So I have this Neo4j database and I made a query like this:

WITH something, collect(field1:aaaa,field2:bbbb) AS d

RETURN something, d

I wanted to get the results and show them with python-tabulate by creating a subtable of the ‘d’ fields and then inserting it into a main table. However, when making the subtable I realized something.

In theory, each record should’ve looked something like this

{
    something: "abcd"
    d: [{
        'field1':"abcd"
        'field2':abcd
    },
    {
        'field1':"abcd"
        'field2':abcd
    } ...etc
    ]
{

However, when I checked at the response in Python, it turns out that field1 and field2 are in the wrong order, like:

{
    something: abcd
    d: [{
        'field2':"abcd"
        'field1':abcd
    },
    {
        'field2':"abcd"
        'field1':abcd
    } ...etc
    ]
{

This kinda matters because I wanted to use python-tabulate to format the output, which should still be possible but I’d have to change the positions of field2 and field 1.

I used print() to look at the records to verify it’s not a thing of the tabulate library, which it’s not. The fields are already swapped from the response.

So my question is: Why is this happening? Does Cypher prioritize numbers over strings or something like that (I guess it’s worth mentioning, field 1 is a string and field 2 an int)? And if so, how do I prevent it?