Ability to Import Spreadsheet Data into Table in Apps added Anytime Soon??

Question

I love HTMLDB so much, and I thought I would never find anything it couldn't do, but I have run headlong into a problem that I've discovered many users have.
The Import Spreadsheet Data into Table functionality available in Data Workshop is such a treasure, and I was hoping with all of my heart that such functionality would be available for use inside the User Applications in HTMLDB 2.0. Alas, it is not so.
I know some brave folks have come up with work-arounds for this lack, and I applaud them. However, none of those will work for the situation I am faced with. I am about to strike out into the no-man's-land of trying to create my own PL/SQL csv file parsing function. Please wish me luck. Also, could you please tell me if this functionality is going to be included anytime soon?

Answer 1

Well I looked at the FLOWS_020000.WWV_FLOW_LOAD_DATA.LOAD_CSV_DATA procedure
From what I can gather(considering nobody cant look at the actual procedural code):
P_FILE_ID seems to be file id you find in FLOWS_FILES.WWV_FLOW_FILE_OBJECTS$
P_CNAMES seems to be the column names of table you are loading into
P_UPLOAD seems to be the names of field from file you are loading
P_SCHEMA seems to be the schema that contains table where data will be loaded
P_TABLE seems to be table that you load data into
P_DATA_TYPE seems to be the datatypes from file to be loaded
P_DATA_FORMAT not sure
P_PARSED_DATA_FORMAT not sure
P_SEPARATOR seems to be the field seperator from file
P_ENCLOSED_BY seems to be what fields will be enclosed with
P_FIRST_ROW_IS_COL_NAME seems to be to tell whether or not first row contains column headings
P_LOAD_TO not sure
P_CURRENCY seems to be the currency symbol that comes from file
P_NUMERIC_CHARS not sure
P_CHARSET not sure
P_LOAD_ID not sure
Essentially you could run this process provided you get all the kinks worked out(that is what all the parameters mean/are expected)...
...or..
you write your own process:
- that reads the blob/clob content line by line
declare
n number;
buffer varchar2(4000);
begin
(do some sort of loop until end of length of lob)
n := dbms_lob.instr( "your lob", ',',"next comma position" );
if ( nvl(n,0) > 0 )
then
buffer := dbms_lob.read( p_lob, "amount of lob length to read", n);
end if;
end;
The above code with some tweaking will read the lob. However you still will have to add a piece that creates sql insert statement. One of the pitfalls I see with custom code is the fact that if you end up with huge lob string you will run into problem of creating sql insert statement. Not the actual statement but assigning the statement to a variable due to datatype lengths.
Hopefully this will help you more in the right direction!