php - json_encode won't encode French characters

Question

I'm trying to work with French characters. For whatever reason, PHP won't output them correctly unless I force the ISO-8859-1 character set (which I really don't want to do); it won't let me use UTF-8.

So doing a print_r() of my array when I force ISO-8859-1 yields the following:

Array
(
[0] => stdClass Object
    (
        [language] => fr
        [langselect] => ?
        [s1next] => Prochain
        [hascodespan] => Avez-vous déjà un code d'enregistrement?
        [s2prev] => Précédent
        [s2next] => Oui, j'ai déjà un code.
        [ecodedescription] => Un E-Code vous donne un accès exclusif à des événements vraiment cool. Si vous êtes dans le besoin d'un code, vous pouvez en acheter un dans notre boutique en ligne en visitant ce lien <a href="#"> boutique en ligne</ a>.
        [purchase] => Cliquez ici pour acheter en ligne billets
        [ecodespan] => S'il vous pla?t entrer votre e-code
        [ecodelocdescription] => Votre code peut être trouvé ci-dessous le code à barres sur votre billet
        [s3prev] => Précédent
        [s3next] => Prochain
        [validationtext] => Validation E-Code ... Un instant.
    )

)

When using UTF-8, the output is as following:

Array
(
[0] => stdClass Object
    (
        [language] => fr
        [langselect] => ?
        [s1next] => Prochain
        [hascodespan] => Avez-vous d?j? un code d'enregistrement?
        [s2prev] => Pr?c?dent
        [s2next] => Oui, j'ai d?j? un code.
        [ecodedescription] => Un E-Code vous donne un acc?s exclusif ? des ?v?nements vraiment cool. Si vous ?tes dans le besoin d'un code, vous pouvez en acheter un dans notre boutique en ligne en visitant ce lien <a href="#"> boutique en ligne</ a>.
        [purchase] => Cliquez ici pour acheter en ligne billets
        [ecodespan] => S'il vous pla?t entrer votre e-code
        [ecodelocdescription] => Votre code peut ?tre trouv? ci-dessous le code ? barres sur votre billet
        [s3prev] => Pr?c?dent
        [s3next] => Prochain
        [validationtext] => Validation E-Code ... Un instant.
    )

)

In both instances, executing a json_encode() yields the following result:

[
    {
        language: "fr",
        langselect: null,
        s1next: "Prochain",
        hascodespan: null,
        s2prev: null,
        s2next: null,
        ecodedescription: null,
        purchase: "Cliquez ici pour acheter en ligne billets",
        ecodespan: null,
        ecodelocdescription: null,
        s3prev: null,
        s3next: "Prochain",
        validationtext: "Validation E-Code ... Un instant."
    }
]

I have my database set to UTF-8, yet for some odd reason every time I insert anything with French characters it reverts itself to some Western European encoding.

Mainly, I really need json_encode() to return valid results so I can use it in my translations. I have tried iconv() and utf8_encode(), but to no avail.

Any help would be greatly appreciated.

Answer 1

Please, besides the database encoding, be sure you check the following:

• utf8 encoding of the FILES (js/php)
• utf8 html content: <meta HTTP-EQUIV="content-type" CONTENT="text/html; charset=utf-8">
• utf8 of your db connection: SET character_set_connection = 'utf8'
• utf8 of your mysql tables: ALTER TABLE table CONVERT TO CHARACTER SET utf8;
• utf8 of your query results: SET character_set_results = 'utf8'
• utf8 of your db client: SET character_set_client = 'utf8'
• utf8 of your db server: SET character_set_database = 'utf8' and SET character_set_server = 'utf8'
• in some cases, forcing utf8 in file is necessary when hardcoded values need encoding. You would need to add a comment on top of your php/js/xml file for instance, with charset=utf-8, so ultra edit or your favorite editor can detect it.

rgds.