akita-har_logger 0.2.0 → 0.2.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3da46a0fcf41a0f5f1f1d88261e7345713fe9487d7a7c4485d85e0ee808c8717
-  data.tar.gz: 8857a59375d849bf4c94b47a131c90a950f5bc06da6979c75815d3a6f1c0c4d9
+  metadata.gz: d1a952e08b6244134dc4498e24b7185a700fe6d60abdc257bae8e5c3f67a57c6
+  data.tar.gz: 205780425d013375017a8677630c68c078b3be55772ebc0bebf0968d8e16601b
 SHA512:
-  metadata.gz: a2a17be7f105d38c3f75b70bd42dd8982c189a129895ca887333fc3609d2e3f1c362fddffd7a77f97bd43358bc907982ea6b718f96515297cf9b65c437073a0d
-  data.tar.gz: 1919850a9c0be28cd22610699a17caec16bafa2b8bf9485ab4b0f7646fbf77a2b4c3294f676033ab6c14258bf52a7ffa08dd9fe3c7e0fe729c90fa228d2acfa2
+  metadata.gz: 0beadf21aa88256c40d0b6bc4c2e0c54dae7656aea1cd472b8ae7a610484969f70ed9239c3a783362eaf68be41c373085cb181eb456bce413510e8f50d5a50ed
+  data.tar.gz: 622997ac27c899f809b1d7c632ca0da6f8916f00a25b9cdb22d8d6861b1107c8226cfd7802c9b03745d774268780d066db5bcb739081a06c66cd89cee235016a

data/.gitignore

@@ -0,0 +1 @@
+/pkg/

data/Gemfile.lock

@@ -1,14 +1,12 @@
 PATH
   remote: .
   specs:
-    akita-har_logger (0.2.0)
-      json (~> 2.3)
+    akita-har_logger (0.2.5)
 
 GEM
   remote: https://rubygems.org/
   specs:
     diff-lcs (1.4.4)
-    json (2.5.1)
     rake (13.0.3)
     rspec (3.10.0)
       rspec-core (~> 3.10.0)
@@ -33,4 +31,4 @@ DEPENDENCIES
   rspec (~> 3.10)
 
 BUNDLED WITH
-   2.2.15
+   2.2.23

data/README.md

@@ -48,6 +48,8 @@ Here is a sample configuration for a test environment that just adds the
 instrumentation.
 
 ```ruby
+# config/environments/test.rb
+
 Rails.application.configure.do
   # Other configuration for the Rails application...
 
@@ -61,11 +63,30 @@ end
 ### `ActionController` filter
 
 To instrument with a filter, add an instance of `Akita::HarLogger::Filter` as
-an `around_action` filter to your `ActionController` implementation. Here is an
-example of a bare-bones `app/controllers/application_controller.rb` with this
-instrumentation.
+an `around_action` filter to your `ActionController` implementation.
+
+For convenience, you can call `Akita::HarLogger::Filter.install` to do this for
+all `ActionController`s in your application. We recommend adding this call to a
+configuration initializer. For example, this initializer adds the filter only
+in the test environment:
 
