Ansible split Filter: Split Strings into Lists (Complete Guide)

The split filter in Ansible (Jinja2) divides a string into a list of substrings based on a delimiter. It's essential for parsing command output, configuration files, and structured text.

Basic Split

- name: Split a comma-separated string
  ansible.builtin.debug:
    msg: "{{ 'apple,banana,cherry' | split(',') }}"
  # Output: ['apple', 'banana', 'cherry']

See also: Ansible regex_search Filter: Search Strings with Regex Patterns (Guide)

Split with Different Delimiters

# Split by space
- debug:
    msg: "{{ 'hello world foo' | split(' ') }}"
  # ['hello', 'world', 'foo']
# Split by newline
- debug:
    msg: "{{ multiline_var | split('\n') }}"
# Split by colon (passwd-style)
- debug:
    msg: "{{ 'root:x:0:0:root:/root:/bin/bash' | split(':') }}"
  # ['root', 'x', '0', '0', 'root', '/root', '/bin/bash']
# Split by multiple characters
- debug:
    msg: "{{ 'key=value' | split('=') }}"
  # ['key', 'value']

Limit Number of Splits

- name: Split only first occurrence
  debug:
    msg: "{{ 'key=value=extra' | split('=', 1) }}"
  # ['key', 'value=extra']

See also: Ansible Split String: Filter Guide for CSV, Delimiters & Lists

Split Command Output

- name: Get disk usage
  ansible.builtin.command: df -h /
  register: df_output
- name: Parse disk usage
  ansible.builtin.set_fact:
    disk_usage: "{{ df_output.stdout_lines[1].split() }}"
    # disk_usage[4] = usage percentage like '45%'

Combine split with Other Filters

Split and select

- name: Get username from email
  debug:
    msg: "{{ 'user@example.com' | split('@') | first }}"
  # 'user'

Split and join (replace)

- name: Replace dashes with underscores
  debug:
    msg: "{{ 'my-variable-name' | split('-') | join('_') }}"
  # 'my_variable_name'

Split and filter empty strings

- name: Handle multiple delimiters
  debug:
    msg: "{{ '  hello   world  ' | split(' ') | reject('equalto', '') | list }}"
  # ['hello', 'world']

See also: Ansible from_json & to_json Filters: Parse & Convert JSON (Guide)

Parse Key-Value Pairs

- name: Parse config line
  ansible.builtin.set_fact:
    config_key: "{{ 'max_connections = 100' | split('=') | first | trim }}"
    config_value: "{{ 'max_connections = 100' | split('=') | last | trim }}"

Split vs regex_findall

For complex patterns, use regex_findall instead:

# Extract all numbers from a string
- debug:
    msg: "{{ 'server1:8080, server2:9090' | regex_findall('\d+') }}"
  # ['1', '8080', '2', '9090']
# Extract IP addresses
- debug:
    msg: "{{ log_line | regex_findall('\d+\.\d+\.\d+\.\d+') }}"

Practical Examples

Parse /etc/hosts

- name: Read hosts file
  ansible.builtin.slurp:
    path: /etc/hosts
  register: hosts_file
- name: Parse host entries
  ansible.builtin.set_fact:
    host_entries: >-
      {{ (hosts_file.content | b64decode).split('\n')
         | reject('match', '^#')
         | reject('equalto', '')
         | map('split', None, 1)
         | list }}

Parse CSV data

- name: Process CSV line
  ansible.builtin.debug:
    msg: "Name: {{ item.split(',')[0] }}, Age: {{ item.split(',')[1] }}"
  loop:
    - "Alice,30"
    - "Bob,25"

FAQ

What happens if the delimiter isn't found?

The split filter returns a list with the original string as the only element: "hello" | split(",") → ["hello"].

How do I split by whitespace (any amount)?

Use split() without arguments — it splits on any whitespace and removes empty strings: " a b c " | split() → ["a", "b", "c"].

Can I split into a fixed number of parts?

Yes, use the maxsplit parameter: "a-b-c-d" | split("-", 2) → ["a", "b", "c-d"].

Basic Split

- vars:
    csv_line: "web1,web2,web3,db1"
  debug:
    msg: "{{ csv_line.split(',') }}"
  # ['web1', 'web2', 'web3', 'db1']

Split with Filter Syntax

- debug:
    msg: "{{ 'hello world' | split(' ') }}"
  # ['hello', 'world']
- debug:
    msg: "{{ '/var/log/nginx/access.log' | split('/') }}"
  # ['', 'var', 'log', 'nginx', 'access.log']

Split Command Output

- command: "df -h / | tail -1"
  register: disk
  changed_when: false
- set_fact:
    disk_usage: "{{ disk.stdout.split() }}"
  # ['/dev/sda1', '50G', '25G', '25G', '50%', '/']
- debug:
    msg: "Disk usage: {{ disk_usage[4] }}"
  # 50%

Split Multi-Line Text

- vars:
    hosts_content: |
      192.168.1.10 web1
      192.168.1.11 web2
      192.168.1.20 db1
  debug:
    msg: "{{ hosts_content.strip().split('\n') }}"

Parse CSV Data

- vars:
    csv_data: "name,email,role\nalice,alice@co.com,admin\nbob,bob@co.com,user"
  set_fact:
    users: >-
      {{ csv_data.split('\n')[1:] | map('split', ',') |
         map('zip', ['name', 'email', 'role']) |
         map('map', 'reverse') | map('community.general.dict') | list }}

Simpler approach:

- vars:
    line: "alice,alice@example.com,admin"
    fields: "{{ line.split(',') }}"
  debug:
    msg: "User: {{ fields[0] }}, Email: {{ fields[1] }}, Role: {{ fields[2] }}"

