Using Offset on Time ?

scott74scott74 Member Posts: 35
in Apps & Integrations
I am using integromat and used their time now function to pull sales for today's date -5 minutes so it pulls only sales within the last 5 minutes on every pull.

The time came through at 19:55:24 instead of 13:55:24 which would be around 1:55 pm Central time so guessing the light speed server is 7 hours ahead of me?

Was reading on this page to apply an offset. I am not sure how to add this to this URl
Any help by passing an example URL with US Central Time with the link I have below back would be very much appreciated.
https://developers.lightspeedhq.com/retail/introduction/parameters/


https://api.lightspeedapp.com/API/Account/{{Account_ID}}/Sale.json?load_relations=["Customer","Customer.Contact"]&timeStamp=><,2018-12-21T00:00:00+,2018-12-21T23:59:00
· Share on Google+

14 comments

  • scott74scott74 Member Posts: 35
    edited December 2018
    BTW.. Tried this way but did not work. Still about 6 Hours off from what the output should have been when testing

    https://api.lightspeedapp.com/API/Account/{{Account_ID}}/Sale.json?load_relations=["Customer","Customer.Contact"]&timeStamp=><,2018-12-27T20:33:08.946Z+,2018-12-27T20:38:08.946Z&offset=UTC−06:00



    · Share on Google+
  • gregaricangregarican Member Posts: 382 
    Maybe try something like this:

    https://api.lightspeedapp.com/API/Account/{{Account_ID}}/Sale.json?load_relations=["Customer","Customer.Contact"]&timeStamp=><,2018-12-27T20:33:08.946-6:00,2018-12-27T20:38:08.946-6:00
    · Share on Google+
  • scott74scott74 Member Posts: 35
    edited December 2018
    gregarican said:
    Maybe try something like this:

    https://api.lightspeedapp.com/API/Account/{{Account_ID}}/Sale.json?load_relations=["Customer","Customer.Contact"]&timeStamp=><,2018-12-27T20:33:08.946-6:00,2018-12-27T20:38:08.946-6:00
    Hmmmm.. This resulted in this. Not sure what this is its tagging on the end 17.865Z

    https://api.lightspeedapp.com/API/Account/{{Account_ID}}/Sale.json?load_relations=["Customer","Customer.Contact"]&timeStamp=><,2018-12-27T22:04:17.863Z-6:00+,2018-12-27T22:09:17.865Z-6:00

    Time is 22 and should say 16 I believe for 4 pm Central




    · Share on Google+
  • gregaricangregarican Member Posts: 382 
    edited December 2018
    You don't include the Z in there. Here is a capture of a successful request/response pair I just ran. I commented out my account ID and the auth token. And truncated all the sales data, but look at the URL I used. This would pull all sales from 12/27/2018 between 11 and noon ET.

    GET https://api.merchantos.com/API/Account/{AccountId}/Sale.json?archived=false&load_relations=[%22SaleLines%22,%22SaleLines.Discount%22,%22SaleLines.Note%22,%22SaleLines.Item%22,%22SalePayments%22,%22SalePayments.PaymentType%22,%22SalePayments.SaleAccounts%22,%22Customer%22,%22Customer.Contact%22,%22Discount%22]&offset=0&limit=100&timeStamp=%3E%3C,2018-12-27T11:00:00-5:00,2018-12-27T12:00:00-5:00 HTTP/1.1
    cache-control: no-cache
    Postman-Token: 4a63e894-0a34-4880-a840-5381a8378817
    Authorization: Bearer {AuthToken}
    User-Agent: PostmanRuntime/7.4.0
    Accept: */*
    Host: api.merchantos.com
    cookie: __cfduid=dc92c208fe990511a3e7b3f1c7a60be141540561895
    accept-encoding: gzip, deflate
    Connection: close


    HTTP/1.1 200 OK
    Date: Fri, 28 Dec 2018 13:07:18 GMT
    Content-Type: application/json
    Content-Length: 19266
    Connection: close
    x-frame-options: SAMEORIGIN
    X-XSS-Protection: 1; mode=block
    X-Content-Type-Options: nosniff
    X-LS-Acct-Id: 164019
    X-LS-OAuth-Client-Id: 64577
    X-LS-API-Bucket-Level: 1/100
    X-LS-Shard-Id: 18
    X-LS-API-Drip-Rate: 5
    X-LS-Master-System: true
    X-LS-Master-Account: true
    X-LS-Master-Catalog: false
    Vary: Accept-Encoding
    Expect-CT: max-age=604800, report-uri="https://report-uri.cloudflare.com/cdn-cgi/beacon/expect-ct"
    Server: cloudflare
    CF-RAY: 49043fc6adfb9f17-ORD

    {"@attributes":{"count":"6","offset":"0","limit":"100"},"Sale": [...all the data is here...]}
    · Share on Google+
  • scott74scott74 Member Posts: 35
    edited December 2018
    Thanks.. Think I got working. Having it run every 5 min but pulling only sales within the last 5 minutes.. Have to figure out what that Z is but seems to be working. I will have to create a fake sale to check.

    I guess when there is no sale you get a data output like so? and you mentioned something about only allowing some 100 limit. Does this mean it will only pull 100 sales from each pull? Just want to make sure I don't run into any hiccups later or hit that bucket limit or anything.

    Thx for your input.





    · Share on Google+
  • gregaricangregarican Member Posts: 382 
    Correct. The most records per "page pull" that will be passed back are 100. I think that's for pretty much any API pull request in the LS Retail API endpoint collection --- products, sales, customers, etc.

