Creating an invalid tx when trying to mint too large amount of tokens

[nielstron]

Describe the bug
Trying to mint too many tokens produces and invalid transaction (basically minting 1231234145232534000000 here for example)
This results in ogmios failing to evaluate/submit the transaction with a very obscure error message.

To Reproduce
TBD

Logs

ogmios.errors.ResponseError: Ogmios responded with error: {'jsonrpc': '2.0', 'method': 'evaluateTransaction', 'error': {'code': -32602, 'message': "Invalid transaction; It looks like the given transaction wasn't well-formed. Note that I try to decode the transaction in every possible era and it was malformed in ALL eras. Yet, I can't pinpoint the exact issue for I do not know in which era / format you intended the transaction to be. The 'data' field, therefore, contains errors for each era.", 'data': {'babbage': "invalid or incomplete value of type 'Transaction': expected word", 'conway': "invalid or incomplete value of type 'Transaction': expected word", 'alonzo': "invalid or incomplete value of type 'Transaction': expected list len or indef", 'mary': "invalid or incomplete value of type 'Transaction': Size mismatch when decoding Object / Array. Expected 4, but found 3.", 'allegra': "invalid or incomplete value of type 'Transaction': Size mismatch when decoding Object / Array. Expected 4, but found 3.", 'shelley': "invalid or incomplete value of type 'Transaction': Size mismatch when decoding Object / Array. Expected 4, but found 3."}}, 'id': None}

Expected behavior
Ogmios could report that the integer is breaking/incorrectly serialized. PyCardano should abort the building or raise an error/warning this.

Environment and software version (please complete the following information):

• OS: Ubuntu 22
• PyCardano Version: 0.10.0

Additional context
Looping in @KtorZ for ogmios, but the error is really with PyCardano here.

[KtorZ]

Indeed, the mint value must fit into a word which is a u64 or u32 depending on OS arch, but most likely u64 (so 18_446_744_073_709_551_615 max).

[KtorZ]

Note that: this is as good as it gets when it comes to deserialization errors here. I cannot get anything better without re-writing the CBOR parsers which I am slightly reluctant to at the moment 😅 ...

[nielstron]

So it should be possible to just check that mint is smaller than sys.maxsize.

[cffls]

From the ledger spec, it looks like mint field should be int64, and uint should be used in all other cases, which doesn't have an upper bound (at least I couldn't find its definition in the spec). Maybe we should only limit mint field to int64. What do you think?

[nielstron]

If the spec sais i64 I agree we should restrict to that, not sys.maxsize