fix!(types): v24/v28 fixes

[storopoli]

A bunch of issues when porting a custom types.rs module from bitcoind-async-client to corepc-types.

• v24::GetRawMempoolVerbose should use the same version MempoolEntry and not the v17.
• v28::SubmitPackageTxResult might not have effective-includes field.
• v24::ListUnspentItem might not have a label field.

[tcharding]

Nice! Thanks man. Can you add an into_model method too please. Copy from v17 also, or here it is.

impl GetRawMempoolVerbose {
    /// Converts version specific type to a version nonspecific, more strongly typed type.
    pub fn into_model(self) -> Result<model::GetRawMempoolVerbose, MapMempoolEntryError> {
        use MapMempoolEntryError as E;

        let mut map = BTreeMap::new();
        for (k, v) in self.0.into_iter() {
            let txid = k.parse::<Txid>().map_err(E::Txid)?;
            let relative = v.into_model().map_err(E::MempoolEntry)?;
            map.insert(txid, relative);
        }
        Ok(model::GetRawMempoolVerbose(map))
    }
}

[storopoli]

Done in a5f9033

[tcharding]

The need for this change proves what we suspect, our test coverage is not good enough (see #58).

cc @jamillambert

[tcharding]

And this one too. Thanks for finding and fixing these for us @storopoli.

[tcharding]

ACK d8e2ac1

[tcharding]

Woops I acked forgetting that we need the into_model still.

[storopoli]

Woops I acked forgetting that we need the into_model still.

Try now :)

[jamillambert]

ACK 5bb3b31

All looks good to me, thanks for the fix.

[tcharding]

ACK 5bb3b31

[tcharding]

FTR it came to me this morning that during review I did not check if theses modified types had any newer counterparts in more recent Coer versions. Checked just now before merging.

[tcharding]

Thanks for this @storopoli, super useful.

[storopoli]

No worries. Can we cut a new release of corepc-types?

[tcharding]

Yeah mate, sure thing. I'll do it on Monday for you.

[storopoli]

Don't forget the tag :)