Mongodb Field Aliases

[Anoop Madhusudanan]

How to represent a field name alias in Mongodb JSON query? Is it supported in the C# driver? Thanks

Anoop Madhusudanan
Microsoft MVP, Client App Dev
http://amazedsaint.com

[craiggwilson]

Could you give an example of what you are talking about?  I could go a number of different ways with an answer.

[craiggwilson]

This is a cross-post from the mongodb-csharp forum. https://groups.google.com/d/topic/mongodb-csharp/wnhsxqvMShI/discussion

[Robert Stam]

Suppose you have a C# class and you are aliasing some of the field
names:

public class Employee {
public ObjectId Id { get; set; }
[BsonElement("nm")]
public string Name { get; set; }
}

then your documents are going to look like:

{
_id : ObjectId("4db8571e16203a6fccb7d5d9"),
nm : "John Doe"
}

So in your query you need to use the alias name:

var query = Query.EQ("nm", "John Doe");

Let me know if this is not what you were asking about.

Robert

[amazedsaint]

Robert,

Thank you for the reply.

I'm afraid we can't do it on the class level, as we don't have a typed
model and we are always serializing dictionaries. We are constructing
our queries dynamically and parse it to a BsonDocument query. How to
represent the alias in a json query level?

Thanks

[Nat]

If you have a limited set of field names you use, you can simply
define a class with string constants for aliasing.

[Robert Stam]

If you don't have a C# class and are just serializing dictionaries
then you don't have aliases, do you? The keys in the dictionary become
the element names in your BSON document (I'm assuming the dictionary
keys are strings).

So in this case when constructing the queries dynamically you would
use the keys from the dictionary as the names of the elements you are
querying.

Am I missing something?

[amazedsaint]

You are right. How ever, our purpose is to rename the fields at the time of selecting the fields back, for the purpose. For eg, if a dictionary is stored as {{"key1":"value"}{"key2":"value"}}, we need to fetch it back as {{"k1":"value"}{"k2":"value"}} while constructing the dynamic query via field aliases, with out committing the additional cost of renaming the dictionary fields manually after the fetch.

We are looking for a SQL "AS" equivalent in Mongodb JSON querying. For eg, what is the JSON query equivalent of SELECT column_name AS alias_name

FROM table_name?

Thanks in advance

[Sam Millman]

There is a JIRA somewhere for this (maybe some one who knows how to search the JIRA DB can find it? I suck at searching it) but unfortunately this functionality does not yet exist.

The only way is to code this directly into your application within your models (basically as most frameworks do).

"For eg, if a dictionary is stored as {{"key1":"value"}{"key2":"value"}}, we need to fetch it back as {{"k1":"value"}{"k2":"value"}} while constructing the dynamic query via field aliases, with out committing the additional cost of renaming the dictionary fields manually after the fetch."

Even with an AS keyword it would be hard to do that and maybe laggy since you have specified the structure of a subdocument there and if you go through, say, 10,000 subdocuments assigning this shortened name it wont be long before you get a massive speed impact due to the DB having to traverse this array and do this.

--
You received this message because you are subscribed to the Google Groups "mongodb-user" group.
To post to this group, send email to mongod...@googlegroups.com.
To unsubscribe from this group, send email to mongodb-user...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/mongodb-user?hl=en.

[Scott Hernandez]

Maybe you are thinking of this one?

https://jira.mongodb.org/browse/SERVER-863

Also, I find using google to search jira works well, with a site restriction. It also works well for searching the groups.

http://www.google.com/search?q=site:jira.mongodb.org+name+field+mapping

[Nat]

Sam,

That's https://jira.mongodb.org/browse/SERVER-863. Anyway, this kind
of thing sounds like a premature optimization and it's something that
MongoDB can handle this better in the future without having to do it
manually in the code or api layer.

On May 5, 4:16 pm, Sam Millman <sam.mill...@gmail.com> wrote:
> There is a JIRA somewhere for this (maybe some one who knows how to search
> the JIRA DB can find it? I suck at searching it) but unfortunately this
> functionality does not yet exist.
>
> The only way is to code this directly into your application within your
> models (basically as most frameworks do).
>
> "For eg, if a dictionary is stored as {{"key1":"value"}{"key2":"value"}}, we
> need to fetch it back as {{"k1":"value"}{"k2":"value"}} while constructing
> the dynamic query via field aliases, with out committing the additional cost
> of renaming the dictionary fields manually after the fetch."
>
> Even with an AS keyword it would be hard to do that and maybe laggy since
> you have specified the structure of a subdocument there and if you go
> through, say, 10,000 subdocuments assigning this shortened name it wont be
> long before you get a massive speed impact due to the DB having to traverse
> this array and do this.
>

[Anoop Madhusudanan]

Yep, I too think this is something that MongoDB should handle on the query level, rather than depending on driver/api to do the naming. The disadvantage is for static languages, you enforce that the objects should be typed while persisting/retrieving them to apply the attributes, and this restricts usage of untyped dictionaries when dealing with field aliases.

Thanks

[RubyTuesdayDONO]

another compelling use-case is aliasing field names to conform documents to some downstream "interface". for example, i can define a function:

Long increment(DBObject obj){
return ((Long)obj.get("target"))+1;
}

... and feed it with a query that aliases "mySpecialField" to "target", decoupling the definition of increment() from the definition of the documents i want to process.

https://jira.mongodb.org/browse/SERVER-863?focusedCommentId=94404#comment-94404

[Isuru Liyanage]

Since the answers are old, thought of updating it.

You can use the aggregate framework as below

db.mycollection.aggregate([{$project:{'alias':"$fieldName"}}])