 ```ruby
+# config/initializers/har_logging.rb
+
+# Add the HAR logger as an `around_action` filter to all `ActionControllers`
+# that are loaded by the application in the test environment. Optionally give
+# an output HAR file to save your trace. If not specified, this defaults to
+# `akita_trace_{timestamp}.har`.
+Akita::HarLogger::Filter.install("akita_trace.har") if Rails.env.test?
+```
+
+You can also selectively instrument your `ActionController` implementations by
+adding the filter manually. Here is a bare-bones `ActionController`
+implementation that adds the filter only in the test environment.
+
+```ruby
+# app/controllers/application_controller.rb
+
 class ApplicationController < ActionController::API
   include Response
   include ExceptionHandler
@@ -73,7 +94,7 @@ class ApplicationController < ActionController::API
   # Add the HAR logger as an `around_action` filter. Optionally give an output
   # HAR file to save your trace. If not specified, this defaults to
   # `akita_trace_{timestamp}.har`.
-  around_action Akita::HarLogger::Filter.new("akita_trace.har")
+  around_action Akita::HarLogger::Filter.new("akita_trace.har") if Rails.env.test?
 end
 ```
 

data/akita-har_logger.gemspec

@@ -31,7 +31,5 @@ Gem::Specification.new do |spec|
   spec.executables   = spec.files.grep(%r{\Aexe/}) { |f| File.basename(f) }
   spec.require_paths = ['lib']
 
-  spec.add_dependency 'json', '~> 2.3'
-
   spec.add_development_dependency 'rspec', '~> 3.10'
 end

data/lib/akita/har_logger.rb

@@ -29,14 +29,10 @@ module Akita
         @app = app
 
         if out_file_name == nil then
-          out_file_name = "akita_trace_#{Time.now.to_i}.har"
+          out_file_name = HarLogger.default_file_name
         end
 
-        # This queue is used to ensure that event logging is thread-safe. The
-        # main thread will enqueue HarEntry objects. The HAR writer thread
-        # below dequeues these objects and writes them to the output file.
-        @entry_queue = Queue.new
-        WriterThread.new out_file_name, @entry_queue
+        @entry_queue = HarLogger.get_queue(out_file_name)
       end
 
       def call(env)
@@ -58,16 +54,21 @@ module Akita
     class Filter
       def initialize(out_file_name = nil)
         if out_file_name == nil then
-          out_file_name = "akita_trace_#{Time.now.to_i}.har"
+          out_file_name = HarLogger.default_file_name
         end
 
-        # This queue is used to ensure that event logging is thread-safe. The
-        # main thread will enqueue HarEntry objects. The HAR writer thread
-        # below dequeues these objects and writes them to the output file.
-        @entry_queue = Queue.new
-        WriterThread.new out_file_name, @entry_queue
+        @entry_queue = HarLogger.get_queue(out_file_name)
       end
 
+      # Registers an `on_load` initializer to add a logging filter to any
+      # ActionController that is created.
+      def self.install(out_file_name = nil, hook_name = :action_controller)
+        ActiveSupport.on_load(hook_name) do
+          around_action Filter.new(out_file_name)
+        end
+      end
+
+      # Implements the actual `around` filter.
       def around(controller)
         start_time = Time.now
 
@@ -84,5 +85,35 @@ module Akita
                                       [response.body])
       end
     end
+
+    @@default_file_name = "akita_trace_#{Time.now.to_i}.har"
+    def self.default_file_name
+      @@default_file_name
+    end
+
+    # Maps the name of each output file to a queue of entries to be logged to
+    # that file. The queue is used to ensure that event logging is thread-safe.
+    # The main thread will enqueue HarEntry objects. A HAR writer thread
+    # dequeues these objects and writes them to the output file.
+    @@entry_queues = {}
+    @@entry_queues_mutex = Mutex.new
+
+    # Returns the entry queue for the given file. If an entry queue doesn't
+    # already exist, one is created and a HAR writer thread is started for the
+    # queue.
+    def self.get_queue(out_file_name)
+      queue = nil
+      @@entry_queues_mutex.synchronize {
+        if @@entry_queues.has_key?(out_file_name) then
+          return @@entry_queues[out_file_name]
+        end
+
+        queue = Queue.new
+        @@entry_queues[out_file_name] = queue
+      }
+
+      WriterThread.new out_file_name, queue
+      return queue
+    end
   end
 end

data/lib/akita/har_logger/har_entry.rb

@@ -1,6 +1,5 @@
 # frozen_string_literal: true
 
-require 'json'
 require_relative 'http_request'
 require_relative 'http_response'
 

data/lib/akita/har_logger/har_utils.rb

@@ -3,17 +3,46 @@
 module Akita
   module HarLogger
     class HarUtils
+      # Rack apparently uses 8-bit ASCII for everything, even when the string
+      # is not 8-bit ASCII. This reinterprets 8-bit ASCII strings as UTF-8.
