retab 0.1.0 → 0.1.2

data/lib/retab/jobs.rb

@@ -2,7 +2,7 @@
 
 # This file is auto-generated by oagen. Do not edit.
 
-require 'json'
+require "json"
 
 module Retab
   class Jobs
@@ -32,12 +32,12 @@ module Retab
     # @param include_request [Boolean, nil] Whether to include the full original request body in each listed job.
     # @param include_response [Boolean, nil] Whether to include full response payloads in each listed job.
     # @param request_options [Hash] (see Retab::Types::RequestOptions)
-    # @return [Retab::Types::ListStruct<Retab::Job>]
+    # @return [Retab::PaginatedList<Retab::Job>]
     def list(
       before: nil,
       after: nil,
       limit: 20,
-      order: 'desc',
+      order: "desc",
       job_id: nil,
       status: nil,
       endpoint: nil,
@@ -57,35 +57,35 @@ module Retab
       request_options: {}
     )
       params = {
-        'before' => before,
-        'after' => after,
-        'limit' => limit,
-        'order' => order,
-        'job_id' => job_id,
-        'status' => status,
-        'endpoint' => endpoint,
-        'source' => source,
-        'project_id' => project_id,
-        'workflow_id' => workflow_id,
-        'workflow_block_id' => workflow_block_id,
-        'model' => model,
-        'filename_regex' => filename_regex,
-        'filename_contains' => filename_contains,
-        'document_type' => document_type,
-        'from_date' => from_date,
-        'to_date' => to_date,
-        'metadata' => metadata,
-        'include_request' => include_request,
-        'include_response' => include_response
+        "before" => before,
+        "after" => after,
+        "limit" => limit,
+        "order" => order,
+        "job_id" => job_id,
+        "status" => status,
+        "endpoint" => endpoint,
+        "source" => source,
+        "project_id" => project_id,
+        "workflow_id" => workflow_id,
+        "workflow_block_id" => workflow_block_id,
+        "model" => model,
+        "filename_regex" => filename_regex,
+        "filename_contains" => filename_contains,
+        "document_type" => document_type,
+        "from_date" => from_date,
+        "to_date" => to_date,
+        "metadata" => metadata,
+        "include_request" => include_request,
+        "include_response" => include_response
       }.compact
       response = @client.request(
         method: :get,
-        path: '/v1/jobs',
+        path: "/v1/jobs",
         auth: true,
         params: params,
         request_options: request_options
       )
-      fetch_next = ->(cursor) {
+      fetch_next = -> (cursor) {
         list(
           before: before,
           after: cursor,
@@ -110,10 +110,30 @@ module Retab
           request_options: request_options
         )
       }
-      Retab::Types::ListStruct.from_response(
+      Retab::PaginatedList.from_response(
         response,
         model: Retab::Job,
-        filters: { before: before, limit: limit, order: order, job_id: job_id, status: status, endpoint: endpoint, source: source, project_id: project_id, workflow_id: workflow_id, workflow_block_id: workflow_block_id, model: model, filename_regex: filename_regex, filename_contains: filename_contains, document_type: document_type, from_date: from_date, to_date: to_date, metadata: metadata, include_request: include_request, include_response: include_response },
+        filters: {
+          before: before,
+          limit: limit,
+          order: order,
+          job_id: job_id,
+          status: status,
+          endpoint: endpoint,
+          source: source,
+          project_id: project_id,
+          workflow_id: workflow_id,
+          workflow_block_id: workflow_block_id,
+          model: model,
+          filename_regex: filename_regex,
+          filename_contains: filename_contains,
+          document_type: document_type,
+          from_date: from_date,
+          to_date: to_date,
+          metadata: metadata,
+          include_request: include_request,
+          include_response: include_response
+        },
         fetch_next: fetch_next
       )
     end
@@ -131,19 +151,23 @@ module Retab
       request_options: {}
     )
       body = {
-        'endpoint' => endpoint,
-        'request' => request,
-        'metadata' => metadata
+        "endpoint" => endpoint,
+        "request" => request,
+        "metadata" => metadata
       }.compact
       response = @client.request(
         method: :post,
-        path: '/v1/jobs',
+        path: "/v1/jobs",
         auth: true,
         body: body,
         request_options: request_options
       )
       result = Retab::Job.new(response.body)
-      result.last_response = Retab::Types::ApiResponse.new(http_status: response.code.to_i, http_headers: response.each_header.to_h, request_id: response["x-request-id"])
+      result.last_response = Retab::Types::ApiResponse.new(
+        http_status: response.code.to_i,
+        http_headers: response.each_header.to_h,
+        request_id: response["x-request-id"]
+      )
       result
     end
 
