Hi folks: I started to code a bit the large support for SqueakDBX to large objects and I want to ask you for help. In a first step, I will start with the basic large objects: CLOB, NCLOB, XML, BLOB.
Now, I need to map each of them, to an Smalltalk object. I though about this:
CLOB -> String
NCLOB -> String
XML -> String and the the user do whatever they want with the string representing the XML.
BLOB -> ByteArray
what do you think? are there suitable objects in a core image ? (I don't want to depend in another package)
Now, do you know which is the faster why to create a ByteArray from a String ? and vice-versa ? just asString ?
Another question....to get a field value of a row, I call an OpenDBX using FFI like this:
apiQueryFieldValue: handle index: index
"const char* odbx_field_value( odbx_result_t* result, unsigned long pos )"
<cdecl: char* 'odbx_field_value' (ulong ulong) module: 'opendbx'>
^self externalCallFailed
As you can see that function of OpenDBX answer me in a char* the value of the field. The problem is...suppose that field is a large object (a CLOB for example). A Clob can be 1gb or even more....of course, that will be bad for my image (freeze ? crash ? I don't know...). I would like to validate the size of that before bring it into the image, so that not to crash or whatever. I want to put like a limit. The problem is that I don't know how to do that from Pharo side....is that possible ? or the only way is to change OpenDBX side ?
Thanks a lot in advance,
Mariano
_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Locked
