During my work flow, after extracting the data from API, the JSON has the following structure:
[
{
"fields":
[
{
"meta": {
"app_type": "ios"
},
"name": "app_id",
"value": 100
},
{
"meta": {},
"name": "country",
"value": "AE"
},
{
"meta": {
"name": "Top"
},
"name": "position",
"value": 1
}
],
"metrics": {
"click": 1,
"price": 1,
"count": 1
}
}
]
Then it is store as .jsonl and put on GCS. However, when I load it onto BigQuery for further extraction, the automatic schema inference return the following error:
Error while reading data, error message: JSON parsing error in row starting at position 0: Could not convert value to string. Field: value; Value: 100
I want to convert it in to the following structure:
app_type
app_id
country
position
click
price
count
ios
100
AE
Top
1
1
1
Is there a way to define manual schema on BigQuery to achieve this result? Or do I have to preprocess the jsonl file before put it to BigQuery?
One of the limitations in loading JSON data from GCS to BigQuery is that it does not support maps or dictionaries in JSON.
A invalid example would be:
"metrics": {
"click": 1,
"price": 1,
"count": 1
}
Your jsonl file should be something like this:
{"app_type":"ios","app_id":"100","country":"AE","position":"Top","click":"1","price":"1","count":"1"}
I already tested it and it works fine.
So wherever you process the conversion of the json files to jsonl files and storage to GCS, you will have to do some preprocessing.
Probably you have to options:
precreate target table with an app_id field as an INTEGER
preprocess jsonfile and enclose 100 into quotes like "100"
Related
What am I trying to achieve:
I would like to have a time series chart showing the total number of members in my club at any time. This member count should be calculated by using the field "Eintrittsdatum" (joining-date) and "Austrittsdatum" (leaving-date). I’m thinking of it as a running sum - every filled field with a joining-date means +1 on the member count, every leaving-date entry is a -1.
Data structure
I’m calling the API of webling.ch with a secret key. This is my data structure with sample data per member:
[
{
"type": "member",
"meta": {
"created": "2020-03-02 11:33:00",
"createuser": {
"label": "Joana Doe",
"type": "user"
},
"lastmodified": "2022-12-06 16:32:56",
"lastmodifieduser": {
"label": "Joana Doe",
"type": "user"
}
},
"readonly": true,
"properties": {
"Mitglieder ID": 99,
"Anrede": "Dear",
"Vorname": "Jon",
"Name": "Doe",
"Strasse": "Doeington Street",
"Adresszusatz": null,
"PLZ": "9999",
"Ort": "Doetown",
"E-Mail": "jon.doe#doenet.net",
"Telefon Privat": null,
"Telefon Geschäft": null,
"Mobile": "099 877 54 54",
"Geschlecht": "m",
"Geburtstag": "1966-03-10",
"Mitgliedschaftstyp": "Aktivmitgliedschaft",
"Eintrittsdatum": "2020-03-01",
"Austrittsdatum": null,
"Passfoto": null,
"Wordpress Benutzername": null,
"Wohnhaft im Glarnerland": false,
"Lat": "43.1563379",
"Long": "6.0474622"
},
"parents": [
240
],
"children": {
},
"links": {
"debitor": [
2124,
3056,
3897
],
"attendee": [
2576
]
},
"id": 1815
}
]
Grafana data source
I am using the “JSON API” by Marcus Olsson: GitHub - grafana/grafana-json-datasource: A data source plugin for loading JSON APIs into Grafana.
Grafana v9.3.1 (89b365f8b1) on Linux
My current approach
Queries:
Query C - uses a filter on the source-API to only show entries with "Eintrittsdatum" IS NOT EMPTY
Field 1 (alias "datum") has a JSONata-Query of:
properties.Eintrittsdatum
Field 2 (alias "names") should return the full name and has a query of:
$map($.properties, function($v) {(
($v.Vorname&" "&$v.Name);
)})
Field 3 (alias "value") should return "1" for every entry and has a query of:
$map($.properties, function($v) {(
(1);
)})
Query D - uses a filter on the source-API to only show entries with "Austrittsdatum" IS NOT EMPTY
Field 1 (alias "datum") has a JSONata-Query of:
properties.Austrittsdatum
Field 2 (alias "names") should return the full name and has a query of:
$map($.properties, function($v) {(
($v.Vorname&" "&$v.Name);
)})
Field 3 (alias "value") should return "1" for every entry and has a query of:
$map($.properties, function($v) {(
(1);
)})
Here's a screenshot to clarify things
(https://zigerschlitzmakers.ch/wp-content/uploads/2023/01/ScreenshotGrafana-1.png)
Transformations:
My applied transformations
(https://zigerschlitzmakers.ch/wp-content/uploads/2023/01/ScreenshotGrafana-2.png)
What's working
I can correctly gather the number of members added/subtracted per day.
