Prefix removal is unintuitive

[drebelsky]

SEP-51 says

Discriminated Union
... In both cases the discriminant name is modified to be snake_case, and truncated removing any shared prefix if there are multiple identifiers in the enum defined as the discriminant. ...
...
Enum
... The string is the name-identifier modified to be snake_case, and truncated removing any shared prefix if there are multiple identifiers in the enum. ...

In practice, it seems like this repo truncates shared prefix based on _-delineations. For example,

enum TransactionResultCode
{
    txFEE_BUMP_INNER_SUCCESS = 1, // fee bump inner transaction succeeded
    txSUCCESS = 0,                // all operations succeeded

    txFAILED = -1, // one of the operations failed (none were applied)

    txTOO_EARLY = -2,         // ledger closeTime before minTime
    txTOO_LATE = -3,          // ledger closeTime after maxTime
    txMISSING_OPERATION = -4, // no operation was specified
    txBAD_SEQ = -5,           // sequence number does not match source account

    txBAD_AUTH = -6,             // too few valid signatures / wrong network
    txINSUFFICIENT_BALANCE = -7, // fee would bring account below reserve
    txNO_ACCOUNT = -8,           // source account not found
    txINSUFFICIENT_FEE = -9,     // fee is too small
    txBAD_AUTH_EXTRA = -10,      // unused signatures attached to transaction
    txINTERNAL_ERROR = -11,      // an unknown error occurred

    txNOT_SUPPORTED = -12,          // transaction type not supported
    txFEE_BUMP_INNER_FAILED = -13,  // fee bump inner transaction failed
    txBAD_SPONSORSHIP = -14,        // sponsorship not confirmed
    txBAD_MIN_SEQ_AGE_OR_GAP = -15, // minSeqAge or minSeqLedgerGap conditions not met
    txMALFORMED = -16,              // precondition is invalid
    txSOROBAN_INVALID = -17         // soroban-specific preconditions were not met
};

$ echo 'AAAAAA==' | stellar xdr decode --type TransactionResultCode
"tx_success"
$ 'AAAAAQ==' | stellar xdr decode --type TransactionResultCode
"tx_fee_bump_inner_success"

Note that each result code is prefixed with tx, but that they are kept after the transformation.

[drebelsky]

xdrgen does indeed split by _: https://github.com/stellar/xdrgen/blob/9796d627a739533d0c158ce87fde349b5f764f8b/lib/xdrgen/ast/concerns/named.rb#L16

Updating to do a true prefix removal likely has more downstream effects, but keeping as is should probably receive more documentation (in SEP-51/xdrgen/here?). Relevant lines from SEP-0051 include

Security Concerns

Changed with Protocols

…any use of XDR-JSON should be limited to short-lived developer facing user
interfaces. Other uses requiring more stability can use embedding the $schema
as a way to signal schema version and fallback to past versions of the
implementation for passing them appropriately.

Breaking Changes

Two types of breaking changes can occur in relation to XDR-JSON:

• As a result of XDR structural changes.

  These changes will not result in a change or version change to this document.
  While the structure of the XDR will cause the schema of XDR-JSON to change,
  the XDR-JSON specification itself will not. Applications none-the-less still
  need to coordinate these breakages if the format is used for anything more
  than a short-lived use case.

• As a result of the XDR-JSON specification defined in this proposal.

  These changes will result in a change to this document and an update to the
  specifications version, because the specification itself is changing. These
  will be rarer but will still be coordinated with new protocol releases where
  practical to reduce the need for additional coordination points for the
  Stellar community.