Thread

Is there a documented fix for the character encoding problems that result from the differing encodings between pmachine db structure and EE structures?

Not that I know of- it’s more of a db issue.  Did you keep the encodings consistent?  latin->latin?  If so- can you link to an example page where it’s borking?

I just followed the step by step in the CP Home ›  Admin ›  pMachine Pro Import Utility.  I *know* it’s a problem with latin->utf8.

Quite a few � on this page:

http://spl.dfbills.com/P30/

Yep- you can’t just change the encode setting- that doesn’t actually convert the existing data.  If it was latin- need to keep it latin.  Or- actuall convert the file.  Which isn’t necessarily easy.  I think Ingmar linked to a script before that did it.

BUT- can you just keep it latin?  When I force the browser to use latin/wester, page seems to render fine.  Try it in your browser.  So basically- keep it latin, switch the header, which is now:

<meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>

If you’re using the global to output that, then switching it in the cp will automatically switch it in the header.  And add new entries in latin- assuming you haven’t changed the db settings.

But to go latin-utf?  You actually have to convert the data.  Looks to me like the old data are still latin.

That make any sense?  This stuff is tricky.

Robin’s spot on, of course. I suspect your data was Latin-1 (nothing wrong with that), and now your website uses utf-8. That works only for characters where it’s never an issue anyway, ie 7-bit ASCII. So, to reiterate what she said: try changing utf-8 to latin-1 in your control panel—or take the plunge, convert your database and re-upload again. Your call 😊

I didn’t have time to test this evening, but are you saying that I shouldn’t bother with the script you posted previously?

I think it boils down to what you want to do.  Do you want/need to move to utf-8?  If so, then you’ll need to convert from the latin->utf-8, then import.

If you don’t need/want to?  Then just leaving it latin should solve the issue.

Not really a right/wrong per se.  Leaving it be in latin is the easy option.  But it’s not the only option.

Make sense?

Finally attempting this now-

I’ve get EE set to ISO 8859-1 encoding and I’m importing from the pMachine database with a Latin-1 encoding.  But as you said, it is really all in the display settings.

<meta http-equiv=“Content-Type” content=“text/html; charset=charset=ISO-8859-1”/>

instead of

<meta http-equiv=“Content-Type” content=“text/html; charset=utf-8”/>

Seems to do the trick.

Glad to hear it. Does that mean you’re all set?

I guess.. until I need to use extended characters 😊

OK, I’m closing that one out then, for now. Unless you need, say, Chinese or Russian characters, using either Latin-1 or Unicode should be fine. If anything comes up, just start a new thread. Thanks.