Return to join.1 CVS log

Up to [Research Unix] / researchv10dc / man / adm / man1

Annotation of researchv10dc/man/adm/man1/join.1, revision 1.1.1.1

1.1       root        1: .TH JOIN 1
                      2: .CT 1 files
                      3: .SH NAME
                      4: join \(mi relational database operator
                      5: .SH SYNOPSIS
                      6: .B join
                      7: [
                      8: .I options
                      9: ]
                     10: .I file1 file2
                     11: .SH DESCRIPTION
                     12: .I Join
                     13: forms, on the standard output,
                     14: a join
                     15: of the two relations specified by the lines of
                     16: .I file1
                     17: and
                     18: .IR file2 .
                     19: If
                     20: .I file1
                     21: is 
                     22: .LR - ,
                     23: the standard input is used.
                     24: .PP
                     25: .I File1
                     26: and
                     27: .I file2
                     28: must be sorted in increasing
                     29: .SM ASCII 
                     30: collating
                     31: sequence on the fields
                     32: on which they are to be joined,
                     33: normally the first in each line.
                     34: .PP
                     35: There is one line in the output
                     36: for each pair of lines in
                     37: .I file1
                     38: and
                     39: .I file2
                     40: that have identical join fields.
                     41: The output line normally consists of the common field,
                     42: then the rest of the line from
                     43: .IR file1 ,
                     44: then the rest of the line from
                     45: .IR file2 .
                     46: .PP
                     47: Input fields are normally separated by blank, tab or newline;
                     48: output fields by blank.
                     49: In this case, multiple separators count as one, and
                     50: leading separators are discarded.
                     51: .PP
                     52: These options are recognized:
                     53: .TP
                     54: .BI -a n
                     55: In addition to the normal output,
                     56: produce a line for each unpairable line in file
                     57: .IR n ,
                     58: where
                     59: .I n
                     60: is 1 or 2.
                     61: .TP
                     62: .BI -e \ s
                     63: Replace empty output fields by string
                     64: .IR s .
                     65: .TP
                     66: .BI -j n\ m
                     67: Join on the
                     68: .IR m th
                     69: field of file
                     70: .IR n .
                     71: If
                     72: .I n
                     73: is missing, use the
                     74: .IR m th
                     75: field in each file.
                     76: .TP
                     77: .BI -o \ field \ \fR...
                     78: Each output line comprises the named fields, each designated
                     79: as 0, meaning the join field, or as
                     80: .IR n . m ,
                     81: where
                     82: .I n
                     83: is a file number and
                     84: .I m
                     85: is a field number.
                     86: .PP
                     87: .TP
                     88: .BI -t c
                     89: Use character
                     90: .I c
                     91: as the only separator (tab character) on input and output.
                     92: Every appearance of
                     93: .I c
                     94: in a line is significant.
                     95: .SH EXAMPLES
                     96: .TP
                     97: .L
                     98: sort /etc/password | join -t: -a1 -e "" - bdays
                     99: Add birthdays to password information, leaving unknown
                    100: birthdays empty.
                    101: The layout of 
                    102: .F /etc/passwd
                    103: is given in
                    104: .IR passwd (5);
                    105: .B bdays
                    106: contains sorted lines like
                    107: .LR "ken:Feb\ 4" .
                    108: .TP
                    109: .L
                    110: tr : ' ' </etc/passwd | sort +2 >temp
                    111: .br
                    112: .ns
                    113: .TP
                    114: .L
                    115: join -j1 3 -j2 3 -o 1.1 2.1 temp temp | awk '$1 < $2'
                    116: Print all pairs of users with identical userids.
                    117: .SH "SEE ALSO"
                    118: .IR sort (1), 
                    119: .IR comm (1), 
                    120: .IR awk (1)
                    121: .SH BUGS
                    122: With default field separation,
                    123: the collating sequence is that of
                    124: .LR "sort\ -b" ;
                    125: with
                    126: .BR -t ,
                    127: the sequence is that of a plain sort.
                    128: .br
                    129: The field conventions of
                    130: .I join, sort, comm, uniq, look
                    131: and
                    132: .IR awk (1)
                    133: are wildly incongruent.

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.