@@ -160,8 +184,8 @@ module Retab
       request_options: {}
     )
       params = {
-        'include_request' => include_request,
-        'include_response' => include_response
+        "include_request" => include_request,
+        "include_response" => include_response
       }.compact
       response = @client.request(
         method: :get,
@@ -171,7 +195,11 @@ module Retab
         request_options: request_options
       )
       result = Retab::Job.new(response.body)
-      result.last_response = Retab::Types::ApiResponse.new(http_status: response.code.to_i, http_headers: response.each_header.to_h, request_id: response["x-request-id"])
+      result.last_response = Retab::Types::ApiResponse.new(
+        http_status: response.code.to_i,
+        http_headers: response.each_header.to_h,
+        request_id: response["x-request-id"]
+      )
       result
     end
 
@@ -190,7 +218,11 @@ module Retab
         request_options: request_options
       )
       result = Retab::Job.new(response.body)
-      result.last_response = Retab::Types::ApiResponse.new(http_status: response.code.to_i, http_headers: response.each_header.to_h, request_id: response["x-request-id"])
+      result.last_response = Retab::Types::ApiResponse.new(
+        http_status: response.code.to_i,
+        http_headers: response.each_header.to_h,
+        request_id: response["x-request-id"]
+      )
       result
     end
 
@@ -209,7 +241,11 @@ module Retab
         request_options: request_options
       )
       result = Retab::Job.new(response.body)
-      result.last_response = Retab::Types::ApiResponse.new(http_status: response.code.to_i, http_headers: response.each_header.to_h, request_id: response["x-request-id"])
+      result.last_response = Retab::Types::ApiResponse.new(
+        http_status: response.code.to_i,
+        http_headers: response.each_header.to_h,
+        request_id: response["x-request-id"]
+      )
       result
     end
   end

data/lib/retab/mime.rb

@@ -9,9 +9,10 @@
 # MimeData input handling from the Python (`prepare_mime_document`), Node
 # (`coerceMimeData`), Go (`InferMIMEData`), and Rust (`From<T>` impls) SDKs.
 
-require 'base64'
-require 'pathname'
-require 'uri'
+require "base64"
+require "pathname"
+require "stringio"
+require "uri"
 
 module Retab
   # Wire-shape MimeData. Mirrors the spec's `MIMEData` component schema.
@@ -29,7 +30,7 @@ module Retab
   #   - `String` containing a path on disk (falls back to file-read)
   #   - `Hash` with `:filename` + `:url` keys (already-built wire shape)
   class MimeData
-    attr_accessor :filename, :url
+    attr_accessor :filename, :url, :mime_type
 
