First, identify the packages you want to transfer from PyPI. You can use Python to download these packages directly using the `pip download` command. This will download the package files to your local directory without installing them.

```bash
pip download
```

After downloading the package files (typically in `.tar.gz` or `.whl` formats), you need to extract the metadata. Use Python’s built-in libraries like `tarfile` or `zipfile` to extract files and read metadata from `PKG-INFO` or `METADATA` files within the package.

```python
import tarfile

with tarfile.open('package-name-version.tar.gz', 'r:gz') as tar:
tar.extractall(path='extracted_path')
```

Once extracted, parse the necessary metadata such as package name, version, author, license, and dependencies. This can be done by reading and processing the `METADATA` file, which is typically formatted similarly to an email message.

```python
from email import message_from_file

with open('extracted_path/package_name/METADATA') as f:
metadata = message_from_file(f)
```

Ensure you have access to an Oracle database where you have the necessary permissions to create tables and insert data. Use SQL to create a table structure that will store the package metadata.

```sql
CREATE TABLE pypi_packages (
package_name VARCHAR2(255),
version VARCHAR2(50),
author VARCHAR2(255),
license VARCHAR2(255),
dependencies CLOB
);
```

Use Python's `cx_Oracle` library to establish a connection to your Oracle database. Make sure you have the Oracle Instant Client installed and configured correctly.

```python
import cx_Oracle

connection = cx_Oracle.connect('username/password@hostname/SID')
cursor = connection.cursor()
```

With the connection established, write a script to insert the parsed metadata into the Oracle database. Ensure you handle exceptions and commit transactions properly.

```python
insert_query = """
INSERT INTO pypi_packages (package_name, version, author, license, dependencies)
VALUES (:package_name, :version, :author, :license, :dependencies)
"""

cursor.execute(insert_query, {
'package_name': metadata['Name'],
'version': metadata['Version'],
'author': metadata.get('Author', 'Unknown'),
'license': metadata.get('License', 'Unknown'),
'dependencies': ', '.join(metadata.get_all('Requires-Dist', []))
})
connection.commit()
```

Finally, verify that the data has been successfully transferred. Query the Oracle table to ensure the records match the expected data from the PyPI packages.

```python
cursor.execute("SELECT FROM pypi_packages")
for row in cursor.fetchall():
print(row)

cursor.close()
connection.close()
```

By following these steps, you can manually move data from PyPI to Oracle without relying on third-party connectors or integrations.