Finally out of MySQL encoding hell [en]

[fr] Description de comment je me suis sortie des problèmes d'encodage qui résultaient en l'affichage de hiéroglyphes sur tous les sites hébergés sur mon serveur.

It took weeks, mainly because I was busy with a car accident and the end of school, but it also took about two real whole days of head-banging on the desk to get it fixed.

Here’s what happened: remember, a long time ago, I had trouble with stuff in my database which was supposed to be UTF-8 but seemed to be ISO-8859-1? And then, sometime later, I had a weird mixture of UTF-8 and ISO-8859-1 in the same database?

Well, somewhere along the line this is what I guess happened: my database installation must have been serving UTF-8 content as ISO-8859-1, leading me to believe it was ISO-8859-1 when it was in fact UTF-8. That led me to try to convert it to UTF-8 — meaning I took UTF-8 strings and ran them through a converter supposed to turn ISO-8859-1 into UTF-8. The result? Let’s call it “double-UTF-8” (doubly encoded UTF-8), for want of a better name.

Anyway, that’s what I had in my database. When we upgraded MySQL and PHP on the server, I suddenly started seeing a load of junk instead of my accented characters:

encoding-problem-2

What I was seeing looked furiously like UTF-8 looks when your server setup is messed up and serves it as ISO-8859-1 instead. But, as you can see on the picture above, this page was being served as UTF-8 by the server. How did I know it wasn’t ISO-8859-1 in my database instead of this hypothetical “double-UTF-8”? Well, for one, I knew the page was served as UTF-8, and I also know that ISO-8859-1 (latin-1) served as UTF-8 makes accented characters look like question marks. Then, if I wanted to be sure, I could just change the page encoding in Firefox to ISO-8859-1 (that should make it look right if it was ISO-8859-1, shouldn’t it?) Well, it made it look worse.

Another indication was that when the MySQL connection encoding (in my.cnf) was set back to latin-1 (ISO-8859-1), the pages seemed to display correctly, but WordPress broke.

The first post on the picture I’m showing here looks “OK”, because it was posted after the setup was changed. It really is UTF-8.

Now how did we solve this? My initial idea was to take the “double-UTF-8” content of the database (and don’t forget it was mixed with the more recent UTF-8 content) and convert it “from UTF-8 to ISO-8859-1”. I had a python script we had used to fix the last MySQL disaster which converted everything to UTF-8 — I figured I could reverse it. So I rounded up a bunch of smart people (dda_, sbp, bonsaikitten and Blackb|rd — and countless others, sorry if I forgot you!) and got to work.

It proved a hairier problem than expected. What also proved hairy was explaining the problem to people who wanted to help and insisted in misunderstanding the situation. In the end, we produced a script (well, “they” rather than “we”) which looked like it should work, only… it did nothing. If you’re really interested in looking at it, here it is — but be warned, don’t try it.

We tried recode. We tried iconv. We tried changing my.cnf settings, dumping the databases, changing them back, and importing the dumps. Finally, the problem was solved manually.

  1. Made a text file listing the databases which needed to be cured (dblist.txt).
  2. Dumped them all: for db in $(cat dblist.txt); do mysqldump --opt -u user -ppassword ${db} > ${db}-20060712.sql; done
  3. Sent them over to Blackb|rd who did some search and replace magic in vim, starting with this list of characters (just change the browser encoding to latin-1 to see what they look like when mangled)
  4. Imported the corrected dumps back in: for db in $(cat dblist.txt); do mysql -u user -ppassword ${db} < ${db}-20060712.sql; done

Blackb|rd produced a shell script for vim (?) which I’ll link to as soon as I lay my hands on the URL again. The list of characters to convert was produced by trial and error, knowing that corrupted characters appeared in the text file as A tilde or A circonflexe followed by something else. I’d then change the my.cnf setting back to latin-1 to view the character strings in context and allow Blackbr|d to see what they needed to be replaced with.

Thanks. Not looking forward to the next MySQL encoding problem. They just seem to get worse and worse. (And yes, I do use UTF-8 all over the place.)

Problèmes d'encodage MySQL [fr]

Un joli mélange de latin-1 et d’utf-8 dans ma base de données. Un script python pour nettoyer tout ça.

[en] I've been to MySQL encoding hell and back. The little question marks you may have seen in place of accented characters a few weeks back were caused by a lovely mix-up of utf-8 and latin-1 inside my databases. Dda_ from #joiito kindly helped me by writing a python script to identify fields with non-utf-8 characters in them, and convert them back.

