I need to modify a column in a SQLite database but I have to do it programatically due to the database already being in production. From my research I have found that in or
When I ran "CREATE TABLE tmp_table AS SELECT id, name FROM src_table", I lost all the column type formatting (e.g., time field turned into a integer field
As initially stated seems like it should be easier, but here is what I did to fix. I had this problem b/c I wanted to change the Not Null field in a column and Sqlite doesnt really help there.
Using the 'SQLite Manager' Firefox addon browser (use what you like). I created the new table by copying the old create statement, made my modification, and executed it. Then to get the data copied over, I just highlighted the rows, R-click 'Copy Row(s) as SQL', replaced "someTable" with my table name, and executed the SQL.
As said here, these kind of features are not implemented by SQLite.
As a side note, you could make your two first steps with a create table with select:
CREATE TABLE tmp_table AS SELECT id, name FROM src_table
That's one of the better-known drawbacks of SQLite (no MODIFY COLUMN
support on ALTER TABLE
), but it's on the list of SQL features that SQLite does not implement.
edit: Removed bit that mentioned it may being supported in a future release as the page was updated to indicate that is no longer the case
If the modification is not too big (e.g. change the length of a varchar), you can dump the db, manually edit the database definition and import it back again:
echo '.dump' | sqlite3 test.db > test.dump
then open the file with a text editor, search for the definition you want to modify and then:
cat test.dump | sqlite3 new-test.db