context.dev 2.17.0 → 2.18.0

data/lib/context_dev/models/web_web_scrape_html_params.rb

@@ -38,6 +38,19 @@ module ContextDev
       #   @return [Array<String>, nil]
       optional :exclude_selectors, ContextDev::Internal::Type::ArrayOf[String], nil?: true
 
+      # @!attribute extract_rules
+      #   Optional CSS extraction rules applied to the returned HTML after selector and
+      #   main-content filtering. Use selector strings ("h1", "a@href") or objects with
+      #   selector, type (item or list), and output (text, html, @attribute, or nested
+      #   rules). Text whitespace is normalized; html includes the matched element;
+      #   attributes are returned as written. Missing items are null and missing lists are
+      #   empty. CSS only; XPath is not supported. Maximum: 100 fields across 5 levels.
+      #   Send a JSON-encoded string in the extractRules query parameter.
+      #
+      #   @return [Hash{Symbol=>String, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1}, nil]
+      optional :extract_rules,
+               -> { ContextDev::Internal::Type::HashOf[union: ContextDev::WebWebScrapeHTMLParams::ExtractRule] }
+
       # @!attribute headers
       #   Optional outbound HTTP headers forwarded only to the target URL, sent as
       #   deep-object query params such as headers[X-Custom]=value. When provided, caching
@@ -116,14 +129,15 @@ module ContextDev
 
       # @!attribute zdr
       #   Set to enabled to bypass shared caches and omit request and response content
-      #   from retained usage logs. Requires zero data retention to be enabled for your
-      #   organization (contact support@context.dev), otherwise the request fails with
-      #   ZDR_NOT_ENABLED. Successful ZDR responses include X-Context-ZDR: true.
+      #   from retained usage logs. Asset uploads are skipped, so hosted image URLs are
+      #   omitted. Requires zero data retention to be enabled for your organization
+      #   (contact support@context.dev), otherwise the request fails with ZDR_NOT_ENABLED.
+      #   Successful ZDR responses include X-Context-ZDR: true.
       #
       #   @return [Symbol, ContextDev::Models::WebWebScrapeHTMLParams::Zdr, nil]
       optional :zdr, enum: -> { ContextDev::WebWebScrapeHTMLParams::Zdr }
 
-      # @!method initialize(url:, actions: nil, country: nil, exclude_selectors: nil, headers: nil, include_frames: nil, include_selectors: nil, max_age_ms: nil, pdf: nil, settle_animations: nil, tags: nil, timeout_opts: nil, use_main_content_only: nil, wait_for_ms: nil, zdr: nil, request_options: {})
+      # @!method initialize(url:, actions: nil, country: nil, exclude_selectors: nil, extract_rules: nil, headers: nil, include_frames: nil, include_selectors: nil, max_age_ms: nil, pdf: nil, settle_animations: nil, tags: nil, timeout_opts: nil, use_main_content_only: nil, wait_for_ms: nil, zdr: nil, request_options: {})
       #   Some parameter documentations has been truncated, see
       #   {ContextDev::Models::WebWebScrapeHTMLParams} for more details.
       #
@@ -135,6 +149,8 @@ module ContextDev
       #
       #   @param exclude_selectors [Array<String>, nil] CSS selectors to remove from the result. Applied after includeSelectors. Exclusi
       #
+      #   @param extract_rules [Hash{Symbol=>String, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1}] Optional CSS extraction rules applied to the returned HTML after selector and ma
+      #
       #   @param headers [Hash{Symbol=>String}] Optional outbound HTTP headers forwarded only to the target URL, sent as deep-ob
       #
       #   @param include_frames [Boolean] When true, iframes are rendered inline into the returned HTML.
@@ -522,6 +538,168 @@ module ContextDev
         #   @return [Array<Symbol>]
       end
 
