Fixed bug #2352: yaz-marcdump crashes for certain record.

diff --git a/src/marcdisp.c b/src/marcdisp.c
index a11f9dc..a5084ce 100644 (file)
--- a/src/marcdisp.c
+++ b/src/marcdisp.c
@@ -467,7 +467,7 @@ int yaz_marc_write_line(yaz_marc_t mt, WRBUF wr)
                    the code is a single character .. However we've
                    seen multibyte codes, so see how big it really is */
                 size_t using_code_len = 
                    the code is a single character .. However we've
                    seen multibyte codes, so see how big it really is */
                 size_t using_code_len = 

-                    (identifier_length != 2) ? identifier_length - 1
+                    (identifier_length > 2) ? identifier_length - 1

                     :
                     cdata_one_character(mt, s->code_data);
                 
                     :
                     cdata_one_character(mt, s->code_data);
                 

@@ -616,7 +616,7 @@ static int yaz_marc_write_marcxml_ns1(yaz_marc_t mt, WRBUF wr,
                    the code is a single character .. However we've
                    seen multibyte codes, so see how big it really is */
                 size_t using_code_len = 
                    the code is a single character .. However we've
                    seen multibyte codes, so see how big it really is */
                 size_t using_code_len = 

-                    (identifier_length != 2) ? identifier_length - 1
+                    (identifier_length > 2) ? identifier_length - 1

                     :
                     cdata_one_character(mt, s->code_data);
                 
                     :
                     cdata_one_character(mt, s->code_data);
                 

@@ -786,7 +786,7 @@ int yaz_marc_write_xml(yaz_marc_t mt, xmlNode **root_ptr,
                    the code is a single character .. However we've
                    seen multibyte codes, so see how big it really is */
                 size_t using_code_len = 
                    the code is a single character .. However we've
                    seen multibyte codes, so see how big it really is */
                 size_t using_code_len = 

-                    (identifier_length != 2) ? identifier_length - 1
+                    (identifier_length > 2) ? identifier_length - 1

                     :
                     cdata_one_character(mt, s->code_data);
 
                     :
                     cdata_one_character(mt, s->code_data);