format_parser 2.0.0.pre → 2.0.0.pre.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d7c965b7783ecaea4802f7e585861b4400b2210fee4cb90388757530880fa074
-  data.tar.gz: fc8b7cc3f00825fa054c948a7ae817b1eee6457ffaec9e5a6b5bdd9a0b92d126
+  metadata.gz: 01730fd580c14a2c35b87be1f62384b06f9b2683bd07ba01ed21d21a4211b1f5
+  data.tar.gz: c27ba63f09c8251ccaf237040cb6f01c495686f6a025cdb7207aa0bfb2fe0098
 SHA512:
-  metadata.gz: 73f774ebe540dfd54e87f89cedecfc0fabf4a97f4e2ef72afcd94edc5e0fbc344c7c67b365942e3bb915dfe76f94f038072671c259c2d366a69d64a73cbde960
-  data.tar.gz: bc1405329d521487ec4d0738c258fb12c3acdb37b6b8ecebf7451a866d5f1072cfc23774e2ecc3d7d297095ff280320756fb4cd9000de3eac447a105cf87028b
+  metadata.gz: 427c1f3652e33592a7bd3a63c0db676a865390b1947732e452905d0fe64f54df82af2b9a339dfdf962d9586c73fe7ac8a2b19da099c502a187efd8215993ddbf
+  data.tar.gz: 3652ca2c1ae42ddb5e29811dcae13e8bbb0fcf964c4f47547638e992d21612235bdf489d908b5c4bdf3aa2859768eb0dc98c6fc58527ce5031396ee341207354

data/.github/workflows/main.yml

@@ -8,7 +8,7 @@ env:
 jobs:
   lint:
     name: Code Style
-    runs-on: ubuntu-18.04
+    runs-on: ubuntu-latest
     if: github.event_name == 'push' || github.event.pull_request.head.repo.full_name != github.repository
     strategy:
       matrix:
@@ -54,7 +54,7 @@ jobs:
         run: bundle exec rubocop
   test:
     name: Specs
-    runs-on: ubuntu-18.04
+    runs-on: ubuntu-latest
     if: github.event_name == 'push' || github.event.pull_request.head.repo.full_name != github.repository
     strategy:
       matrix:

data/CHANGELOG.md

@@ -1,4 +1,4 @@
-## 2.0.0.pre (Prerelease)
+## 2.0.0.pre.2 (Prerelease)
 * Drop support for Ruby `<2.7`.
 * Drop faraday dependencies.
 * Loosen version constraints on other dependencies.

data/format_parser.gemspec

@@ -39,6 +39,7 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency 'rake'
   spec.add_development_dependency 'rspec'
   spec.add_development_dependency 'simplecov'
+  spec.add_development_dependency 'webmock'
   spec.add_development_dependency 'wetransfer_style', '1.0.0'
   spec.add_development_dependency 'yard'
 end

data/lib/format_parser/version.rb

@@ -1,3 +1,3 @@
 module FormatParser
-  VERSION = '2.0.0.pre'
+  VERSION = '2.0.0.pre.2'
 end

data/lib/remote_io.rb

@@ -122,7 +122,7 @@ class FormatParser::RemoteIO
         error_message = [
           "We requested #{requested_range_size} bytes, but the server sent us more",
           "(#{response_size} bytes) - it likely has no `Range:` support.",
-          "The error occurred when talking to #{uri})"
+          "The error occurred when talking to #{uri}"
         ]
         raise InvalidRequest.new(response.code, error_message.join("\n"))
       end
@@ -141,10 +141,13 @@ class FormatParser::RemoteIO
       # to be 206
       [size, response.body]
     when Net::HTTPMovedPermanently, Net::HTTPFound, Net::HTTPSeeOther, Net::HTTPTemporaryRedirect, Net::HTTPPermanentRedirect
-      raise RedirectLimitReached(uri) if redirects == 0
+      raise RedirectLimitReached, uri if redirects == 0
       location = response['location']
       if location
-        request_range(range, redirect_uri(location, uri), redirects - 1)
+        new_uri = redirect_uri(location, uri)
+        # Clear the Authorization header if the new URI has a different host.
+        @headers.delete('Authorization') unless [@uri.scheme, @uri.host, @uri.port] == [new_uri.scheme, new_uri.host, new_uri.port]
+        request_range(range, new_uri, redirects - 1)
       else
         raise InvalidRequest.new(response.code, "Server at #{uri} replied with a #{response.code}, indicating redirection; however, the location header was empty.")
       end

data/spec/remote_io_spec.rb

@@ -3,187 +3,270 @@ require 'spec_helper'
 describe FormatParser::RemoteIO do
   it_behaves_like 'an IO object compatible with IOConstraint'
 
-  it 'returns the partial content when the server supplies a 206 status' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPPartialContent.new('2', '206', 'Partial Content')
-    response['Content-Range'] = '10-109/2577'
-    allow(response).to receive(:body).and_return('Response body')
+  # 2XX
+
+  context 'when the response code is 200 (OK)' do
+    context 'when the response size does not exceed the requested range' do
+      it 'returns the entire response body' do
+        url = 'https://images.invalid/img.jpg'
+        body = 'response body'
+
+        stub = stub_request(:get, url)
+          .with(headers: { 'range' => 'bytes=10-109' })
+          .to_return(body: body, status: 200)
+
+        rio = described_class.new(url)
+        rio.seek(10)
+        read_result = rio.read(100)
+
+        expect(read_result).to eq(body)
+        expect(stub).to have_been_requested
+      end
+    end
+
+    context 'when the response size exceeds the requested range' do
+      it 'raises an error' do
+        url = 'https://images.invalid/img.jpg'
+        body = 'This response is way longer than 10 bytes.'
+
+        stub = stub_request(:get, url)
+          .with(headers: { 'range' => 'bytes=10-19' })
+          .to_return(body: body, status: 200)
+
+        rio = described_class.new(url)
+        rio.seek(10)
+
+        expect { rio.read(10) }.to raise_error(
+          "We requested 10 bytes, but the server sent us more\n"\
+          "(42 bytes) - it likely has no `Range:` support.\n"\
+          "The error occurred when talking to #{url}"
+        )
+        expect(stub).to have_been_requested
+      end
+    end
+  end
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+  context 'when the response status code is 206 (Partial Content)' do
+    context 'when the Content-Range header is present' do
+      it 'returns the response body' do
+        url = 'https://images.invalid/img.jpg'
+        body = 'response body'
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| URI::HTTPS === uri && uri.to_s == url },
-      a_hash_including('range' => 'bytes=10-109')
-    )
+        stub = stub_request(:get, url)
+          .with(headers: { 'range' => 'bytes=10-109' })
+          .to_return(body: body, headers: { 'Content-Range' => '10-109/2577' }, status: 206)
 
-    rio = described_class.new(url)
-    rio.seek(10)
-    read_result = rio.read(100)
+        rio = described_class.new(url)
+        rio.seek(10)
+        read_result = rio.read(100)
 
-    expect(read_result).to eq(response.body)
-  end
+        expect(read_result).to eq(body)
+        expect(stub).to have_been_requested
+      end
 
-  it 'returns the entire content when the server supplies the Content-Range response but sends a 200 status' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPOK.new('2', '200', 'OK')
-    allow(response).to receive(:body).and_return('Response body')
+      it 'maintains and exposes pos' do
+        url = 'https://images.invalid/img.jpg'
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+        stub = stub_request(:get, url)
+          .with(headers: { 'range' => 'bytes=0-0' })
+          .to_return(body: 'a', headers: { 'Content-Range' => '0-0/13' }, status: 206)
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| URI::HTTPS === uri && uri.to_s == url },
-      a_hash_including('range' => 'bytes=10-109')
-    )
+        rio = described_class.new(url)
 
-    rio = described_class.new(url)
-    rio.seek(10)
-    read_result = rio.read(100)
+        expect(rio.pos).to eq(0)
 
-    expect(read_result).to eq(response.body)
-  end
+        rio.read(1)
 
-  it 'raises a specific error for all 4xx responses except 416' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPForbidden.new('2', '403', 'Forbidden')
+        expect(rio.pos).to eq(1)
+        expect(stub).to have_been_requested
+      end
+    end
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+    context 'when the Content-Range header is not present' do
+      it 'raises an error' do
+        url = 'https://images.invalid/img.jpg'
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| uri.to_s == url },
-      a_hash_including('range' => 'bytes=100-199')
-    )
+        stub = stub_request(:get, url)
+          .with(headers: { 'range' => 'bytes=10-109' })
+          .to_return(status: 206)
 
-    rio = described_class.new(url)
-    rio.seek(100)
+        rio = described_class.new(url)
+        rio.seek(10)
 
-    expect { rio.read(100) }.to raise_error(/replied with a 403 and refused/)
+        expect { rio.read(100) }.to raise_error("The server replied with 206 status but no Content-Range at #{url}")
+        expect(stub).to have_been_requested
+      end
+    end
   end
 
-  it 'returns nil on a 416 response' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPRangeNotSatisfiable.new('2', '416', 'Range Not Satisfiable')
+  # 3XX
+
+  [301, 302, 303, 307, 308].each do |code|
+    context "when the response code is #{code}" do
+      context 'when the location header is present and the redirect limit is not exceeded' do
+        context 'when the location is absolute' do
+          it 'redirects to the specified location, without the Authorization header' do
+            redirecting_url = 'https://my_images.invalid/my_image'
+            destination_url = 'https://images.invalid/img.jpg'
+            body = 'response body'
+
+            redirect_stub = stub_request(:get, redirecting_url)
+              .with(headers: { 'Authorization' => 'token', 'range' => 'bytes=10-109' })
+              .to_return(headers: { 'location' => destination_url }, status: code)
+            destination_stub = stub_request(:get, destination_url)
+              .with { |request| request.headers['Range'] == 'bytes=10-109' && !request.headers.key?('Authorization') }
+              .to_return(body: body, status: 200)
+
+            rio = described_class.new(redirecting_url, headers: { 'Authorization' => 'token' })
+            rio.seek(10)
+            read_result = rio.read(100)
+
+            expect(read_result).to eq(body)
+            expect(redirect_stub).to have_been_requested
+            expect(destination_stub).to have_been_requested
+          end
+        end
+
+        context 'when the location is relative' do
+          it 'redirects to the specified location under the same host, with the same Authorization header' do
+            host = 'https://images.invalid'
+            redirecting_path = '/my_image'
+            redirecting_url = host + redirecting_path
+            destination_path = '/img.jpg'
+            destination_url = host + destination_path
+            body = 'response body'
+
+            redirect_stub = stub_request(:get, redirecting_url)
+              .with(headers: { 'Authorization' => 'token', 'range' => 'bytes=10-109' })
+              .to_return(headers: { 'location' => destination_path }, status: code)
+            destination_stub = stub_request(:get, destination_url)
+              .with(headers: { 'Authorization' => 'token', 'range' => 'bytes=10-109' })
+              .to_return(body: body, status: 200)
+
+            rio = described_class.new(redirecting_url, headers: { 'Authorization' => 'token' })
+            rio.seek(10)
+            read_result = rio.read(100)
+
+            expect(read_result).to eq(body)
+            expect(redirect_stub).to have_been_requested
+            expect(destination_stub).to have_been_requested
+          end
+        end
+      end
+
+      context 'when the location header is not present' do
+        it 'raises an error' do
+          url = 'https://images.invalid/my_image'
+
+          stub = stub_request(:get, url)
+            .with(headers: { 'range' => 'bytes=10-109' })
+            .to_return(status: code)
+
+          rio = described_class.new(url)
+          rio.seek(10)
+
+          expect { rio.read(100) }.to raise_error("Server at #{url} replied with a #{code}, indicating redirection; however, the location header was empty.")
+          expect(stub).to have_been_requested
+        end
+      end
+
+      context 'when the redirect limit is exceeded' do
+        it 'raises an error' do
+          redirecting_url = 'https://images.invalid/my_image'
+          destination_url = 'https://images.invalid/img.jpg'
+
+          stub = stub_request(:get, /https:\/\/images\.invalid.*/)
+            .with(headers: { 'range' => 'bytes=10-109' })
+            .to_return(headers: { 'location' => destination_url }, status: code)
+
+          rio = described_class.new(redirecting_url)
+          rio.seek(10)
+
+          expect { rio.read(100) }.to raise_error("Too many redirects; last one to: #{destination_url}")
+          expect(stub).to have_been_requested.times(4)
+        end
+      end
+    end
+  end
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+  # 4XX
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| uri.to_s == url },
-      a_hash_including('range' => 'bytes=100-199')
-    )
+  context 'when the response status code is 416 (Range Not Satisfiable)' do
+    it 'returns nil' do
+      url = 'https://images.invalid/img.jpg'
 
-    rio = described_class.new(url)
-    rio.seek(100)
+      stub = stub_request(:get, url)
+        .with(headers: { 'range' => 'bytes=100-199' })
+        .to_return(status: 416)
 
-    expect(rio.read(100)).to be_nil
-  end
+      rio = described_class.new(url)
+      rio.seek(100)
 
-  it 'sets the status_code of the exception on a 4xx response from upstream' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPForbidden.new('2', '403', 'Forbidden')
+      expect(rio.read(100)).to be_nil
+      expect(stub).to have_been_requested
+    end
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+    it 'does not change pos or size' do
+      url = 'https://images.invalid/img.jpg'
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| uri.to_s == url },
-      a_hash_including('range' => 'bytes=100-199')
-    )
+      stub = stub_request(:get, url)
+        .with(headers: { 'range' => 'bytes=0-0' })
+        .to_return(body: 'response body', headers: { 'Content-Range' => 'bytes 0-0/13' }, status: 206)
 
-    rio = described_class.new(url)
-    rio.seek(100)
-    expect { rio.read(100) }.to(raise_error { |e| expect(e.status_code).to eq(403) })
-  end
+      rio = described_class.new(url)
+      rio.read(1)
 
-  it 'returns a nil when the range cannot be satisfied and the response is 416' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPRangeNotSatisfiable.new('2', '416', 'Range Not Satisfiable')
+      expect(rio.size).to eq(13)
+      expect(stub).to have_been_requested
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+      stub = stub_request(:get, url)
+        .with(headers: { 'range' => 'bytes=100-199' })
+        .to_return(status: 416)
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| uri.to_s == url },
-      a_hash_including('range' => 'bytes=100-199')
-    )
+      rio.seek(100)
+      rio.read(100)
 
-    rio = described_class.new(url)
-    rio.seek(100)
-
-    expect(rio.read(100)).to be_nil
-  end
-
-  it 'does not overwrite size when the range cannot be satisfied and the response is 416' do
-    url = 'https://images.invalid/img.jpg'
-    response_1 = Net::HTTPPartialContent.new('2', '206', 'Partial Content')
-    response_1['Content-Range'] = 'bytes 0-0/13'
-    allow(response_1).to receive(:body).and_return('Response body')
-    response_2 = Net::HTTPRangeNotSatisfiable.new('2', '416', 'Range Not Satisfiable')
-
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response_1, response_2)
-    allow(Net::HTTP).to receive(:request_get).and_return(response_1, response_2)
-
-    expect(Net::HTTP).to receive(:request_get)
-      .with(
-        an_object_satisfying { |uri| uri.to_s == url },
-        a_hash_including('range' => 'bytes=0-0')
-      )
-      .ordered
-    expect(Net::HTTP).to receive(:request_get)
-      .with(
-        an_object_satisfying { |uri| uri.to_s == url },
-        a_hash_including('range' => 'bytes=100-199')
-      )
-      .ordered
-
-    rio = described_class.new(url)
-    rio.read(1)
-
-    expect(rio.size).to eq(13)
-
-    rio.seek(100)
-
-    expect(rio.read(100)).to be_nil
-    expect(rio.size).to eq(13)
+      expect(rio.pos).to eq(100)
+      expect(rio.size).to eq(13)
+      expect(stub).to have_been_requested
+    end
   end
 
