Skip to content

petere/pguri

Repository files navigation

uri type for PostgreSQL

https://twitter.com/pvh/status/567395527357001728

This is an extension for PostgreSQL that provides a uri data type. Advantages over using plain text for storing URIs include:

  • URI syntax checking
  • functions for extracting URI components
  • human-friendly sorting

The actual URI parsing is provided by the uriparser library, which supports URI syntax as per RFC 3986.

Note that this might not be the right data type to use if you want to store user-provided URI data, such as HTTP referrers, since they might contain arbitrary junk.

Installation

You need to have the above-mentioned uriparser library installed. It is included in many operating system distributions and package management systems. pkg-config will be used to find it. I recommend at least version 0.8.0. Older versions will also work, but they apparently contain some bugs and might fail to correctly accept or reject URI syntax corner cases. This is mainly a problem if your application needs to be robust against junk input.

To build and install this module:

make
make install

or selecting a specific PostgreSQL installation:

make PG_CONFIG=/some/where/bin/pg_config
make PG_CONFIG=/some/where/bin/pg_config install

And finally inside the database:

CREATE EXTENSION uri;

Using

This module provides a data type uri that you can use like a normal type. For example:

CREATE TABLE links (
    id int PRIMARY KEY,
    link uri
);

INSERT INTO links VALUES (1, 'https://github.com/petere/pguri');

A number of functions are provided to extract parts of a URI:

  • uri_scheme(uri) returns text

    Extracts the scheme of a URI, for example http or ftp or mailto.

  • uri_userinfo(uri) returns text

    Extracts the user info part of a URI. This is normally a user name, but could also be of the form username:password. If the URI does not contain a user info part, then this will return null.

  • uri_host(uri) returns text

    Extracts the host of a URI, for example www.example.com or 192.168.0.1. (For IPv6 addresses, the brackets are not included here.) If there is no host, the return value is null.

  • uri_host_inet(uri) returns inet

    If the host is a raw IP address, then this will return it as an inet datum. Otherwise (not an IP address or no host at all), the return value is null.

  • uri_port(uri) returns integer

    Extracts the port of a URI as an integer, for example 5432. If no port is specified, the return value is null.

  • uri_path(uri) returns text

    Extracts the path component of a URI. Logically, a URI always contains a path. The return value can be an empty string but never null.

  • uri_path_array(uri) returns text[]

    Returns the path component of a URI as an array, with the path split at the slash characters. This is probably not as useful as the uri_path function, but it is provided here because the uriparser library exposes it.

  • uri_query(uri) returns text

    Extracts the query part of a URI (roughly speaking, everything after the ?). If there is no query part, returns null.

  • uri_fragment(uri) returns text

    Extracts the fragment part of a URI (roughly speaking, everything after the #). If there is no fragment part, returns null.

Other functions:

  • uri_normalize(uri) returns uri

    Performs syntax-based normalization of the URI. This includes case normalization, percent-encoding normalization, and removing redundant . and .. path segments. See RFC 3986 section 6.2.2 for the full details.

    Note that this module (and similar modules in other programming languages) compares URIs for equality in their original form, without normalization. If you want to consider distinct URIs without regard for mostly irrelevant syntax differences, pass them through this function.

  • uri_escape(text, space_to_plus boolean DEFAULT false, normalize_breaks boolean DEFAULT false) returns text

    Percent-encodes all reserved characters from the text. This can be useful for constructing URIs from strings.

    If space_to_plus is true, then spaces are replaced by plus signs. If normalize_breaks is true, then line breaks are converted to CR LF pairs (and subsequently percent-encoded). Note that these two conversions come from the HTML standard for encoding form data but are not part of the specification for URIs.

  • uri_unescape(text, plus_to_space boolean DEFAULT false, break_conversion boolean DEFAULT false) returns text

    Decodes all percent-encodings in the text.

    If plus_to_space is true, then plus signs are converted to spaces. If break_conversion is true, then CR LF pairs are converted to simple newlines (\n).