使用 awk 或 sed 从 bash 变量中提取动态信息的正则表达式

Question

If the value changes, ie it could be

project-crm-backend-build-pipeline-run-98b6r-slack-notific-8fzvf

or

backend-build-pipeline-run-aaaaa-test-eeee

and I only need to extract the following from both examples above

project-crm-backend-build-pipeline-run-98b6r
backend-build-pipeline-run-aaaaa

ie the only thing that is constant is "pipeline-run", everything else could be changing, so I need to include everything before the pipeline-run and the single token right after the pipeline-run

Answer 1

With Parameter Expansion :

$ t='build-pipeline'

$ s='project-crm-backend-build-pipeline-run-98b6r-slack-notific-8fzvf'
$ echo "${s%-${s#*${t}-*-*-}}"
project-crm-backend-build-pipeline-run-98b6r

$ s='backend-build-pipeline-run-aaaaa-test-eeee'
$ echo "${s%-${s#*${t}-*-*-}}"
backend-build-pipeline-run-aaaaa

${s#*${t}-*-*-} will delete minimally from start of string, matching contents of variable t and then till 3rd - after.

The remaining string that is obtained is then used (after prefixing with a - character) to be deleted from end of string, thus giving the portion required.

With sed

echo "$s" | sed -E 's/('"$t"'-run-[^-]+).*/\1/'

Both these solution assume contents of variable t don't have characters that can effect the expansion/regex.

Answer 2

okay this is my working solution, just in case someone gets into the same issue

TEKTON_PIPELINE_NAME=build-pipeline
echo $POD_NAME | awk -v search="[0-9a-zA-Z-]+$TEKTON_PIPELINE_NAME-run-[a-z0-9]+" 'match($0, search) { print substr( $0, RSTART, RLENGTH )}'