From dc5874f5aa891dcb94b3afd38e25c0ae8a5ab56c Mon Sep 17 00:00:00 2001
From: Yorick Peterse <yorickpeterse@gmail.com>
Date: Wed, 3 Sep 2014 22:52:59 +0200
Subject: [PATCH] Show XML parsing errors when using IO as input.

Previously this wouldn't display anything due to the IO object being exhausted.
To fix this the input has to be wound back to the start, which means re-reading
it. Sadly I can't think of a way around this that doesn't require buffering
lines while parsing them (which massively increases memory usage).
---
 lib/oga/xml/parser.y              | 44 +++++++++++++++++--------------
 spec/oga/xml/parser/error_spec.rb | 16 ++++++++++-
 2 files changed, 39 insertions(+), 21 deletions(-)

diff --git a/lib/oga/xml/parser.y b/lib/oga/xml/parser.y
index ff4f4c9..afddde6 100644
--- a/lib/oga/xml/parser.y
+++ b/lib/oga/xml/parser.y
@@ -228,31 +228,35 @@ end
   # @raise [Racc::ParseError]
   #
   def on_error(type, value, stack)
-    name  = token_to_str(type)
-    index = @line - 1
-    lines = @data.each_line.to_a
-    code  = ''
+    name        = token_to_str(type)
+    index       = @line - 1
+    index_range = (index - 5)..(index + 5)
+    code        = ''
+
+    # For IO we sadly have to re-read the input :<
+    if @data.respond_to?(:rewind)
+      @data.rewind
+    end
 
     # Show up to 5 lines before and after the offending line (if they exist).
-    (-5..5).each do |offset|
-      line   = lines[index + offset]
-      number = @line + offset
+    @data.each_line.with_index do |line, line_index|
+      next unless index_range.cover?(line_index)
 
-      if line and number > 0
-        if offset == 0
-          prefix = '=> '
-        else
-          prefix = '   '
-        end
+      number = line_index + 1
 
-        line = line.strip
-
-        if line.length > 80
-          line = line[0..79] + ' (more)'
-        end
-
-        code << "#{prefix}#{number}: #{line}\n"
+      if line_index == index
+        prefix = '=> '
+      else
+        prefix = '   '
       end
+
+      line = line.strip
+
+      if line.length > 80
+        line = line[0..79] + ' (more)'
+      end
+
+      code << "#{prefix}#{number}: #{line}\n"
     end
 
     raise Racc::ParseError, <<-EOF.strip
diff --git a/spec/oga/xml/parser/error_spec.rb b/spec/oga/xml/parser/error_spec.rb
index dbbd07b..617864b 100644
--- a/spec/oga/xml/parser/error_spec.rb
+++ b/spec/oga/xml/parser/error_spec.rb
@@ -16,7 +16,7 @@ describe Oga::XML::Parser do
       expect { parse(@invalid_xml) }.to raise_error(Racc::ParseError)
     end
 
-    example 'display a more meaningful error message' do
+    example 'include the offending input when using String as input' do
       # Racc basically reports errors at the last moment instead of where they
       # *actually* occur.
       partial = <<-EOF.strip
@@ -29,5 +29,19 @@ describe Oga::XML::Parser do
 
       parse_error(@invalid_xml).should =~ /#{partial}/
     end
+
+    example 'include the offending input when using IO as input' do
+      # Racc basically reports errors at the last moment instead of where they
+      # *actually* occur.
+      partial = <<-EOF.strip
+   1. <person>
+   2. <name>Alice</name>
+   3. <age>25
+   4. <nationality>Dutch</nationality>
+=> 5. </person>
+      EOF
+
+      parse_error(StringIO.new(@invalid_xml)).should =~ /#{partial}/
+    end
   end
 end