Query on array on a nested document with pymongodb - python-3.x

Question

I am building a chat program
And in the backend, I have
chatcol.insert({ "chatid": "133235", "messages": [ {"from": "user3", "content": "Hello", "time": "20101213T172215"}, {"from": "user2", "content": "Hi", "time": "20101214T172215"} ] })
chatcol.insert({ "chatid": "134735", "messages": [ {"from": "user2", "content": "Hello", "time": "20101217T172215"}, {"from": "user12", "content": "Hi", "time": "20101213T172215"} ] })
Since there can be a lot of messages, I want the server only give the new message the client didnt see.
The client will give me a lastuptime, the time where they last logon.
I want to find chats with only the new message only.
How do I write such query?

Answer 1

I have considered time as ISODate()
Also, lastuptime given by the client wil be ISODate()
collection:
{
"_id" : ObjectId("5da31a63c35040d7fbf1db3c"),
"chatid" : "133235",
"messages" : [
{
"from" : "user3",
"content" : "Hello",
"time" : ISODate("2016-05-01T00:00:00Z")
},
{
"from" : "user2",
"content" : "Hi",
"time" : ISODate("2018-05-01T00:00:00Z")
}
]
}
{
"_id" : ObjectId("5da31ab2c35040d7fbf1db3d"),
"chatid" : "133235",
"messages" : [
{
"from" : "user2",
"content" : "Hello",
"time" : ISODate("2010-05-01T00:00:00Z")
},
{
"from" : "user12",
"content" : "Hi",
"time" : ISODate("2019-05-01T00:00:00Z")
}
]
}
There are two scenarios. Depending on your use-case, select the appropriate:
1. Fetch chats where any one of the messages has time >= lastuptime
db.chatcol.find({'messages.time': {"$gte": ISODate("2017-10-01T00:00:00.000Z")}})
Output:
{
"_id" : ObjectId("5da31a63c35040d7fbf1db3c"),
"chatid" : "133235",
"messages" : [
{
"from" : "user3",
"content" : "Hello",
"time" : ISODate("2016-05-01T00:00:00Z")
},
{
"from" : "user2",
"content" : "Hi",
"time" : ISODate("2018-05-01T00:00:00Z")
}
]
}
{
"_id" : ObjectId("5da31ab2c35040d7fbf1db3d"),
"chatid" : "133235",
"messages" : [
{
"from" : "user2",
"content" : "Hello",
"time" : ISODate("2010-05-01T00:00:00Z")
},
{
"from" : "user12",
"content" : "Hi",
"time" : ISODate("2019-05-01T00:00:00Z")
}
]
}
2. Fetch messages where time >= lastuptime
This one flattens the message array
db.chatcol.aggregate([
{ $unwind :'$messages'},
{ $match : {"messages.time": { "$gte": ISODate("2017-10-01T00:00:00.000Z") }}}
])
Output:
{
"_id" : ObjectId("5da31a63c35040d7fbf1db3c"),
"chatid" : "133235",
"messages" : {
"from" : "user2",
"content" : "Hi",
"time" : ISODate("2018-05-01T00:00:00Z")
}
}
{
"_id" : ObjectId("5da31ab2c35040d7fbf1db3d"),
"chatid" : "133235",
"messages" : {
"from" : "user12",
"content" : "Hi",
"time" : ISODate("2019-05-01T00:00:00Z")
}
}
Note: Map the $gte query according to your lastuptime