Merge pull request #420 from simonx1/upload-file-support-for-file-object

Add support for File and StringIO `file` parameter in `upload`.

Author: alexrudall

README.md

@@ -599,7 +599,7 @@ Put your data in a `.jsonl` file like this:
 {"prompt":"@lakers disappoint for a third straight night ->", "completion":" negative"}
 ```
 
-and pass the path to `client.files.upload` to upload it to OpenAI, and then interact with it:
+and pass the path (or a StringIO object) to `client.files.upload` to upload it to OpenAI, and then interact with it:
 
 ```ruby
 client.files.upload(parameters: { file: "path/to/sentiment.jsonl", purpose: "fine-tune" })

lib/openai/files.rb

@@ -1,5 +1,11 @@
 module OpenAI
   class Files
+    PURPOSES = %w[
+      assistants
+      batch
+      fine-tune
+    ].freeze
+
     def initialize(client:)
       @client = client
     end
@@ -9,12 +15,17 @@ def list
     end
 
     def upload(parameters: {})
-      validate(file: parameters[:file]) if parameters[:file].include?(".jsonl")
+      file_input = parameters[:file]
+      file = prepare_file_input(file_input: file_input)
+
+      validate(file: file, purpose: parameters[:purpose], file_input: file_input)
 
       @client.multipart_post(
         path: "/files",
-        parameters: parameters.merge(file: File.open(parameters[:file]))
+        parameters: parameters.merge(file: file)
       )
+    ensure
+      file.close if file.is_a?(File)
     end
 
     def retrieve(id:)
@@ -31,12 +42,33 @@ def delete(id:)
 
     private
 
-    def validate(file:)
-      File.open(file).each_line.with_index do |line, index|
+    def prepare_file_input(file_input:)
+      if file_input.is_a?(String)
+        File.open(file_input)
+      elsif file_input.respond_to?(:read) && file_input.respond_to?(:rewind)
+        file_input
+      else
+        raise ArgumentError, "Invalid file - must be a StringIO object or a path to a file."
+      end
+    end
+
+    def validate(file:, purpose:, file_input:)
+      raise ArgumentError, "`file` is required" if file.nil?
+      unless PURPOSES.include?(purpose)
+        raise ArgumentError, "`purpose` must be one of `#{PURPOSES.join(',')}`"
+      end
+
+      validate_jsonl(file: file) if file_input.is_a?(String) && file_input.end_with?(".jsonl")
+    end
+
+    def validate_jsonl(file:)
+      file.each_line.with_index do |line, index|
         JSON.parse(line)
       rescue JSON::ParserError => e
         raise JSON::ParserError, "#{e.message} - found on line #{index + 1} of #{file}"
       end
+    ensure
+      file.rewind
     end
   end
 end

lib/openai/http.rb

@@ -99,10 +99,14 @@ def multipart_parameters(parameters)
       parameters&.transform_values do |value|
         next value unless value.respond_to?(:close) # File or IO object.
 
+        # Faraday::UploadIO does not require a path, so we will pass it
+        # only if it is available. This allows StringIO objects to be
+        # passed in as well.
+        path = value.respond_to?(:path) ? value.path : nil
         # Doesn't seem like OpenAI needs mime_type yet, so not worth
         # the library to figure this out. Hence the empty string
         # as the second argument.
-        Faraday::UploadIO.new(value, "", value.path)
+        Faraday::UploadIO.new(value, "", path)
       end
     end
 

spec/fixtures/cassettes/files_upload.yml renamed to spec/fixtures/cassettes/files_upload_file.yml

@@ -11,18 +11,38 @@
     let(:upload_id) { upload["id"] }
 
     describe "#upload" do
-      let(:upload_cassette) { "files upload" }
+      let(:upload_cassette) { "files upload #{cassette_label}" }
+
+      context "with an invalid file" do
+        let(:cassette_label) { "unused" }
+        let(:filename) { File.join("errors", "missing_quote.jsonl") }
+
+        it { expect { upload }.to raise_error(JSON::ParserError) }
+      end
+
+      context "with an invalid purpose" do
+        let(:cassette_label) { "unused" }
+        let(:upload_purpose) { "invalid" }
+
+        it { expect { upload }.to raise_error(ArgumentError) }
+      end
+
+      context "with a `File` instance content" do
+        let(:cassette_label) { "file" }
+        let(:file) { File.open(File.join(RSPEC_ROOT, "fixtures/files", filename)) }
 
-      context "with a valid JSON lines file" do
         it "succeeds" do
           expect(upload["filename"]).to eq(filename)
         end
       end
 
-      context "with an invalid file" do
-        let(:filename) { File.join("errors", "missing_quote.jsonl") }
+      context "with a `StringIO` instance content" do
+        let(:cassette_label) { "stringio" }
+        let(:file) { StringIO.new(File.read(File.join(RSPEC_ROOT, "fixtures/files", filename))) }
 
-        it { expect { upload }.to raise_error(JSON::ParserError) }
+        it "succeeds" do
+          expect(upload["filename"]).to eq("local.path")
+        end
       end
     end