     # Set by resource methods when MimeData is returned from the API (see
     # `Files#complete_upload`). Mirrors the spec-derived models' BaseModel
@@ -44,20 +45,24 @@ module Retab
     #     `MIMEData` field via `Retab::MimeData.new(hash[:document])`):
     #       Retab::MimeData.new({ filename: 'x.pdf', url: 'data:...' })
     #       Retab::MimeData.new('{"filename": "x.pdf", "url": "data:..."}')
-    def initialize(arg = nil, filename: nil, url: nil)
+    def initialize(arg = nil, filename: nil, url: nil, mime_type: nil)
       case arg
       when nil
         @filename = filename
         @url = url
+        @mime_type = mime_type
       when Hash
         sym = arg.transform_keys(&:to_sym)
         @filename = sym[:filename]
         @url = sym[:url]
+        @mime_type = sym[:mime_type] || sym[:mimeType]
       when String
-        require 'json'
+        require "json"
+
         parsed = JSON.parse(arg, symbolize_names: true)
         @filename = parsed[:filename]
         @url = parsed[:url]
+        @mime_type = parsed[:mime_type] || parsed[:mimeType]
       else
         raise ArgumentError, "cannot initialize Retab::MimeData from #{arg.class}"
       end
@@ -77,20 +82,33 @@ module Retab
       when String
         from_string(input)
       else
-        raise ArgumentError,
+        raise(
+          ArgumentError,
           "cannot coerce #{input.class} to Retab::MimeData; supply a Pathname, IO, String (URL or path), Hash, or MimeData"
+        )
+      end
+    end
+
+    def self.coerce_document_map(input)
+      unless input.respond_to?(:transform_values)
+        raise ArgumentError, "cannot coerce #{input.class} to a Retab::MimeData document map"
+      end
+
+      input.transform_values do |document|
+        file_ref_like?(document) ? document : coerce(document)
       end
     end
 
     def to_h
-      { filename: filename, url: url }
+      {filename: filename, url: url, mime_type: mime_type}
     end
 
     # JSON serialization. Body construction in resource methods passes the
     # MimeData object straight into `JSON.generate`; without `to_json` it
     # would fall back to `to_s` and ship the inspect string.
     def to_json(*args)
-      require 'json'
+      require "json"
+
       to_h.to_json(*args)
     end
 
@@ -101,61 +119,74 @@ module Retab
     private_class_method def self.from_pathname(path)
       bytes = path.binread
       filename = path.basename.to_s
-      new(filename: filename, url: bytes_to_data_url(bytes, filename))
+      mime_type = guess_mime_type(filename) || "application/octet-stream"
+      new(filename: filename, url: bytes_to_data_url(bytes, filename, mime: mime_type), mime_type: mime_type)
     end
 
     private_class_method def self.from_io(io)
       io.binmode if io.respond_to?(:binmode)
       bytes = io.read
-      filename = io.respond_to?(:path) && io.path ? ::File.basename(io.path) : 'document'
-      new(filename: filename, url: bytes_to_data_url(bytes, filename))
+      filename = io.respond_to?(:path) && io.path ? ::File.basename(io.path) : "document"
+      mime_type = guess_mime_type(filename) || "application/octet-stream"
+      new(filename: filename, url: bytes_to_data_url(bytes, filename, mime: mime_type), mime_type: mime_type)
     end
 
     private_class_method def self.from_hash(hash)
       sym = hash.transform_keys(&:to_sym)
-      new(filename: sym.fetch(:filename, 'document'), url: sym.fetch(:url))
+      new(
+        filename: sym.fetch(:filename, "document"),
+        url: sym.fetch(:url),
+        mime_type: sym[:mime_type] || sym[:mimeType]
+      )
+    end
+
+    private_class_method def self.file_ref_like?(input)
+      return true if defined?(Retab::FileRef) && input.is_a?(Retab::FileRef)
+      return false unless input.is_a?(Hash)
+
+      input.key?(:id) || input.key?("id")
     end
 
