Conduit transaction.search: document supported values for objectType
Closed, ResolvedPublic
Actions

Assigned To

Authored By

	valerio.bozzolan
	Sun, May 4, 10:18

Description

At the moment the Conduit page /conduit/method/transaction.search/ does not have practical documentation about the supported values for the parameter objectType. The only available documentation is:

You can quickly find an unknown type constant by looking at the PHID of an object of that type.

I think the above line is useful but people may not have the time and the patience to do the above check (that, by the way, that attempt is a waste of resources).

It's true that there are some exception messages already available that suggest the already-existing objectType(s) in case of some mistakes, but:

Note that these exception messages showing available values, also mention unsupported values. This may be considered a bug:

EXCEPTION: (Exception) In call to "transaction.search", specified "objectType" ("FOOB") is unknown. Valid object types are: ABND, ACNT, ADEV, AEML, AFTR, AINT, AINV, AKEY, AMSG, ANAM, ANET, ANSW, APAS, APPE, APPS, ASRV, ATOM, AUTH, BDGE, BLOG, BOOK, BULK, CART, CDTL, CDWN, CEVT, CEXP, CHAL, CHRG, CIMP, CMIT, CONF, CONP, CTNM, CXNV, DCNG, DIFF, DREV, DRYA, DRYB, DRYL, DRYO, DRYR, DSHB, DSHP, EADR, FBAK, FILE, FITV, FORM, FPRV, HLXS, HMBA, HMBB, HMBD, HMBT, HMCL, HMCP, HMCS, HRUL, HWBH, HWBR, LEGD, LEGS, MCRO, MOCK, MTAM, NUAC, NUAI, NUAQ, NUAS, OASA, OASC, OPKG, PANL, PAYM, PCOL, PDCT, PHPR, PHRL, PIMG, PLCY, PMRC, POLL, POST, PPAK, PPUB, PRCH, PROJ, PRTL, PSET, PSHE, PSHL, PSTE, PSUB, PULE, PVAR, PVER, QUES, REPO, RIDT, RREF, RURI, SPCE, SSSN, SYNE, TASK, TOKN, TRIG, USER, WIKI, WRDS, WTRG, XACT, XIDT, XOBJ, XUSR. at [<phorge>/src/applications/transactions/conduit/TransactionSearchConduitAPIMethod.php:382]

In fact, the XACT is not really supported - T16054: Conduit API transaction.search with objectType = XACT causes internal error "get_class() expects parameter 1 to be object, null given".

So, from a documentation perspective, it would be nice to fix the current nonsense exception suggestions, and add these types in the documentation itself, so, you don't have anymore to cause an exception to actually discover the values.