+      #
+      # If we are unable to do this reinterpretation, return the string
+      # unchanged, but log a warning that points to the caller.
+      def self.fixEncoding(v)
+        if v != nil && v.encoding == Encoding::ASCII_8BIT then
+          forced = String.new(v).force_encoding(Encoding::UTF_8)
+          if forced.valid_encoding? then
+            v = forced
+          else
+            Rails.logger.warn "[#{caller_locations(1, 1)}] Unable to fix encoding: not a valid UTF-8 string. This will likely cause JSON serialization to fail."
+          end
+        end
+
+        v
+      end
+
       # Converts a Hash into a list of Hash objects. Each entry in the given
       # Hash will be represented in the output by a Hash object that maps
       # 'name' to the entry's key and 'value' to the entry's value.
       def self.hashToList(hash)
         hash.reduce([]) { |accum, (k, v)|
           accum.append({
-            name: k,
-            value: v,
+            name: fixEncoding(k),
+            value: fixEncoding(v),
           })
         }
       end
+
+      # Determines whether all values in a Hash are strings.
+      def self.allValuesAreStrings(hash)
+        hash.each do |_, value|
+          if !(value.is_a? String) then
+            return false
+          end
+        end
+
+        return true
+      end
     end
   end
 end

data/lib/akita/har_logger/http_request.rb

@@ -1,6 +1,5 @@
 # frozen_string_literal: true
 
-require 'json'
 require_relative 'har_utils'
 
 module Akita
@@ -12,7 +11,7 @@ module Akita
 
         @self = {
           method: getMethod(env),
-          url: req.url,
+          url: HarUtils.fixEncoding(req.url),
           httpVersion: getHttpVersion(env),
           cookies: getCookies(env),
           headers: getHeaders(env),
@@ -34,7 +33,7 @@ module Akita
 
       # Obtains the client's request method from an HTTP environment.
       def getMethod(env)
-        (Rack::Request.new env).request_method
+        HarUtils.fixEncoding (Rack::Request.new env).request_method
       end
 
       # Obtains the client-requested HTTP version from an HTTP environment.
@@ -42,7 +41,9 @@ module Akita
         # The environment doesn't have HTTP_VERSION when running with `rspec`;
         # assume HTTP/1.1 when this happens. We don't return nil, so we can
         # calculate the size of the headers.
-        env.key?('HTTP_VERSION') ? env['HTTP_VERSION'] : 'HTTP/1.1'
+        env.key?('HTTP_VERSION') ?
+          HarUtils.fixEncoding(env['HTTP_VERSION']) :
+          'HTTP/1.1'
       end
 
       # Builds a list of cookie objects from an HTTP environment.
@@ -72,23 +73,71 @@ module Akita
         HarUtils.hashToList paramMap
       end
 
+      # Obtains the character set of the posted data from an HTTP environment.
+      def getPostDataCharSet(env)
+        req = Rack::Request.new env
+        if req.content_charset != nil then
+          return req.content_charset
+        end
+
+        # RFC 2616 says that "text/*" defaults to ISO-8859-1.
+        if env['CONTENT_TYPE'].start_with?('text/') then
+          return Encoding::ISO_8859_1
+        end
+
+        Encoding::ASCII_8BIT
+      end
+
       # Obtains the posted data from an HTTP environment.
       def getPostData(env)
         if env.key?('CONTENT_TYPE') && env['CONTENT_TYPE'] then
           result = { mimeType: env['CONTENT_TYPE'] }
 
           # Populate 'params' if we have URL-encoded parameters. Otherwise,
-          # populate 'text.
+          # populate 'text'.
           req = Rack::Request.new env
           if env['CONTENT_TYPE'] == 'application/x-www-form-urlencoded' then
-            # Decoded parameters can be found as a map in req.params. Convert
-            # this map into an array.
+            # Decoded parameters can be found as a map in req.params.
             #
-            # XXX Spec has space for files, but are file uploads ever
-            # URL-encoded?
-            result[:params] = HarUtils.hashToList req.params
+            # Requests originating from specs can be malformed: the values in
+            # req.params are not necessarily strings. Encode all of req.params
+            # in JSON and pretend the content type was "application/json".
+            if HarUtils.allValuesAreStrings req.params then
+              # Convert req.params into an array.
+              #
+              # XXX Spec has space for files, but are file uploads ever
+              # URL-encoded?
+              result[:params] = HarUtils.hashToList req.params
+            else
+              result[:mimeType] = 'application/json'
+              result[:text] = req.params.to_json
+            end
           else
-            result[:text] = req.body.string
+            # Rack has been observed to use ASCII-8BIT encoding for the request
+            # body when the request specifies UTF-8. Reinterpret the content
+            # body according to what the request says it is, and re-encode into
+            # UTF-8.
+            #
+            # Gracefully handle any characters that are invalid in the source
+            # encoding and characters that have no UTF-8 representation by
+            # replacing with '?'. Log a warning when this happens.
+            source = req.body.string.force_encoding(getPostDataCharSet(env))
+            utf8EncodingSuccessful = false
+            if source.valid_encoding? then
+              begin
+                result[:text] = source.encode(Encoding::UTF_8)
+                utf8EncodingSuccessful = true
+              rescue Encoding::UndefinedConversionError
+                Rails.logger.warn "[#{caller_locations(0, 1)}] Unable to losslessly convert request body from #{source.encoding} to UTF-8. Characters undefined in UTF-8 will be replaced with '?'."
+              end
+            else
+              Rails.logger.warn "[#{caller_locations(0, 1)}] Request body is not valid #{source.encoding}. Invalid characters and characters undefined in UTF-8 will be replaced with '?'."
+            end
+
+            if !utf8EncodingSuccessful then
+              result[:text] = source.encode(Encoding::UTF_8,
+                  invalid: :replace, undef: :replace, replace: '?')
+            end
           end
 
           result

data/lib/akita/har_logger/http_response.rb

@@ -1,6 +1,5 @@
 # frozen_string_literal: true
 
-require 'json'
 require_relative 'har_utils'
 
 module Akita
@@ -26,7 +25,7 @@ module Akita
 
       # Obtains the status text corresponding to a status code.
       def getStatusText(status)
-        Rack::Utils::HTTP_STATUS_CODES[status]
+        HarUtils.fixEncoding(Rack::Utils::HTTP_STATUS_CODES[status])
       end
 
       # Obtains the HTTP version in the response.
@@ -37,7 +36,9 @@ module Akita
         # The environment doesn't have HTTP_VERSION when running with `rspec`;
         # assume HTTP/1.1 when this happens. We don't return nil, so we can
         # calculate the size of the headers.
-        env.key?('HTTP_VERSION') ? env['HTTP_VERSION'] : 'HTTP/1.1'
+        env.key?('HTTP_VERSION') ?
+          HarUtils.fixEncoding(env['HTTP_VERSION']) :
+          'HTTP/1.1'
       end
 
       def getCookies(headers)
@@ -64,8 +65,8 @@ module Akita
           if match then cookie_value = match[1] end
 
           result << {
-            name: cookie_name,
-            value: cookie_value,
+            name: HarUtils.fixEncoding(cookie_name),
+            value: HarUtils.fixEncoding(cookie_value),
           }
         }
 
@@ -78,14 +79,14 @@ module Akita
         text = +""
         body.each { |part|
           # XXX Figure out how to join together multi-part bodies.
-          text << part;
+          text << (HarUtils.fixEncoding part);
         }
 
         {
           size: getBodySize(body),
 
           # XXX What to use when no Content-Type is given?
-          mimeType: headers['Content-Type'],
+          mimeType: HarUtils.fixEncoding(headers['Content-Type']),
 
           text: text,
         }
@@ -94,7 +95,9 @@ module Akita
       def getRedirectUrl(headers)
         # Use the "Location" header if it exists. Otherwise, based on some HAR
         # examples found online, it looks like an empty string is used.
-        headers.key?('Location') ? headers['Location'] : ''
+        headers.key?('Location') ?
+          HarUtils.fixEncoding(headers['Location']) :
+          ''
       end
 
       def getHeadersSize(env, status, headers)

data/lib/akita/har_logger/version.rb

@@ -2,6 +2,6 @@
 
 module Akita
   module HarLogger
-    VERSION = "0.2.0"
+    VERSION = "0.2.5"
   end
 end

data/lib/akita/har_logger/writer_thread.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require 'json'
-
 module Akita
   module HarLogger
     # A thread that consumes HarEntry objects from a queue and writes them to a

metadata

@@ -1,29 +1,15 @@
 --- !ruby/object:Gem::Specification
 name: akita-har_logger
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.5
 platform: ruby
 authors:
 - Jed Liu
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2021-04-17 00:00:00.000000000 Z
+date: 2021-07-12 00:00:00.000000000 Z
 dependencies:
-- !ruby/object:Gem::Dependency
-  name: json
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '2.3'
-  type: :runtime
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '2.3'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -46,6 +32,7 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- ".gitignore"
 - Gemfile
 - Gemfile.lock
 - LICENSE
@@ -82,7 +69,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.2.15
+rubygems_version: 3.2.21
 signing_key:
 specification_version: 4
 summary: Rails middleware for HAR logging