sourCEntral - mobile manpages




dbroweval - evaluate code for each row of a fsdb file


dbroweval [-f CodeFile] code [code...]


Evaluate code for each row of the data.

Typical actions are things like reformatting and other data transformations.

Code can include embedded column names preceded by underscores; these result in the value of that column for the current row.

The values of the last row’s columns are retrieved with _last_foo where foo is the column name.

Even more perverse, _columname(N) is the value of the Nth column after columnname [so _columnname(0) is the also the column’s value.



Run CODE before reading any data (like awk BEGIN blocks).


Run CODE at the end of all data (like awk END blocks).


Read code from the FILE.

-n or --no-output

no output except for comments and what is in the provided code

-N or --no-output-even-comments

no output at all, except for what is in the provided code

-m or --manual-output

The user must setup output, allowing arbitrary comments. See example 2 below for details.

-w or --warnings

Enable warnings in user supplied code.

--saveoutput $OUT_REF

Save output writer (for integration with other fsdb filters).

This module also supports the standard fsdb options:


Enable debugging output.

-i or --input InputSource

Read from InputSource, typically a file name, or "-" for standard input, or (if in Perl) a IO::Handle, Fsdb::IO or Fsdb::BoundedQueue objects.

-o or --output OutputDestination

Write to OutputDestination, typically a file name, or "-" for standard output, or (if in Perl) a IO::Handle, Fsdb::IO or Fsdb::BoundedQueue objects.

--autorun or --noautorun

By default, programs process automatically, but Fsdb::Filter objects in Perl do not run until you invoke the run() method. The "--(no)autorun" option controls that behavior within Perl.

--header H

Use H as the full Fsdb header, rather than reading a header from then input.


Show help.


Show full manual.


Typically dbroweval outputs a line in the same schema for each input line. For advanced usage, one can violate each of these assumptions.

Some fun:
omitting a line

Add the code "next row if ($your condition);"

outputting an extra line

Call "&$write_fastpath_sub($fref)". You may find $fref, the input row, useful.

changing the schema

See the examples below in "Command 2: Changing the Schema"


#fsdb size mean stddev pct_rsd
1024 1.4962e+06 2.8497e+05 19.047
10240 5.0286e+06 6.0103e+05 11.952
102400 4.9216e+06 3.0939e+05 6.2863
# | dbsetheader size bw
# | /home/johnh/BIN/DB/dbmultistats size bw
# | /home/johnh/BIN/DB/dbcol size mean stddev pct_rsd

cat data.fsdb | dbroweval '_mean = sprintf("%8.0f", _mean); _stddev = sprintf("%8.0f", _stddev);'

#fsdb size mean stddev pct_rsd
1024 1496200 284970 19.047
10240 5028600 601030 11.952
102400 4921600 309390 6.2863
# | dbsetheader size bw
# | /home/johnh/BIN/DB/dbmultistats size bw
# | /home/johnh/BIN/DB/dbcol size mean stddev pct_rsd
# | /home/johnh/BIN/DB/dbroweval { _mean = sprintf("%8.0f", _mean); _stddev = sprintf("%8.0f", _stddev); }

Command 2: Changing the Schema
By default, dbroweval reads and writes the same format file. The recommended method of adding and removing columns is to do so before or after dbroweval. I.e.,

cat data.fsdb |
dbcolcreate divisible_by_ten |
dbroweval '_divisible_by_ten = (_size % 10 == 0);' |
dbrow '_divisible_by_ten == 1' |
dbcol size mean divisible_by_ten

Another approach is to use the "next row" command to skip output of a row. I.e., the equivalent:

cat data.fsdb |
dbcolcreate divisible_by_ten |
dbroweval '_divisible_by_ten = (_size % 10 == 0); next row if (!_divisible_by_ten);' |
dbcol size mean divisible_by_ten

However, neither of these approachs work very well when the output is a completely different schema.

The recommended method for schema-changing commands is to write a full filter, but a full filter is a bit heavy weight. As an alternative, one can use the "-m" option to request manual configuration of the output, then use @out_args to define the output schema (it specifies the "Fsdb::IO::Writer" arguments), and $ofref is the output row. It may also reference <$in>, the input "Fsdb::IO::Reader" argument, and <$fref> as an aref to the current line. Note that newly created columns do not have underscore-names

Thus a third equivalent is:

cat data.fsdb | \
dbroweval -m -b '@out_args = ( -clone => $in, \
-cols => ($in->cols, divisible_by_ten); ' \
'my $div_by_10 = (_size % 10 == 0); \
$ofref = [ @$fref, $div_by_10 ] if ($div_by_ten);' |
dbcol size mean divisible_by_ten


cat data.fsdb | \
dbroweval -m -b '@out_args = ( -clone => $in, \
-cols => [qw(size mean divisible_by_ten)] ); ' \
'my $div_by_10 = (_size % 10 == 0); \
$ofref = [ _mean, _size, $div_by_10 ] if ($div_by_ten);'

Finally, one can write different a completely different schema, although it’s more work:

cat data.fsdb | \
dbroweval -m -b '@out_args = (-cols => [qw(size n)]);' \
'$ofref = [ _size, 1 ];'

writes different columns, and

cat data.fsdb | \
dbroweval -n -m -b '@out_args = (-cols => [qw(n)]); \
my $count = 0;' -e '$ofref = [ $count ];' '$count++;'

Is a fancy way to count lines.

The begin code block should setup @out_args to be the arguments to a "Fsdb::IO::Writer::new" call, and whatever is in $ofref (if anything) is written for each input line, and once at the end.

Command 3: Fun With Suppressing Output
The "-n" option suppresses default output. Thus, a simple equivalent to tail -1 is:

dbroweval -n -e '$ofref = $lfref;'

Where $ofref is the output fields, which are copied from $lfref, the hereby documented internal representation of the last row. Yes, this is a bit unappetizing, but, in for a penny with $ofref, in for a pound.

Command 4: Extra Ouptut
Calling "&$write_fastpath_sub($fref)" will do extra output, so this simple program will duplicate each line of input (one extra output, plus one regular output for each line of input):

dbroweval '&$write_fastpath_sub($fref)'


Handling of code in files isn’t very elegant.




$filter = new Fsdb::Filter::dbroweval(@arguments);


Internal: set up defaults.

Not a method; but an internal routine to make sure code compiles.


Internal: parse options


Internal: setup, parse headers.


Internal: run over all IO


Internal: write trailer.

$log = $filter->figure_program_log();

Override compute_program_log to do pretty-printed arguments.


Copyright (C) 1991-2018 by John Heidemann <johnh AT isi DOT edu>

This program is distributed under terms of the GNU general public license, version 2. See the file COPYING with the distribution for details.