>>>>> "Miguel" == Miguel Beltran R <yourpadre@gmail.com> writes:


    Miguel> After search a while, someone can confirm if I undestand it well? the
    Miguel> problem is because it is trying to decode to "ascii" instead of "iso-8859-1"

    Miguel> if it's true, why is doing it? I set to use iso-8859-1 in my conf file

    Miguel> 2016-06-28 16:56 GMT-05:00 Miguel Beltran R. <yourpadre@gmail.com>:

    >> Using the ZMI if I check the "Z SQL Method" the data is displayed
    >> correctly, but using a "page template" it gives me this error
    >>
    >>
    >> *Expression: <StringExpr u' ${repeat/items/number}. ${items/nombre}'>*
    >>
    >> - Module zope.tales.expressions, line 263, in __call__UnicodeDecodeError:
    >> 'ascii' codec can't decode byte 0xf3 in position 10: ordinal not in
    >> range(128)
    >>

this is due probably because the template engine tries to convert a
string into an unicode, but it doesn't know the charset of your data, so
it's trying to use the default encoding as set in
/usr/lib/python2.7/site.py during interpreter bootstrap. If you are
using a good database you can trust your data and and let the database
connector (the encoding should be settable somewhere in the connector or
Z SQL method conf, i don't remember since i don't use them anymore)
convert the strings to unicodes for you.

If you have something like MySQL you can't always trust that the data in
the db is in the expected encoding because some of its table engine
simply don't do any check on the input. If this is your case, it will be
more problematic.

In my experience, setting a default encoding in the config file didn't
always fix the problem.

As a last resort you can tweak the site.py, but really if there issues
with data's encoding this doesn't mean you solved the problem. Likely
the user will see wrong charachters where non-ascii ones are supposed to
be.