• Separate content-types in different indexes (ex: usernames, article titles...)
• Can index multiple terms in the same string
• Attach a payload/document to the indexed string
• Scoring mechanism :
  • by indexing many times the same document
  • by indexing a score
• Indexation protected by a basic authentication

• Sorted results (by score)
• Case insensitive
• Pretty mode (add a tag around the terms)
• Return a payload
• Go + Redis == Very fast (Elasticsearch run out of time)

• "Incremental search API"
• "Autocompletion API"
• "As you type suggestions API"
• "Instant results as you type API"

Many values can be set by the environment :

• redis.hostname
• redis.port
• redis.port
• http.port
• http.basic.username
• http.basic.password

Example: """js { "redis": { "hostname": "$REDIS_HOSTNAME", "port": "$REDIS_PORT", "password": "$REDIS_PASSWORD" }, "http": { "port": "$API_PORT", "basic": { "username": "$API_BASIC_USERNAME", "password": "$API_BASIC_PASSWORD" } }, "index": { "minSizeToken": 2, "prefix": "users" } } """

Build image : """sh docker build -t autocompletion_api . """

Run container : """sh $ docker run -d -p 6379:6379 --name redis redis $ docker run -d -p 8080:8080 --link redis:redis autocompletion_api """

Update config.json setting redis hostname to "localhost"

"""sh $ make vendor_get $ make run """

PUT /index/#content-type/store

• content-type: "username", "article-title" or whatever...
• request body (json):
  • id (string): ID of the document to index. If used multiple time, the score increase of 1.
  • terms ([]string): You can index a string with many world. Example, my twitter can be indexed with ["@SamuelBerthe", "Samuel", Berthe"].
  • payload (json, optional): whatever you want to store with the string.
  • score (float, optional, default=1): the search result can be sorted by score

"""js { results: [...] } """

"""sh curl -X PUT "http://autocompletion:autocompletion@127.0.0.1:8080/index/twitter_accounts/store" -H "Content-Type: application/json" -d '{"id": "0001", "terms": ["@SamuelBerthe", "Samuel", "BERTHE"], "label": "Just put online a new side project: http://pipe.grep.so. Try it out !", "Payload": {"RT": 42, "favorites": 1337}, "score":42.0}' """

GET /index/#content-type/search?seed,score,pretty,payload

• content-type: "username", "article-title" or whatever...
• request query:
  • seed (string): Term to autocomplete (can include URL-encoded spaces)
  • ordered (bool, optional, default=true): return result ordered by score
  • score (bool, optional, default=false): for each result, return the score
  • pretty (bool, optional, default=false): for each label, add tags around the terms that match with the seed parameter
  • payload (bool, optional, default=false): for each result, return the payload, if provided

"""js { results: [...] } """

"""sh curl -X GET "http://127.0.0.1:8080/index/twitter_accounts/search?seed=sam%20ber&pretty=true&payload=true&score=true" """

GET /index/#content-type/#id?payload

• content-type: "username", "article-title" or whatever...
• id: content ID to retrieve
• request query:
  • payload (bool, optional, default=true): return the payload, if provided

"""js { doc: {...} } """

"""sh curl -X GET "http://127.0.0.1:8080/index/twitter_accounts/00042 """

• Add a tokenizer
• Add a english anayliser