---json
{
    "aliases": [
        { "path": ":howto:generator:csv" }
    ],
    "description": "This article shows you how to generate a csv file from a generator tabular file.",
    "low_quality_page": "false",
    "name": "Generate a CSV",
    "page_id": "dldtsswriy1bmcinb28vy"
}
---
====== Tabulify - How to create a CSV File with generated data ======

===== About =====
This article shows you how to generate a <icon name="carbon:csv" /> [[docs:resource:csv|csv file]] from a [[docs:resource:generator|generator]] [[docs:resource:tabular_file|tabular file]]

<note important>
As a [[docs:resource:generator|generator]] is a [[docs:resource:file|file]], if you use any [[docs:op:op|data operation]] against a [[docs:system:file|file system]] data store, the target created would be a [[docs:resource:file|file]].

Therefore, this is mandatory to use the [[docs:op:fill|fill operation]]
</note>
===== Steps by Step =====

==== Prerequisites ====
  * [[howto:getting_started:1_install|A Tabulify Installation]]
<unit display="none">
<code bash>
tabul data drop --no-strict-selection date_dim.csv@tmp
</code>
</unit>

==== Create your generator ====
You first need to create your [[docs:resource:generator|generator data resource]] to define the data that should be generated.

We will take as example the ''d_date--generator.yml'' generator that generates a ''date dimension''.
<note info>
This generator is explained in this [[..:sql_table:date_dimension|howto]]
</note>

<unit >
<file yaml date_dim--generator.yml>
kind: generator
spec:
  Comment: An example of date dimension generator based on the `date_dim` table of TPCDS
  primary-columns: [ "d_date_sk" ]
  Columns:
    - name: d_date_sk
      comment: A surrogate key
      Type: integer
      data-supplier:
        type: sequence
    - name: d_date
      comment: A business key in date format
      Type: date
      data-supplier:
        type: sequence
        arguments:
          start: 2025-05-13
    - name: d_date_id
      comment: A business key in string
      Type: varchar
      data-supplier:
        type: expression
        arguments:
          column-variable: d_date
          expression: "d_date.toISOString().substring(0,10)"
    - name: d_month_seq
      comment: An ascendant sequence for the month
      Type: integer
      data-supplier:
        type: expression
        arguments:
          column-variable: d_date
          expression: "function pad(number) {if (number < 10) { return '0' + number; } return number; }; d_date.getFullYear()+''+(pad(d_date.getMonth()+1))"
    - name: d_day_name
      comment: The name of the day
      Type: varchar
      data-supplier:
        type: expression
        arguments:
          column-variable: d_date
          expression: "var days = ['Sunday','Monday','Tuesday','Wednesday','Thursday','Friday','Saturday']; days[d_date.getDay()]"
    - name: d_moy
      comment: the month number in year
      Type: Integer
      data-supplier:
        type: expression
        arguments:
          column-variable: d_date
          expression: "d_date.getMonth()+1"
    - name: d_year
      comment: The year number
      Type: Integer
      data-supplier:
        type: expression
        arguments:
          column-variable: d_date
          expression: "d_date.getFullYear()"

</file>
</unit>
==== Run the fill command ====

The below [[docs:tabul:data:copy|data copy]] command will create a [[docs:resource:csv|CSV file]] named ''d_date.csv'' in the [[docs:connection:tmp|temporary directory]].
<unit >
<code bash>
tabul data copy date_dim--generator.yml@howto date_dim.csv@tmp
</code>
<console>
Transfer results
input             target             latency   record_count   error_code   error_message
---------------   ----------------   -------   ------------   ----------   -------------
date_dim@memgen   date_dim.csv@tmp   0.831s             100
</console>
</unit>

==== Check the result ====

What is the data of the CSV file with the [[docs:tabul:data:head|data head command]]?
<unit >
<code bash>
tabul data head --limit 30 date_dim.csv@tmp
</code>
<console>
The first 30 rows of the data resource (date_dim.csv@tmp):
d_date_sk   d_date       d_date_id    d_month_seq   d_day_name   d_moy   d_year
---------   ----------   ----------   -----------   ----------   -----   ------
1           2025-05-12   2025-05-12   202505        Monday       5       2025
2           2025-05-11   2025-05-11   202505        Sunday       5       2025
3           2025-05-10   2025-05-10   202505        Saturday     5       2025
4           2025-05-09   2025-05-09   202505        Friday       5       2025
5           2025-05-08   2025-05-08   202505        Thursday     5       2025
6           2025-05-07   2025-05-07   202505        Wednesday    5       2025
7           2025-05-06   2025-05-06   202505        Tuesday      5       2025
8           2025-05-05   2025-05-05   202505        Monday       5       2025
9           2025-05-04   2025-05-04   202505        Sunday       5       2025
10          2025-05-03   2025-05-03   202505        Saturday     5       2025
11          2025-05-02   2025-05-02   202505        Friday       5       2025
12          2025-05-01   2025-05-01   202505        Thursday     5       2025
13          2025-04-30   2025-04-30   202504        Wednesday    4       2025
14          2025-04-29   2025-04-29   202504        Tuesday      4       2025
15          2025-04-28   2025-04-28   202504        Monday       4       2025
16          2025-04-27   2025-04-27   202504        Sunday       4       2025
17          2025-04-26   2025-04-26   202504        Saturday     4       2025
18          2025-04-25   2025-04-25   202504        Friday       4       2025
19          2025-04-24   2025-04-24   202504        Thursday     4       2025
20          2025-04-23   2025-04-23   202504        Wednesday    4       2025
21          2025-04-22   2025-04-22   202504        Tuesday      4       2025
22          2025-04-21   2025-04-21   202504        Monday       4       2025
23          2025-04-20   2025-04-20   202504        Sunday       4       2025
24          2025-04-19   2025-04-19   202504        Saturday     4       2025
25          2025-04-18   2025-04-18   202504        Friday       4       2025
26          2025-04-17   2025-04-17   202504        Thursday     4       2025
27          2025-04-16   2025-04-16   202504        Wednesday    4       2025
28          2025-04-15   2025-04-15   202504        Tuesday      4       2025
29          2025-04-14   2025-04-14   202504        Monday       4       2025
30          2025-04-13   2025-04-13   202504        Sunday       4       2025
</console>
</unit>