README.md
# validate-website
## Description
Web crawler for checking the validity of your documents
## Installation
### Debian
```
apt install ruby-dev libxslt1-dev libxml2-dev
```
If you want complete local validation look [tidy
packages](https://binaries.html-tidy.org/)
### RubyGems
```
gem install validate-website
```
## Synopsis
```
validate-website [OPTIONS]
validate-website-static [OPTIONS]
```
## Examples
```
validate-website -v -s https://www.ruby-lang.org/
validate-website -v -x tidy -s https://www.ruby-lang.org/
validate-website -v -x nu -s https://www.ruby-lang.org/
validate-website -h
```
## Description
validate-website is a web crawler for checking the markup validity with XML
Schema / DTD and not found urls (more info [doc/validate-website.adoc](https://github.com/spk/validate-website/blob/master/doc/validate-website.adoc)).
validate-website-static checks the markup validity of your local documents with
XML Schema / DTD (more info [doc/validate-website-static.adoc](https://github.com/spk/validate-website/blob/master/doc/validate-website-static.adoc)).
HTML5 support with [libtidy5](http://www.html-tidy.org/) or [Validator.nu Web
Service](https://checker.html5.org/).
## Exit status
* 0: Markup is valid and no 404 found.
* 64: Not valid markup found.
* 65: There are pages not found.
* 66: There are not valid markup and pages not found.
## On your application
``` ruby
require 'validate_website/validator'
body = '<!DOCTYPE html><html></html>'
v = ValidateWebsite::Validator.new(Nokogiri::HTML(body), body)
v.valid? # => false
```
## Jekyll static site validation
You can add this Rake task to validate a
[jekyll](https://github.com/jekyll/jekyll) site:
``` ruby
desc 'validate _site with validate website'
task validate: :build do
Dir.chdir("_site") do
system("validate-website-static",
"--verbose",
"--exclude", "examples",
"--site", HTTP_URL)
exit($?.exitstatus)
end
end
end
```
## More info
### HTML5
#### Tidy5
If the libtidy5 is found on your system this will be the default to validate
your html5 document. This does not depend on a tier service everything is done
locally.
#### nokogiri
nokogiri can validate html5 document without tier service but reports less
errors than tidy.
#### Validator.nu web service
When `--html5-validator nu` option is used HTML5 support is done by using the
Validator.nu Web Service, so the content of your webpage is logged by a tier.
It's not the case for other validation because validate-website use the XML
Schema or DTD stored on the data/ directory.
Please read <http://about.validator.nu/#tos> for more info on the HTML5
validation service.
##### Use validator standalone web server locally
You can download [validator](https://github.com/validator/validator) jar and
start it with:
```
java -cp PATH_TO/vnu.jar nu.validator.servlet.Main 8888
```
Then you can use validate-website option:
```
--html5-validator-service-url http://localhost:8888/
# or
export VALIDATOR_NU_URL="http://localhost:8888/"
```
This will prevent you to be blacklisted from validator webservice.
## Tests
With standard environment:
```
bundle exec rake
```
## Credits
* Thanks tenderlove for Nokogiri, this tool is inspired from markup_validity.
* And Chris Kite for Anemone web-spider framework and postmodern for Spidr.
## Contributors
See [GitHub](https://github.com/spk/validate-website/graphs/contributors).
## License
The MIT License
Copyright (c) 2009-2022 Laurent Arnoud <laurent@spkdev.net>
---
[](https://gitlab.com/spkdev/validate-website/-/commits/master)
[](https://gitlab.com/spkdev/validate-website/-/commits/master)
[](https://rubygems.org/gems/validate-website)
[](http://www.rubydoc.info/gems/validate-website)
[](http://opensource.org/licenses/MIT "MIT")
[](http://inch-ci.org/github/spk/validate-website)