A simple command line application to interactively search Zendesk ticket, user and organization data supplied in json format.
The code is written in Scala and requires
- JDK 11
- Scala 2.13
To build run
sbt assembly
Unit tests can be run via
sbt test
and integration tests via
sbt it:test
Note that the integration tests include a performance test that can run for quite a long time.
To run the app use bin/zendesk-search.sh or bin/zendesk-search.bat from the project directory
bin/zendesk-search.sh --help will display the command line options
When the app runs it will display help with some examples
- Data schemas are the same as the sample file provided. No need to cater for data arriving in a different format
- Ids for org and user are always integers & don't exceed integer max value of 2147483647
- Ids are unique across data set
- Example data is representative of the required and optional fields
- Fields that are collections (e.g tags) will be searched by the elements of the collection i.e. a tag field will match if any tag in the set is the same as the search term
-
To speed up the search, all the fields of the entities are indexed. This greatly increases speed but also increases memory usage
-
There is a limit set on the number of results that will be shown with their related entities (e.g. the users & tickets of an org).
This is to prevent the search time scaling linearly when the search result set is large -
All fields are searched via exact match
- If data does not include all the required fields then the app will error while parsing the file.
This can be fixed by changing the appropriate field in the case class to Option[X] and updating the get function in the corresponding SearchableField (also should update the output in the ConsoleWriter class)
-
Better error handling of file parsing errors
-
Support fuzzy matching on some types of fields (e.g. datetimes could match on just the date part)