Vous avez peut-être remarqué, il y a une semaine ou deux, que les accents de ce site avaient été remplacés subrepticement par de vilains points d’interrogation. Une fois de plus, je me trouvais dans la situation où je croyais avoir de l’utf-8 dans mes bases de données, pour réaliser ensuite qu’il s’agissait en fait de latin-1. Et cette fois, c’était encore bien pire qu’avant: j’avais un mélange d’utf-8 et de latin-1.

Dda_ a eu la grande gentillesse de passer plusieurs heures à  me pondre un script en python qui fait le tour de tous les champs de toutes les tables de toutes les bases de données, et les convertit en utf-8 s’il y détecte des caractères non-utf-8 (ce qui signifierait, dans mon cas, qu’on se trouve en présence de latin-1). Une fois que c’est fait, le script va changer l’encodage des tables pour que tout nouveau contenu y soit stocké en utf-8.

Bref, voici l’explication et le script.

Converting MySQL Database Contents to UTF-8 [en]

I finally managed to convert my WordPress database content to UTF-8. It’s easy to do, but it wasn’t easy to figure out.

[fr] Voici comment j'ai converti le contenu de ma base de données WordPress en UTF-8. C'est assez simple en soi, mais ça m'a pris longtemps pour comprendre comment le faire!

A few weeks ago, I discovered (to my horror) that my site was not UTF-8, as I thought it was. Checking my pages in the validator produced this error:

The character encoding specified in the HTTP header (iso-8859-1) is different from the value in the XML declaration (utf-8). I will use the value from the HTTP header (iso-8859-1).

In all likeliness, my server adds a default header (iso-8859-1) to the pages it serves. When I switched to WordPress, I was careful to save all my import files as UTF-8, and I honestly thought that everything I had imported into the database was UTF-8. Somewhere in the process, it got switched back to iso-8859-1 (latin-1).

The solution to make sure the pages served are indeed UTF-8, as specified in the meta tags of my HTML pages, is to add the following line to .htaccess:

AddDefaultCharset OFF

(If one wanted to force UTF-8, AddDefaultCharset UTF-8 would do it, but actually, it’s better to leave the possibility to serve pages with different encodings, isn’t it?)

Now, when I did that, of course, all the accented characters in my site went beserk — proof if it was needed that my database content was not UTF-8. Here is the story of what I went through (and it took many days to find the solution, believe me, although it takes only 2 minutes to do once everything is ready) to convert my database content from ISO-8859-1 to UTF-8. Thanks a lot to all those who helped me through this — and they are many!

First thing, dump the database. I always forget the command for dumps, so here it is:

mysqldump --opt -u root -p wordpress > wordpress.sql

As we’re going to be doing stuff, it might be wise to make a copy of the working wordpress database. I did that by creating a new database in PhpMyAdmin, and importing my freshly dumped database into it:

mysql -u root -p wordpress-backup < wordpress.sql

Then, conversion. I tried a PHP script, I tried BBEdit, and they seemed to mess up. (Though as I had other issues elsewhere, they may well have worked but I mistakenly thought the problem was coming from there.) Anyway, command-line conversion with iconv is much easier to do:

iconv -f iso-8859-15 -t utf8 wordpress.sql > wordpress-iconv.sql

Then, import into the database. I first imported it into another database, edited wp-config.php to point to the new database, and checked that everything was ok:

mysql -u root -p wordpress-utf8 < wordpress-iconv.sql

Once I was happy that it was working, I imported my converted dump into the WordPress production database:

mysql -u root -p wordpress < wordpress-iconv.sql

On the way there, I had some trouble with MySQL. The MySQL dump more or less put the content of all my weblog posts on one line. For some reason, it didn’t cause any problems when importing the dump before conversion, to create the backup database, but it didn’t play nice after conversion.

I got this error when trying to import:

ERROR 1153 at line 378: Got a packet bigger than 'max_allowed_packet'

Line 378 contained half my weblog posts… and was obviously bigger than the 1Mb limit for max_allowed_packet (the whole dump is around 2Mb).

I had to edit my.cnf (/etc/mysql/my.cnf on my system) and change the value for max_allowed_packet in the section titled [mysqld]. I set it to 8Mb. Then, I had to stop mysql and restart it: mysqladmin -u root -p shutdown to stop it, and mysqld_safe & to start it again (as root).

This is not necessarily the best way to do it, and it might not work like that on your system, but it’s what I did and the site is now back up again. Comments welcome, and hope this can be useful to others!