instrument | Python observability & instrumentation primitives | Analytics library

A few notes on this issue:

They query for browsers can detect non-browser applications, which will lead the an ActivityNotFoundException when launching the Custom Tab. See this issue for an example.

When querying for packages that can handle browser intents, I recommend using the below:

It seems that your HomeFragment doesn't have no parameters constructor. If you want to use fragment that requires constructor parameters you need to provide FragmentFactory to navigation component.

Add empty constructor HomeFragment() to HomeFragment, and it should work.

As it was described in the reference:

When creating the NavHostFragment using FragmentContainerView or if manually adding the NavHostFragment to your activity via a FragmentTransaction, attempting to retrieve the NavController in onCreate() of an Activity via Navigation.findNavController(Activity, @IdRes int) will fail. You should retrieve the NavController directly from the NavHostFragment instead.

Looks like you should use

As of 6.0 you can simply append the CSVs to same table one by one given the table has designated timestamp and partitioned it will work.

If your CSVs are huge I think batching them in transactions with few million rows will be better than offloading billions at once.

Depending of how much data you have and your box memory you need to partition in a way that single partition fits memory several times. So you choose if you want daily or monthly partitions.

Once you decide with partitioning you can speed up the upload if you able to upload day by day batches (or month by month) from all CSVs.

You will not need to rebuild the table every time you add an instrument, table will be rewritten automatically partition by partition when you insert records out of order.

If you use string literals in the start-pattern, you must use printf-style format string.

Thus, this will not work:

Issues

You have a UNIQUE index on the NUM_JUG column (perhaps implicit if NON_JUG is defined with PRIMARY KEY) and you are using the standard update method which uses the default conflict action of ABORT and therefore fails if an attempt is made to duplicate a NOM_JUG value.

As secondary issue is that the SQLiteDatabase update method returns the number of updates (see extract and link below) (same for updateWithOnConflict). The number returned will never be -1 it will be 0 or more (0 indicating that no updates have been applied).

As per SQLite Database - Update

Returns

int the number of rows affected

Fix

To fix the main issue you should use the updateWithOnConflict method. This takes a 4th parameter a conflict and you probably want IGNORE so you could use :-

This line:

As per the documentation on launch():

This method throws ActivityNotFoundException if there was no Activity found to run the given Intent.

While any of the ActivityResultContracts (such as the GetContent one you're using) should be available on every device, users may be running a custom build of Android that removes the apps / system utilities that handle these common intents or the user may have manually disabled the app (this is more common with things like a Browser or Camera app than this particular case).

Therefore you should consider surrounding your call to launch() with a try/catch block that catches an ActivityNotFoundException and informs the user that their device does not support this functionality.

A possible and promising approach is to generate the Pydantic model starting from the corresponding JSON Schema of the STIX object.

Luckily enough the JSON schemas for all the STIX 2 objects have been defined by the OASIS Open organization on the GitHub repository CTI-STIX2-JSON-Schemas.

In particular, the JSON Schema for the Attack-Pattern is available here.