Hi,

i've found a weird behavior in domain extraction;

$extract = new Extract();
$result = $extract->parse('dkim._domainkey.phea.fr');
print_r($result->toArray());

$result = $extract->parse('dkim.domainkey.phea.fr');
print_r($result->toArray());

result

Array
(
    [subdomain] => dkim._domainkey.phea
    [hostname] => fr
    [suffix] => 
)
Array
(
    [subdomain] => dkim.domainkey
    [hostname] => phea
    [suffix] => fr
)

the problem come from _ character.

this regex fix the problem

    const HOSTNAME_PATTERN = '#^((?!-)[a-z0-9_-]{0,62}[a-z0-9]\.)+[a-z]{2,63}|[xn\-\-a-z0-9]]{6,63}$#';

I use this library inside a symfony project where intl is not required because symfony give a polyfill for it.

Currently its not possible to use this library without intl installed (composer will error). I would recommend to use the symfony intl polyfill https://github.com/symfony/polyfill-intl-icu instead of require ext-intl.

"symfony/polyfill-intl-icu": "^1.0",

Thanks for this extension.

I get this error: Cannot put TLD list to cache file /xxxxx/vendor/layershifter/tld-extract/src/../cache/.tld_set, check writes rights on directory of file

So, I should change permission cache directory in vendor that this is not good job.

I suggest modify code for check permission and writable with is_writable() and if don't have permission, try to change with chmod(). or more better solution is auto created cache directory with mkdir('cache', 0777); command.

Hi, I'm getting the error:

Stack trace:
Message: Uncaught OutOfRangeException: Unknown field "errors" in xxx/vendor/layershifter/tld-extract/src/Result.php:220
File: xxx/vendor/layershifter/tld-extract/src/Result.php
Line: 220

thrown
#0 LayerShifter\TLDExtract\Result->__get('errors')

My code is straight from the example:

$extract = new LayerShifter\TLDExtract\Extract();
$result = $extract->parse('shop.github.com');
$result->getRegistrableDomain();

Composer install details:

Using version ^2.0 for layershifter/tld-extract
./composer.json has been updated
Loading composer repositories with package information
Updating dependencies (including require-dev)
Package operations: 5 installs, 0 updates, 0 removals
  - Installing symfony/polyfill-php72 (v1.10.0): Downloading (100%)         
  - Installing symfony/polyfill-intl-idn (v1.10.0): Downloading (100%)         
  - Installing layershifter/tld-support (1.1.1): Downloading (100%)         
  - Installing layershifter/tld-database (1.0.65): Downloading (100%)         
  - Installing layershifter/tld-extract (2.0.1): Downloading (100%)   

Am I missing something or doing something wrong?

Many thanks,

jkns

As part of some validation requirements, I wanted to loop over subdomain labels, and since an array was the expected return of the getSubdomains() function, I assumed it would return an empty array if none were found.

Turns out it returns a null, meaning it broke any attempts to use it as the expected return type. I ended up working around this using a null coalescence like the following:

$subdomainLabels = $hostnameExtract->getSubdomains() ?? array();

So, should this function return an empty array by default? I'd personally prefer if it did, since it keeps the return type consistent - but others might have differing opinions.

I have a loop that extracts registerable domains. However, despite most of them running fine, a still have a lot of empty responses. What could be the cause that it does not find a registerable domain for the following examples :

mail-sor-f69.google.com (you would expect google.com) static.vnpt.vn (you would expect vnpt.vn) mx1.sub5.homie.mail.dreamhost.com (you would expect dreamhost.com)

and there are many , many, many more.

It seems that IPv6 addresses are not recognized properly:

>>> $w = new LayerShifter\TLDExtract\Extract()
=> LayerShifter\TLDExtract\Extract {#197}
>>> $res = $w->parse('2bf2:eaa0::7:314:5474')
=> LayerShifter\TLDExtract\Result {#205}
>>> $res->isIp()
=> false
>>> $res->isValidDomain()
=> false

For "http://example.com?foo=bar", getRegistrableDomain() returns "example.com?foo=bar" rather than "example.com".

My temporary quick fix in my code is to replace all "?" characters with "/" before passing the URL to TLDExtract.

Currently the suffixExists method on the Extract class is private, it would be useful if it was protected, so you could override default behavior and provide an alternate data source

$result = $extract->parse('test..com');
$result->isValidDomain();  // gives true

Registrable Domain ist .com in this case... I think this domain name should be invalid.

Hi,

My code is as follows: $line = 'http://www.upfile.co.il/xxxxxxxx.html'; $extract = new LayerShifter\TLDExtract\Extract(); $domain = $extract->parse($line); $domain->getRegistrableDomain();

I get www.upfile.co.il instead of upfile.co.il.

Any idea what is going wrong?

I'm running some domain names through TLDExtract and came across a domain not being properly parsed.

The URL is called blogspot.com

$url = 'blogspot.com';
$domain = tld_extract($url);
var_dump($domain);

Returns: 
object(LayerShifter\TLDExtract\Result)[9]
  private 'subdomain' => null
  private 'hostname' => string 'blogspot.com' (length=12)
  private 'suffix' => null

Weirdly the URL 'flogspot.com' works fine and returns:

object(LayerShifter\TLDExtract\Result)[9]
  private 'subdomain' => null
  private 'hostname' => string 'flogspot' (length=8)
  private 'suffix' => string 'com' (length=3)

The URL logspot.com also works and returns:

object(LayerShifter\TLDExtract\Result)[9]
  private 'subdomain' => null
  private 'hostname' => string 'logspot' (length=7)
  private 'suffix' => string 'com' (length=3)

Any idea why the TLD in 'blogspot.com' is not being added to the suffix? Is this a bug?

Some reading for improve or fix in case it’s needed :

https://en.wikipedia.org/wiki/Domain_name introduce https://tools.ietf.org/html/rfc1034 https://tools.ietf.org/html/rfc1035

domain name are URL https://en.wikipedia.org/wiki/Url introduce https://url.spec.whatwg.org/

should permit to close #45

$extract = new Extract();
$result = $extract->parse("test@[email protected]")->isValidDomain();

This results in true, when it should be false ("@" is not allowed in domain names)

The parser fails for the following:

// If the subdomain has "-"
$url = 'https://s3-ap-southeast-2.amazonaws.com/blabla/blabla/wp-content/uploads/media/2019/03/16860571424_31c94205de_b.jpg';

// Extract domain parts
$extract = new \LayerShifter\TLDExtract\Extract();
$domainParser = $extract->parse($url);

parse_url($url, PHP_URL_HOST); // s3-ap-southeast-2.amazonaws.com
$domainParser->getSubdomain(); // null 

I have tested with hundred of domains, but for the domain test.ru i get the following result:

Result {#1711 ▼
  -subdomain: null
  -hostname: "test.ru"
  -suffix: null
}

Hello There is a bug in the libary where it mishandles the backslash character

For example https://1337.karimrahal.com.bla.com would give bla.com as the registered domain

where as the browser would lead to karimrahal.com domain