When an appointment was triggered a flag was set in the Watcher, and removed later on in the Responder if the transaction ended up being rejected. That's pretty annoying. Since we have information about whether a transaction has made it to the mempool or not via the Carrier's receipt, this can be all done in the Watcher, which makes more sense and reduces the interaction with the db (1 write if succeeds, 0 otherwise instead of 1 write if succeeds, 2 otherwise).
Data used to be stored both in memory and disk (db). This commits modifies the Watcher, Responder and Cleaner so they only keep the needed maps and load information from disk when necessary.
ChainMonitor is the actor in charge of checking new blocks. It improves from the previous zmq_subscriber by also doing polling and, therefore, making sure that no blocks are missed.
Documentation and tests are still required. Partially covers #31
The Watcher used to receive a secret key file path ion the __init__ to load a secret key for signing. That made testing the Watcher hard, since the file needed to be present. Changing it so the main (pisad) loads the file from disk and passes the data the Watcher on init.
We were passing some unnecessary parameters to the Cleaner (locator) that could be derived from other data (uuid and appointments). Also standarises the order of the parameters to match the rest of the methods
- All encryption/decryption and signing/verifying calls are performed by the cryptographer now.
- The current signature format is temporal. We should define something not base on json.
- Some Cryptographer tests are still missing.
- The cli tests should be modified to fit this too.
Brings back check_potential_matches and check_matches to the Watcher.
The methods have now been renamed to check_matches (old check_potential_matches) and filter_valid_matches (old check_matches) to provide a better description of their purpose.
decode_raw_transaction has been added to BlockProcessor
The difference between master and the old data-persistence branch was huge, seems more effective to manually cherry pick the changes and integrate them