     private_class_method def self.from_string(str)
       if str.match?(%r{\A(https?|gs|data):}i)
         # It's a URL.
-        filename =
-          begin
-            uri = URI.parse(str)
-            ::File.basename(uri.path.to_s.empty? ? 'document' : uri.path)
-          rescue URI::InvalidURIError
-            'document'
-          end
-        new(filename: filename, url: str)
+        filename = begin
+          uri = URI.parse(str)
+          ::File.basename(uri.path.to_s.empty? ? "document" : uri.path)
+        rescue URI::InvalidURIError
+          "document"
+        end
+
+        new(filename: filename, url: str, mime_type: guess_mime_type(filename))
       elsif ::File.exist?(str)
         from_pathname(Pathname.new(str))
       else
         # Treat as raw text payload — base64-encode as a data URL.
-        new(filename: 'document', url: bytes_to_data_url(str, 'document', mime: 'text/plain'))
+        new(filename: "document", url: bytes_to_data_url(str, "document", mime: "text/plain"), mime_type: "text/plain")
       end
     end
 
     private_class_method def self.bytes_to_data_url(bytes, filename, mime: nil)
-      mime ||= guess_mime_type(filename) || 'application/octet-stream'
+      mime ||= guess_mime_type(filename) || "application/octet-stream"
       "data:#{mime};base64,#{Base64.strict_encode64(bytes)}"
     end
 
     EXTENSION_MIME_MAP = {
-      '.pdf' => 'application/pdf',
-      '.png' => 'image/png',
-      '.jpg' => 'image/jpeg',
-      '.jpeg' => 'image/jpeg',
-      '.gif' => 'image/gif',
-      '.webp' => 'image/webp',
-      '.txt' => 'text/plain',
-      '.csv' => 'text/csv',
-      '.json' => 'application/json',
-      '.xml' => 'application/xml',
-      '.html' => 'text/html',
-      '.md' => 'text/markdown',
-      '.docx' => 'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
-      '.xlsx' => 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
-      '.pptx' => 'application/vnd.openxmlformats-officedocument.presentationml.presentation',
+      ".pdf" => "application/pdf",
+      ".png" => "image/png",
+      ".jpg" => "image/jpeg",
+      ".jpeg" => "image/jpeg",
+      ".gif" => "image/gif",
+      ".webp" => "image/webp",
+      ".txt" => "text/plain",
+      ".csv" => "text/csv",
+      ".json" => "application/json",
+      ".xml" => "application/xml",
+      ".html" => "text/html",
+      ".md" => "text/markdown",
+      ".docx" => "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+      ".xlsx" => "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+      ".pptx" => "application/vnd.openxmlformats-officedocument.presentationml.presentation"
     }.freeze
 
     private_class_method def self.guess_mime_type(filename)

