Merge of all arrays with append strategy fails.

[dnj12345]

Hi, Not sure if this is an issue or if there is a lack of understanding on my part. I am trying to merge two complicated json objects. The merge fails with the error:

No element of 'oneOf' validates both base and head: #

I am using a slightly modified schema mentioned in in issue #28 and https://www.tablix.org/~avian/blog/articles/talks/tomaz_solc_jsonmerge.pdf. That is, merge all arrays with append strategy.

The merge works great for some JSON documents. However, it failing for some other documents. I've pasted a sample python script below that causes the problem. Could someone let me know if there is an error in my coding or if this is not possible with JSON merge?

Thanks in advance for any feedback.

regards,

#!/usr/bin/env python3

import json
import sys

from jsonmerge import Merger

merge_schema = """
{
  "oneOf": [
    { "type": "number" },
    { "type": "string" },
    {
      "type": "array",
      "mergeStrategy": "arrayMergeById",
      "mergeOptions": {
        "idRef": "/"
      }
    },
    {
      "type": "object",
      "additionalProperties": {
        "$ref": "#"
      }
    }
  ]
}
"""

base = """
{
  "version": "1.0",
  "student": {
    "name": "Jane",
    "dob": "1-1-2020",
    "attribute1": {
      "size": 16777216,
      "name": "abc-xyz"
    },
    "class": {
      "type": "custom",
      "exams": [
        "final"
      ],
      "book": {
        "isbn": 1234,
        "name": "ABC Book",
        "author": "JohnDoe"
      }
    },
    "log": [
      "file"
    ]
  },
  "system": {
    "update": true,
    "update-path": "/tmp/file.json",
    "store1": {
      "store-url": "http://www.test.com/students/store1.json",
      "polling-interval": 5,
      "client-id": "abc-112233",
      "expire": {
        "batch-size": 1000,
        "scan-interval": 15
      }
    },
    "store2": {
      "store-url": "http://www.test.com/students/store2.json",
      "polling-interval": 5,
      "client-id": "abc-112233",
      "expire": {
        "batch-size": 1000,
        "scan-interval": 15
      }
    },
    "store3": {
      "store-url": "http://www.test.com/students/store3.json",
      "polling-interval": 5,
      "client-id": "abc-112233"
    },
    "report": {
      "polling-interval": 5,
      "batch-size": 5000,
      "report-type": "file",
      "exports": [
        {
          "name": "export1",
          "server": "server1:8443",
          "topic": "topic1",
          "resonse" : "required",
          "metadata": {
            "meta1": "Meta 1",
            "meta2": "Meta 2",
            "meta3": "Meta 3"
          }
        }
      ]
    }
  }
}
"""

new = """
{
  "student": {
    "class": {
      "type": "regular",
      "name": "no-name-class",
      "exams": [
        "mid-term1",
        "mid-term2"
      ],
      "book": {
        "isbn": 1234,
        "name": "ABC Book",
        "author": "JohnDoe"
      }
    }
  }
}
"""

def JsonMerge(base, new_obj):
  schema = json.loads(merge_schema)
  merger = Merger(schema)
  return merger.merge(base, new_obj, schema)

if __name__ == "__main__":
  bjson = None
  try:
    bjson = json.loads(base)
  except Exception as err:
    print('Base JSON load error. %s' % err)
    sys.exit(-1)

  njson = None
  try:
    njson = json.loads(new)
  except Exception as err:
    print('New JSON load error. %s' % err)
    sys.exit(-1)

  merged = None
  try:
    merged = JsonMerge(bjson, njson)
  except Exception as err:
    print('JSON merge error. %s' % err)
    sys.exit(-1)

  print(json.dumps(merged, indent=2, separators=(',', ': ')))

[avian2]

Try passing your data through jsonschema.validate before calling merge.

[dnj12345]

@avian2 Thank you.

After running it through jsconschema.validate, it looks like I was missing merge strategy for boolean in my schema. After adding this { "type": "boolean" } to schema, it worked. Thanks again.