What's not working
I can't get the graph to display the way i want: I'd like to have a running sum of these numbers instead of the following two graphs.
Time series graph with merged queries
(https://zigerschlitzmakers.ch/wp-content/uploads/2023/01/ScreenshotGrafana-3.png)
Time series graph with unmerged queries
(https://zigerschlitzmakers.ch/wp-content/uploads/2023/01/ScreenshotGrafana-4.png)
I can't get the names to display within the tooltip of the data points (really not THAT necessary).
I am banging my head head here for the past 2 hours with all the available JSON_... functions in BigQuery. I've read quite a few questions here but no matter why I try, I never succeed in extracting the "amounts" from my JSON below.
This is my JSON stored in a BQ column:
{
"lines": [
{
"id": "70223039-83d6-463d-a482-7ce4d50bf0fc",
"charges": [
{
"type": "price",
"amount": 50.0
},
{
"type": "discount",
"amount": -40.00
}
]
},
{
"id": "70223039-83d6-463d-a482-7ce4d50bf0fc",
"charges": [
{
"type": "price",
"amount": 20.00
},
{
"type": "discount",
"amount": 0.00
}
]
}
]
}
Imagine the above being an order containing multiple items.
I am trying to get a sum of all amounts => 50-40+20+0. The result needs to be 30 = the total order price.
Is it possible to pull all the amount values and then have them summed up just via SQL without any custom JS functions? I guess the summing is the easy part - getting the amounts into an array is the challenge here.
Use below
select (
select sum(cast(json_value(charge, '$.amount') as float64))
from unnest(json_extract_array(order_as_json, '$.lines')) line,
unnest(json_extract_array(line, '$.charges')) charge
) total
from your_table
if applied to sample data in y our question - output is
I`m new to nifi and I want to convert big amount of json data to csv format.
This is what I am doing at the moment but it is not the expected result.
These are the steps:
processes to create access_token and send request body using InvokeHTTP(This part works fine I wont name the processes since this is the expected result) and getting the response body in json.
Example of json response:
[
{
"results":[
{
"customer":{
"resourceName":"customers/123456789",
"id":"11111111"
},
"campaign":{
"resourceName":"customers/123456789/campaigns/222456422222",
"name":"asdaasdasdad",
"id":"456456546546"
},
"adGroup":{
"resourceName":"customers/456456456456/adGroups/456456456456",
"id":"456456456546",
"name":"asdasdasdasda"
},
"metrics":{
"clicks":"11",
"costMicros":"43068982",
"impressions":"2079"
},
"segments":{
"device":"DESKTOP",
"date":"2021-11-22"
},
"incomeRangeView":{
"resourceName":"customers/456456456/incomeRangeViews/456456546~456456456"
}
},
{
"customer":{
"resourceName":"customers/123456789",
"id":"11111111"
},
"campaign":{
"resourceName":"customers/123456789/campaigns/222456422222",
"name":"asdasdasdasd",
"id":"456456546546"
},
"adGroup":{
"resourceName":"customers/456456456456/adGroups/456456456456",
"id":"456456456546",
"name":"asdasdasdas"
},
"metrics":{
"clicks":"11",
"costMicros":"43068982",
"impressions":"2079"
},
"segments":{
"device":"DESKTOP",
"date":"2021-11-22"
},
"incomeRangeView":{
"resourceName":"customers/456456456/incomeRangeViews/456456546~456456456"
}
},
....etc....
]
}
]
Now I am using:
===>SplitJson ($[].results[])==>JoltTransformJSON with this spec:
[{
"operation": "shift",
"spec": {
"customer": {
"id": "customer_id"
},
"campaign": {
"id": "campaign_id",
"name": "campaign_name"
},
"adGroup": {
"id": "ad_group_id",
"name": "ad_group_name"
},
"metrics": {
"clicks": "clicks",
"costMicros": "cost",
"impressions": "impressions"
},
"segments": {
"device": "device",
"date": "date"
},
"incomeRangeView": {
"resourceName": "keywords_id"
}
}
}]
==>> MergeContent( here is the problem which I don`t know how to fix)
Merge Strategy: Defragment
Merge Format: Binary Concatnation
Attribute Strategy Keep Only Common Attributes
Maximum number of Bins 5 (I tried 10 same result)
Delimiter Strategy: Text
Header: [
Footer: ]
Demarcator: ,
What is the result I get?
