infos = Information about the csvstorage plugin is in keys below
infos/author = Thomas Waser thoma.nosp@m.s.wa.nosp@m.ser@l.nosp@m.ibel.nosp@m.ektra.nosp@m..org
infos/licence = BSD
infos/provides = storage/csv
infos/needs =
infos/recommends =
infos/placements = getstorage setstorage
infos/status = unittest nodep libc configurable limited
infos/description = parses CSV files

This plugin allows you to read and write CSV files using Elektra. It aims to be compatible with RFC 4180. Rows and columns are written using Elektra's arrays (#0, #1,..). By configuring the plugin you can give columns a name.

delimiter Tells the plugin what delimiter is used in the file. The default delimiter is , and will be used if delimiter is not set.

header Tells the plugin to use the first line as a header if it is set to colname. The columns will get the corresponding names. Skip the first line if it is set to skip or treat the first line as a record if it is set to record. If header is not set, or set to record, the columns get named #0,#1,... (array key naming)

columns If this key is set the plugin will yield an error for every file that doesn't have exactly the amount of columns as specified in columns.

columns/names Sets the column names. Only usable in combination with the columns key. The number of subkeys must match the number of columns. Conflicts with usage of header.

columns/index Specifies which column should be used to index records instead of the record number.

export=,export/<column name>= Export column column name:

The key export must be present, additionally to export/<column name>
Also multiple column names can be given for different columns to export.
- Then delimiter will be used as delimiter (, as default).
- The order depends on the alphabetic order of the column names. Use `awk -F',' 'BEGIN{OFS=","} {print $2, $1, $3}'` or similar to reorder.
- Unknown column names are ignored.

First line should determine the headers:

kdb mount test.csv /csv csvstorage \

"delimiter=;,header=colname,columns=2,columns/names,columns/names/#0=col0Name,columns/names/#1=col1Name"

The example below shows how you can use this plugin to read and write CSV files.

`` @section autotoc_md128 Mount plugin touser/tests/csv`

We use the column names from the first line of the

config file as key names

sudo kdb mount config.csv user/tests/csv csvstorage "header=colname,columns/names/#0=col0Name,columns/names/#1=col1Name"

Add some data

printf 'band,album
' >> kdb file user/tests/csv printf 'Converge,All We Love We Leave Behind
' >> kdb file user/tests/csv printf 'mewithoutYou,Pale Horses
' >> kdb file user/tests/csv printf 'Kate Tempest,Everybody Down
' >> kdb file user/tests/csv

kdb ls user/tests/csv #> user/tests/csv/#0 #> user/tests/csv/#0/album #> user/tests/csv/#0/band #> user/tests/csv/#1 #> user/tests/csv/#1/album #> user/tests/csv/#1/band #> user/tests/csv/#2 #> user/tests/csv/#2/album #> user/tests/csv/#2/band #> user/tests/csv/#3 #> user/tests/csv/#3/album #> user/tests/csv/#3/band

The first array element contains the column names

kdb get user/tests/csv/#0/band #> band kdb get user/tests/csv/#0/album #> album

Retrieve data from the last entry

kdb get user/tests/csv/#3/album #> Everybody Down kdb get user/tests/csv/#3/band #> Kate Tempest

Change an existing item

kdb set user/tests/csv/#1/album 'You Fail Me'

Retrieve the new item

kdb get user/tests/csv/#1/album #> You Fail Me

The plugin stores the index of the last column

in all of the parent keys.

kdb get user/tests/csv/#0 #> #1 kdb get user/tests/csv/#1 #> #1 kdb get user/tests/csv/#2 #> #1 kdb get user/tests/csv/#3 #> #1

The configuration file reflects the changes

kdb file user/tests/csv | xargs cat #> album,band #> You Fail Me,Converge #> Pale Horses,mewithoutYou #> Everybody Down,Kate Tempest

Undo changes to the key database

kdb rm -r user/tests/csv sudo kdb umount user/tests/csv

# Column as index

kdb mount config.csv /tests/csv csvstorage "delimiter=;,header=colname,columns/index=IMDB"

printf 'IMDB;Title;Year
' >> kdb file /tests/csv printf 'tt0108052;Schindler´s List;1993
' >> kdb file /tests/csv printf 'tt0110413;Léon: The Professional;1994
' >> kdb file /tests/csv

kdb ls /tests/csv #> user/tests/csv/tt0108052 #> user/tests/csv/tt0108052/IMDB #> user/tests/csv/tt0108052/Title #> user/tests/csv/tt0108052/Year #> user/tests/csv/tt0110413 #> user/tests/csv/tt0110413/IMDB #> user/tests/csv/tt0110413/Title #> user/tests/csv/tt0110413/Year

kdb get /tests/csv/tt0108052/Title #> Schindler´s List

kdb rm -r /tests/csv sudo kdb umount /tests/csv

# Export filter

kdb mount config.csv /tests/csv csvstorage "delimiter=;,header=colname,columns/index=IMDB"

printf 'IMDB;Title;Year
' >> kdb file /tests/csv printf 'tt0108052;Schindler´s List;1993
' >> kdb file /tests/csv printf 'tt0110413;Léon: The Professional;1994
' >> kdb file /tests/csv

kdb export /tests/csv csvstorage -c "delimiter=;,header=colname,columns/index=IMDB,export=,export/IMDB=,export/Title=" #> IMDB;Title #> tt0108052;Schindler´s List #> tt0110413;Léon: The Professional

kdb export /tests/csv csvstorage -c "delimiter=;,header=colname,columns/index=IMDB,export=,export/IMDB=,export/Year=" #> IMDB;Year #> tt0108052;1993 #> tt0110413;1994

kdb rm -r /tests/csv sudo kdb umount /tests/csv

```

Limitations

Does not work on file streams (e.g. kdb import without file)
When using csvstorage for exporting, all parent keys must be present (see https://issues.libelektra.org/2304)