databricks 1.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: c27151990dcc087e396f387f45f6909f54bb6eb8529b8835706ce84d3ac399ab
+  data.tar.gz: 0ddfaf4cd06981894bddc1ff2ede2bf2372aa6d08521699458502cdf4b83f308
+SHA512:
+  metadata.gz: 747bf19db30cc24740b101aaacd14d22eeb7051995b7d1707917835a34fb90bc096304d2e11c18ea4c04c28049e5ead90961f1fcdb319642bbd54409feb580cf
+  data.tar.gz: dae0b046280172f770df37ce37440d0311f5e65896fe35209e3d0e900c4961d1a8396d195b570dac122c23b373aeb0be364f9c94178d54da888fab20dd81a819

data/lib/databricks.rb

@@ -0,0 +1,17 @@
+require 'databricks/domains/root'
+require 'databricks/resource'
+
+module Databricks
+
+  # Get an API connector based on a Databricks URL and API token
+  #
+  # Parameters::
+  # * *host* (String): Host to connect to
+  # * *token* (String): Token to be used in th API
+  # Result::
+  # * Domains::Root: The root domain of the API
+  def self.api(host, token)
+    Domains::Root.new(Resource.new(host, token))
+  end
+
+end

data/lib/databricks/domain.rb

@@ -0,0 +1,47 @@
+module Databricks
+
+  # Encapsulate a part of the API for better organization
+  class Domain
+
+    # Declare sub-domain accessors in the current domain.
+    # This will make sure the current domain has methods named after the sub-domain identifiers that will instantiate the corresponding domains at will.
+    #
+    # Parameters::
+    # * *domains* (Array<Symbol>): Domains to instantiate
+    def self.sub_domains(*domains)
+      domains.flatten.each do |domain|
+        self.define_method(domain) do
+          sub_domain(domain)
+        end
+      end
+    end
+
+    # Instantiate a sub-domain.
+    # Keep a cache of it.
+    #
+    # Parameters::
+    # * *domain* (Symbol): Sub-domain identifier.
+    # Result::
+    # * Domain: Corresponding sub-domain
+    def sub_domain(domain)
+      unless @sub_domains.key?(domain)
+        require "#{__dir__}/domains/#{domain}.rb"
+        @sub_domains[domain] = Domains.const_get(domain.to_s.split('_').collect(&:capitalize).join.to_sym).new(@resource)
+      end
+      @sub_domains[domain]
+    end
+
+    # Constructor
+    #
+    # Parameters::
+    # * *resource* (Resource): Resource handling API calls
+    def initialize(resource)
+      @resource = resource
+      # Keep a map of sub-domains instantiated, per domain identifier.
+      # Hash< Symbol, Domain >
+      @sub_domains = {}
+    end
+
+  end
+
+end

data/lib/databricks/domains/dbfs.rb

@@ -0,0 +1,46 @@
+require 'databricks/domain'
+
+module Databricks
+
+  module Domains
+
+    # Provide the DBFS API
+    # cf. https://docs.databricks.com/dev-tools/api/latest/dbfs.html
+    class Dbfs < Domain
+
+      # List a path
+      #
+      # Parameters::
+      # * *path* (String): Path to be listed
+      # Result::
+      # * Array<String>: List of DBFS paths
+      def list(path)
+        @resource.get_json(
+          'dbfs/list',
+          {
+            path: path
+          }
+        )['files'].map { |file_info| file_info['path'] }
+      end
+
+      # Put a new file
+      #
+      # Parameters::
+      # * *path* (String): Path to the file to create
+      # * *local_file* (String): Path to the local file to put
+      def put(path, local_file)
+        @resource.post(
+          'dbfs/put',
+          {
+            path: path,
+            contents: File.new(local_file, 'rb'),
+            overwrite: true
+          }
+        )
+      end
+
+    end
+
+  end
+
+end

data/lib/databricks/domains/jobs.rb

@@ -0,0 +1,31 @@
+require 'databricks/domain'
+
+module Databricks
+
+  module Domains
+
+    # Provide the Jobs API
+    # cf. https://docs.databricks.com/dev-tools/api/latest/jobs.html
+    class Jobs < Domain
+
+      # List a path
+      #
+      # Result::
+      # * Array<Hash>: List of jobs information
+      def list
+        @resource.get_json('jobs/list')['jobs']
+      end
+
+      # Create a new job
+      #
+      # Parameters::
+      # * *settings* (Hash<Symbol,Object>): Settings to create the job
+      def create(**settings)
+        @resource.post_json('jobs/create', settings)
+      end
+
+    end
+
+  end
+
+end

