Text index required for $text query

Moralis General
#1

I want to seach an item in db by text and implement this in pipeline (because I need to join tables).
My function is

Moralis.Cloud.define("test", async (request) => {
  const query = new Moralis.Query("itemCreatedData");
  const pipeline = [ 
  	{ match: { $text: { $search: "queen" } } }
  ]
  const response = await query.aggregate(pipeline);
  return response
})

But I get an error

{
  "code": 141,
  "error": "text index required for $text query"
}

Does the $text works in moralis mongoDB or the only option I have is to use query.fullText

#2

There is another option to search a substring, directly without pipeline. I think that it is with march

query.matches

#3

There is the option only without pipeline?
This is not I’d like, since I need in pipeline lookup to join two tables.
I want that my function finds the most relevant object by text and then it joins with another table.

#4

I don’t know of an option with pipeline, it may be there an option too.
the syntax should be specific to mongo db

#5

Okay, then maybe someone else could help because I don’t understand.
I use $text for searching by text in db how it’s said in documentation https://www.mongodb.com/docs/v3.2/text-search/ .
But it just throws an error "error": "text index required for $text query"

#6

Have you created a text index? Text Indexes — MongoDB Manual

#7

Oh, No I don’t. How can I create a text index in moralis table?
In documentation I found that I can do it only when I work in node.js

#8

You’ll have to do it when connecting to your MongoDB instance directly.

Syntax according to the docs should be:

const table = client.db('admin').collection('itemCreatedData');
const result = await table.createIndex({ fieldhere: 'text' });
console.log(result);
#9

Excuse me, but to connect to MongoDB I have to provide in client MONGO_HOST and MONGO_PORT
but I see only IP for MONGO_HOST. Where is MONGO_PORT

image
image1507×1273 82.5 KB

#10

The port is after the :. E.g. ipaddress:12345, the port is 12345.

#11

Okay, I’m now on a half way but now I can’t get data. Before createIndex and wanted to check if everything works and for that I just asked get all data in table

const MONGO_HOST = "xxx"
const MONGO_PORT = "xxx"

const uri = `mongodb://${MONGO_HOST}:${MONGO_PORT}`
const client = new MongoClient(uri)

await client.connect()
const database = client.db("admin")

const collection = database.collection("itemCreatedData")
const data = await collection.find({}).toArray() //  = []

My code works but I just get empty data. Maybe it’s a problem with permissions but I checked and read is enabled

image
image926×101 7.56 KB

#12

try with parse as db there instead of admin

#13

Finally it works correctly. Ty everyone for help!!

1 Like
#14

can you share your solution, what index did you add?

#15 
await client.connect()

const database = client.db("parse")
const collection = database.collection("itemCreatedData")

await collection.createIndex({ title: "text" })

const datas = await collection.aggregate([{ $match: { $text: { $search: "queen" } } }]).toArray()
1 Like
How can I use query.fullText()?
How can I use query.fullText()?
#16

You can also achieve it by writing a regex like this.
PS: The example below is meant to check if the username starts with request.params.username

const pipeline = [
        {
            match: {
                username: { $regex: `^${request.params.username}` },
            },
        }
]
1 Like
#17

Hm, ye, your implementation is much simpler than mine.
But it has drawback that it doesn’t ignore case sensitive. Maybe it can through /i/ but I can’t seem to do it

#18

Regex worked decently for me. Index caused pain. Interested to see how well it holds up with large data sets.

Out of curiosity, what’s the best way to $or the regex searches for multiple fields. I tried match: $or: [] and it didn’t like it. :thinking:

UPDATE:
Never mind, got it. It didn’t like string interpolation for my regex+options, so I just used $options for the options.