Update About This Fork
This fork was made in early September of 2019. When attempting to install npm audit
indicated that the open
package had serious security risks.
Trying to use --open
to view the dashboard in a browser doesn't work anyway (on Ubuntu 18.04) so there may have been an OS level patch of some kind...not that this is secure, apparently phantomjs
also has some vulnerabilities. Buyer beware.
I don't have Java installed so I don't know about the Nu HTML Checker
and I haven't tried it via Docker but, I updated the packages in package.json
to refelect the currently installed and working version--at least, if it still works now that I have removed open
from the list.
I was able to get it working on OpenLitespeed, which acts as a reverse proxy to itself so that I can server the app on port 80 and visitors can use it without appending a port to the URL.
The A11y Machine (or a11ym
for short, spelled “alym”) is an automated
accessibility testing tool which crawls and tests pages of any Web
application to produce detailed reports. It validates pages against the
following specifications/laws:
- W3C Web Content Accessibility Guidelines (WCAG) 2.0, including A, AA and AAA levels (understanding levels of conformance),
- U.S. Section 508 legislation,
- W3C HTML5 Recommendation.
If privacy matters for you, you're likely to install The A11y Machine over any SaaS services: It runs locally so you don't need to send your code somewhere, you can test all parts of your application including the ones which require an authentification (like a checkout, a back-office etc.)…
Here are some pros and cons compared to SaaS solutions:
Properties | The A11y Machine | SaaS services |
---|---|---|
Can run locally | yes | no |
Can test each patch | yes | no (except if deployed) |
Reduce the test loop | yes | no (the loop is longer) |
Can test private code | yes | no (you must send your code) |
Can test auth-required parts | yes | no |
Can crawl all your pages | yes | yes (but it can be pricey) |
Accessibility is not only a concern for disabled people. Bots can be considered as such, like DuckDuckGo, Google or Bing. By respecting these standards, you're likely to have a better ranking. Also it helps to clean your code. Accessibility issues are often left unaddressed for budget reasons. In fact most of the cost is spent looking for errors on your website. The A11y Machine greatly help with this task, you can thus focus on fixing your code and reap the benefits.
NPM is required. Then, execute the following lines:
$ npm install -g the-a11y-machine
If you would like to validate your pages against the HTML5 recommendation, then you need to install Java.
As an alternative you can run a Docker image instead, which will ensure the image is available locally:
$ docker build -t liip/the-a11y-machine .
$ docker run liip/the-a11y-machine --help
To get access to a report you will need to:
- Mount a path into the container,
- Specifify that internal path in your
a11ym
CLI options.
For example:
$ docker run -v $PWD:/var/output liip/the-a11y-machine -o /var/output http://example.org
As a prelude, see the help:
Usage: a11ym [options] url …
Options:
-h, --help output usage information
-b, --bootstrap <path> Bootstrap file, i.e. the configuration file. All CLI options will overwrite options defined in the configuration file.
-e, --error-level <error_level> Minimum error level: In ascending order, `notice` (default), `warning`, and `error` (e.g. `warning` includes all warnings and errors).
-c, --filter-by-codes <codes> Filter results by comma-separated WCAG codes (e.g. `H25,H91,G18`).
-C, --exclude-by-codes <codes> Exclude results by comma-separated WCAG codes (e.g. `H25,H91,G18`).
-d, --maximum-depth <depth> Explore up to a maximum depth (hops).
-m, --maximum-urls <maximum_urls> Maximum number of URLs to compute.
-o, --output-directory <output_directory> Output directory.
-r, --report <report> Report format: `cli`, `csv`, `html` (default), `json` or `markdown`.
-s, --standards <standards> Standard to use: `WCAG2A`, `WCAG2AA` (default), ` WCAG2AAA`, `Section508`, _`HTML`_ or your own (see `--sniffers`). `HTML` can be combined with any other by a comma.
-S, --sniffers <sniffers> Path to the sniffers file, e.g. `resource/sniffers.js` (default).
-u, --filter-by-urls <urls> Filter URL to test by using a regular expression without delimiters (e.g. 'news|contact').
-U, --exclude-by-urls <urls> Exclude URL to test by using a regular expression without delimiters (e.g. 'news|contact').
-w, --workers <workers> Number of workers, i.e. number of URLs computed in parallel.
