Document -jq option and benchmark

[jvmvik]

Could you provide more documentation on the -jq option?

The code appears to be written in Go. Could you clarify that?
Add a few example.
“.organic_results[0:3]|[.[]|{title,link}]”

Additionally, it would be great if you could share the SKILLS.md section related to this.
https://github.com/serpapi/skills/blob/master/skills/serpapi-web-search/SKILL.md

[ilyazub]

Yes, sure. jq filters and transforms JSON output. The --jq option is similar to piping jq or the way gh (GitHub) CLI handles this.

Example:

# --jq filter
λ serpapi search engine=search_index q=serpapi --jq '.organic_results[0:3] | [.[] | {title,link} ]' 2>/dev/null
[{"link":"https://github.com/serpapi","title":"SerpApi"},{"link":"https://serpapi.com/","title":"SerpApi: Google Search API"},{"link":"https://medium.com/tag/serpapi","title":"The most insightful stories about Serpapi"}]

# pipe jq
λ serpapi search engine=search_index q=serpapi | jq '.organic_results[0:3] | [.[] | {title,link}]' 2>/dev/null
[
  {
    "title": "SerpApi",
    "link": "https://github.com/serpapi"
  },
  {
    "title": "SerpApi: Google Search API",
    "link": "https://serpapi.com/"
  },
  {
    "title": "The most insightful stories about Serpapi",
    "link": "https://medium.com/tag/serpapi"
  }
]

[jvmvik]

With this explanation, everything becomes crystal clear. I’ve created a small documentation pull request,