We have added the WorkunitId. In order to keep the meaning of each id clear, the id field previously added has been renamed to ResultId. Here is an example output:

{
           "AppName": "zika",
           "ClaimedCredit": 0.0,
           "CpuTime": 0.0,
           "ElapsedTime": 0.0,
           "ExitStatus": 0,
           "GrantedCredit": 0.0,
           "DeviceId": 3074575,
           "DeviceName": "Nimbus",
           "ModTime": 1484140157,
           "WorkunitId": 1913094363,
           "ResultId": 1191202595,
           "Name": "ZIKA_000181133_x2m9q_9NMR_DengNS3pr_wPepAnalog_0384_0",
           "Outcome": 0,
           "ReportDeadline": "2017-01-21T13:09:17",
           "SentTime": "2017-01-11T13:09:17",
           "ServerState": 4,
           "ValidateState": 0,
           "FileDeleteState": 0
}

Arggh, Id rename to ResultId has blown all Xml maps so arduously regenerated, spelling the word 'unhappy', wasted hours of my time.

(Just adding Id cut the archive piece of the period run from 7-8 minutes to under 2, on an archive of 340000 on 45000 active results.)

1	ResultsAvailable		
2	ResultsReturned		
3	Offset		
4	AppName		           "AppName": "zika",
5	ClaimedCredit		           "ClaimedCredit": 0.0,
6	CpuTime		           "CpuTime": 0.0,
7	ElapsedTime		           "ElapsedTime": 0.0,
8	ExitStatus		           "ExitStatus": 0,
9	GrantedCredit		           "GrantedCredit": 0.0,
10	DeviceId		           "DeviceId": 3074575,
11	DeviceName		           "DeviceName": "Nimbus",
12	ModTime		           "ModTime": 1484140157,
13	WorkunitId		           "WorkunitId": 1913094363,
14	ResultId		           "ResultId": 1191202595,
15	Name		           "Name": "ZIKA_000181133_x2m9q_9NMR_DengNS3pr_wPepAnalog_0384_0",
16	Outcome		           "Outcome": 0,
17	ReceivedTime		?????????????????
18	ReportDeadline		           "ReportDeadline": "2017-01-21T13:09:17",
19	SentTime		           "SentTime": "2017-01-11T13:09:17",
20	ServerState		           "ServerState": 4,
21	ValidateState		           "ValidateState": 0,
22	FileDeleteState		           "FileDeleteState": 0

Field 17 took an escape \o^o/

Except, when specifying SS 4, the Received Time is also in the list of mapped fields, that is when doing a pull with office query. What I know is not relevant, what is documented IS!

This morning had some bwilliance occuring to me [see Hunting Tool thread] which again halved the time from previous, this afternoon while cycling 30 clicks in the cold, uphill to get warm, another 'moment' happened, and revisited some old John Nash stuff, returned to the original claim it could be done 1000x faster, and indeed, after application of the 'moment', it was.

Instead of this on chronologically sorted data [and ModTime is far from unique]

=IFNA(MATCH([@ResultId];ArchResId;0);0)

this, on ResultId small to large sorted data

=IFNA(IF(INDEX(Archive[ResultId];MATCH([@ResultId];Archive[ResultId];1))=[@ResultId];MATCH([@ResultId];Archive[ResultId];1);NA());0)

Yes, this looks like lots more work, but it isnt. Now, the part that needs to find an exact match and fails on unsorted key [limited range data, as set in ArchResId], finds the exact match, if true, without specifying an exact match in the initial search, and low and behold, so fast, the part of the cycle that took minutes to find if any of current 6000 results were already on archive, went to 'don't blink as you'll miss it', 0.3 seconds according the programmed stopwatch, and this against an archive of now 627000 results.

How it works... if there's a number returned in an 'approximate search', search switch 1, like looking for ResId 1263712642 then compare the closest returned value to the original target value and if true, 1263712642, then look again for that 'approximate value' and register the record to be used to find other values of that record, such as was there credit grant or not, else, it's 100% sure to not be on archive, i.e. 0 = false

So, goodoos again on adding the ability to do binary searching instead of the painfully slow match search on the result name [the formerly only unique piece of a result].



(Just fizzy water with a suspect color in a champagne glass ;P)