Hi Jon, [jd wrote]
[snipped some lengthy description :-)] > > This is where the effort jumped off the track. > > Using repmcli I was able to import data into a table from > a csv file providing the file only had one field. Once I added > a second field repmcli complained that it couldn't find data for > the second field. After scouring the archives I stumbled upon the > information that I needed to have all fields encased in double > quotes whether they were text fields or not. Hmmm. that's a pain > but I can deal with it. Once I was able to add more than one column > via repmcli I decided to make a table with all the data types I'd > have to deal with. > Using CSV files you have the choice to either enclose all data in delimiters or none. In the latter case you need to specify this in the command as follows (example using DATALOAD): DATALOAD TABLE <name> <some columns descriptions> INFILE '<name>' DELIMITER '' <some other file descriptions if necessary> > Here's how I figured I'd have to convert from MS Access types to > SAPDB types: number(long) -> integer, date/time -> date, text -> varchar, > memo -> long(ascii), currency -> float. Do you see any potential problems > with this? I'm particularly nervous about the data/time fields and memo > field conversions. > I built a table using all these types. I created a csv file with all > fields enclosed in double quotes by changing all of my Access fields > to type text so Access would export the data with the double quotes. > The data file looked to me exactly as I'd want it. Despite this I could > never get repmcli to get past the first line of data as I kept getting > the No data supplied error message on one of the memo fields. Figuring > I was now creating more problems than I was solving with all these > gyrations I decided to abandon repmcli and search for another method. > What a pity! It's not that bad :-). May be an earlier mail would have prevented some annoyance. Ok. Lets come to the problem. With repmcli the LONG data must be saved in an extra file not in the data file itself. The data file will then contain only offsets of the LONG data in the extra file. Suppose you have a table with 2 cols where the second is a LONG. Your command could look like this (example with DATALOAD again and no delimiters): DATALOAD TABLE <name> col1 1 col2 2 INFILE '<name>' DELIMITER '' <some other file descriptions if necessary> LONGFILE col2 '<long file name>' The corresponding data file (INFILE) could have entries like this one: firstdata,1-567889 ... The corresponding LONGFILE would contain the LONG data. The first 567889 Bytes would be occupied by the first LONG, the following by the subsequent ones. Thats it. Let me know if I can help using RepMan to migrate. I'll do anything I can to help you to come to SAP DB ;-)). Regards, Steffen -- Steffen Schildberg SAP DB Team SAP Labs Berlin _______________________________________________ sapdb.general mailing list [EMAIL PROTECTED] http://listserv.sap.com/mailman/listinfo/sapdb.general
