Repository files navigation

This is the elixir implementation of Algolia search API, it is purely functional

Add to your dependencies

defpdepsdo[{:algolia,"~> 0.8.0"}]end

(Pre-Elixir-1.4) Add :algolia to your applications

defapplicationdo[applications:[:algolia]]end

ALGOLIA_APPLICATION_ID=YOUR_APPLICATION_IDALGOLIA_API_KEY=YOUR_API_KEY

config :algolia,  application_id: YOUR_APPLICATION_ID,  api_key: YOUR_API_KEY

NOTE: You must use ADMIN API_KEY instead of SEARCH API_KEY to enable write access

You don't need to initiate an index with this client unlike other OO Algolia clients.However, Most of the client search/write functions all use the syntax

operation(index, args....)

So you can easy emulate the index.function() syntax using piping

"my_index" |> operation(args)

All functions are serialized into maps before returning these responses

• {:ok, response}
• {:error, error_code, response}
• {:error, "Cannot connect to Algolia"}: The client implements retrystrategy on all Algolia hosts with increasing timeout, It should onlyreturn this error when it has tried all 4 hosts.More Details here.

"my_index"|>search("some query")

With Options

"my_index"|>search("some query",[attributesToRetrieve:"firstname",hitsPerPage:20])

See all available search optionshere

multi([%{index_name=>"my_index1",query:"search query"},%{index_name=>"my_index2",query:"another query",hitsPerPage:3,},%{index_name=>"my_index3",query:"3rd query",tagFilters:"promotion"}])

You can specify a strategy to optimize your multiple queries

• :none: Execute the sequence of queries until the end.
• stop_if_enough_matches: Execute the sequence of queries until the number of hits is reached by the sum of hits.

multi([query1,query2],strategy::stop_if_enough_matches)

Allsave_* operations will overrides the object at the objectID

Save a single object to index without specifying objectID, must have objectIDinside object, or use theid_attribute option (see below)

"my_index"|>save_object(%{objectID:"1"})

Save a single object with a given objectID

"my_index"|>save_object(%{title:"hello"},"12345")

Save multiple objects to an index

"my_index"|>save_objects([%{objectID:"1"},%{objectID:"2"}])

Partially updates a single object

"my_index"|>partial_update_object(%{title:"hello"},"12345")

Update multiple objects, must have objectID in each object, or use theid_attribute option (see below)

"my_index"|>partial_update_objects([%{objectID:"1"},%{objectID:"2"}])

Partial update by default creates a new object if an object does not exist at theobjectID, you can turn this off by passingfalse to the:upsert? option

"my_index"|>partial_update_object(%{title:"hello"},"12345",upsert?:false)"my_index"|>partial_update_objects([%{id:"1"},%{id:"2"}],id_attribute::id,upsert?:false)

All write functions such assave_object andpartial_update_object comes with anid_attribute option that lets the you specifying an objectID from an existing field in the object, so you do nothave to generate it yourself

"my_index"|>save_object(%{id:"2"},id_attribute::id)

It also works for batch operations, such assave_objects andpartial_update_objects

"my_index"|>save_objects([%{id:"1"},%{id:"2"}],id_attribute::id)

However, this cannot be used together with an ID specifying argument together

"my_index"|>save_object(%{id:"1234"},"1234",id_attribute::id)>Error

All write operations can be waited on by simply piping the response into wait/1

"my_index"|>save_object(%{id:"123"})|>wait

Since the client polls the server to check for publishing status,You can specify a time between each tick of the poll, the default is 1000 ms

"my_index"|>save_object(%{id:"123"})|>wait(2_000)

You can also use the underlying wait_task function explicitly

{:ok,%{"taskID"=>task_id,"indexName"=>index}}="my_index"|>save_object(%{id:"123"}    wait(index,task_id)

or with option

wait(index,task_id,2_000)

list_indexes()

Moves an index to a new one

move_index(source_index,destination_index)

Copies an index to a new one

copy_index(source_index,destination_index)

clear_index(index)

get_settings(index)

Example response

{:ok,%{"minWordSizefor1Typo"=>4,"minWordSizefor2Typos"=>8,"hitsPerPage"=>20,"attributesToIndex"=>nil,"attributesToRetrieve"=>nil,"attributesToSnippet"=>nil,"attributesToHighlight"=>nil,"ranking"=>["typo","geo","words","proximity","attribute","exact","custom"],"customRanking"=>nil,"separatorsToIndex"=>"","queryType"=>"prefixAll"}}

set_settings(index,%{"hitsPerPage"=>20})>%{"updatedAt"=>"2013-08-21T13:20:18.960Z","taskID"=>10210332."indexName"=>"my_index"}

• get_object
• save_object
• save_objects
• update_object
• partial_update_object
• partial_update_objects
• delete_object
• delete_objects
• list_indexes
• clear_index
• wait_task
• wait (convenience function for piping response into wait_task)
• set_settings
• get_settings
• list_user_keys
• get_user_key
• add_user_key
• update_user_key
• delete_user_key