JSON is not saving in the right structure in python-CodePudding

I'm having a little problem with a class activity of python. On this activity, i should create a I should create a system for a "small store" to be able to register products (in a JSON file) and then be able to load the file and make queries with the same loaded. Basically, the job is 90% completed, but i'm stuck trying to figure out why my products records isn't saving on the right structure of JSON. I don´t know why the JSON is not saving the items with the "[ ]" before and after the records.

global estoque
global contador_sinal
global carregar_flag
estoque = []
lista_produtos = []   # <--- HERE I CREATE THE LIST THAT RECEIVES PRODUCT REGISTRATIONS
contador_sinal = 0
contador = 0
carregar_flag = 0

class Produto: # <-- CREATING THE CLASS OF PRODUCTS
    def __init__(self, codigo, nome, descricao, valor):
        self.codigo = codigo
        self.nome = nome
        self.descricao = descricao
        self.valor = valor

def gravar(): # <-- AND THIS IS THE FUNCTION THAT SAVES THE LIST PRODUCTS IN JSON
    global contador_sinal
    print("="*50)
    print(" ","\nGravando Arquivo Externo dos Produtos...")
    with open('data.json', 'w') as fp:
        for item in lista_produtos:
            if type(item) == Produto:
                json.dump(item.__dict__, fp, indent=4)    
                fp.write("\n")
            elif type(item) == dict:
                json.dump(item, fp, indent=4)    
    time.sleep(1)
    print(" ")
    print("\033[32mARQUIVO SALVO COM SUCESSO!\033[0m\n")
    fp.close()

    contador_sinal = 1
    
    return menu()

Here is how the information appears on the JSON after saved:

{
    "codigo": 1,
    "nome": "test1",
    "descricao": "description",
    "valor": 10.0
}
{
    "codigo": 2,
    "nome": "test2",
    "descricao": "description",
    "valor": 20.0
}
{
    "codigo": 3,
    "nome": "test3",
    "descricao": "description",
    "valor": 30.0
}

If someone knows how to fix this structure and the reason for being saved wrong, it would be a tremendous help.

CodePudding user response：

import json


class Produto:  # <-- CREATING THE CLASS OF PRODUCTS
    def __init__(self, codigo, nome, descricao, valor):
        self.codigo = codigo
        self.nome = nome
        self.descricao = descricao
        self.valor = valor


lista_produtos = [Produto(1, "test1", "description", 10), Produto(2, "test2", "description", 20), Produto(3, "test3", "description", 30)]


def gravar():  # <-- AND THIS IS THE FUNCTION THAT SAVES THE LIST PRODUCTS IN JSON
    global contador_sinal
    print("=" * 50)
    print(" ", "\nGravando Arquivo Externo dos Produtos...")
    data = [item.__dict__ if type(item) == Produto else item for item in lista_produtos] # <-- NON PRODUTO ITEMS ARE EXPECTED TO BE A DICT
    with open('data.json', 'w') as fp:
        json.dump(data, fp)
    print(" ")
    print("\033[32mARQUIVO SALVO COM SUCESSO!\033[0m\n")


def main():
    gravar()


if __name__ == "__main__":
    main()

Does this give the result you want?

CodePudding user response：

It is not possible to have valid JSON, append more data and get a still valid JSON. That is the problem of saving each dict (object in JSON speak) separately in a loop.

There are few ways to get around it:

Create a list of dicts in memory and save it as JSON just once.
Just for completness, because it is a hack. Add [ and , and ] characters at the beginning, as a separator and at the end of file respectively, to create a JSON list.
My recommendation: Change your file format from JSON to newline delimited JSON. In the code posted all you need to change is to remove the indent=4 parameter. Of course, the program reading such file must read it line-by-line and load each JSON line separately. That should be no problem.