Convert SQLite3 to Oracle
Upload a .sqlite or .db or .sqlite3 file and get back a Oracle SQL script you can load straight into your database. Types are preserved rather than flattened to text. Free, no registration, up to 10MB.
Click to upload or drag and drop
CSV, Excel, SQLite, SQL, DBF (max 10 MB)
How your data is typed in Oracle
Generated from the converter's own Oracle type mapping — this is what it actually emits, not an example.
| Your data | Oracle type |
|---|---|
| Whole number | NUMBER(19) |
| Decimal (money) | NUMBER(10,2) |
| Floating point | BINARY_DOUBLE |
| True / false | NUMBER(1) |
| Date | DATE |
| Date and time | TIMESTAMP |
| Short text | VARCHAR2(255) |
| Long text | CLOB |
| Binary data | BLOB |
Working from a .sqlite3 file
The 3 in .sqlite3 is the version of the file format, not of your file, and it has not changed since 2004. A .sqlite3 and a .sqlite are the same kind of thing under two naming habits, and both are read the same way.
- →db.sqlite3 is Django's default database name, and a Django database converts whole: django_migrations, auth_user and django_session are ordinary tables here and come across with everything else.
- →Nothing is filtered on the way through, so if you only want the application tables it is easier to drop the framework ones in the target database after loading.
- →The extension does not change how the file is read - it only has to be one the uploader recognises, and .sqlite, .sqlite3 and .db all are.
Oracle-specific edges in the script
Oracle diverges further from ordinary SQL than any other target here, and the script is shaped around that. Three of its choices are worth reading before you load, because each is a place where the source can hold something Oracle has no room for.
- →Oracle's reserved list is full of words that legacy tables use as column names, so those get a trailing underscore: a column called DATE becomes DATE_, NUMBER becomes NUMBER_, LEVEL becomes LEVEL_.
- →Binary values are written as HEXTORAW('...'), which a plain SQL script can carry only up to 2000 hex characters - 1000 bytes. Anything larger needs PL/SQL chunking that no portable script can express, so a table of real attachments moves better through the SQLite or DuckDB target.
- →A string column of unknown length is declared VARCHAR2(4000) and anything longer becomes a CLOB. Under the default settings VARCHAR2 counts bytes rather than characters, so 4000 characters of Cyrillic or CJK will not fit in one.
- →Rows are batched a hundred to a statement, each block closed off with SELECT 1 FROM DUAL, so a 100,000-row table arrives as a thousand statements rather than a hundred thousand.
Reading SQLite3
- →Declared types, primary keys, NOT NULL constraints and blobs all carry through.
- →SQLite reports INTEGER PRIMARY KEY as nullable because it aliases the rowid; the conversion corrects that.
- →SQLite type affinity is loose, so a NUMERIC column with no declared scale is given room for decimals rather than being rounded to whole numbers.
Writing Oracle
- →Uses INSERT ALL rather than multi-row VALUES, which Oracle rejects outright — the most common reason a generated Oracle script fails to load.
- →Dates are written with TO_DATE and an explicit format mask.
- →Booleans become NUMBER(1), because Oracle had no BOOLEAN column type before 23c.
- →Identifiers are capped at 30 characters so the script loads on Oracle 12.1 and earlier too.
- →Note that Oracle itself treats an empty string as NULL — that is Oracle behaviour, not a conversion error.
What you get
A ZIP containing dump.sql, plus a README with the exact command to load it. If anything about your data needed a judgement call, a _warnings.txt explains it.
Load this dump into Oracle:
sqlplus USER/PASSWORD@//HOST:1521/SERVICE @dump.sql
Notes specific to Oracle:
- Oracle treats an empty string as NULL. A column that held ''
in the source will read back as NULL here. This is Oracle
behaviour, not a conversion error, and it cannot be avoided.
- Booleans are NUMBER(1), 1 or 0: Oracle had no BOOLEAN column
type before 23c.
- Identifiers are capped at 30 characters for compatibility
with 12.1 and earlier.
- Table and column names are quoted, so Oracle keeps their exact
case. Unquoted Oracle identifiers fold to upper case, which
means you must quote them in your queries too:
SELECT * FROM "MyTable"; not SELECT * FROM MyTable;Questions
- Is converting SQLite3 to Oracle free?
- Yes. Files up to 10MB convert free with no account. There is a limit of five conversions per day per IP address.
- Will my SQLite3 data keep its types in Oracle?
- Yes — that is the point. Whole numbers, decimals, dates and booleans are mapped to real Oracle types rather than everything becoming text. The mapping table above shows exactly what you get.
- Do you store my file?
- No. The file is written to a temporary directory, converted, and the directory is deleted as soon as your download has been sent.