investigating build failure

I updated the diff a bit too fast, so some of these recent errors are from previous versions of the diff.

improve for loop as per clippy's advice

remove unneeded rehash and wrong comment

In D17366#394869, @teamcity wrote:

Build Bitcoin ABC Diffs / Diff Testing (ecash-lib-integration-tests) failed.

Unrelated. Looks like a permission error when creating a directory

custom error code and test for wrong type for boolean arg

i missed one

feedback

TIL chage

Note that I also tried a form for the macro that check both min and max check_number_of_params!(params, 1, 2); but then I ran into a compiler warning when I added check_number_of_params!(params, 0, 0); in the server.ping handler

line length nit

address clippy's complaint

feedback

so it looks like we don't need to use serde_json::Value; in the module that implement try_from(value: &'_ serde_json::Value), but we still need to have the crate as a depencency (Cargo.[toml/lock])

Not sure why the build claims serde_json:: Value is unused in txid.rs, will investigate

use is_none() isntead of match

take a reference for TryFrom, use it also in transaction_get (minor behavior change in the error message, because now we have one less line that can raise a variant on the same error message)

rebase after reducing the port range in D17314

rebase after reducing the port range in D17314

rebase after reducing the port range in D17314

@bot build-linux64 build-linux-aarch64 build-chronik build-chronik-plugins

reduce port range to 3000 to keep the same effective range of ports

Collision on port 33380, Lets try to reduce the port range first to not conflict with docker ports (see D16048)

do this without the macro from D17355 (need to study macros more to make it work), fix clippy::needless-question-mark

rebase

use consistent error message "Unknown transaction id"

test all errors responses

Let's wait to see if tests pass on CI or we get port collisions. The direct parent diff is a draft, this one was supposed to be one too.

rebase

@bot build-chronik build-chronik-plugins

These issues are likely port collisions

2024-12-13T20:49:51.570827Z [chronik/chronik-lib/src/bridge.rs:181] [<chronik unknown>] Starting Chronik Electrum interface bound to [(127.0.0.1:33252, Tcp)]
2024-12-13T20:49:51.574474Z [chronik/chronik-indexer/src/indexer.rs:1491] [<chronik unknown>] Chronik Electrum failed serving: Address already in use (os error 98)
2024-12-13T20:49:51.574503Z [/work/src/shutdown.cpp:22] [AbortNode] *** ERROR Chronik in ChronikElectrumServer::serve
2024-12-13T20:49:51.574521Z [/work/src/noui.cpp:42] [noui_ThreadSafeMessageBox] Error: Chronik Electrum failed serving: Address already in use (os error 98)

There seems to be some kind if intermittent issue introduced by this diff or it's parent. To be investigated.

remove the small intermediate macros that only have one callsite. They just make the logic harder to follow.

c'est cassé

rebase

raise an error if response data is too large, improve doc for ChronikElectrumClient with code samples, remove unneed id param for MethodNameProxy

rework the API, use a client object similar to ChronikClient, set a timeout for the socket so that we get an error if the server never replies (or never sends the final \n)

Benchmark on my machine: