ghostcrawl 2.2.1 → 2.2.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 49310560652f43f54a4ce7d69b0792e22675e6d5d23923ec7960f9f4be0cd52d
-  data.tar.gz: 6245e25155ae69a24bb442775639f8872526f5988b69b1db3a35541ba388f822
+  metadata.gz: ff09081da25b9c78b261437d5bc26c2a4b25f3e727a53e8393f1adb00aa6536c
+  data.tar.gz: ad0869fe147e1c33a49fadd364bc93ce5857fd73eb03b584d4c1ce618e1092a9
 SHA512:
-  metadata.gz: b2c7fcaced06f2628f77d2c574ad34714f064d91660037fdcb5f06bda80e33e446dba4717c169216a3585762ee83ed7a94c708dd316902cde0a21bcfd2b9ff29
-  data.tar.gz: 618777d4e590acf5828ca2999028af7f3c43b71469ba65caf6e6970a62c551fa8aeba35ddda7e5febef22b75249cc77b5e7ad380e23515b25bbd392c6be89b94
+  metadata.gz: '085820b153fb6b70ea26f6e866787388835cfec7e4cce86498f83551abe6b4c8800b5eb14f86203c59ca77b7e70cffbd459d0d84f7a75fc69f5337914bcf60e6'
+  data.tar.gz: 11a00207fc7469d02bec976018c8e627ea11c0da0ce78d4dc5a6d5b76fced17a5f3a37053159e0bde94dd0c44e4d3e273e7cf57cdac1d652a410bcba4da814ff

data/lib/ghostcrawl/client.rb

@@ -668,7 +668,7 @@ module Ghostcrawl
     # @param extract_schema [Hash, nil] JSON Schema for structured extraction
     # @param raise_on_result_error [Boolean] raise {Ghostcrawl::ScrapeError} on a
     #   target-side (HTTP-200) failure instead of returning the raw hash (default true)
-    # @return [Hash] response with +markdown+, +status+, and other fields
+    # @return [Hash] response with +content+, +markdown+, +status+, and other fields
     def scrape(url:, format: "markdown", engine: "auto", javascript: true, extract_schema: nil,
                raise_on_result_error: true, **opts)
       # Use AdditionalDataBody to send only the fields we specify — the generated
@@ -678,6 +678,7 @@ module Ghostcrawl
                "javascript_enabled" => javascript }.merge(opts.transform_keys(&:to_s))
       data["extract_schema"] = extract_schema unless extract_schema.nil?
       hash = ResponseHelper.to_hash(@v1.scrape.post(AdditionalDataBody.new(data)))
+      normalize_scrape_content(hash)
       raise_on_result_error ? ResponseHelper.raise_on_result_error!(hash) : hash
     end
 
@@ -757,5 +758,42 @@ module Ghostcrawl
       request_info = @v1.me.to_get_request_information(nil)
       ResponseHelper.to_hash(@adapter.send_async(request_info, Ghostcrawl::V1::Binary, {}))
     end
+
+    private
+
+    # Normalizes a +"content"+ key onto a decoded scrape response.
+    #
+    # The API returns the rendered page under a FORMAT-SPECIFIC key
+    # (+"markdown"+, +"html"+, or +"text"+), but the documented quickstart reads
+    # +result["content"]+. This mirrors that value onto +"content"+ in place,
+    # KEEPING the format-specific key intact (backward compatible).
+    #
+    # No-op unless +result+ is a Hash that does not already carry +"content"+.
+    # The value chosen is: the field named by +result["format"]+ when that field
+    # is a String, else the first String among +"markdown"+, +"html"+, +"text"+.
+    #
+    # @param result [Object] the decoded response (only mutated when a Hash)
+    # @return [Object] the same +result+, unchanged reference
+    # @api private
+    def normalize_scrape_content(result)
+      return result unless result.is_a?(Hash) && !result.key?("content")
+
+      fmt = result["format"]
+      value = result[fmt] if fmt.is_a?(String)
+      value = nil unless value.is_a?(String)
+
+      unless value.is_a?(String)
+        %w[markdown html text].each do |key|
+          candidate = result[key]
+          if candidate.is_a?(String)
+            value = candidate
+            break
+          end
+        end
+      end
+
+      result["content"] = value if value.is_a?(String)
+      result
+    end
   end
 end

data/lib/ghostcrawl/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Ghostcrawl
-  VERSION = "2.2.1"
+  VERSION = "2.2.2"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ghostcrawl
 version: !ruby/object:Gem::Version
-  version: 2.2.1
+  version: 2.2.2
 platform: ruby
 authors:
 - GhostCrawl