I get a json file that has parts of the json data
Example: I have 50k customer_ids in 1 json file so I would like to send this data into big query table and have all the ids under the same field "customer_id".
The MergeContent uses the split json files and combines them but I will still get 10k customer_ids for each file i.e. I have 5 files and not 1 file with 50k customer_ids
After the MergeContent I use ==>>ConvertRecord with these settings:
Record Reader JsonTreeReader (Schema Access Strategy: InferSchema)
Record Writer CsvRecordWriter (
Schema Write Strategy: Do Not Write Schema
Schema Access Strategy: Inherit Record Schema
CSV Format: Microsoft Excel
Include Header Line: true
Character Set UTF-8
)
==>>UpdateAttribute (custom prop: filename: ${filename}.csv) ==>> PutGCSObject(and put the data into the google bucket (this step works fine- I am able to put files there))
With this approach I am UNABLE to send data to big query(After MergeContent I tried using PutBigQueryBatch and used this command in bq sheel to get the schema I need:
bq show --format=prettyjson some_data_set.some_table_in_that_data_set | jq '.schema.fields'
I filled all the fields as needed and Load file type: I tried NEWLINE_DELIMITED_JSON or CSV if I converted it to CSV (I am not getting errors but no data is uploaded into the table)
)
What am I doing wrong? I basically want to map the data in such a way that each fields data will be under the same field name
The trick you are missing is using Records.
Instead of using X>SplitJson>JoltTransformJson>Merge>Convert>X, try just X>JoltTransformRecord>X with a JSON Reader and a CSV Writer. This skips a lot of inefficiency.
If you really need to split (and you should avoid splitting and merging unless totally necessary), you can use MergeRecord instead - again with a JSON Reader and CSV Writer. This would make your flow X>Split>Jolt>MergeRecord>X.
I have been struggling with this problem for a long time. I need to create a new JSON flowfile using QueryRecord by taking an array (field ref) from input JSON field refs and skip the object field as shown in example below:
Input JSON flowfile
{
"name": "name1",
"desc": "full1",
"refs": {
"ref": [
{
"source": "source1",
"url": "url1"
},
{
"source": "source2",
"url": "url2"
}
]
}
}
QueryRecord configuration
JSONTreeReader setup as Infer Schema and JSONRecordSetWriter
select name, description, (array[rpath(refs, '//ref[*]')]) as sources from flowfile
Output JSON (need)
{
"name": "name1",
"desc": "full1",
"references": [
{
"source": "source1",
"url": "url1"
},
{
"source": "source2",
"url": "url2"
}
]
}
But got error:
QueryRecord Failed to write MapRecord[{references=[Ljava.lang.Object;#27fd935f, description=full1, name=name1}] with schema ["name" : "STRING", "description" : "STRING", "references" : "ARRAY[STRING]"] as a JSON Object due to java.lang.ClassCastException: null
Try the following approach, in your case it shoud work:
1) Read your JSON field fully (I imitated it with GenerateFlowFile processor with your example)
2) Add EvaluateJsonPath processor which will put 2 header fileds (name, desc) into the attributes:
3) Add SplitJson processor which will split your JSON byt refs/ref/ groups (split by "$.refs.ref"):
4) Add ReplaceText processor which will add you header fields (name, desc) to the split lines (replace "[{]" value with "{"name":"${json.name}","desc":"${json.desc}","):
5) It`s done:
Full process in my demo case:
Hope this helps.
Solution!: use JoltTransformJSON to transform JSON by Jolt specification. About this specification.
I am trying to load json with the following schema into BigQuery:
{
key_a:value_a,
key_b:{
key_c:value_c,
key_d:value_d
}
key_e:{
key_f:value_f,
key_g:value_g
}
}
The keys under key_e are dynamic, ie in one response key_e will contain key_f and key_g and for another response it will instead contain key_h and key_i. New keys can be created at any time so I cannot create a record with nullable fields for all possible keys.
