Re: [PATCH 3/3] perf-test: add memory test for notmuch-insert

[Tomi Ollila <tomi.ollila@iki.fi>]

On Fri, Aug 18 2017, David Bremner wrote:

> In the future it might be worthwhile selecting corpus messages to
> insert, but that seems a bit overcomplicated for now
> ---
>  performance-test/M06-insert.sh | 16 ++++++++++++++++
>  1 file changed, 16 insertions(+)
>  create mode 100755 performance-test/M06-insert.sh
>
> diff --git a/performance-test/M06-insert.sh b/performance-test/M06-insert.sh
> new file mode 100755
> index 00000000..9bca199e
> --- /dev/null
> +++ b/performance-test/M06-insert.sh
> @@ -0,0 +1,16 @@
> +#!/bin/bash
> +
> +test_description='search'
> +
> +. ./perf-test-lib.sh || exit 1
> +
> +memory_start
> +
> +mkdir -p "$MAIL_DIR"/{cur,new,tmp}
> +
> +for count in $(seq 1 20); do

Since we use bash, {1..20} works. And IIRC some systems did not have
seq(1) (but some alternative).

LGTM (whole series) in any case.

Tomi


> +    generate_message "[file]=\"insert-$count\"" "[dir]='tmp/'"
> +    memory_run "insert $count" "notmuch insert < $gen_msg_filename"
> +done
> +
> +memory_done
> -- 
> 2.14.1