data/lib/retab/paginated_list.rb

@@ -0,0 +1,86 @@
+# frozen_string_literal: true
+
+# This file is auto-generated by oagen. Do not edit.
+
+# @oagen-ignore-file
+#
+# Cross-language paginated-list wrapper. See
+# .notes/blueprints/sdk-pagination-contract.md for the full contract.
+#
+# Every `def list` method on a Retab service returns one of these. The
+# class includes `Enumerable`, so callers can iterate every page
+# transparently with `#each`, `#map`, `#select`, etc. — the closure
+# captured at construction time fetches each subsequent page lazily.
+#
+# Pages built from JSON without a closure (e.g. fixtures rehydrated from
+# disk) iterate just the current `data` and stop, mirroring the WorkOS
+# AutoPaginatable behaviour from the upstream emitter family.
+
+require "json"
+
+module Retab
+  class PaginatedList
+    include Enumerable
+
+    attr_reader :data, :list_metadata, :last_response
+
+    # @param data [Array]
+    # @param list_metadata [Hash] envelope with `before` / `after` cursors
+    # @param last_response [Retab::Types::ApiResponse, nil]
+    # @param fetch_next [#call, nil] closure that takes the next cursor and
+    #   returns the next page; `nil` for terminal pages or rehydrated fixtures
+    def initialize(data: [], list_metadata: {}, last_response: nil, fetch_next: nil)
+      @data = data || []
+      @list_metadata = list_metadata || {}
+      @last_response = last_response
+      @fetch_next = fetch_next
+    end
+
+    # Iterate every page lazily. Each page's `data` is yielded item-by-item;
+    # when the cursor advances, the captured closure is invoked to fetch the
+    # next page and iteration continues. Terminal pages stop after the last
+    # in-memory item.
+    def each
+      return enum_for(:each) unless block_given?
+      page = self
+      loop do
+        page.data.each { |item| yield item }
+        break unless page.has_next? && @fetch_next
+        cursor = page.list_metadata[:after] || page.list_metadata["after"]
+        page = @fetch_next.call(cursor)
+        break unless page
+      end
+    end
+
+    def has_next?
+      cursor = @list_metadata.is_a?(Hash) ? (@list_metadata[:after] || @list_metadata["after"]) : nil
+      !cursor.nil? && !cursor.to_s.empty?
+    end
+
+    def next_page
+      return nil unless has_next? && @fetch_next
+      cursor = @list_metadata[:after] || @list_metadata["after"]
+      @fetch_next.call(cursor)
+    end
+
+    # Construct a PaginatedList from an HTTP response with
+    # `{data: [...], list_metadata: {...}}`.
+    def self.from_response(response, model: nil, filters: {}, fetch_next: nil)
+      _ = filters
+      body = response.respond_to?(:body) ? response.body : response
+      parsed = body.is_a?(String) ? JSON.parse(body) : (body || {})
+      raw_items = parsed["data"] || parsed[:data] || []
+      items = raw_items.map { |item| model ? model.new(item) : item }
+      meta = parsed["list_metadata"] || parsed[:list_metadata] || {}
+      api_response = if response.respond_to?(:code)
+        Types::ApiResponse.new(
+          http_status: response.code.to_i,
+          http_headers: response.respond_to?(:each_header) ? response.each_header.to_h : {},
+          request_id: response.respond_to?(:[]) ? response["x-request-id"] : nil
+        )
+      end
+
+      new(data: items, list_metadata: meta, last_response: api_response, fetch_next: fetch_next)
+    end
+  end
+end

data/lib/retab/parses/parse.rb

@@ -17,7 +17,7 @@ module Retab
       created_at: :created_at
     }.freeze
 
-    attr_accessor \
+    attr_accessor(
       :id,
       :file,
       :model,
@@ -27,8 +27,10 @@ module Retab
       :output,
       :usage,
       :created_at
+    )
 
     def initialize(json)
+      super()
       hash = self.class.normalize(json)
       @id = hash[:id]
       @file = hash[:file] ? Retab::FileRef.new(hash[:file]) : nil

data/lib/retab/parses/parse_output.rb

@@ -10,11 +10,13 @@ module Retab
       text: :text
     }.freeze
 
-    attr_accessor \
+    attr_accessor(
       :pages,
       :text
+    )
 
     def initialize(json)
+      super()
       hash = self.class.normalize(json)
       @pages = (hash[:pages] || [])
       @text = hash[:text]

data/lib/retab/parses/parse_request.rb

@@ -14,15 +14,17 @@ module Retab
       bust_cache: :bust_cache
     }.freeze
 
-    attr_accessor \
+    attr_accessor(
       :document,
       :model,
       :table_parsing_format,
       :image_resolution_dpi,
       :instructions,
       :bust_cache
+    )
 
     def initialize(json)
+      super()
       hash = self.class.normalize(json)
       @document = hash[:document] ? Retab::MimeData.new(hash[:document]) : nil
       @model = hash[:model]

data/lib/retab/parses.rb

@@ -2,7 +2,7 @@
 
 # This file is auto-generated by oagen. Do not edit.
 
-require 'json'
+require "json"
 
