akita-har_logger 0.2.4 → 0.2.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 51fcd2da35cb17ecb365988d7975f746a717f12b002d3a20f1d0c3eab245b10c
-  data.tar.gz: 4972d841dfc7c8dcd41085137d45faedeb450836d46e6d3f14d584785d997a24
+  metadata.gz: d1a952e08b6244134dc4498e24b7185a700fe6d60abdc257bae8e5c3f67a57c6
+  data.tar.gz: 205780425d013375017a8677630c68c078b3be55772ebc0bebf0968d8e16601b
 SHA512:
-  metadata.gz: 8fdffa25d122a573e8c72526934e5e713fbdc573c0d4ebfa13383d749ea856f9b01d2e5bc43e89b1c152df64f97cd965ad54bd5a4ebab924d222a1829bbbec60
-  data.tar.gz: 8ca06a98ce89ec4e9c3e6ac7f3b10ec98f73c44aacab1af0bac8c81649f342ec721017ee57ff6590c93a24234cce9d40c9b341aa8277a39a59bc86a3a9c3484e
+  metadata.gz: 0beadf21aa88256c40d0b6bc4c2e0c54dae7656aea1cd472b8ae7a610484969f70ed9239c3a783362eaf68be41c373085cb181eb456bce413510e8f50d5a50ed
+  data.tar.gz: 622997ac27c899f809b1d7c632ca0da6f8916f00a25b9cdb22d8d6861b1107c8226cfd7802c9b03745d774268780d066db5bcb739081a06c66cd89cee235016a

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    akita-har_logger (0.2.4)
+    akita-har_logger (0.2.5)
 
 GEM
   remote: https://rubygems.org/

data/lib/akita/har_logger/har_utils.rb

@@ -5,12 +5,20 @@ module Akita
     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
-          v
-        else
-          String.new(v).force_encoding(Encoding::UTF_8)
+        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

data/lib/akita/har_logger/http_request.rb

@@ -85,7 +85,7 @@ module Akita
           return Encoding::ISO_8859_1
         end
 
-        Encoding.default_external
+        Encoding::ASCII_8BIT
       end
 
       # Obtains the posted data from an HTTP environment.
@@ -117,8 +117,27 @@ module Akita
             # 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.
-            result[:text] = req.body.string.encode(Encoding::UTF_8,
-                getPostDataCharSet(env))
+            #
+            # 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/version.rb

@@ -2,6 +2,6 @@
 
 module Akita
   module HarLogger
-    VERSION = "0.2.4"
+    VERSION = "0.2.5"
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: akita-har_logger
 version: !ruby/object:Gem::Version
-  version: 0.2.4
+  version: 0.2.5
 platform: ruby
 authors:
 - Jed Liu
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2021-07-11 00:00:00.000000000 Z
+date: 2021-07-12 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec