fix string lookups when cells have complicated formatting (#6)

if a cell has multiple different formats, this shows up as multiple
different rich text parts in the shared string table, so make sure to
only count each shared string table entry once

1 files changed, 5 insertions, 2 deletions

diff --git a/lib/Spreadsheet/ParseXLSX.pm b/lib/Spreadsheet/ParseXLSX.pm
index 1aa14eb..6efc3b8 100644
--- a/lib/Spreadsheet/ParseXLSX.pm
+++ b/lib/Spreadsheet/ParseXLSX.pm
@@ -238,8 +238,11 @@ sub _parse_shared_strings {
 
     return [
         map {
-            { Text => $_->text } # XXX are Unicode, Rich, or Ext important?
-        } $strings->find_nodes('//t')
+            my $node = $_;
+            # XXX this discards information about formatting within cells
+            # not sure how to represent that
+            { Text => join('', map { $_->text } $node->find_nodes('t')) }
+        } $strings->find_nodes('//si')
     ];
 }