Hi all, I've a few hundreds sqlite3 database files, all almost equals except for some constraint that has changed during time. As I know, there is no way to alter constraint (e.g., unique indexes), and therefore I have to migrate data to a new schema version. That is possible because data will fit into the same table schema, but I'm looking for a way to do it massively (and remotely). So far I was thinking to wrap a shell script to dump the content of a single database, create the new schema and reload the dump in it. The problem is that .dump provides data and schema, while I'd like to have data only. Other commands like .clone and .backup works pretty much the same, as far as I understand.
This leads me to either use awk/sed to manipulate the dump or to hard-code single select statements into the script to extract data. Is there any smarter way to dump only data in a loadable form? Thanks Luca
