Python uriref

version

0.0.3-dev-20170106

last-update

2020-10-11

description

URL and URN parser written in regular expressions. Based on RFC 2396 BNF terms, update to RFC 3986 planned but not started.

license

FreeBSD

status

This is an experimental library. Do not use it in production unless you are prepared to put in time for testing that it does what you need.

uriref reference matching, compared to stdlib urlparse for several iteration-counts. The implementations are not tested for identical operation.

The diagram shows constant times for each iteration count. The regex implementation outperforms stdlib's urlparse module by almost 100%. The latter runs at slighty above 6e-4 seconds, with the former at ~3.5e-4 seconds (at my machine).

Memory profiling remains to be done, I expect regex is taking a lot more.

There are almost 100 tests, a good bunch of which need to be reviewed (33 failures). The modules has 34% test coverage.

Installation of library and parseuri.py is via:

python setup.py install

For lib source see uriref.

Usage

Examples:

uriref-cli.py -qs absolute <URL> # Quietly validate URL reference
uriref-cli.py --pretty <URL>     # Parse URI parts to human readable table
uriref-cli.py -O yaml <URL>      # Parse URI parts to YAML
uriref-cli.py --pretty -O json <URL> # etc.

See also bin/examples.py.

Tests

python test/py/main.py

Or:

make test

Coverage report and test results are available in html.

There are tests that show for which sort of URLs uriref is compatible with stdlib urlparse.

TODO The setup should be fixed by splitting up the expected test results to function. Currently there is one set of parameters for all test methods.