Benchmark

Simple (and quite complex) operations are currently ~2x faster in jstr than in similar tools such as jq.node:

hyperfine --warmup 3 \
  "cat tmp/users.json | jqn 'filter(has(\"email\")) | groupBy(flow(get(\"email\"), split(\"@\"), get(1)))'" \
  "cat tmp/users.json | jstr -i \"flow(A.filter(D.get('email')), A.groupBy(flow(D.get('email'), S.split('@'), A.get(1))))\""

Package	Version	Time (mean ± σ)	Range (min … max)	Runs
@jliocsar/jstr	0.3.6	110.8 ms ± 4.5 ms	104.7 ms … 124.8 ms	26
jq.node	3.0.0	217.7 ms ± 9.2 ms	204.3 ms … 237.2 ms	13

Tested on:

• Node v20.7.0

• Linux 6.2.0-32-generic x86_64 x86_64

• 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz