47 lines
1.4 KiB
Ruby
47 lines
1.4 KiB
Ruby
|
# frozen_string_literal: true
|
||
|
|
|
||
|
|
RSpec.describe DiscourseAi::Summarization::Strategies::TruncateContent do
|
||
|
|
subject(:strategy) { described_class.new(model) }
|
||
|
|
|
||
|
|
before { SiteSetting.ai_summarization_discourse_service_api_endpoint = "https://test.com" }
|
||
|
|
|
||
|
|
let(:summarize_text) { "This is a text" }
|
||
|
|
let(:full_text) { "(1 asd said: #{summarize_text} " }
|
||
|
|
let(:model_tokens) { ::DiscourseAi::Tokenizer::BertTokenizer.size(full_text) - 5 }
|
||
|
|
|
||
|
|
let(:model) do
|
||
|
|
DiscourseAi::Summarization::Models::Discourse.new(
|
||
|
|
"flan-t5-base-samsum",
|
||
|
|
max_tokens: model_tokens,
|
||
|
|
)
|
||
|
|
end
|
||
|
|
|
||
|
|
let(:content) { { contents: [{ poster: "asd", id: 1, text: summarize_text }] } }
|
||
|
|
|
||
|
|
let(:summarized_text) { "this is a single summary" }
|
||
|
|
|
||
|
|
let(:user) { User.new }
|
||
|
|
|
||
|
|
describe "#summary" do
|
||
|
|
it "truncates the content and requests a summary" do
|
||
|
|
truncated =
|
||
|
|
::DiscourseAi::Tokenizer::BertTokenizer.truncate(
|
||
|
|
"(1 asd said: This is a text ",
|
||
|
|
model_tokens,
|
||
|
|
)
|
||
|
|
|
||
|
|
WebMock
|
||
|
|
.stub_request(
|
||
|
|
:post,
|
||
|
|
"#{SiteSetting.ai_summarization_discourse_service_api_endpoint}/api/v1/classify",
|
||
|
|
)
|
||
|
|
.with(body: JSON.dump(model: model.model, content: truncated))
|
||
|
|
.to_return(status: 200, body: JSON.dump({ summary_text: summarized_text }))
|
||
|
|
|
||
|
|
summary = strategy.summarize(content, user).dig(:summary)
|
||
|
|
|
||
|
|
expect(summary).to eq(summarized_text)
|
||
|
|
end
|
||
|
|
end
|
||
|
|
end
|