Skip to main content

Indexing Content

A very important part of making a good search experience is to select only the relevant content for indexing. Eg. avoid adding text which is repeated on multiple pages like menus, footers etc.

tip

Findkit crawler automatically extracts the relevant page content on best effort basis but for production use cases it is recommended to explicitly select the content for indexing with the following methods.

Data Attribute​

If you are just building the site you should just mark the content right in the html to be indexed using the data-fdk-content data attribute.

Example

<div class="content" data-fdk-content>The content text...</div>

For full list of data-attributes see data-attributes.

You can use the attribute multiple times on a single page. The crawler will select the text content from all elements with it and combine them to the page index. You can use the findkit inspect <url> CLI command to view currently indexed content and findkit crawl test <url> to see what content would be selected to the index.

Custom CSS Selector​

If you cannot add the data attributes you can use the content_selector field in the findkit.toml to select content using CSS selectors.

Example

[[targets]]
host = "example.com"
content_selector = ".content, .post"

You can define multiple selectors by separating them with commas. If multiple elements are matched the content is indexed from all of them.

Cleaning Content​

Sometimes the selected content can contain some inner elements that you don't want to index. These elements can be removed from indexed content with the skip data attribute data-fdk-skip or the clean up selector.

<div class="content" data-fdk-content>
    The content text...
    <div class="author-bio" data-fdk-skip>
        Author bio that is also on the dedicated author page.
    </div>
</div>

or in the findkit.toml

[[targets]]
host = "example.com"
content_selector = ".content, .post"
cleanup_selector = ".author-bio"

Title Selector​

Findkit tries to determine page title automatically using the built-in default selectors, but its behaviour can be customized via options listed below.

Default title selectors in the order of importance:

  • findkit data attribute data-fdk-title
  • <meta og:title></meta> element
  • <title></title> element

Meta Tag​

Title can be defined directly in page meta

Meta tag CSS Selector​

Title CSS-selector can be passed in page meta

Data Attribute​

Title can be defined with default data attributes.

<h1 data-fdk-title>Custom Title</h1>

CSS Selector in findkit.toml​

You can pass CSS selector in findkit.toml. First elements content matching the selector is chosen as title.

Example

[[targets]]
host = "example.com"
title_selector = "h1"

Modifying selected title​

After choosing the title, the title can be modified using one of the options below.

Meta tag titleSelectorRegex​

Chosen title can be modified with regex defined in page meta

Config title_selector_regex​

Chosen title can be modified with regex defined in findkit.toml

Example

<h1>Title - Unwanted</h1>
[[targets]]
host = "example.com"
title_selector = "h1"
title_selector_regex = "([^ ]+)"

Skipping Pages​

Your website might include pages that you do not want to be included in the search index. For this purpose you can use one of the options listed below

Robots.txt​

Place /robots.txt to website root. For more info refer to documentation.

You can instruct Findkit Crawler to not to respect /robots.txt rules with respect_robots_txt configuration option.

Meta Robots Tag​

By placing robots meta tag on an individual page you can instruct crawlers, findkit included, to not index the page or not to follow links on the page.

For more information see Google's documentation.

You can instruct Findkit Crawler to not to respect robots meta rules respect_robots_meta.

Deny Patterns​

You can define patterns for crawler to skip in findkit.toml. Patterns can be strings or regexes in string staring with reg:. Regexes need to be written as strings without the encapsulating //

Strings are matched to start of any URL pathname. Regular Expression patterns are matched against the whole URL pathname.

You can define as many deny patterns as you need, but there are better options available for skipping individual urls.

Example deny pattern string:

[[targets]]
host = "example.com"
content_selector = ".content, .post"
deny_patterns = ["/secret/"]

Would cause crawler to skip:

  • example.com/secret/
  • example.com/secret/first_secret But would not skip:
  • example.com/secret extraneous / at the end of deny pattern
  • example.com/some_folder/secret/ deny pattern is matched against the start of the URL pathname

Example deny pattern regular expression:

[[targets]]
host = "example.com"
content_selector = ".content, .post"
deny_patterns = ["reg:/secret"]

Would cause crawler to skip:

  • example.com/secret
  • example.com/secret/
  • example.com/secret/first_secret
  • example.com/some_folder/secret/
  • example.com/some_folder/secret/first_secret
  • example.com/secretariat/big_red