Search

ODE framework enables applications to register into the search engine module (see Search Engine)

  • Search engine uses Publish-subscribe and Point-to-Point model. It’s backed by the Vertx Event Bus.

  • Publish-subscribe model is used to push event to registered applications. While the Point-to-Point model allows applications to send their search’s result.

  • Search engine API uses a fulltext model .

  • An API and a defaut implementation are provided for standard search behaviors. You can also override it to offer alternative search’s behavior.

Note
 Search engine module only support Document (MongoDB) and Relational (PostgreSQL) storage. File-type storage is not yet supported.

Subscribe Search Event

In your Application BaseServer (that extends a Verticle) add the following piece of code.

// Subscribe to events published for searching
if (config.getBoolean("searching-event", true)) {
	setSearchingEvents(new $Application$SearchingEvents(new $SearchService()));
}
Note
 $Application$ by convention is the name of the verticule. This class must implement the SearchingEvents interface. $SearchService is a class that implements the SearchService interface

examples:

  1. Mongo example: See

  2. Postgre example: See

Searching Events

See the contract service to SearchingEvents

search contract:

Args Type Description

userId

String

User id

groupIds

List<String>

groups id

returnFields

List<String>

Returned fields by the backend query

searchWords

List<String>

Words to search

page

Integer

current page

limit

Integer

Number of occurrences to return

handler

Handler<Either<String, JsonArray>>

Callback with the result
Warning
 Looks at the Result chapter for the returned values convention

Document

Index

Add a JS script to "/deployment/$applicationName/migration/x.y.z/addTextIndex.js"

Note
 $applicationName is the name of your app. x.y.z is the version of your app. 
db.$collection.createIndex({ "$field1": "text", "$fieldN": "text"});
Note
 $collection is the name of mongo collection, and $field1, $fieldN are attributes to be indexed

Implementation

Create your own $Application$SearchingEvents class with MongoDbSearchService instance. See example : ShareBigFilesSearchingEvents

Relational data

Index

Add a SQL script to "src/main/resources/sql/xxx-$applicationName.sql"

Note
 $applicationName is the name of your app. xxx is the number version of your script. 
 -- specific configuration to language natively create vectors without accents (one configuration per supported language)
 CREATE TEXT SEARCH CONFIGURATION  fr ( COPY = french ) ;
 ALTER TEXT SEARCH CONFIGURATION fr ALTER MAPPING
 FOR hword, hword_part, word WITH unaccent, french_stem;

 -- With evolution of create services, it is possible to manage multiple languages to search for the same application instance
 ALTER TABLE $schema.$resourceTable ADD language VARCHAR(2) NOT NULL DEFAULT('fr');

 ALTER TABLE $schema.$resourceTable ADD COLUMN text_searchable tsvector;

 -- Only if data alreay exists
 UPDATE $schema.$resourceTable SET text_searchable =
      to_tsvector(language::regconfig, coalesce(title,'') || ' ' || coalesce(regexp_replace(content, '<[^>]*>',' ','g'),'')    );

 CREATE INDEX textsearch_idx ON $schema.$resourceTable USING gin(text_searchable);

 CREATE FUNCTION $schema.text_searchable_trigger() RETURNS trigger AS $$
 begin
   new.text_searchable := to_tsvector(new.language::regconfig, coalesce(new.title,'') || ' ' || coalesce(regexp_replace(new.content, '<[^>]*>',' ','g'),''));
   return new;
 end
 $$ LANGUAGE plpgsql;

 CREATE TRIGGER tsvector_update_trigger BEFORE INSERT OR UPDATE
     ON $schema.$resourceTable FOR EACH ROW EXECUTE PROCEDURE $schema.text_searchable_trigger();
Note
 $schema is schema of the application and $resourceTable is replaced by the name of the table that contains the searchable resources
Warning
 text_searchable_trigger function must use fields of your choice. In this case, indexed field are title and html content

Implementation

Create your own $Application$SearchingEvents class with SqlSearchService instance. See example : ActualitesSearchingEvents

Result

The result must be formatted according to the following convention

Json result example:

[
    {
      "title":"resource title",
      "description":"resource description",
      "modified": "modified date",
      "ownerDisplayName": "owner fullname",
      "ownerId": "owner uid",
      "url": "resource link"
    }, ...
]

Reference

https://docs.mongodb.com/manual/reference/operator/query/text https://www.postgresql.org/docs/9.5/static/textsearch.html

results matching ""

    No results matching ""