[TYPO3-core] RFC: Bug 5602: spamProtectEmailAddresses_atSubst is not recognized correctly

[Martin Kutschker]

Oliver Hader schrieb:
> Hi Masi,
> 
> Martin Kutschker schrieb:
>> Oliver Hader schrieb:
>>> This is a SVN patch request.
>>>
>>> Problem:
>>> Imagine a TypoScript like the following one and test it:
>>>     config {
>>>       spamProtectEmailAddresses = 2
>>>       spamProtectEmailAddresses_atSubst = [at]
>>>       spamProtectEmailAddresses_lastDotSubst = [dot]
>>>     }
>>>     10 = TEXT
>>>     10.value = <b>john at doe.com</b>
>>>     10.typolink.parameter = john at doe.com
>>> The result is, the the e-mail address is not "protected" by [at] or
>>> [dot] and is shown unsubstituted. Not remove the wrapping bold tags
>>> (<b>|</b>) from the value property of the TEXT object from above and
>>> test again. Everything fine, as expected...
>>> tslib_cObj:getMailTo($mailAddress,$linktxt,$initP='?') only substituded
>>> the "at" and "dot" if the $mailAddress (here: 10.typolink.parameter) and
>>> the $linktxt (here: 10.value) are the same. This behavior was introduced
>>> in RFC #3985 by... me (sorry!)...
>>>
>>> Solution:
>>> Create a spam protected variant of the $mailAddress and and the end of
>>> tslib_cObj:getMailTo() substitute the unprotected mail address by the
>>> spam protected mail address. In this case it doesn't matter in which
>>> context/wraps the e-mail address is nested.
>>>
>>> Comments:
>>> #RFC 3985 was about wrong substitution of "dot" and "at" in non e-mail
>>> addresses, like filenames or directories. See the following example:
>>>     <a href="mailto:..."><img src="/my.files/logo.jpg" /></a>
>>> was changed to
>>>     <a href="mailto:..."><img src="/my.files/logo[dot]jpg" /></a>
>> Any news on this? The discussion did not end...
> 
> Well, concerning Franz' post, I don't think this is a misbehaviour. On
> using "spamProtectEmailAddresses = ascii", we'll get the following for
> the e-mail address "john at doe.com":
> 
> <a
> href="&#109;&#97;&#105;&#108;&#116;&#111;&#58;&#106;&#111;&#104;&#110;&#64;&#100;&#111;&#101;&#46;&#99;&#111;&#109;">john[at]doe.com</a>
> 
> The 'mailto:' is also converted to ASCII characters. But this doesn't
> matter, since browsers accept this, too.

And any mail harvester too. This is the most pointless option in TYOP3 
ever ;-)

Masi

PS: And now for some nit picking. They are not ASCII characters, they 
are (SGML) entities. And this entities are not ASCII encoded they are in 
Unicode.