regression from https://github.com/spesmilo/electrum/pull/8462
- pr.verify() was called in qml, but not in qt gui
- we now call pr.verify() in get_payment_request(), to make the API less error-prone
- it is now ok to call pr.verify() multiple times, the result is cached
A new config API is introduced, and ~all of the codebase is adapted to it.
The old API is kept but mainly only for dynamic usage where its extra flexibility is needed.
Using examples, the old config API looked this:
```
>>> config.get("request_expiry", 86400)
604800
>>> config.set_key("request_expiry", 86400)
>>>
```
The new config API instead:
```
>>> config.WALLET_PAYREQ_EXPIRY_SECONDS
604800
>>> config.WALLET_PAYREQ_EXPIRY_SECONDS = 86400
>>>
```
The old API operated on arbitrary string keys, the new one uses
a static ~enum-like list of variables.
With the new API:
- there is a single centralised list of config variables, as opposed to
these being scattered all over
- no more duplication of default values (in the getters)
- there is now some (minimal for now) type-validation/conversion for
the config values
closes https://github.com/spesmilo/electrum/pull/5640
closes https://github.com/spesmilo/electrum/pull/5649
Note: there is yet a third API added here, for certain niche/abstract use-cases,
where we need a reference to the config variable itself.
It should only be used when needed:
```
>>> var = config.cv.WALLET_PAYREQ_EXPIRY_SECONDS
>>> var
<ConfigVarWithConfig key='request_expiry'>
>>> var.get()
604800
>>> var.set(3600)
>>> var.get_default_value()
86400
>>> var.is_set()
True
>>> var.is_modifiable()
True
```
given a bip21 uri that has both onchain addr and bolt11,
if we have LN enabled but no LN channels, auto-fallback to paying onchain
we will have to clean up and unify this logic between GUIs. becoming spaghetti :/
rumour has it, Thomas has a branch? :P
The app hangs indefinitely if we try to quit it while one of
these threads is active, because once asyncio has shut down,
futures never return. This was already fixed for lightning
payments in c5dc133, but there are many other cases.
QEInvoice/QEInvoiceParser now properly split for mapping to Invoice type (QEInvoice)
and parsing/resolving of payment identifiers (QEInvoiceParser).
additionally, old, unused QEUserEnteredPayment was removed.
invoices are now never saved with user-entered amount if the original invoice
did not specify an amount (e.g. address-only, no-amount bip21 uri, or no-amount
lightning invoice). Furthermore, QEInvoice now adds an isSaved property so the
UI doesn't need to infer that from the existence of the invoice key.
Payments of lightning invoices are now triggered through QEInvoice.pay_lightning_invoice(),
using the internally kept Invoice instance. This replaces the old call path of
QEWallet.pay_lightning_invoice(invoice_key) which required the invoice to be saved
in the backend wallet before payment.
The LNURLpay flow arriving on InvoiceDialog implicitly triggered payment, this is
now indicated by InvoiceDialog.payImmediately property instead of inferrred from the
QEInvoiceParser isLnurlPay property.
Note that this allows users to save invoices that have an empty
amount, which is not allowed by the Qt GUI. Qt will complain at
pay time about empty amount if a lightning invoice without amount
is saved. With onchain invoices, Qt will create an onchain tx with
a zero output.
and invoice has fallback address and amount can be paid on-chain.
In WalletMainView, follow on-chain payment path if available lighting balance is
insufficient for the invoice amount
(this somehow escaped attention before, as most objects usually don't have multiple instances,
unless multiple wallets are open at the same time.)
Also, move all signal declarations, class constants and variables to the top of class definitions.
