NAME¶
tabmerge - unify delimited files on common fields
SYNOPSIS¶
tabmerge [action] [options] file1 file2 [...]
Actions:
--min Take only fields present in all files [DEFAULT]
--max Take all fields present
-f|--fields=f1[,f2] Take only the fields mentioned in the
comma-separated list
Options:
-l|--list List available fields
--fs=x Use "x" as the field separator
(default is tab "\t")
--rs=x Use "x" as the record separator
(default is newline "\n")
-s|--sort=f1[,f2] Sort data ASCII-betically on field(s)
--stdout Print data in original delimited format
(i.e., not in a table format)
--help Show brief help and quit
--man Show full documentation
DESCRIPTION¶
This program merges the fields -- not the rows -- of delimited text files. That
is, if several files are almost but not quite entirely unlike each other in
their structure (in their field names, numbers or orders), this script allows
you to easily unify the files into one file with all the same fields. The
output can be based on fields as determined by the three "action"
flags.
For the following examples, consider three files that contain the following
fields:
+------------+---------------------------------+
| File | Fields |
+------------+---------------------------------+
| merge1.tab | name, type, position |
| merge2.tab | name, type, position, lod_score |
| merge3.tab | name, position |
+------------+---------------------------------+
To list all available fields in the files and the number of times they are
present:
$ tabmerge --list merge*
+-----------+-------------------+
| Field | No. Times Present |
+-----------+-------------------+
| lod_score | 1 |
| name | 3 |
| position | 3 |
| type | 2 |
+-----------+-------------------+
To merge the files on the minimum overlapping fields:
$ tabmerge merge*
+----------+----------+
| name | position |
+----------+----------+
| RM104 | 2.30 |
| RM105 | 4.5 |
| TX5509 | 10.4 |
| UU189 | 19.0 |
| Xpsm122 | 3.3 |
| Xpsr9556 | 4.5 |
| DRTL | 2.30 |
| ALTX | 4.5 |
| DWRF | 10.4 |
+----------+----------+
To merge the files and include all the fields:
$ tabmerge --max merge*
+-----------+----------+----------+--------+
| lod_score | name | position | type |
+-----------+----------+----------+--------+
| | RM104 | 2.30 | RFLP |
| | RM105 | 4.5 | RFLP |
| | TX5509 | 10.4 | AFLP |
| 2.4 | UU189 | 19.0 | SSR |
| 1.2 | Xpsm122 | 3.3 | Marker |
| 1.2 | Xpsr9556 | 4.5 | Marker |
| | DRTL | 2.30 | |
| | ALTX | 4.5 | |
| | DWRF | 10.4 | |
+-----------+----------+----------+--------+
To merge and extract just the "name" and "type" fields:
$ tabmerge -f name,type merge*
+----------+--------+
| name | type |
+----------+--------+
| RM104 | RFLP |
| RM105 | RFLP |
| TX5509 | AFLP |
| UU189 | SSR |
| Xpsm122 | Marker |
| Xpsr9556 | Marker |
| DRTL | |
| ALTX | |
| DWRF | |
+----------+--------+
To merge the files on just the "name" and "lod_score" fields
and sort on the name:
$ tabmerge -f name,lod_score -s name merge*
+----------+-----------+
| name | lod_score |
+----------+-----------+
| ALTX | |
| DRTL | |
| DWRF | |
| RM104 | |
| RM105 | |
| TX5509 | |
| UU189 | 2.4 |
| Xpsm122 | 1.2 |
| Xpsr9556 | 1.2 |
+----------+-----------+
To do the same but mimic the original tab-delimited input:
$ tabmerge -f name,lod_score -s name --stdout merge*
name lod_score
ALTX
DRTL
DWRF
RM104
RM105
TX5509
UU189 2.4
Xpsm122 1.2
Xpsr9556 1.2
Why would you want to do this? Suppose you have several delimited text files
with nearly the same structure and want to create just one file from them, but
the fields may be in a different order in each file and/or some files may
contain more or fewer fields than others. (As far-fetched as it may seem, it
happens to the author more than he'd like.)
SEE ALSO¶
- •
- Text::RecordParser
- •
- Text::TabularDisplay
AUTHOR¶
Ken Youens-Clark <kclark@cpan.org>.
LICENSE AND COPYRIGHT¶
Copyright (C) 2006-10 Ken Youens-Clark. All rights reserved.
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2.
This program is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
A PARTICULAR PURPOSE. See the GNU General Public License for more
details.