Posted
about 13 years
ago
by
samran
wellnessdd
|
Posted
about 13 years
ago
by
Brett Zamir
@John Supplee, @Josep Sanz, @Luke Scott: The ut8_decode was removed in Git. We should indeed be following PHP and UTF8-friendly behavior.
@Chris Buckley: Fixed comments in Git.
@Yaffle: I see PHP also doesn't depend on the padding (though the MIME
|
Posted
about 13 years
ago
by
Brett Zamir
@ekim: No, var_export in php.js currently does not support recursive references (or flag them as errors). Feel free to add support!
@Hans Henrik: We cannot use JSON.stringify for a few reasons: 1) JSON doesn't allow functions; JSON is only a subset
|
Posted
about 13 years
ago
by
Brett Zamir
@Chris: I have applied your fix to Git. Thanks!
|
Posted
about 13 years
ago
by
Brett Zamir
@Igor Tykhyy: (I should have mentioned that the notes to the function also mention this.)
|
Posted
about 13 years
ago
by
Brett Zamir
@Edward: I have made fixes in Git (also for unpack() which is in the workbench). Please check "raw js source".
@Igor Tykhyy: I am just trying to give a quick answer now, but it looks like each character is examined in the data, with its Unicode
|
Posted
about 13 years
ago
by
Brett Zamir
@Rajiva: In addition to my last question, what browser were you using?
|
Posted
about 13 years
ago
by
Brett Zamir
@Rajiva: Can you put your long sequence somewhere so we can confirm? I am not able to replicate with decodeURIComponent() (which would be giving the URI error) even with a very large sequence:
var arr = decodeURIComponent(new Array(101000).join('%20'));
|
Posted
about 13 years
ago
by
Brett Zamir
@Oscar Broman: Can you integrate at least as comprehensive argument checking as we have in our existing version, following PHP behavior without losing the performance benefits of your new version?
|
Posted
about 13 years
ago
by
Brett Zamir
@Michal: Thanks--If you have any way to shorten the existing function, without losing support for all the existing arguments and functionality, let us know.
|