遍历 JSON 文件中的嵌套字典

Question

I have the following output from a json file

    {
  "Threshold": 0.6,
  "Services": [
    {
      "Name": "Service1",
      "Query": [
        "query1",
        "query2",
        "query3"
      ],
      "Products": [
        {
          "Name": "product1",
          "Query": [
            "query4",
            "query5"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product2",
          "Query": [
            "query6",
            "query7",
            "query8"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product3",
          "Query": [
            "query9",
            "query10"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product4",
          "Query": [
            "query11",
            "query12"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product5",
          "Query": [
            "query13",
            "query14"
          ],
          "Threshold": 0.75
        }
      ]
    },
    {
      "Name": "Service2",
      "Query": [
        "query1",
        "query2",
        "query3"
      ],
      "Products": [
        {
          "Name": "product1",
          "Query": [
            "query4",
            "query5"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product2",
          "Query": [
            "query6",
            "query7",
            "query8"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product3",
          "Query": [
            "query9",
            "query10"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product4",
          "Query": [
            "query11",
            "query12"
          ],
          "Threshold": 0.75
        },
        {
          "Name": "product5",
          "Query": [
            "query13",
            "query14"
          ],
          "Threshold": 0.75
        }
      ]
    }
  ]
}

The structure of the file is as follows: There are two services and each of this service has five similar products. Each service has a list of queries that contain some keywords that describe the service.

The same applies for the products.

Each product as a list of query that contains keywords that describe the individual product.

I want to loop into the services and select the service1. I then want to run the queries in an algorithm against a text to find whether one or all the queries are present in the text. If the query or queries are present, I want to go into the products and start iterating through product1 to product5. If not, it should skip and go to the service2

I want to do the same for the service2 and the respective products.

I can only run it when I run the following code:

for service in configfile["Services"]:
        if service["Name"] == "Service1":

The code must run without hardcoding the "Service1 or "Service2" name.

In essence, I want to take the service1 and service2 and access the queries. I will run the code from Sentence transformers .

# Query sentences:
queries = ['query1', 'query2', 'query3']



top_k = min(5, len(corpus))
for query in queries:
    query_embedding = embedder.encode(query, convert_to_tensor=True)

   # We use cosine-similarity and torch.topk to find the highest 5 scores
    cos_scores = util.pytorch_cos_sim(query_embedding, corpus_embeddings)[0]
    top_results = torch.topk(cos_scores, k=top_k)

    print("\n\n======================\n\n")
    print("Query:", query)
    print("\nTop 5 most similar sentences in corpus:")

    for score, idx in zip(top_results[0], top_results[1]):
         print(corpus[idx], "(Score: {:.4f})".format(score))

I get some scores from this algorithm for each query. I then sort the queries and I take the query with the largest value. If this value is above a threshold, I want to continue with the same algorithm in the queries of product1 to product5 that are in the services1. Take the largest value and if the value is above a threshold, I will add it to a dictionary as "positive". if not, I will add it as "negative".

If the value is below the threshold, I want to skip the iteration to the products and go to Service2 and run the algorithm again for the queries and repeat the process.

Answer 1

It's not clear what you want to do with your products but since Services and Products are lists, you can simply loop through them:

for serv in json_data.get("Services", []):
    if serv.get("Query", None):
        for prod in serv.get("Products", []):
            # do your stuff with prod
            print(serv.get("Name", None), prod.get("Name", None))

Output:

Service1 product1
Service1 product2
Service1 product3
Service1 product4
Service1 product5
Service2 product1
Service2 product2
Service2 product3
Service2 product4
Service2 product5