From 1179278b7800633f40f940c3b2b4bdee1fdc1721 Mon Sep 17 00:00:00 2001
From: Alfonso Uceda Pompa <alfonso@bebanjo.com>
Date: Fri, 8 Apr 2016 17:15:36 +0200
Subject: [PATCH] Removed hyperlinks when parsing files in streaming mode

---
 CHANGELOG.md                | 4 ++++
 lib/roo/excelx.rb           | 1 +
 lib/roo/excelx/sheet_doc.rb | 9 +++++++--
 3 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 3d330b9c..c4f1848e 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,7 @@
+## [Unreleased]
+### Added
+- Discard hiperlinks lookups to allow streaming parsing without loading whole files
+
 ## [2.4.0] 2016-05-14
 ### Fixed  
 - Fixed opening spreadsheets with charts [315](https://github.com/roo-rb/roo/pull/315)
diff --git a/lib/roo/excelx.rb b/lib/roo/excelx.rb
index e120d85a..369d5e38 100644
--- a/lib/roo/excelx.rb
+++ b/lib/roo/excelx.rb
@@ -36,6 +36,7 @@ def initialize(filename_or_stream, options = {})
       cell_max = options.delete(:cell_max)
       sheet_options = {}
       sheet_options[:expand_merged_ranges] = (options[:expand_merged_ranges] || false)
+      sheet_options[:no_hyperlinks] = (options[:no_hyperlinks] || false)
 
       unless is_stream?(filename_or_stream)
         file_type_check(filename_or_stream, %w[.xlsx .xlsm], 'an Excel 2007', file_warning, packed)
diff --git a/lib/roo/excelx/sheet_doc.rb b/lib/roo/excelx/sheet_doc.rb
index 9788c1d2..f8d0cae3 100755
--- a/lib/roo/excelx/sheet_doc.rb
+++ b/lib/roo/excelx/sheet_doc.rb
@@ -39,8 +39,13 @@ def each_row_streaming(&block)
       def each_cell(row_xml)
         return [] unless row_xml
         row_xml.children.each do |cell_element|
-          key = ::Roo::Utils.ref_to_key(cell_element['r'])
-          yield cell_from_xml(cell_element, hyperlinks(@relationships)[key])
+          # If you're sure you're not going to need this hyperlinks you can discard it
+          hyperlinks = unless @options[:no_hyperlinks]
+                         key = ::Roo::Utils.ref_to_key(cell_element['r'])
+                         hyperlinks(@relationships)[key]
+                       end
+
+          yield cell_from_xml(cell_element, hyperlinks)
         end
       end