SPL to APL Translator

Type safety: Fields like status are often stored as strings. Always cast before numeric comparison: toint(status) >= 500, not status >= 500.

Critical Differences

1. Time is explicit in APL: SPL time pickers don't translate — add where _time between (ago(1h) .. now())
2. Structure: SPL index=... | command → APL ['dataset'] | operator
3. Join is preview: limited to 50k rows, inner/innerunique/leftouter only
4. cidrmatch args reversed: SPL cidrmatch(cidr, ip) → APL ipv4isin_range(ip, cidr)

Core Command Mappings

Complete mappings: reference/command-mapping.md

Stats → Summarize

# SPL
| stats count by status

# APL  
| summarize count() by status

Key function mappings

Conditional count pattern

# SPL
| stats count(eval(status>=500)) as errors by host

# APL
| summarize errors = countif(status >= 500) by host

Complete function list: reference/function-mapping.md

Eval → Extend

# SPL
| eval new_field = old_field * 2

# APL
| extend new_field = old_field * 2

Key function mappings

Case statement pattern

# SPL
| eval level = case(
    status >= 500, "error",
    status >= 400, "warning",
    1==1, "ok"
  )

# APL  
| extend level = case(
    status >= 500, "error",
    status >= 400, "warning",
    "ok"
  )

Note: SPL's 1==1 catch-all becomes implicit default in APL.

Rex → Parse/Extract

# SPL
| rex field=message "user=(?<username>\w+)"

# APL - parse with regex
| parse kind=regex message with @"user=(?P<username>\w+)"

# APL - extract function  
| extend username = extract("user=(\\w+)", 1, message)

Simple pattern (non-regex)

# SPL
| rex field=uri "^/api/(?<version>v\d+)/(?<endpoint>\w+)"

# APL
| parse uri with "/api/" version "/" endpoint

Time Handling

SPL time pickers don't translate. Always add explicit time range:

# SPL (time picker: Last 24 hours)
index=logs

# APL
['logs'] | where _time between (ago(24h) .. now())

Timechart translation

# SPL
| timechart span=5m count by status

# APL
| summarize count() by bin(_time, 5m), status

Common Patterns

Error rate calculation

# SPL
| stats count(eval(status>=500)) as errors, count as total by host
| eval error_rate = errors/total*100

# APL
| summarize errors = countif(status >= 500), total = count() by host
| extend error_rate = toreal(errors) / total * 100

Subquery (subsearch)

# SPL
index=logs [search index=errors | fields user_id | format]

# APL
let error_users = ['errors'] | where _time between (ago(1h) .. now()) | distinct user_id;
['logs']
| where _time between (ago(1h) .. now())
| where user_id in (error_users)

Join datasets

# SPL
| join user_id [search index=users | fields user_id, name]

# APL
| join kind=inner (['users'] | project user_id, name) on user_id

Transaction-like grouping

# SPL
| transaction session_id maxspan=30m

# APL (no direct equivalent — reconstruct with summarize)
| summarize 
    start_time = min(_time),
    end_time = max(_time),
    events = make_list(pack("time", _time, "action", action)),
    duration = max(_time) - min(_time)
  by session_id
| where duration <= 30m

String Matching Performance

Prefer has over contains (word-boundary matching is faster). Use _cs variants for case-sensitive (faster).

Reference

• reference/command-mapping.md — complete command list
• reference/function-mapping.md — complete function list
• reference/examples.md — full query translation examples
• APL docs: https://axiom.co/docs/apl/introduction