Split with Limit

- debug:
    msg: "{{ 'one:two:three:four'.split(':', 2) }}"
  # ['one', 'two', 'three:four']  — splits at most 2 times

Related String Operations

| Operation | Syntax | Result | |-----------|--------|--------| | Split | 'a,b,c' | split(',') | ['a','b','c'] | | Join | ['a','b'] | join(',') | 'a,b' | | Replace | 'hello' | replace('l','r') | 'herro' | | Regex split | 'a1b2c' | regex_findall('[a-z]+') | ['a','b','c'] | | First element | 'a/b/c'.split('/') | first | 'a' | | Last element | 'a/b/c'.split('/') | last | 'c' |

Extract from Structured Strings

# Parse "key=value" pairs
- vars:
    config_line: "server=db.example.com port=5432 name=mydb"
  set_fact:
    config: "{{ dict(config_line.split() | map('split', '=')) }}"
  # {server: db.example.com, port: '5432', name: mydb}

FAQ

split vs regex_findall?

split divides by a fixed delimiter. regex_findall extracts matches by pattern:

# Split by comma
"a,b,c".split(',')  →  ['a', 'b', 'c']
# Extract numbers
"age:25 score:98" | regex_findall('\d+')  →  ['25', '98']

How do I split and pick the Nth element?

msg: "{{ 'path/to/file.txt'.split('/')[-1] }}"
# file.txt (last element)

Can I split by multiple delimiters?

Use regex: "a,b;c" | regex_findall('[^,;]+')

Basic Split

- vars:
    csv_line: "alice,bob,charlie,dave"
  debug:
    msg: "{{ csv_line.split(',') }}"
# ["alice", "bob", "charlie", "dave"]

Split by Different Delimiters

# Colon (PATH-style)
- debug:
    msg: "{{ '/usr/bin:/usr/local/bin:/home/user/bin'.split(':') }}"
# Space
- debug:
    msg: "{{ 'one two three'.split() }}"
# Newlines
- debug:
    msg: "{{ multiline_output.split('\n') }}"
# Pipe
- debug:
    msg: "{{ 'field1|field2|field3'.split('|') }}"
# Multiple characters
- debug:
    msg: "{{ 'key=value'.split('=') }}"

Split with Limit

# Split only first N times
- vars:
    line: "key=value=with=equals"
  debug:
    msg: "{{ line.split('=', 1) }}"
# ["key", "value=with=equals"]

Split Command Output

- command: df -h /
  register: disk_result
- set_fact:
    disk_usage: "{{ disk_result.stdout_lines[1].split() }}"
# ["dev/sda1", "50G", "23G", "27G", "46%", "/"]
- debug:
    msg: "Disk usage: {{ disk_usage[4] }}"
# "46%"

Split and Select

# Get filename from path
- vars:
    filepath: "/var/log/myapp/error.log"
  debug:
    msg: "{{ filepath.split('/')[-1] }}"
# "error.log"
# Get directory
- debug:
    msg: "{{ filepath.split('/')[:-1] | join('/') }}"
# "/var/log/myapp"
# Get file extension
- debug:
    msg: "{{ filepath.split('.')[-1] }}"
# "log"
# Get domain from email
- vars:
    email: "user@example.com"
  debug:
    msg: "{{ email.split('@')[1] }}"
# "example.com"

Split + Loop

- vars:
    allowed_ips: "10.0.1.1,10.0.1.2,10.0.1.3"
  ansible.builtin.iptables:
    chain: INPUT
    source: "{{ item }}"
    jump: ACCEPT
  loop: "{{ allowed_ips.split(',') }}"
  become: true

Split + Filter

# Split and remove empty strings
- vars:
    messy: "one,,two,,,three"
  debug:
    msg: "{{ messy.split(',') | select() | list }}"
# ["one", "two", "three"]
# Split, trim, and sort
- vars:
    tags: " web , api , database , cache "
  debug:
    msg: "{{ tags.split(',') | map('trim') | sort | list }}"
# ["api", "cache", "database", "web"]

Split Multiline Output

- shell: cat /etc/passwd
  register: passwd
- set_fact:
    users: "{{ passwd.stdout_lines | map('regex_replace', ':.*', '') | list }}"
# Extracts usernames from each line

Parse Key=Value Pairs

- vars:
    env_string: "DB_HOST=localhost DB_PORT=5432 DB_NAME=mydb"
  set_fact:
    env_dict: "{{ dict(env_string.split() | map('split', '=') | list) }}"
# {"DB_HOST": "localhost", "DB_PORT": "5432", "DB_NAME": "mydb"}

Join (Reverse of Split)

- vars:
    items: ["one", "two", "three"]
  debug:
    msg: "{{ items | join(', ') }}"
# "one, two, three"
# Split and rejoin with different delimiter
- vars:
    path: "one/two/three"
  debug:
    msg: "{{ path.split('/') | join('.') }}"
# "one.two.three"

Related Filters

| Filter | Purpose | |--------|---------| | split() | String → List | | join() | List → String | | regex_replace | Pattern replacement | | trim | Remove whitespace | | map('split') | Split each item in list | | select | Filter empty items |

FAQ

split vs splitlines?

split('\n') and splitlines() are similar, but splitlines() handles \r\n and other line endings. For command output, use stdout_lines (already split).

How to split by regex?

Use regex_findall instead:

msg: "{{ 'one2two3three' | regex_findall('[a-z]+') }}"
# ["one", "two", "three"]

Can I split and get the Nth element in one expression?

Yes: "{{ mystring.split(',')[2] }}" gets the third element.

Related Articles

• rendering Jinja2 templates with Ansible • Ansible Multiline Strings Guide • chdir option in Ansible command • Ansible Loops Guide • Ansible Split Filter Guide