How to parse Google search results?

Question

Google

The question is rather not on the technique of which libraries to use, but on the experience of parsing. There is a need to search regularly (and a lot) for certain queries in the Google SERPs, visiting only the first page and analyzing the snippets of the first search results. Who had a similar experience, tell me what rake they stepped on and how best to carry out this task?

Answer 1

logged in is better

Answer 2

We need an ip pool and their regular change + via API we connect the captcha solving service.
And to adjust empirically timeouts.