+      module ExtractRule
+        extend ContextDev::Internal::Type::Union
+
+        variant String
+
+        variant -> { ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1 }
+
+        class UnionMember1 < ContextDev::Internal::Type::BaseModel
+          # @!attribute selector
+          #
+          #   @return [String]
+          required :selector, String
+
+          # @!attribute output
+          #
+          #   @return [Symbol, String, Hash{Symbol=>String, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1}, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output, nil]
+          optional :output, union: -> { ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output }
+
+          # @!attribute type
+          #
+          #   @return [Symbol, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Type, nil]
+          optional :type, enum: -> { ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Type }
+
+          # @!method initialize(selector:, output: nil, type: nil)
+          #   @param selector [String]
+          #   @param output [Symbol, String, Hash{Symbol=>String, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1}, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output]
+          #   @param type [Symbol, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Type]
+
+          # @see ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1#output
+          module Output
+            extend ContextDev::Internal::Type::Union
+
+            variant const: -> { ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::TEXT }
+
+            variant const: -> { ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTML }
+
+            variant String
+
+            variant -> { ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRuleMap }
+
+            module HTMLExtractionRule
+              extend ContextDev::Internal::Type::Union
+
+              variant String
+
+              variant -> { ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1 }
+
+              class UnionMember1 < ContextDev::Internal::Type::BaseModel
+                # @!attribute selector
+                #
+                #   @return [String]
+                required :selector, String
+
+                # @!attribute output
+                #
+                #   @return [Symbol, String, Object, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Output, nil]
+                optional :output,
+                         union: -> { ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Output }
+
+                # @!attribute type
+                #
+                #   @return [Symbol, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Type, nil]
+                optional :type,
+                         enum: -> { ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Type }
+
+                # @!method initialize(selector:, output: nil, type: nil)
+                #   @param selector [String]
+                #   @param output [Symbol, String, Object, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Output]
+                #   @param type [Symbol, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Type]
+
+                # @see ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1#output
+                module Output
+                  extend ContextDev::Internal::Type::Union
+
+                  variant const: -> { ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Output::TEXT }
+
+                  variant const: -> { ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Output::HTML }
+
+                  variant String
+
+                  variant ContextDev::Internal::Type::Unknown
+
+                  # @!method self.variants
+                  #   @return [Array(Symbol, String, Object)]
+
+                  define_sorbet_constant!(:Variants) do
+                    T.type_alias do
+                      T.any(
+                        ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1::Output::TaggedSymbol,
+                        String,
+                        T.anything
+                      )
+                    end
+                  end
+
+                  # @!group
+
+                  TEXT = :text
+                  HTML = :html
+
+                  # @!endgroup
+                end
+
+                # @see ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1#type
+                module Type
+                  extend ContextDev::Internal::Type::Enum
+
+                  ITEM = :item
+                  LIST = :list
+
+                  # @!method self.values
+                  #   @return [Array<Symbol>]
+                end
+              end
+
+              # @!method self.variants
+              #   @return [Array(String, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1)]
+            end
+
+            # @!method self.variants
+            #   @return [Array(Symbol, String, Hash{Symbol=>String, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::UnionMember1})]
+
+            define_sorbet_constant!(:Variants) do
+              T.type_alias do
+                T.any(
+                  ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::TaggedSymbol,
+                  String,
+                  T::Hash[Symbol, ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule::Variants]
+                )
+              end
+            end
+
+            # @!group
+
+            TEXT = :text
+            HTML = :html
+
+            # @!endgroup
+
+            # @type [ContextDev::Internal::Type::Converter]
+            HTMLExtractionRuleMap =
+              ContextDev::Internal::Type::HashOf[union: -> {
+                ContextDev::WebWebScrapeHTMLParams::ExtractRule::UnionMember1::Output::HTMLExtractionRule
+              }]
+          end
+
+          # @see ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1#type
+          module Type
+            extend ContextDev::Internal::Type::Enum
+
+            ITEM = :item
+            LIST = :list
+
+            # @!method self.values
+            #   @return [Array<Symbol>]
+          end
+        end
+
+        # @!method self.variants
+        #   @return [Array(String, ContextDev::Models::WebWebScrapeHTMLParams::ExtractRule::UnionMember1)]
+      end
+
       class Pdf < ContextDev::Internal::Type::BaseModel
         # @!attribute end_
         #   Last 1-based PDF page to parse. When omitted, parsing ends at the last page.
@@ -580,7 +758,7 @@ module ContextDev
         #   credits. "return-partial" returns usable results collected so far; if none are
         #   available, the request still fails without charging credits. Partial results are
         #   not cached as complete results. "return-partial" requires milliseconds of at
-        #   least 15000.
+        #   least 5000.
         #
         #   @return [Symbol, ContextDev::Models::WebWebScrapeHTMLParams::TimeoutOpts::Behavior, nil]
         optional :behavior, enum: -> { ContextDev::WebWebScrapeHTMLParams::TimeoutOpts::Behavior }
@@ -601,7 +779,7 @@ module ContextDev
         # credits. "return-partial" returns usable results collected so far; if none are
         # available, the request still fails without charging credits. Partial results are
         # not cached as complete results. "return-partial" requires milliseconds of at
-        # least 15000.
+        # least 5000.
         #
         # @see ContextDev::Models::WebWebScrapeHTMLParams::TimeoutOpts#behavior
         module Behavior
@@ -616,9 +794,10 @@ module ContextDev
       end
 
