'How to do fast query on String dataType in MongoDB, where values are of type double
I have a field contractValue and other fields in a collection contract which is of type String . It basically holds double value like 1200 or 1500 but at some places it may contain value like $1200 or $1500.
Sample data from collection:
{ ..
..
contractValue: "1200", //This is the one stored as String. I need
// to perform range query over it
..
..
}
{ ..
..
contractValue: "$1500",
..
..
}
I have requirement where i need to fetch contracts based on contract values. Query can be like below:
{$and: [ {'contractValue': {$gt: 100}}, {'contractValue': {$lt: 1000 }}]}
This query is giving me wrong result. It is also giving me documents having contractValue like 1238999
Also I need to create indexes on contractValue
Is it possible to create index on contract value , so that I can efficiently make range query, so that whenever making any query, it will do < or > on Index and will fetch exact set of documents, rather than making change in schema?
How to handle values like $1200 in index, so index value just contain 1200 as integer rather than $1200
Solution 1:[1]
try this:
https://mongoplayground.net/p/TG3Y5tdh9aK
it assumes string data will be either a quoted number or a quoted number with "$" at the front
db.collection.aggregate([
{
$project: {
"newContractValue": {
"$convert": {
"input": "$contractValue",
"to": "double",
"onError": {
$toDouble: {
"$substr": [
"$contractValue",
1,
{
"$strLenCP": "$contractValue"
}
]
}
}
}
}
}
},
{
$match: {
$and: [
{
"newContractValue": {
$gt: 100
}
},
{
"newContractValue": {
$lt: 1000
}
}
]
}
}
])
This can be used to set a new contractValueNew field as number from the existing contractValue
db.getCollection('yourCollection').find({})
.forEach(function(record) {
if(record.contractValue.toString().substring(0, 1) == '$') {
record.contractValueNew = NumberInt(parseInt(record.contractValue.substring(1, record.contractValue.length)));
} else {
record.contractValueNew = NumberInt(parseInt(record.contractValue))
}
db.getCollection('yourCollection').save(record)
})
Solution 2:[2]
Try:
db.collection.find({'contractValue': {$gt: 100, $lt: 1000 }})
Create index on contractValue , but convert all values as numbers ...
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|---|
| Solution 1 | |
| Solution 2 |