-  it 'raises a specific error for all 5xx responses' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPBadGateway.new('2', '502', 'Bad Gateway')
+  [*400..415, *417..499].each do |code|
+    context "when the response status code is #{code}" do
+      it 'raises an error' do
+        url = 'https://images.invalid/img.jpg'
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+        stub = stub_request(:get, url)
+          .with(headers: { 'range' => 'bytes=100-199' })
+          .to_return(status: code)
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| uri.to_s == url },
-      a_hash_including('range' => 'bytes=100-199')
-    )
+        rio = described_class.new(url)
+        rio.seek(100)
 
-    rio = described_class.new(url)
-    rio.seek(100)
-
-    expect { rio.read(100) }.to raise_error(/replied with a 502 and we might want to retry/)
+        expect { rio.read(100) }.to raise_error("Server at #{url} replied with a #{code} and refused our request")
+        expect(stub).to have_been_requested
+      end
+    end
   end
 
-  it 'maintains and exposes #pos' do
-    url = 'https://images.invalid/img.jpg'
-    response = Net::HTTPPartialContent.new('2', '206', 'Partial Content')
-    response['Content-Range'] = 'bytes 0-0/13'
-    allow(response).to receive(:body).and_return('a')
+  # 5XX
+
+  (500..599).each do |code|
+    context "when the response status code is #{code}" do
+      it 'raises an error' do
+        url = 'https://images.invalid/img.jpg'
 
-    allow(Net::HTTP).to receive(:start).and_yield(Net::HTTP).and_return(response)
-    allow(Net::HTTP).to receive(:request_get).and_return(response)
+        stub = stub_request(:get, url)
+          .with(headers: { 'range' => 'bytes=100-199' })
+          .to_return(status: code)
 
-    expect(Net::HTTP).to receive(:request_get).with(
-      an_object_satisfying { |uri| uri.to_s == url },
-      a_hash_including('range' => 'bytes=0-0')
-    )
+        rio = described_class.new(url)
+        rio.seek(100)
 
-    rio = described_class.new(url)
-    expect(rio.pos).to eq(0)
-    rio.read(1)
-    expect(rio.pos).to eq(1)
+        expect { rio.read(100) }.to raise_error("Server at #{url} replied with a #{code} and we might want to retry")
+        expect(stub).to have_been_requested
+      end
+    end
   end
 end

data/spec/spec_helper.rb

@@ -7,6 +7,10 @@ $LOAD_PATH.unshift File.expand_path('../../lib', __FILE__)
 $LOAD_PATH.unshift(File.dirname(__FILE__))
 
 require 'rspec'
+require 'webmock/rspec'
+
+WebMock.disable_net_connect!(allow_localhost: true)
+
 require 'format_parser'
 
 module SpecHelpers

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: format_parser
 version: !ruby/object:Gem::Version
-  version: 2.0.0.pre
+  version: 2.0.0.pre.2
 platform: ruby
 authors:
 - Noah Berman
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2022-10-13 00:00:00.000000000 Z
+date: 2022-10-31 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: exifr
@@ -123,6 +123,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: webmock
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: wetransfer_style
   requirement: !ruby/object:Gem::Requirement