(chinese) unicode labeling with postgis db

[Christian Sengstock]

Hi,
actually its working exact the way its described in the docs.

@mario: nice way to test these isuses, think i'll use that from now on.

what i missed to add was:

LABEL
  ..
  ENCODING GB18030
END
(i wrote that i tested this already but unfortunatley it seems that i didnt,
sorry)

However i dont really understand why its that way because:

1.) I imported the data from GB18030 (ascii) into the UNICODE postgres
database (actually its saved as UNICODE) 
2.) use truetype and a UNICODE font in mapserver (postgis connection)

I'm a  bit confused that i have to set the GB18030 encoding now in the
mapserver LABEL section, because the data in the db is already UNICODE !?

Maybe someone knows about that,
Chris



On Thu, 15 Feb 2007 00:13:18 +0900, Mario Basa <mario.basa at GMAIL.COM> wrote:

>Hello Christian,
>
>What I usually do if I have label encoding problems (w/c happens a lot
>with japanese characters) is to create an annotation layer with an
>inline feature and then test using different fonts or encoding
>parameters. Something like this:
>
>LAYER
>    NAME "Label Test"
>    STATUS DEFAULT
>
>    TRANSFORM FALSE
>
>    TYPE ANNOTATION
>    FEATURE
>      POINTS 420 420 END
>      TEXT " Japanese_Characters_go_here "
>    END
>
>    CLASS
>      LABEL
>        TYPE TRUETYPE
>        ENCODING SJIS
>        FONT pgothic
>        SIZE 10
>        COLOR 255 255 255
>        BACKGROUNDCOLOR 0 0 255
>      END
>    END
>  END
>
>Also, the ENCODING parameter translates the label into UTF-8,  so if
>the label is in UTF-8 already, ENCODING shouldn't be necessary
>anymore.
>
>Regards,
>
>Mario.
>
>2007/2/14, Christian Sengstock <csengstock at gmail.com>:
>> Hi list,
>> i try to do labeling for chinese features but have some problems.
>> I tried to make it the way i found in some docs in the internet and in
>> the mail archive:
>>
>> - use a postgis unicode db
>> - use unicode truetype fonts
>>
>> But unfourtunately its not working for now.
>>
>>
>> I use a postgres/postgis dbms (PostgreSQL 8.0.9  POSTGIS 1.0.6) and
>> mapserver 4.10.0 (ms4w package).
>>
>> My database is UNICODE encoded (imported with GB18030 client encoding).
>> The unicode chars seem to be ok, i can use these chinese strings in my
>> web application and they work.
>>
>> I have a point layer which i render with UMN mapserver.
>> Now i tried to render the points like this...
>>
>> LAYER
>>   NAME 'poi'
>>   TYPE POINT
>>   CONNECTIONTYPE postgis
>>   CONNECTION '...'
>>   DATA 'geom FROM geo.poi'
>>   METADATA
>>     'wms_title' 'poi'
>>   END
>>   STATUS ON
>>   TRANSPARENCY 100
>>   PROJECTION
>>   'proj=longlat'
>>    'ellps=WGS84'
>>    'datum=WGS84'
>>    'no_defs'
>>   END
>>   LABELITEM 'name'
>>   CLASS
>>      NAME poi
>>      STYLE
>>         SYMBOL 'circle'
>>         SIZE 3
>>         COLOR 255 150 150
>>      END
>>      LABEL
>>        ANTIALIAS true
>>        COLOR 255 255 255
>>        OUTLINECOLOR 0 0 0
>>        POSITION CC
>>        OFFSET 0 2
>>        BUFFER 2
>>        TYPE TRUETYPE
>>        FONT 'arialuni'
>>        ANGLE auto
>>        SIZE 8
>>        #ENCODING UTF-8 | UNICODE ???
>>     END
>>   END
>> END
>>
>> where font 'arialuni' ist the microsoft arial unicode font:
>> arialuni ARIALUNI.TTF
>>
>> If i use ENCODING UNICODE, its no labels at all.
>> If i use ENCODING UTF-8 (or GB18030), its like i have no ENCODING attribute.
>>
>> Attached is a screenshot from the map.
>>
>>
>> Maybe someone can help or can point me to a good doc?
>> Thanx a lot,
>> Chris
>>
>>