On 5/26/19, John Brigham <[email protected]> wrote: > I have an Arduino/Python experiment that generates lots of simple > numerical data; about 7000 records a day. I run it through Python into an > sqlite3 database.I have been running this for 8 months. I start a new > database about every six weeks so every database has about 260k rows. I > have one database that is missing the first three weeks.
The "dec3_database" contains 168601 records from 2018-12-03 10:50:07.640000 to 2018-12-26 21:46:53.346000 in the "first_table" table, but then just 9 more record in the "dec26_table" table and 341 more records in "dec26_table_cont". > There is evidence > that the records are in the file but not appearing: The size of the file > reflects the number of rows that should be present. And furthermore, the > size of the file nicely reflects the number of days. How many rows were you expecting, and covering what dates? > My objectives are > twofold: first, can you fix this faulty file The file looks to be intact to me. I'm not sure what about it needs fixing? > and second, this appears to be > a bug in your software and I want you informed. What makes you think this bug is in SQLite and not in your python script? Do you have an example of SQLite doing something other than what your script specifically asked it to do? And do you know for certain that some other agent (another script, or perhaps a human using a database inspection tool) didn't connect to the database in the middle of the session and change the table names around on you? > I use DBrowser and am > satisfied with it. I have knowledge about CLI SQL, but am not presently > using it. The software can be seen at > https://github.com/mrphysh?tab=repositories I quickly check the link and see > that the example data-base is the exact one.Notice that the file name is > dec3. And notice that the database starts with dec 26. I didn't notice that. The first record in the "first_table" table seems to be dated "2018-12-03 10:50:07.640000" On the other hand, it is unclear to me why you have three separate tables "first_table", "dec26_table", and "dec26_table_cont", all with the same schema and similar data. So perhaps I am misinterpreting the data. I also notice that your columns are named "column_one", "column_two", "column_three", and so forth. Is that deliberate, or have you post-processed the database to obfuscate these column names? If the latter, might the missing rows have been deleted by the post-processing and obfuscation step? -- D. Richard Hipp [email protected] _______________________________________________ sqlite-users mailing list [email protected] http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
