Ignore anchor with h1/lvl1

## Description

https://docsearch.algolia.com/docs/tips/#add-anchors-to-headings and https://docsearch.algolia.com/docs/required-configuration/#use-the-right-classes-as-recordprops makes a lot of sense for h2,h3,h4, etc.

However, I believe that it's harmful to h1. e.g. https://docusaurus.io/ doesn't have an anchor. The value is so that people get a cleaner URL when they copy and paste it to share after a search, which happens quite often in my use cases.

I had to add this code so that the crawler won't pick up an anchor for the lvl1.

```diff
recordExtractor: ({ $, helpers, url }) => {
// Removing DOM elements we don't want to crawl
const toRemove = ".skip-algolia-crawler";
$(toRemove).remove();
+ $("#main-content, #__next").removeAttr("id");
```

## Steps to reproduce

I could reproduce on: https://mui.com/material-ui/react-alert/ with https://crawler.algolia.com/admin/crawlers/739c29c8-99ea-4945-bd27-17a1df391902/configuration/edit.

I get this URL: https://mui.com/material-ui/react-alert/#main-content.

## Expected behavior

`helpers.docsearch` doesn't try to find an anchor for h1.
I would get this URL: https://mui.com/material-ui/react-alert/.

## Environment

- DocSearch version: v3

Start with the helpers.docsearch anchor extraction behavior in DocSearch v3 and reproduce the reported case on https://mui.com/material-ui/react-alert/. Check how h1/lvl1 headings produce #main-content, then verify that the completed behavior leaves the page URL without an anchor while preserving anchors for lower-level headings.

Written by the indexing model from the issue text.