join - relational database operator
join [-a file_number | -v file_number][-e string][-o list][-t char]
[-1 field][-2 field] file1 file2
The join utility shall perform an equality join on the files file1 and file2. The joined files shall be written to the standard output.
The join field is a field in each file on which the files are compared. The join utility shall write one line in the output for each pair of lines in file1 and file2 that have identical join fields. The output line by default shall consist of the join field, then the remaining fields from file1, then the remaining fields from file2. This format can be changed by using the -o option (see below). The -a option can be used to add unmatched lines to the output. The -v option can be used to output only unmatched lines.
The files file1 and file2 shall be ordered in the collating sequence of sort -b on the fields on which they shall be joined, by default the first in each line. All selected output shall be written in the same collating sequence.
The default input field separators shall be <blank>s. In this case, multiple separators shall count as one field separator, and leading separators shall be ignored. The default output field separator shall be a <space>.
The field separator and collating sequence can be changed by using the -t option (see below).
If the same key appears more than once in either file, all combinations of the set of remaining fields in file1 and the set of remaining fields in file2 are output in the order of the lines encountered.
If the input files are not in the appropriate collating sequence, the results are unspecified.
The join utility shall conform to the Base Definitions volume of IEEE Std 1003.1-2001, Section 12.2, Utility Syntax Guidelines.
The following options shall be supported:
- -a file_number
- Produce a line for each unpairable line in file file_number, where file_number is 1 or 2, in addition to the default output. If both -a1 and -a2 are specified, all unpairable lines shall be output.
- -e string
- Replace empty output fields in the list selected by -o with the string string.
- -o list
- Construct the output line to comprise the fields specified in list, each element of which shall have one of the following two forms:
file_number.field, where file_number is a file number and field is a decimal integer field number
0 (zero), representing the join field
The elements of list shall be either comma-separated or <blank>-separated, as specified in Guideline 8 of the Base Definitions volume of IEEE Std 1003.1-2001, Section 12.2, Utility Syntax Guidelines. The fields specified by list shall be written for all selected output lines. Fields selected by list that do not appear in the input shall be treated as empty output fields. (See the -e option.) Only specifically requested fields shall be written. The application shall ensure that list is a single command line argument.
- -t char
- Use character char as a separator, for both input and output. Every appearance of char in a line shall be significant. When this option is specified, the collating sequence shall be the same as sort without the -b option.
- -v file_number
- Instead of the default output, produce a line only for each unpairable line in file_number, where file_number is 1 or 2. If both -v1 and -v2 are specified, all unpairable lines shall be output.
- -1 field
- Join on the fieldth field of file 1. Fields are decimal integers starting with 1.
- -2 field
- Join on the fieldth field of file 2. Fields are decimal integers starting with 1.
The following operands shall be supported:
- file1, file2
- A pathname of a file to be joined. If either of the file1 or file2 operands is '-', the standard input shall be used in its place.
The standard input shall be used only if the file1 or file2 operand is '-'. See the INPUT FILES section.
The input files shall be text files.
The following environment variables shall affect the execution of join:
- Provide a default value for the internationalization variables that are unset or null. (See the Base Definitions volume of IEEE Std 1003.1-2001, Section 8.2, Internationalization Variables for the precedence of internationalization variables used to determine the values of locale categories.)
- If set to a non-empty string value, override the values of all the other internationalization variables.
- Determine the locale of the collating sequence join expects to have been used when the input files were sorted.
- Determine the locale for the interpretation of sequences of bytes of text data as characters (for example, single-byte as opposed to multi-byte characters in arguments and input files).
- Determine the locale that should be used to affect the format and contents of diagnostic messages written to standard error.
- [XSI] Determine the location of message catalogs for the processing of LC_MESSAGES .
The join utility output shall be a concatenation of selected character fields. When the -o option is not specified, the output shall be:"%s%s%s\n", <join field>, <other file1 fields>, <other file2 fields>
If the join field is not the first field in a file, the <other file fields> for that file shall be:<fields preceding join field>, <fields following join field>
When the -o option is specified, the output format shall be:"%s\n", <concatenation of fields>
where the concatenation of fields is described by the -o option, above.
For either format, each field (except the last) shall be written with its trailing separator character. If the separator is the default ( <blank>s), a single <space> shall be written after each field (except the last).
The standard error shall be used only for diagnostic messages.
The following exit values shall be returned:
- All input files were output successfully.
- An error occurred.
Pathnames consisting of numeric digits or of the form string.string should not be specified directly following the -o list.
The -o 0 field essentially selects the union of the join fields. For example, given file phone:!Name Phone Number Don +1 123-456-7890 Hal +1 234-567-8901 Yasushi +2 345-678-9012
and file fax:!Name Fax Number Don +1 123-456-7899 Keith +1 456-789-0122 Yasushi +2 345-678-9011
(where the large expanses of white space are meant to each represent a single <tab>), the command:join -t "<tab>" -a 1 -a 2 -e '(unknown)' -o 0,1.2,2.2 phone fax
would produce:!Name Phone Number Fax Number Don +1 123-456-7890 +1 123-456-7899 Hal +1 234-567-8901 (unknown) Keith (unknown) +1 456-789-0122 Yasushi +2 345-678-9012 +2 345-678-9011
Multiple instances of the same key will produce combinatorial results. The following:fa: a x a y a z fb: a p
will produce:a x p a y p a z p
And the following:fa: a b c a d e fb: a w x a y z a o p
will produce:a b c w x a b c y z a b c o p a d e w x a d e y z a d e o p
The -e option is only effective when used with -o because, unless specific fields are identified using -o, join is not aware of what fields might be empty. The exception to this is the join field, but identifying an empty join field with the -e string is not historical practice and some scripts might break if this were changed.
The 0 field in the -o list was adopted from the Tenth Edition version of join to satisfy international objections that the join in the base documents does not support the "full join" or "outer join" described in relational database literature. Although it has been possible to include a join field in the output (by default, or by field number using -o), the join field could not be included for an unpaired line selected by -a. The -o 0 field essentially selects the union of the join fields.
This sort of outer join was not possible with the join commands in the base documents. The -o 0 field was chosen because it is an upwards-compatible change for applications. An alternative was considered: have the join field represent the union of the fields in the files (where they are identical for matched lines, and one or both are null for unmatched lines). This was not adopted because it would break some historical applications.
The ability to specify file2 as - is not historical practice; it was added for completeness.
The -v option is not historical practice, but was considered necessary because it permitted the writing of only those lines that do not match on the join field, as opposed to the -a option, which prints both lines that do and do not match. This additional facility is parallel with the -v option of grep.
Some historical implementations have been encountered where a blank line in one of the input files was considered to be the end of the file; the description in this volume of IEEE Std 1003.1-2001 does not cite this as an allowable case.
awk, comm, sort, uniq
First released in Issue 2.
The obsolescent -j options and the multi-argument -o option are withdrawn in this issue.
The normative text is reworded to avoid use of the term "must" for application requirements.