data/lib/databricks/domains/root.rb

@@ -0,0 +1,20 @@
+require 'databricks/domain'
+
+module Databricks
+
+  module Domains
+
+    # API entry point
+    # cf. https://docs.databricks.com/dev-tools/api/latest/index.html
+    class Root < Domain
+
+      sub_domains %i[
+        dbfs
+        jobs
+      ]
+
+    end
+
+  end
+
+end

data/lib/databricks/resource.rb

@@ -0,0 +1,79 @@
+require 'json'
+require 'rest-client'
+
+module Databricks
+
+  # Underlying resource making API calls
+  class Resource
+
+    # Constructor
+    #
+    # Parameters::
+    # * *host* (String): Host to connect to
+    # * *token* (String): Token to be used in th API
+    def initialize(host, token)
+      @host = host
+      @token = token
+    end
+
+    # Issue a GET request to the API with JSON payload
+    #
+    # Parameters::
+    # * *api_path* (String): API path to query
+    # * *json_payload* (Object): JSON payload to include in the query [default = {}]
+    # Result::
+    # * Object: JSON result
+    def get_json(api_path, json_payload = {})
+      JSON.parse(
+        RestClient::Request.execute(
+          method: :get,
+          url: "#{@host}/api/2.0/#{api_path}",
+          payload: json_payload.to_json,
+          headers: {
+            Authorization: "Bearer #{@token}",
+            'Content-Type': 'application/json'
+          }
+        ).body
+      )
+    end
+
+    # Issue a POST request to the API with JSON payload
+    #
+    # Parameters::
+    # * *api_path* (String): API path to query
+    # * *json_payload* (Object): JSON payload to include in the query [default = {}]
+    # Result::
+    # * Object: JSON result
+    def post_json(api_path, json_payload = {})
+      JSON.parse(
+        RestClient::Request.execute(
+          method: :post,
+          url: "#{@host}/api/2.0/#{api_path}",
+          payload: json_payload.to_json,
+          headers: {
+            Authorization: "Bearer #{@token}",
+            'Content-Type': 'application/json'
+          }
+        ).body
+      )
+    end
+
+    # Issue a POST request to the API with multipart form data payload
+    #
+    # Parameters::
+    # * *api_path* (String): API path to query
+    # * *form_payload* (Hash): Form payload to include in the query [default = {}]
+    def post(api_path, form_payload = {})
+      RestClient::Request.execute(
+        method: :post,
+        url: "#{@host}/api/2.0/#{api_path}",
+        payload: form_payload.merge(multipart: true),
+        headers: {
+          Authorization: "Bearer #{@token}"
+        }
+      )
+    end
+
+  end
+
+end

data/lib/databricks/version.rb

@@ -0,0 +1,5 @@
+module Databricks
+
+  VERSION = '1.0.0'
+
+end

metadata

@@ -0,0 +1,106 @@
+--- !ruby/object:Gem::Specification
+name: databricks
+version: !ruby/object:Gem::Version
+  version: 1.0.0
+platform: ruby
+authors:
+- Muriel Salvan
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2021-01-20 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rest-client
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.1'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.1'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.10'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.10'
+- !ruby/object:Gem::Dependency
+  name: webmock
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.11'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.11'
+- !ruby/object:Gem::Dependency
+  name: sem_ver_components
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.0'
+description: Access the Databricks API using the simple Ruby way
+email:
+- muriel@x-aeon.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/databricks.rb
+- lib/databricks/domain.rb
+- lib/databricks/domains/dbfs.rb
+- lib/databricks/domains/jobs.rb
+- lib/databricks/domains/root.rb
+- lib/databricks/resource.rb
+- lib/databricks/version.rb
+homepage: https://github.com/Muriel-Salvan/databricks
+licenses:
+- BSD-3-Clause
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.2.3
+signing_key: 
+specification_version: 4
+summary: Rubygem wrapping the Databricks REST API
+test_files: []