|Newbie questions about Pliant
Question about unicode possibilities
|Message posted by szamo on 2004/09/14 15:42:58
|While i was reading docs of pliant, i found description of using ASCII
everywhere. Probably it's enough to support English, maybe French, however
there are languages that can be used correctly only with unicode. ASCII
is a good, and old standard, but nowadays it's not enough.
Is there an unicode support in Pliant somewhere hidden, not mentioned?
What are the chances of adding unicode feature to the whole Pliant?
In my case, i wanted to use the Pliant database, but since it doesn't support
unicode, i had to choose another alternative, MySQL (it supports unicode
recently). Since the database is now not the Pliant's one, Pliant lost its
advantage of not using glue code. Thus it easier to implement my server program
in Php or Java now.
I can imagine Pliant can or could use external databases, just don't see how.
Can someone suggest me how i can use pliant with an unicode featured database?
|Message posted by maybe Boris Reitman on 2004/09/14 17:56:44
|If you adventurous you may try to use the MySQL database through
perl's DBI facility that can be accessed with my pliant-perl binding.
See example at http://pliant-perl.sourceforge.net
|Message posted by hubert.tonneau on 2004/09/14 18:32:25
|Unicode is partially supported in Pliant.
We have iStr defined in /pliant/language/type/text/istr.pli
and Str32 defined in /pliant/language/type/text/str32.pli
The Unicode support will be right from the beginning in the Pliant web browser.
It is right from the beginning the the Pliant vector library text rendering.
Also the integration in the database engine and HTTP server is not done yet,
and the problem I have in this area is double:
. I need not to break existing production applications
. My language does not need Unicode, so it's hard for me to test
So you're very right to select another tool at the moment if Unicode support
is mandatory for your application.
If you want to speed up Unicode integration in Pliant, you can help me with
very simple advises:
How do I type an Unicode behond 256 character on my french keyboard ?
What does happen in a Mozilla input field ? Do I have to change the page encoding
to UTF8 or use &#xxx; encoding ? How should I encode the <input value="xxx"> ?
How should I tel Mozilla to encode the form answer, and what encoding should
I expect ?
|Message posted by szamo on 2004/09/25 23:18:28
|To type Unicode behind 256 characters on your french keyboard is possible if
you install another keyboard scheme (for example Hungarian in my case). Then
you can switch to it, and try what the key right to P (ő) and the key left to
backspace (ű) results. I am curious what will appear in the brackets, because
those are unicoded characters.
-=( I typed this after checked preview of this page. It can be seen that my
unicode charaters are converted to &#xxx format. Now i am curious how it will
look like in the forum. Another note, when i edited this note again, i found
my unicode charaters again, and not the &#xxx format. )=-
I don't know exactly how unicode schemes work. When i create webpages for
in Hungarian language, i prefer setting the encoding to iso-8859-2, and not
using those &... codes. Maybe that's wrong anyway. Since we probably don't
want only Hungarian characters, probably the UTF-8 is suitable for
unicode support. (See for example: http://www.elfdata.com/plugin/storage.html)
Well, should think the topic through, but i have not much free time for this
right now, since i am working on the project i mentioned in first message.
I can imagine a field in database/browser/etc, where one can specify an encoding
type, else default (current one) would be used. I am not sure each language can
be handled universally with one encoding, lookint at that link above.
|Message posted by hubert.tonneau on 2004/12/08 00:31:17
|You may want to give a spin to Pliant release 93i:
it is assuming strings (Str data type) are UTF8 encoded.
So, starting from release 94, Pliant will use Unicode instead of ISO8859-1.
Moving back to the old behaviour where strings are using ISO8859-1 charset
is possible through setting 'default_charset_is_utf8' to 'false'