--http-auth-user <http_auth_user> Username to authenticate all HTTP requests.
--http-auth-password <http_auth_password> Password to authenticate all HTTP requests.
--http-tls-disable Disable TLS/SSL when crawling or downloading pages.
-V, --no-verbose Make the program silent.
--ignore-robots-txt Ignore robots.txt file.
Thus, the simplest use is to run a11ym
with a URL:
$ ./a11ym http://example.org/
All URLs accessible from http://example.org/
will be tested against the
WCAG2AA standard. See the --maximum-urls
options to reduce the number of
URLs to test.
Then open a11ym_output/index.html
and browser the result!
You can compute several URLs by adding them to the command-line, like this:
$ ./a11ym http://example.org/A http://example.org/B http://example.org/C
Alternatively, this is possible to read URLs from STDIN, as follows:
$ cat URLs.lists | ./a11ym -
Note the -
: It means “Read URLs from STDIN please”.
When reading several URLs, the --maximum-depth
option will be forced to 1.
The index of the reports:
Report of a specific URL:
The dashboard of all reports:
As mentionned, the following standards are supported:
- W3C WCAG,
- U.S. Section 508 legislation,
- W3C HTML5 recommendation.
You cannot combine standards between each other, except HTML5 that can be
combined with any other. So for instance, to run WCAG2AAA
:
$ ./a11ym --standards WCAG2AAA http://example.org/
To run WCAG2AA
along with HTML
:
$ ./a11ym --standards WCAG2AA,HTML http://example.org/
The pipe looks like this:
- The
node-simplecrawler
tool is used to crawl a Web application based on the given URLs, with our own specific exploration algorithm to provide better results quickly, in addition to support parallelism, - For each URL found, 2 kind of tests are applied:
- Accessibility: PhantomJS runs and
HTML_CodeSniffer
is injected in order to check the page conformance; This step is semi-automated by the help ofpa11y
, which is a very thin layer of code wrapping PhantomJS andHTML_CodeSniffer
, - HTML: The Nu Html Checker (v.Nu) is run on the same URL.
- Accessibility: PhantomJS runs and
- Finally, results from different tools are normalized, and enhanced and easy to use reports are produced.
PhantomJS and HTML_CodeSniffer
are widely-used, tested and precise tools.
pa11y
simplifies the use of these two latters. The Nu Html Checker is the tool
used by the W3C to validate documents online. However, in this case, we do all
validations offline! Nothing is sent over the network. Again, privacy.
HTML_CodeSniffer
is build in a way that allows you to extend existing rules or
write your own. A rule is represented as a sniffer (this is another
terminology). The resource/sniffers/
directory contains an example of
a custom sniffer.
The A11y Machine comes with a default file containing all the sniffers:
resource/sniffers.js
. You can provide your own by using the --sniffers
option. To build your own sniffers, simply copy the resource/sniffers/
somewhere as a basis, complete it, then compile it with the a11ym-sniffers
utility:
$ ./a11ym-sniffers --directory my/sniffers/ --output-directory my_sniffers.js
Then, to effectively use it:
$ ./a11ym --sniffers my_sniffers.js --standard MyStandard http://example.org/
Run the a11ym-dashboard
command to serve the dashboard. The dashboard is an
overview of several reports generated by the a11ym
command. The command can
serves the dashboard over HTTP, or over static files. In addition to requiring
a root directory, it requires: In the first case, an address, and a port, in the
second, nothing more than just a flag. For instance, if the reports are
generated with the following command:
$ ./a11ym --output-directory my_reports/`date +%s`/ http://example.org/A http://example.org/B
Then, the root directory is my_reports/
and thus the dashboard will be started
over HTTP with the following command:
$ ./a11ym-dashboard --root my_reports
Browse 127.0.0.1:8080
(by default) to see the dashboard!
Or to generate static files:
$ ./a11ym-dashboard --root my_reports --static-output
Open my_reports/index.html
, and do the same!
Bonus: Use the --open
option to automatically open the dashboard in your
favorite browser.
The roadmap is public:
The board is publicly available at the following URL: https://waffle.io/liip/TheA11yMachine.
Original author is Ivan Enderlin, accompagnied by Gilles Crettenand and David Jeanmonod. This software is backed by Liip.
Copyright (c), Ivan Enderlin and Liip All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.