    It's a shame that pushing new records into LS Retail via the API won't allow batch recordsets like that. Those pushes are typically one record at a time. Which makes importing data into LS Retail a very expensive operation. Worst case is one record push every 10 seconds I believe due to API throttling.
    · Share on Google+
  • scott74scott74 Member Posts: 35
    Oh wow... So for a big multi location chain, pulling records once a day would not work if say they were doing 500 retail sales a day, you would not get the other 400?

    This is a 23 location supplement store so every 5 minutes should be sufficient. Can't imagine there would be 100 sales every 5 minutes but want to send customers other data fairly quickly after leaving.

    Thanks for the incite on that.

    Can you refresh me on how that bucket limit works though. If you are trying to stay clear of the max 100 record limit that might get you in trouble with your bucket limit correct? Surely this is only the case for 100's of stores with a boatload of sales hopefully lol.
    · Share on Google+
  • gregaricangregarican Member Posts: 382 
    That's 100 records max per individual pull. The API limits you to pulling worst case 1 API call per second. So you can call the first 100 record page, get your response and parse it, then a second later call the second 100 record page, and so on. That's where the offset comes in. That specifies the "page" you are requesting. You look at the @attributes portion of the response body and can calculate how many pages you will need to pull, and use the offset to specify which page you need each iteration.

    You need to revisit the rate limits and the parameters in the API documentation.


    These will show you how you can determine how long to wait before your next API call, as well as how you can specify which page you are requesting. With your example of 500 sales per day, that's only 5 API calls. Which should take only a minute or so to pull pages and parse, even with loaded relations!
    · Share on Google+
  • scott74scott74 Member Posts: 35
    Geezus.. just keeps getting more complicated lol

    Well I guess since I have it pulling every 5 minutes and I am dynamically changing the time to only pull sales within the last 5 minutes then I should not have to worry about getting any second pages unless they have over 100 sales within any 5 min period.

    If that were the case I guess I would have to set an if state of If over 100 records then start a new request with a new https request with the offset added in the URL?

    Think that's what you are saying.


    · Share on Google+
  • gregaricangregarican Member Posts: 382 
    Correct. Let's say we had 105 sales rung yesterday between 11 and noon ET. And I wanted to pull them. My first pull with offset=0 would tell me in the @attributes of the response body that I would need to pull a second page. Here's the first page pull:

    GET https://api.merchantos.com/API/Account/{AccountId}/Sale.json?archived=false&load_relations=[%22SaleLines%22,%22SaleLines.Discount%22,%22SaleLines.Note%22,%22SaleLines.Item%22,%22SalePayments%22,%22SalePayments.PaymentType%22,%22SalePayments.SaleAccounts%22,%22Customer%22,%22Customer.Contact%22,%22Discount%22]&offset=0&limit=100&timeStamp=%3E%3C,2018-12-27T11:00:00-5:00,2018-12-27T12:00:00-5:00 

    And here would be the second page pull:

    GET https://api.merchantos.com/API/Account/{AccountId}/Sale.json?archived=false&load_relations=[%22SaleLines%22,%22SaleLines.Discount%22,%22SaleLines.Note%22,%22SaleLines.Item%22,%22SalePayments%22,%22SalePayments.PaymentType%22,%22SalePayments.SaleAccounts%22,%22Customer%22,%22Customer.Contact%22,%22Discount%22]&offset=100&limit=100&timeStamp=%3E%3C,2018-12-27T11:00:00-5:00,2018-12-27T12:00:00-5:00

    The first (default) pull is offset=0 to skip no records. Once I see that there are 105 sales in that first response, I would then instruct my second pull to have offset=100 to skip over the first 100 records. So my second pull would grab records 101 through 105 and complete the routine.

    Any code you have should evaluate the @attributes, set a variable to the total record count, and then you can perform simple ceiling math to determine the iteration block of how many offsets you will need to specify in separate API calls...


    · Share on Google+
  • scott74scott74 Member Posts: 35
    edited December 2018
    Good explanation and good info.. Would have definitely needed that probably if pulling sales just once a day for 23 locations  


    Guess could just set up a router with a filter if offset = 1 then the https request for that would be offset=100

    if offset response = 2  https request offset=200  and so on... This would poll 201 through 300 correct.

    Then once you know your average daily sales... Throw a couple extra request on there just in case which would cover another 200 over your average...
    · Share on Google+
  • gregaricangregarican Member Posts: 382 
    The offset isn't specifying which page to pull, it's specifying how many records to skip. So offset=0 skips no records, offset=100 skips 100 records, etc.
    · Share on Google+
  • scott74scott74 Member Posts: 35
    Got ya
    · Share on Google+
  • scott74scott74 Member Posts: 35
    gregarican said:
    Maybe try something like this:

    https://api.lightspeedapp.com/API/Account/{{Account_ID}}/Sale.json?load_relations=["Customer","Customer.Contact"]&timeStamp=><,2018-12-27T20:33:08.946-6:00,2018-12-27T20:38:08.946-6:00
    Hmmmm.. This resulted in this. Not sure what this is its tagging on the end 17.865Z


    https://api.lightspeedapp.com/API/Account/{{Account_ID}}/Sale.json?load_relations=["Customer","Customer.Contact"]&timeStamp=><,2018-12-27T22:04:17.863Z-6:00+,2018-12-27T22:09:17.865Z-6:00


    · Share on Google+
Sign In or Register to comment.