Getting on and off on the project…each time, there is progress: Using the program I developed (and dropped) in 2008, I was able to correct a number of mistakes in the program, up to a point that accessing data did actually work. But the record definition (keys etcetera) were not correct, and there was no text – just a number of index lines. So I had to rework the file definitions, once that was done, reading index was file, but since textdata was missing, it stopped there.
Next step was obviously a matter of getting texts loaded. Since I already use a much limited version on the home page, and all these texts (since 2008) are still on the server, I created a procedure to list them, converted their date and create a list that I can sort on that date. Next a program to read that file, read the textfile itself, add the data into the datafile and add a line into the index.
That also meant some adaptations in the file acces code – since adding lines wasn’t tested at all, not even in 2008, but finally, all 100 files are now in the database of this type.
Though there is one more larger change to be done in the file containing the texts: Currently record size is fixed to 8K (minus 4 bytes for the key) but most posts are far less: So either I have to create (and handle) multiple records, of use records of variable size. Just to find out how to handle these, in the access routines of this file.