Instead I want to create a column with JSON datatype that can then be queried using the JSON_EXTRACT() function. I have tried loading key_e as a column with STRING datatype but value_e is analysed as JSON and so fails.
How can I load a section of JSON into a single BigQuery column when there is no JSON datatype?
Having your JSON as a single string column inside BigQuery is definitelly an option. If you have large volume of data this can end up with high query price as all your data will end up in one column and actually querying logic can become quite messy.
If you have luxury of slightly changing your "design" - I would recommend considering below one - here you can employ REPEATED mode
Table schema:
[
{ "name": "key_a",
"type": "STRING" },
{ "name": "key_b",
"type": "RECORD",
"mode": "REPEATED",
"fields": [
{ "name": "key",
"type": "STRING"},
{ "name": "value",
"type": "STRING"}
]
},
{ "name": "key_e",
"type": "RECORD",
"mode": "REPEATED",
"fields": [
{ "name": "key",
"type": "STRING"},
{ "name": "value",
"type": "STRING"}
]
}
]
Example of JSON to load
{"key_a": "value_a1", "key_b": [{"key": "key_c", "value": "value_c"}, {"key": "key_d", "value": "value_d"}], "key_e": [{"key": "key_f", "value": "value_f"}, {"key": "key_g", "value": "value_g"}]}
{"key_a": "value_a2", "key_b": [{"key": "key_x", "value": "value_x"}, {"key": "key_y", "value": "value_y"}], "key_e": [{"key": "key_h", "value": "value_h"}, {"key": "key_i", "value": "value_i"}]}
Please note: it should be newline delimited JSON so each row must be on one line
You can't do this directly with BigQuery, but you can make it work in two passes:
(1) Import your JSON data as a CSV file with a single string column.
(2) Transform each row to pack your "any-type" field into a string. Write a UDF that takes a string and emits the final set of columns you would like. Append the output of this query to your target table.
Example
I'll start with some JSON:
{"a": 0, "b": "zero", "c": { "woodchuck": "a"}}
{"a": 1, "b": "one", "c": { "chipmunk": "b"}}
{"a": 2, "b": "two", "c": { "squirrel": "c"}}
{"a": 3, "b": "three", "c": { "chinchilla": "d"}}
{"a": 4, "b": "four", "c": { "capybara": "e"}}
{"a": 5, "b": "five", "c": { "housemouse": "f"}}
{"a": 6, "b": "six", "c": { "molerat": "g"}}
{"a": 7, "b": "seven", "c": { "marmot": "h"}}
{"a": 8, "b": "eight", "c": { "badger": "i"}}
Import it into BigQuery as a CSV with a single STRING column (I called it 'blob'). I had to set the delimiter character to something arbitrary and unlikely (thorn -- 'þ') or it tripped over the default ','.
Verify your table imported correctly. You should see your simple one-column schema and the preview should look just like your source file.
Next, we write a query to transform it into your desired shape. For this example, we'd like the following schema:
a (INTEGER)
b (STRING)
c (STRING -- packed JSON)
We can do this with a UDF:
// Map a JSON string column ('blob') => { a (integer), b (string), c (json-string) }
bigquery.defineFunction(
'extractAndRepack', // Name of the function exported to SQL
['blob'], // Names of input columns
[{'name': 'a', 'type': 'integer'}, // Output schema
{'name': 'b', 'type': 'string'},
{'name': 'c', 'type': 'string'}],
function (row, emit) {
var parsed = JSON.parse(row.blob);
var repacked = JSON.stringify(parsed.c);
emit({a: parsed.a, b: parsed.b, c: repacked});
}
);
And a corresponding query:
SELECT a, b, c FROM extractAndRepack(JsonAnyKey.raw)
Now you just need to run the query (selecting your desired target table) and you'll have your data in the form you like.
Row a b c
1 0 zero {"woodchuck":"a"}
2 1 one {"chipmunk":"b"}
3 2 two {"squirrel":"c"}
4 3 three {"chinchilla":"d"}
5 4 four {"capybara":"e"}
6 5 five {"housemouse":"f"}
7 6 six {"molerat":"g"}
8 7 seven {"marmot":"h"}
9 8 eight {"badger":"i"}
One way to do it, is to load this file as CSV instead of JSON (and quote the values or eliminate newlines in the middle), then it will become single STRING column inside BigQuery.
P.S. You are right that having a native JSON data type would have made this scenario much more natural, and BigQuery team is well aware of it.