       # Set to enabled to bypass shared caches and omit request and response content
-      # from retained usage logs. Requires zero data retention to be enabled for your
-      # organization (contact support@context.dev), otherwise the request fails with
-      # ZDR_NOT_ENABLED. Successful ZDR responses include X-Context-ZDR: true.
+      # from retained usage logs. Asset uploads are skipped, so hosted image URLs are
+      # omitted. Requires zero data retention to be enabled for your organization
+      # (contact support@context.dev), otherwise the request fails with ZDR_NOT_ENABLED.
+      # Successful ZDR responses include X-Context-ZDR: true.
       module Zdr
         extend ContextDev::Internal::Type::Enum
 

data/lib/context_dev/models/web_web_scrape_html_response.rb

@@ -81,13 +81,23 @@ module ContextDev
       #   @return [Boolean, nil]
       optional :actions_html_stale, ContextDev::Internal::Type::Boolean, api_name: :actionsHtmlStale
 
+      # @!attribute extracted
+      #   Present only when extractRules is supplied. Keys match the requested fields.
+      #   Values are normalized text, raw attribute strings, outer HTML, nested objects,
+      #   or lists. Missing items are null; lists with no matches are empty. Rules run on
+      #   the returned HTML after filtering.
+      #
+      #   @return [Hash{Symbol=>String, Object, Array<String, Object, nil>, nil}, nil]
+      optional :extracted,
+               -> { ContextDev::Internal::Type::HashOf[union: ContextDev::Models::WebWebScrapeHTMLResponse::Extracted, nil?: true] }
+
       # @!attribute key_metadata
       #   Credit usage, included whenever a valid API key is provided.
       #
       #   @return [ContextDev::Models::WebWebScrapeHTMLResponse::KeyMetadata, nil]
       optional :key_metadata, -> { ContextDev::Models::WebWebScrapeHTMLResponse::KeyMetadata }
 
-      # @!method initialize(cache_metadata:, final_dom_state:, html:, metadata:, request_id:, success:, type:, url:, actions_applied: nil, actions_html_stale: nil, key_metadata: nil)
+      # @!method initialize(cache_metadata:, final_dom_state:, html:, metadata:, request_id:, success:, type:, url:, actions_applied: nil, actions_html_stale: nil, extracted: nil, key_metadata: nil)
       #   Some parameter documentations has been truncated, see
       #   {ContextDev::Models::WebWebScrapeHTMLResponse} for more details.
       #
@@ -111,6 +121,8 @@ module ContextDev
       #
       #   @param actions_html_stale [Boolean] True when an action was applied but the returned content could not be refreshed
       #
+      #   @param extracted [Hash{Symbol=>String, Object, Array<String, Object, nil>, nil}] Present only when extractRules is supplied. Keys match the requested fields. Val
+      #
       #   @param key_metadata [ContextDev::Models::WebWebScrapeHTMLResponse::KeyMetadata] Credit usage, included whenever a valid API key is provided.
 
