The files are stored in well... the browser. Specifically in the IndexDB
or localStorage
. This means that the physical location on the disk will depend entirely on the browser that you use, rather than on the operating system, and will likely be inaccessible (for an average user) without decoding binary blobs.
For example, in Chrome you can check the path to the application data using chrome://version/
(under Profile Path
) and in that directory there should be IndexedDB
folder. Then you need to find a sub-folder depending on the domain in which you accessed JupyterLite, for example https_jupyterlite.readthedocs.io_0.indexeddb.leveldb
, and there you will find a LevelDB database file with .ldb
extension and a MANIFEST
file (with the pointer to the current version in the CURRENT
file. The details of how to extract the blobs are outside of scope for this answer, but have a look at How to access Google Chrome's IndexedDB/LevelDB files?.
However, you can use files from your file system directly in JupyterLite without worrying about in-browser technologies with the jupyterlab-filesystem-access extension which uses File System Access API however this API is not available on Firefox yet.
As noted by @Wayne all of this is still quite experimental (both as in "using the newest browser APIs" and "the team of developers is still figuring way forward, please help by providing kind feedback and contributing").