++ed by:
1 non-PAUSE user
Author image Timo Kluck
and 1 contributors

NAME

Text::CSV::Flatten - Perl extension for transforming hierarchical data (nested arrays/hashes) to comma-separated value (csv) output according to a compact, readable, user-specified pattern.

SYNOPSIS

  use Text::CSV::Flatten;
  Text::CSV::Flatten->new(
    '.<index>.*',
    data    => [{ a => 1, b => 2 }, { a => 3, b => 4 }],
  )->csv();

DESCRIPTION

This module transforms hierarchical data (nested arrays/hashes) to comma-separated value (csv) output according to a compact, readable, user-specified pattern.

For example, the pattern .<index>.* transforms a data structure of the form

    [{ a => 1, b => 2 }, { a => 3, b => 4 }]

to the CSV output

    a,b,index
    1,2,0
    3,4,1

The pattern .*.* applied to the same data gives the output

    0_a,0_b,1_a,1_b
    1,2,3,4

The pattern .*.<key> gives the output

    0,1,key
    1,3,a
    2,4,b

It is hoped that the pattern specification is sufficiently powerful for this module to replace a lot of simple boiler-plate data transformations.

PATTERN SPECIFICATION

The dot-separated components represent the following:

  • <name> represents that the keys at that position should be put in a column named name in the csv output. This column will be considered a primary key, and the values belonging to those keys become rows;

  • * represents that the keys at that position in the pattern should be interpreted as column names; their values should be the values for that column, all beloning to the same row;

  • {column_name} or {column_name_1,column_name_2,...} is similar to *, but instead of capturing all the keys at that level of the hierarchy, it only captures the named columns.

  • anything else represents a literal key name.

  • If your pattern does not contain * or {...}, you need to pass an additional column_name => parameter to the constructor to specify the name for the single column where the value will go.

For the purposes of this description, an array should be seen as a collection of index => value pairs.

It is possible to specify several dot-separated paths in a single pattern, separated by spaces. In that case, all the paths need to have the same primary key (that is, the same set of names in <...>). Rows will be formed by joining the columns resulting from the different paths.

SEE ALSO

  Text::CSV

AUTHOR

Timo Kluck, <tkluck@infty.nl>

COPYRIGHT AND LICENSE

Copyright (C) 2016 by Timo Kluck

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself, either Perl version 5.22.1 or, at your option, any later version of Perl 5 you may have available.