Functions for working with URLs

Overview {#overview}

:::note The functions mentioned in this section are optimized for maximum performance and for the most part do not follow the RFC-3986 standard. Functions which implement RFC-3986 have RFC appended to their function name and are generally slower. :::

You can generally use the non-RFC function variants when working with publicly registered domains that contain neither user strings nor @ symbols. The table below details which symbols in a URL can (✔) or cannot (✗) be parsed by the respective RFC and non-RFC variants:

Symbol	non-RFC	RFC
' '	✗	✗
\t	✗	✗
<	✗	✗
>	✗	✗
%	✗	✔*
{	✗	✗
}	✗	✗
|	✗	✗
\\	✗	✗
^	✗	✗
~	✗	✔*
[	✗	✗
]	✗	✔
;	✗	✔*
=	✗	✔*
&	✗	✔*

symbols marked * are sub-delimiters in RFC 3986 and allowed for user info following the @ symbol.

There are two types of URL functions:

• Functions that extract parts of a URL. If the relevant part isn't present in a URL, an empty string is returned.
• Functions that remove part of a URL. If the URL does not have anything similar, the URL remains unchanged.

:::note The functions below are generated from the system.functions system table. :::

<!-- The inner content of the tags below are replaced at doc framework build time with docs generated from system.functions. Please do not modify or remove the tags. See: https://github.com/ClickHouse/clickhouse-docs/blob/main/contribute/autogenerated-documentation-from-source.md --> <!--AUTOGENERATED_START--> <!--AUTOGENERATED_END-->