MongoDB: Perform a text-search in a document field (using high-level API)

问题

It may be related to this question Basic GROUP BY statement using OPA MongoDB high level API.

I want to be able to retrieve a list of documents which "name" field value contains a given string.

Here's my documents list :

{name: "Charles-Hugo"}, {name: "Jean Pierre"}, {name: "Pierre Dupont"},

I want to be able to only retrieve documents which name contains the "Pierre" string: Jean Pierre, and Pierre Dupont.

I know this isn't possible with the MongoDB high-level API. I've looked in the low-level API functions but I don't know what's the easiest way to retrieve my documents in safe Opa type.

Also I'd like to add skip/limit options to my query.

Any idea ?

回答1:

The DbGen automation mechanism in Opa has support for this:

DbSet.iterator(/path/data[name =~ pattern])

回答2:

As @Henri pointed out there is regular expression searching support in Opa since commit [enhance] DbGen: add case insensitive regex operator =~ what is very nice.

Mind that it is using $regex operator, not the full-text index and it may result with some performance loss :( As MongoDB documentation says $regex operator uses indexes in limited way - only for prefix search: pattern ^Jean. Searching for Jean anywhere in text will require full scan.

Personally, I am using full-text index feature of Mongo with Opa's "low-level" API for the $text command like this:

  function list({float score, Article.id id}) textSearch(string query) {
    function onfailure(failure) {
      cat.error("textSearch({{~query}}): {failure}");
      [];
    }
    function onsuccess(success) {
      function aux(~{name,value}) {
        name == "results";
      }
      match (List.filter(aux, success)) {
      | [] :
        // `results` field not found - error
        onfailure(success);
      | results:
        cat.debug("textSearch({~{query}}): {results}");
        function ({~score, obj: ~{id}}) {
          ~{score, id}
        }
        |> List.map(_, Bson.doc2opa(results) ? []);
      }
    }

    opts = [H.str("search", query), H.doc("project", [H.i32("_id",0), H.i32("id",1)])];
    //  { search: query, project: {_id:0, id:1}, }
    //  |> Bson.opa2doc
    outcome = MongoCommands.simple_str_command_opts(ll_db, db_name, "text", coll_name, opts);
    MongoCommon.outcome_map(outcome, onsuccess, onfailure)
  }

Feature is available in Mongo since 2.4 as experimental (you have to turn it on by special configuration option) and in 2.6 as stable (turned on by default).

来源：https://stackoverflow.com/questions/26717851/mongodb-perform-a-text-search-in-a-document-field-using-high-level-api

标签

mongodb

Opa