From dfaec6654b14591b40b8e1b90d7ae3e41dacd04a Mon Sep 17 00:00:00 2001
From: Christian Stimming <stimming@tuhh.de>
Date: Tue, 15 Aug 2006 20:00:14 +0000
Subject: [PATCH] Add gnc_utf8_strip_invalid_strdup() that returns a stripped
 copy instead of working in-place. BP

git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@14679 57a11ea4-9604-0410-9ed3-97b8803252fd
---
 ChangeLog                       |  4 ++++
 src/core-utils/gnc-glib-utils.c |  8 ++++++++
 src/core-utils/gnc-glib-utils.h | 14 ++++++++++++++
 3 files changed, 26 insertions(+)

diff --git a/ChangeLog b/ChangeLog
index 5af23ef0f4..2d193daabc 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,9 @@
 2006-08-15  Christian Stimming <stimming@tuhh.de>
 
+	* src/core-utils/gnc-glib-utils.[hc]: Add
+	gnc_utf8_strip_invalid_strdup() that returns a stripped copy
+	instead of working in-place.
+
 	* src/import-export/hbci/dialog-hbcitrans.c (blz_changed_cb):
 	Ensure the encoding of the ktoblzcheck library is used
 	correctly. Their encoding is still always ISO-8859-1, but in
diff --git a/src/core-utils/gnc-glib-utils.c b/src/core-utils/gnc-glib-utils.c
index 5d5c9f6f1b..3acaaf1955 100644
--- a/src/core-utils/gnc-glib-utils.c
+++ b/src/core-utils/gnc-glib-utils.c
@@ -214,3 +214,11 @@ gnc_utf8_strip_invalid (gchar *str)
     memmove(end, end+1, len);	/* shuffle the remainder one byte */
   } while (!gnc_utf8_validate(str, -1, (const gchar **)&end));
 }
+
+gchar *
+gnc_utf8_strip_invalid_strdup(const gchar* str)
+{
+  gchar *result = g_strdup (str);
+  gnc_utf8_strip_invalid (result);
+  return result;
+}
diff --git a/src/core-utils/gnc-glib-utils.h b/src/core-utils/gnc-glib-utils.h
index e1f03b0bbf..c848dfca6b 100644
--- a/src/core-utils/gnc-glib-utils.h
+++ b/src/core-utils/gnc-glib-utils.h
@@ -67,6 +67,20 @@ int safe_utf8_collate (const char *str1, const char *str2);
  *  characters. */
 void gnc_utf8_strip_invalid (gchar *str);
 
+/** Returns a newly allocated copy of the given string but with any
+ * non-utf8 character stripped from it.
+ *
+ * Note that it also removes some subset of invalid XML characters,
+ * too.  See http://www.w3.org/TR/REC-xml/#NT-Char linked from bug
+ * #346535
+ *
+ * @param str A pointer to the string to be copied and stripped of
+ * non-utf8 characters.
+ *
+ * @return A newly allocated string that has to be g_free'd by the
+ * caller. */
+gchar *gnc_utf8_strip_invalid_strdup (const gchar* str);
+
 
 /** @} */