The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Conflicts

No conflicts as of last run.

that's cool utACK

forgot to add description in RPC help, added.

utACK c772728

rebased

You need to update the help-text for the RPCs affected by this change (decoderawtransaction, decodescript, getrawtransaction) to include the additional field

Travis found a real problem here (for a change !!! 🎉 )

/usr/bin/python3.6 ../test/util/bitcoin-util-test.py

2019-09-10 01:34:27,547 - ERROR - Output data mismatch for txcreateoutpubkey1.json (format json)

2019-09-10 01:34:27,548 - ERROR - Output formatting mismatch for txcreateoutpubkey1.json:

*** txcreateoutpubkey1.json

--- returned

***************

*** 15,22 ****

              "scriptPubKey": {

                  "asm": "02a5613bd857b7048924264d1e70e08fb2a7e6527d32b7ab1bb993ac59964ff397 OP_CHECKSIG",

                  "hex": "2102a5613bd857b7048924264d1e70e08fb2a7e6527d32b7ab1bb993ac59964ff397ac",

!                 "type": "pubkey",

!                 "desc": "pk(02a5613bd857b7048924264d1e70e08fb2a7e6527d32b7ab1bb993ac59964ff397)#rk5v7uqw"

              }

          }

      ],

--- 15,21 ----

              "scriptPubKey": {

                  "asm": "02a5613bd857b7048924264d1e70e08fb2a7e6527d32b7ab1bb993ac59964ff397 OP_CHECKSIG",

                  "hex": "2102a5613bd857b7048924264d1e70e08fb2a7e6527d32b7ab1bb993ac59964ff397ac",

!                 "type": "pubkey"

              }

          }

      ],

2019-09-10 01:34:30,728 - ERROR - FAILED_TESTCASES:

['Creates a new transaction with a single pay-to-pubkey output (output as json)']

fixed rebase error, now all outputs are inferred. added description to decodescript

pushed update fixing @ariard nits

Being pedantic here, but wouldn't it be more logical to put the inferred output descriptor "desc" before "asm" within the "scriptPubKey" json object, to keep the abstraction level order strictly decreasing (address -> script -> raw bytes)? From a user perspective I'd find it nice to find the higher-level interpretation of a scriptPubKey right in the first line, and I only need to dig deeper into the next two lines if I want more low-level representation.

It would be nice to have this, as it compensates somewhat for the (weird) functionality lost in #20286. There are a bunch of unaddressed comments here though. And also, it'd be good if the decodescript output inside the "p2sh" and "segwit" blocks would report a descriptor that takes the structure into account (i.e., report "sh(...)" instead of "addr(...)" if it's a recognizable script).

Ah, thought I closed this PR. I'll take a fresh look.

@sipa Looks like it already does the native segwit inference in decodescript, p2sh doesn't have a block per se, just a single field with the p2sh address itself. Could add it I guess? Name suggestion for the object?

@instagibbs I was confusing decodescript's "p2sh" and "segwit" fields with the "embedded" field in getaddressinfo (thinking it'd give an Object with fields for the p2sh/segwit version of the specified script, instead of just the resulting address). Perhaps there is a use for that, but not in this PR.

Addressed all concerns, I think. Rebased on master since this is pretty ancient.

Desc for string. Update is good. Need known rebase that.

Concept ACK, sorry I missed this

kk will rebase

@meshcollider ready for rereview

rebased and formatting issues fixed

ACK 6498ba1

I don't think we should be using output descriptors to describe redeem scripts and witness scripts. See #24636