       # @see ContextDev::Models::WebWebScrapeHTMLResponse#cache_metadata
@@ -562,6 +574,34 @@ module ContextDev
         end
       end
 
+      module Extracted
+        extend ContextDev::Internal::Type::Union
+
+        variant String
+
+        variant ContextDev::Internal::Type::Unknown
+
+        variant -> { ContextDev::Models::WebWebScrapeHTMLResponse::Extracted::UnionMember2Array }
+
+        module UnionMember2
+          extend ContextDev::Internal::Type::Union
+
+          variant String
+
+          variant ContextDev::Internal::Type::Unknown
+
+          # @!method self.variants
+          #   @return [Array(String, Object)]
+        end
+
+        # @!method self.variants
+        #   @return [Array(String, Object, Array<String, Object, nil>)]
+
+        # @type [ContextDev::Internal::Type::Converter]
+        UnionMember2Array =
+          ContextDev::Internal::Type::ArrayOf[union: -> { ContextDev::Models::WebWebScrapeHTMLResponse::Extracted::UnionMember2 }, nil?: true]
+      end
+
       # @see ContextDev::Models::WebWebScrapeHTMLResponse#key_metadata
       class KeyMetadata < ContextDev::Internal::Type::BaseModel
         # @!attribute credits_consumed

data/lib/context_dev/models/web_web_scrape_images_params.rb

@@ -80,7 +80,17 @@ module ContextDev
       #   @return [Integer, nil]
       optional :wait_for_ms, Integer, nil?: true
 
-      # @!method initialize(url:, actions: nil, dedupe: nil, enrichment: nil, headers: nil, max_age_ms: nil, tags: nil, timeout_opts: nil, wait_for_ms: nil, request_options: {})
+      # @!attribute zdr
+      #   Set to enabled to bypass shared caches and omit request and response content
+      #   from retained usage logs. Asset uploads are skipped, so hosted image URLs are
+      #   omitted. Requires zero data retention to be enabled for your organization
+      #   (contact support@context.dev), otherwise the request fails with ZDR_NOT_ENABLED.
+      #   Successful ZDR responses include X-Context-ZDR: true.
+      #
+      #   @return [Symbol, ContextDev::Models::WebWebScrapeImagesParams::Zdr, nil]
+      optional :zdr, enum: -> { ContextDev::WebWebScrapeImagesParams::Zdr }
+
+      # @!method initialize(url:, actions: nil, dedupe: nil, enrichment: nil, headers: nil, max_age_ms: nil, tags: nil, timeout_opts: nil, wait_for_ms: nil, zdr: nil, request_options: {})
       #   Some parameter documentations has been truncated, see
       #   {ContextDev::Models::WebWebScrapeImagesParams} for more details.
       #
@@ -102,6 +112,8 @@ module ContextDev
       #
       #   @param wait_for_ms [Integer, nil] Optional browser wait time in milliseconds after initial page load before collec
       #
+      #   @param zdr [Symbol, ContextDev::Models::WebWebScrapeImagesParams::Zdr] Set to enabled to bypass shared caches and omit request and response content fro
+      #
       #   @param request_options [ContextDev::RequestOptions, Hash{Symbol=>Object}]
 
