Search Addon questions

Support
#1

I am using search addon with opensearch. I have following quesries-

  1. quartz job - EnqueueingSessionProcessing, what is use of it?
  2. why index processing job automatically get paused after sometime.
  3. How can I format the result set, like sequencing of attributes?
  4. How to open edit view in modal when clicked to resultset. How to pass search string to this edit view?
  5. How to rebuilt index from scratch.
  6. How can I hide {…} button in search field so as user can’t select the stratgy i.e entity. (I want to provide different views for different entities)

Thanks in advance.

#3

Other things are clear only these two are not clear-

  1. How to open edit view in modal when clicked to resultset. How to pass search string to this edit view so as to highlight the search text in textbox?
  2. Can I do indexing in one project and show sear results in another project? (second project has the database as addional datasource?

Thanks

#4

Hi Umesh,

Opening edit view from result set

  • How to open edit view in modal dialog. Currently edit view is opened in dialog only if entire result set view is opened in dialog (openMode = DIALOG on SearchField UI component). If result set view is opened in NAVIGATION mode edit views will be opened via navigation also.
    As a possible solution you can override result set view class (io.jmix.searchflowui.view.result.SearchResultsView) and implement method openEntityView the way you want (force open in dialog).
  • How to pass search string to this edit view so as to highlight the search text in textbox. I’m not sure I understand these statements. Result set view opens the default editor - it doesn’t know about search-specific things.

Also there is a feature of custom result handler. If you don’t want/can’t use the default result set view you can set the custom handler via SearchField#setSearchCompletedHandler.
This is a consumer of SearchCompletedEvent that contains SearchResult so you can render it (or do some process without rendering) the way you need.

Can I do indexing in one project and show search results in another project

Searcher uses data from index definitions (interfaces with @JmixEntitySearchIndex annotation) of current project - it allowed to search that it allowed to index. So if your “search” project doesn’t have them - you can’t search.
Maybe we will implement it later as a “read-only index definition” feature.

Regards,
Ivan

#5

Thanks @i.gavrilov for clarification.

For first item , i will try and comeback if there any issue.

For second item, how can i update indexes in second project (as data is being modified in first project). As the entity being indexed is available in additional datasource, initial automatic indexing is working. But when i modify entity in first project indexes are not getting updated as they are being done in second project.

Please suggest easiest method which can work automatically (once in day reindexing of all data by quarts job)

This can work as i know the entity i have updated in first project.

  • deleteByEntityId(Id entityId) - deletes entity instance from the index by a provided id.
  • indexByEntityId(Id entityId) - stores entity instance to index by a provided id.
  1. Why quarts job for indexing is getting paused automatically in second project after somw time (additional datastore), whereas it keeps running in first project.
    Thanks in advance.
#6

Hi.

To be able to search in some indexes you need to have index definition interfaces for each of them in second project - basically the same you have in the first project.

But that causes the second project will try to manage index structure and track changes of data. And that, as I understand, should be the task for the first project (indexer), second project (searcher) should work in read only mode as far as possible.

Addon provides few application properties that disable automatic management operations:

  • jmix.search.index-schema-management-strategy - set to NONE to disable management of index mapping. That means indexes should be created some external way (in your case - via the first project).
  • jmix.search.changed-entities-indexing-enabled - set to false to disable runtime tracking of changes in indexed entities. No more data will be added to the indexing queue.
  • jmix.search.enqueue-index-all-on-startup-index-recreation-enabled - set to false to disable automatic reindexing if index has been created/recreated.

Using these properties you can make you second project some kind of read only (operations in JMX console are still available).

About paused job: don’t have enough info about this specific case. What is the last/next fire time of their triggers?
And as I understand - your second project should not modify data in index, only search.

Regards,
Ivan