Annex RDS forum

Posted: **Fri Apr 09, 2021 4:07 pm**

I'm trying to extract JSON data from ThingSpeak into an array using the (stripped down) code below and am seeing the first element as the correct value but elements 2 and 3 are read as zero. Is my JSON format request correct for filling this array?

Code: [Local Link Removed for Guests]

for lcnt = 1 to 3
  IntTemp(lcnt) = val(json$(NewData$,"feeds[" + str$(lcnt) + "].field4"))
  wlog "Query string : feeds[" + str$(lcnt) + "].field4" + "     Response " +    str$(IntTemp(lcnt) )
next lcnt

Resulting in the wlog output :
Query string : feeds[1].field4 Response 12.6
Query string : feeds[2].field4 Response 0
Query string : feeds[3].field4 Response 0

UPDATE - SORTED!!!
It works if I move the array reference to the end :
json$(NewData$,"feeds.field4[1]")
json$(NewData$,"feeds.field4[2]")
json$(NewData$,"feeds.field4[3]")
Which may be at odds with correct JSON structure and the example in the help file?

This is the JSON data received from ThingSpeak after running through a formatter.

Code: [Local Link Removed for Guests]

{
   "channel":{
      "id":1283507,
      "name":"Greenhouse Monitor",
      "description":"Monitoring internal and external greenhouse environment with ESP32 and 2 x BME280.\r\nESP32 is running Annex RDS. ",
      "latitude":"0.0",
      "longitude":"0.0",
      "field1":"Battery",
      "field2":"Wifi RSSI",
      "field3":"Atm. Pressure",
      "field4":"Internal Temp.",
      "field5":"Internal Hum.",
      "field6":"External Temp.",
      "field7":"External Hum.",
      "created_at":"2021-01-15T22:52:34Z",
      "updated_at":"2021-03-14T18:18:06Z",
      "last_entry_id":7654
   },
   "feeds":[
      {
         "created_at":"2021-04-09T15:29:20Z",
         "entry_id":7652,
         "field1":"3.7",
         "field2":"45",
         "field3":"1014",
         "field4":"12.6",
         "field5":"82",
         "field6":"8.4",
         "field7":"84"
      },
      {
         "created_at":"2021-04-09T15:39:18Z",
         "entry_id":7653,
         "field1":"3.7",
         "field2":"45",
         "field3":"1014",
         "field4":"12.6",
         "field5":"82",
         "field6":"8.5",
         "field7":"84"
      },
      {
         "created_at":"2021-04-09T15:49:17Z",
         "entry_id":7654,
         "field1":"3.7",
         "field2":"45",
         "field3":"1014",
         "field4":"12.5",
         "field5":"82",
         "field6":"8.4",
         "field7":"83"
      }
   ]
}

Posted: **Sat Apr 10, 2021 7:25 am**

Yes, the JSON parser included in Annex is not a "true" json parser but more a "search engine".
I did it many time ago mainly with the objective to consume as less RAM as possible.

I fact, it works looking for the occurrence of each elements defined in the json$ command in sequence.
Because "feeds" is defined only once in the data, 'feeds[1]' will works but "feeds[2]" and so on will not work.

In your case, you found that it works with json$(NewData$,"feeds.field4[1]")

This is because the JSON$ function looks before for the first occurrence of the element "feed" and then it looks for the first occurrence of the element "field4".
Using json$(NewData$,"feeds.field4[2]") looks before for the first occurrence of the element "feed" and then it looks for the second occurrence of the element "field4".

This can considered as a different way to specify the JSON search pattern.

I already tried to implement a "real" JSON parser but each time I found that the impact in terms of RAM and memory space was not effective considering the resources of the chips (in particular for the ESP8266).

Posted: **Sat Apr 10, 2021 8:29 am**

That makes perfect sense - Thanks for the explanation, Cicciocb.

Your way may not be technically exact, but it does the job and I find it more logical than the 'correct' way.
You may want to update the manual example to mirror the way you've implemented it.

Annex RDS forum

Help with JSON and arrays - Resolved.

Help with JSON and arrays - Resolved.

Re: Help with JSON and arrays - Resolved.

Re: Help with JSON and arrays - Resolved.