       # Browser action discriminated by `do`. Each variant exposes only its applicable
@@ -262,7 +274,7 @@ module ContextDev
 
         # @!attribute hosted_url
         #   Host materializable images on the Brand.dev CDN and return their URL and MIME
-        #   type.
+        #   type. Ignored when zero data retention is enabled.
         #
         #   @return [Boolean, nil]
         optional :hosted_url, ContextDev::Internal::Type::Boolean, api_name: :hostedUrl
@@ -307,7 +319,7 @@ module ContextDev
         #   credits. "return-partial" returns usable results collected so far; if none are
         #   available, the request still fails without charging credits. Partial results are
         #   not cached as complete results. "return-partial" requires milliseconds of at
-        #   least 15000.
+        #   least 5000.
         #
         #   @return [Symbol, ContextDev::Models::WebWebScrapeImagesParams::TimeoutOpts::Behavior, nil]
         optional :behavior, enum: -> { ContextDev::WebWebScrapeImagesParams::TimeoutOpts::Behavior }
@@ -328,7 +340,7 @@ module ContextDev
         # credits. "return-partial" returns usable results collected so far; if none are
         # available, the request still fails without charging credits. Partial results are
         # not cached as complete results. "return-partial" requires milliseconds of at
-        # least 15000.
+        # least 5000.
         #
         # @see ContextDev::Models::WebWebScrapeImagesParams::TimeoutOpts#behavior
         module Behavior
@@ -341,6 +353,21 @@ module ContextDev
           #   @return [Array<Symbol>]
         end
       end
+
+      # Set to enabled to bypass shared caches and omit request and response content
+      # from retained usage logs. Asset uploads are skipped, so hosted image URLs are
+      # omitted. Requires zero data retention to be enabled for your organization
+      # (contact support@context.dev), otherwise the request fails with ZDR_NOT_ENABLED.
+      # Successful ZDR responses include X-Context-ZDR: true.
+      module Zdr
+        extend ContextDev::Internal::Type::Enum
+
+        ENABLED = :enabled
+        DISABLED = :disabled
+
+        # @!method self.values
+        #   @return [Array<Symbol>]
+      end
     end
   end
 end

data/lib/context_dev/models/web_web_scrape_md_params.rb

@@ -143,9 +143,10 @@ module ContextDev
 
       # @!attribute zdr
       #   Set to enabled to bypass shared caches and omit request and response content
-      #   from retained usage logs. Requires zero data retention to be enabled for your
-      #   organization (contact support@context.dev), otherwise the request fails with
-      #   ZDR_NOT_ENABLED. Successful ZDR responses include X-Context-ZDR: true.
+      #   from retained usage logs. Asset uploads are skipped, so hosted image URLs are
+      #   omitted. Requires zero data retention to be enabled for your organization
+      #   (contact support@context.dev), otherwise the request fails with ZDR_NOT_ENABLED.
+      #   Successful ZDR responses include X-Context-ZDR: true.
       #
       #   @return [Symbol, ContextDev::Models::WebWebScrapeMdParams::Zdr, nil]
       optional :zdr, enum: -> { ContextDev::WebWebScrapeMdParams::Zdr }
@@ -615,7 +616,7 @@ module ContextDev
         #   credits. "return-partial" returns usable results collected so far; if none are
         #   available, the request still fails without charging credits. Partial results are
         #   not cached as complete results. "return-partial" requires milliseconds of at
-        #   least 15000.
+        #   least 5000.
         #
         #   @return [Symbol, ContextDev::Models::WebWebScrapeMdParams::TimeoutOpts::Behavior, nil]
         optional :behavior, enum: -> { ContextDev::WebWebScrapeMdParams::TimeoutOpts::Behavior }
@@ -636,7 +637,7 @@ module ContextDev
         # credits. "return-partial" returns usable results collected so far; if none are
         # available, the request still fails without charging credits. Partial results are
         # not cached as complete results. "return-partial" requires milliseconds of at
-        # least 15000.
+        # least 5000.
         #
         # @see ContextDev::Models::WebWebScrapeMdParams::TimeoutOpts#behavior
         module Behavior
@@ -651,9 +652,10 @@ module ContextDev
       end
 
       # Set to enabled to bypass shared caches and omit request and response content
-      # from retained usage logs. Requires zero data retention to be enabled for your
-      # organization (contact support@context.dev), otherwise the request fails with
-      # ZDR_NOT_ENABLED. Successful ZDR responses include X-Context-ZDR: true.
+      # from retained usage logs. Asset uploads are skipped, so hosted image URLs are
+      # omitted. Requires zero data retention to be enabled for your organization
+      # (contact support@context.dev), otherwise the request fails with ZDR_NOT_ENABLED.
+      # Successful ZDR responses include X-Context-ZDR: true.
       module Zdr
         extend ContextDev::Internal::Type::Enum