 module Retab
   class Parses
@@ -19,34 +19,34 @@ module Retab
     # @param from_date [String, nil]
     # @param to_date [String, nil]
     # @param request_options [Hash] (see Retab::Types::RequestOptions)
-    # @return [Retab::Types::ListStruct<Retab::Parse>]
+    # @return [Retab::PaginatedList<Retab::Parse>]
     def list(
       before: nil,
       after: nil,
       limit: 10,
-      order: 'desc',
+      order: "desc",
       filename: nil,
       from_date: nil,
       to_date: nil,
       request_options: {}
     )
       params = {
-        'before' => before,
-        'after' => after,
-        'limit' => limit,
-        'order' => order,
-        'filename' => filename,
-        'from_date' => from_date,
-        'to_date' => to_date
+        "before" => before,
+        "after" => after,
+        "limit" => limit,
+        "order" => order,
+        "filename" => filename,
+        "from_date" => from_date,
+        "to_date" => to_date
       }.compact
       response = @client.request(
         method: :get,
-        path: '/v1/parses',
+        path: "/v1/parses",
         auth: true,
         params: params,
         request_options: request_options
       )
-      fetch_next = ->(cursor) {
+      fetch_next = -> (cursor) {
         list(
           before: before,
           after: cursor,
@@ -58,10 +58,17 @@ module Retab
           request_options: request_options
         )
       }
-      Retab::Types::ListStruct.from_response(
+      Retab::PaginatedList.from_response(
         response,
         model: Retab::Parse,
-        filters: { before: before, limit: limit, order: order, filename: filename, from_date: from_date, to_date: to_date },
+        filters: {
+          before: before,
+          limit: limit,
+          order: order,
+          filename: filename,
+          from_date: from_date,
+          to_date: to_date
+        },
         fetch_next: fetch_next
       )
     end
@@ -86,22 +93,26 @@ module Retab
     )
       document = Retab::MimeData.coerce(document) unless document.nil?
       body = {
-        'document' => document,
-        'model' => model,
-        'table_parsing_format' => table_parsing_format,
-        'image_resolution_dpi' => image_resolution_dpi,
-        'instructions' => instructions,
-        'bust_cache' => bust_cache
+        "document" => document,
+        "model" => model,
+        "table_parsing_format" => table_parsing_format,
+        "image_resolution_dpi" => image_resolution_dpi,
+        "instructions" => instructions,
+        "bust_cache" => bust_cache
       }.compact
       response = @client.request(
         method: :post,
-        path: '/v1/parses',
+        path: "/v1/parses",
         auth: true,
         body: body,
         request_options: request_options
       )
       result = Retab::Parse.new(response.body)
-      result.last_response = Retab::Types::ApiResponse.new(http_status: response.code.to_i, http_headers: response.each_header.to_h, request_id: response["x-request-id"])
+      result.last_response = Retab::Types::ApiResponse.new(
+        http_status: response.code.to_i,
+        http_headers: response.each_header.to_h,
+        request_id: response["x-request-id"]
+      )
       result
     end
 
@@ -120,7 +131,11 @@ module Retab
         request_options: request_options
       )
       result = Retab::Parse.new(response.body)
-      result.last_response = Retab::Types::ApiResponse.new(http_status: response.code.to_i, http_headers: response.each_header.to_h, request_id: response["x-request-id"])
+      result.last_response = Retab::Types::ApiResponse.new(
+        http_status: response.code.to_i,
+        http_headers: response.each_header.to_h,
+        request_id: response["x-request-id"]
+      )
       result
     end
 
@@ -132,7 +147,7 @@ module Retab
       parse_id:,
       request_options: {}
     )
-      response = @client.request(
+      @client.request(
         method: :delete,
         path: "/v1/parses/#{Retab::Util.encode_path(parse_id)}",
         auth: true,

data/lib/retab/partitions/partition.rb

@@ -19,7 +19,7 @@ module Retab
       created_at: :created_at
     }.freeze
 
-    attr_accessor \
+    attr_accessor(
       :id,
       :file,
       :model,
@@ -31,8 +31,10 @@ module Retab
       :consensus,
       :usage,
       :created_at
+    )
 
     def initialize(json)
+      super()
       hash = self.class.normalize(json)
       @id = hash[:id]
       @file = hash[:file] ? Retab::FileRef.new(hash[:file]) : nil