not_analyzed parameter is not applied when indexing fields about mongoosastic

[wedgemartin]

Also seeing this with latest ES 5.4 and mongoosastic 4.3.0. Just seems to igonre it, and digging into the source, I don't see where this would've ever worked.

[guumaster]

This is part of my working schema. With analized and not_analized indexed data.

MessageSchema = new mongoose.Schema({
  ts: {type: Date, default: Date.now, required: true},  
  details: {
    title: {type: String, trim: true, match: /\S+/, required: true, es_indexed: true },
  },
  scope: {
    nbhd: {
      city: {type: String, trim: true, es_indexed: true, es_index: 'not_analyzed'},
      zipCode: {type: String, trim: true, es_indexed: true, es_index: 'not_analyzed'},
   }
 }
}, {versionKey: false});

MessageSchema.plugin(mongoosastic);

var model = mongoose.model('Message', MessageSchema);

[guumaster]

All Elasticsearch properties must be prefixed with "es_". Try changing index: 'not_analyzed' to es_index: 'not_analyzed'

[hengkiardo]

@guumaster i also have been using like what you mention, but it not working.

[guumaster]

@Aredo I'm currently using this feature and works properly after adding es_index: 'not_analyzed'. Try recreating the index and check your elasticsearch.

[davidsandoz]

@guumaster Thanks for your answer. I tried your suggestion but it's not working for me. The parameter is still not recognized by Elasticsearch, even after completely erasing the index and recreating it.

@Aredo So, you are facing the same problem? Are you also manually entering the index to make it work or do you have another workaround?

[davidsandoz]

@guumaster Do you have an idea of where else I should try to look to fix that problem?

[guumaster]

I've created a simple test and was able to reproduce this problem. It seems that we need to check the "magic" behind mapping-generator.js and see why sometimes it doens't work as expected.

[chakravarthysm]

I am also facing the same problem, I've tried both index: 'not_analyzed' and es_index: 'not_analyzed' , but still its not working.

[guumaster]

I'm using it and it works, but it seems in some cases it doesn't. In my working example, I assure you that the correct name is es_index.

We have to compare and figure out why sometimes it doesn't add the not_analyzed property.

[hengkiardo]

@guumaster can you past to us how you define your Schema when using mongoosastic

[eferesen]

Greeting,
I am having a similar issue. Perhaps my mapping is incorrect. I am trying to index an array field with not_analyzed. I followed this example: https://www.elastic.co/guide/en/elasticsearch/guide/current/aggregations-and-analysis.html

The mapping of the field instStat after indexing is as follow:

"instStat": {
    "type": "string"
}

This should not be the case. It is analyzing and breaking up words. I want it to aggregate by the entire phrase. How can I get around this?

Here is the schema:

var UserSchema = new Schema({
  email: String,
  acadPlan: [],
  institutionCount: Number,
  courseCount: Number,
  lastPage: String,
  instStat:  {type: String,
             es_type: 'multi_field',
             es_fields : {
                name : { es_type: 'string', es_index: 'not_analyzed'}
            }
        },
  created: { type: Date, default: Date.now }
});

UserSchema.plugin(mongoosastic, {
  index : 'transfer',
  type : 'users'
});
var User = mongoose.model('User', UserSchema, 'users');

[prasadbhosale]

I was also facing this problem but finally I got a work around for it. You can add 'not_analyzed' in mongosastic library.

ex.
go node_modules->mongoosastic->lib->mapping-generator.js

In a function getCleanTree
find and replace this code

else if (value === String || value === Object || value === Date || value === Number || value === Boolean || value === Array) {
cleanTree[field] = {};
cleanTree[field].type = type;
if (cleanTree[field].type === 'string') {
cleanTree[field].fields = {
raw: {
type: 'string',
index: 'not_analyzed'
}
};
}
}

[francesconero]

I managed to get it working by calling createMapping after the schema definition. I don't know if this should be required or not, I can't figure it out from the docs.

[nschloe]

I've got the same issue, and tried to work around it the same way that @francesconero did: By calling createMapping after schema definition. I'm getting

[Error: [index_already_exists_exception] already exists, with { index=books }]

there though.

[michelem09]

Is it possible since ES 2.2 the multi_field in Mongoosastic doesn't work anymore due to this breaking change? https://www.elastic.co/guide/en/elasticsearch/reference/current/_multi_fields.html

[anton6]

Thanks guumaster, worked for me. I am also calling createMapping after schema definition by the way.

[jonnadams]

I seem to be getting this issue to even after calling createMapping. Does anyone have a full working code sample that I could take a look at?

[oussama]

I used lang:{type:String,es_indexed:true,es_index:'not_analyzed'}, if you previously created the index without not_analyzed, delete the index using curl -XDELETE 'http://localhost:9200/indexname/'

[dmr07]

Why is 'not analyzed' necessary? According to the documentation:

specifying es_indexed: true :

In this case only the name field will be indexed for searching.

Also, you guys are saying to use es_index but the npm docs is saying es_indexed, which is it?

[d4goxn]

@danielmeng07 es_indexed seems to be correct, not es_index.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale]

Closing this issue due to lack of activity. Thank you for your contributions.