Can someone show a working example of how to iterate over the rows of a csv file?

[Nico Sabbi]

HI,

I find Ansible  docs *really* frustrating. Never a full example  that is not trivial or misleading.

Please, can someone provide a working example of a play to iterate over the rows of a csvfile?

Thanks.

[Brian Coca]

there are plenty of specific examples, just your specific case is not
covered, but it is simple enough to inffer from existing ones,
something like this should work:

- name: show the row
debug: var=item
loop: " {{ lookup('file', '/path/tocsv') }}"


you might need to split on ',' if you want the columns, there is a
csvfile lookup it it retrieves specific column, not rows


--
----------
Brian Coca

[Nico Sabbi]

Sorry, but  it's not clear at all how debug iterates over the rows. 

Running a simple debug doesn't help.

For example, if I had a playbook like this (that of course doesn't work), how should I fix it

to iterate over every single row and print the value of the 3 columns?

---

  - hosts: db

    become: true

    remote_user: root

    vars:

      csvfile: "{{ lookup('csvfile', 'file=file.csv delimiter=,') }}"

      

    tasks:

      - name: cat

        debug: msg="{{ item.a }} - {{ item.a }} - {{ item.c }}"  #for every row print the value of a, b and c columns

        with_items: csvfile

  

Thanks.

[Claudia de Luna]

Hi Nico,

Not sure if this is what you are looking for but for a sample CSV file

# CSV File sample.csv

host, user, password, comment,

1.1.1.1, claudia, mypassword, the first one

1.1.1.2, elvira, epassword, the second one

1.1.1.3, billy, bpassword, the third one

with this playbook sample_csv.yml

---

- hosts: localhost

  connection: local

  gather_facts: False

  ignore_errors: yes

  vars:

    contents: "{{ lookup('file', './sample.csv') }}"

    lines: "{{ contents.split('\n') }}"

  tasks:

    - debug: msg="Iterating over CSV lines = {{ item }}"

      with_items: 

        - "{{ lines }}"

you can get this output:

root@127c868b9dd3:/ansible/ansible2_4_base# ansible-playbook -i hosts sample_csv.yml

PLAY [localhost] ****************************************************************************************************************************

TASK [debug] ********************************************************************************************************************************

ok: [localhost] => (item=# CSV File sample.csv) => {

    "item": "# CSV File sample.csv",

    "msg": "Iterating over CSV lines = # CSV File sample.csv"

}

ok: [localhost] => (item=host, user, password, comment,) => {

    "item": "host, user, password, comment,",

    "msg": "Iterating over CSV lines = host, user, password, comment,"

}

ok: [localhost] => (item=1.1.1.1, claudia, mypassword, the first one) => {

    "item": "1.1.1.1, claudia, mypassword, the first one",

    "msg": "Iterating over CSV lines = 1.1.1.1, claudia, mypassword, the first one"

}

ok: [localhost] => (item=1.1.1.2, elvira, epassword, the second one) => {

    "item": "1.1.1.2, elvira, epassword, the second one",

    "msg": "Iterating over CSV lines = 1.1.1.2, elvira, epassword, the second one"

}

ok: [localhost] => (item=1.1.1.3, billy, bpassword, the third one# CSV File) => {

    "item": "1.1.1.3, billy, bpassword, the third one# CSV File",

    "msg": "Iterating over CSV lines = 1.1.1.3, billy, bpassword, the third one# CSV File"

}

PLAY RECAP **********************************************************************************************************************************

localhost                  : ok=1    changed=0    unreachable=0    failed=0

root@127c868b9dd3:/ansible/ansible2_4_base#

you can further split each line by "," but I'm wondering if what you are really looking for is a data structure you can put in group_vars or host_vars.

This ultimately becomes string parsing and as a general rule you want to stay away from that as much as you possibly can.

If you turn your CSV iinto a YAML file you have a bit more control your items and you can refer to them explicitly.

Food for thought but it all depends on what you are trying to do.

Hope this helps!

Claudia

[bill paxton]

Take a look at csv_to_facts module

I found this to be quick and dirty way to access data from a spreadsheet -> converted to csv first.  Look inside the .py for extra info on how to use it.

[bill paxton]

Oh, and you can drop the csv_to_facgts.py in ./library in your project folder...your playbook should then be able to use the module.

---


  - name: import csv
    csv_to_facts:
      src: "{{ some_csv_file }}"  
      table: mainTable              # optionally name the table....helpful if your playbook later imports add'l csv's


   - name: "loop through 'Spine' rows and create directories and switch configs"
    include: core_loop_rows.yml
      destination={{ base_dir }}/{{ item1['zone'] }}
      zone={{ item1['zone'] }}
      building={{ item1['building'] }}
      location={{ item1['locationName']}}
      host_name={{ item1['hostName'] }}
      device_type={{ item1['deviceType'] }}
      core_link_1={{ item1['int1'] }}
      core_link_2={{ item1['int2'] }}
      farend_core_link_1={{ item1['int1'] }}
      farend_core_link_2={{ item1['int2'] }}
      default_iface_mask={{ item1['mask'] }}
      core_link_1_ip={{ item1['int1_ip'] }}
      core_link_2_ip={{ item1['int2_ip'] }}
      core_1_name='core1'
      core_2_name='core2'
      loop0_ip={{ item1['Loop0'] }}
    with_items: "{{ mainTable }}"     # Now we can loop through the rows of the csv and access cells by the column header name
                                      # e.g. my csv file has column headers "Loop0", "locationName", "zone", etc.
    loop_control:
      loop_var: item1