc/passwd
112 cfa37a7b 2004-04-10 devnull file, leaving unknown
113 cfa37a7b 2004-04-10 devnull birthdays empty.
114 cfa37a7b 2004-04-10 devnull The layout of 
115 cfa37a7b 2004-04-10 devnull .B /adm/users
116 cfa37a7b 2004-04-10 devnull is given in
117 d32deab1 2020-08-16 rsc .MR passwd (5) ;
118 cfa37a7b 2004-04-10 devnull .B bdays
119 cfa37a7b 2004-04-10 devnull contains sorted lines like
120 cfa37a7b 2004-04-10 devnull .LR "ken:Feb\ 4,\ 1953" .
121 cfa37a7b 2004-04-10 devnull .TP
122 cfa37a7b 2004-04-10 devnull .L
123 058b0118 2005-01-03 devnull tr : ' ' </etc/passwd | sort -k 3 3 >temp
124 cfa37a7b 2004-04-10 devnull .br
125 cfa37a7b 2004-04-10 devnull .ns
126 cfa37a7b 2004-04-10 devnull .TP
127 cfa37a7b 2004-04-10 devnull .L
128 cfa37a7b 2004-04-10 devnull join -1 3 -2 3 -o 1.1,2.1 temp temp | awk '$1 < $2'
129 cfa37a7b 2004-04-10 devnull Print all pairs of users with identical userids.
130 cfa37a7b 2004-04-10 devnull .SH SOURCE
131 c3674de4 2005-01-11 devnull .B \*9/src/cmd/join.c
132 cfa37a7b 2004-04-10 devnull .SH "SEE ALSO"
133 cfa37a7b 2004-04-10 devnull .IR sort (1), 
134 cfa37a7b 2004-04-10 devnull .IR comm (1), 
135 d32deab1 2020-08-16 rsc .MR awk (1)
136 cfa37a7b 2004-04-10 devnull .SH BUGS
137 cfa37a7b 2004-04-10 devnull With default field separation,
138 cfa37a7b 2004-04-10 devnull the collating sequence is that of
139 cfa37a7b 2004-04-10 devnull .BI "sort -b"
140 cfa37a7b 2004-04-10 devnull .BI -k y , y\f1;
141 cfa37a7b 2004-04-10 devnull with
142 cfa37a7b 2004-04-10 devnull .BR -t ,
143 cfa37a7b 2004-04-10 devnull the sequence is that of
144 cfa37a7b 2004-04-10 devnull .BI "sort -t" x
145 cfa37a7b 2004-04-10 devnull .BI -k y , y\f1.
146 c8b6342d 2005-01-13 devnull .PP
147 cfa37a7b 2004-04-10 devnull One of the files must be randomly accessible.