MARC checkout output: No ()s in comments
[yaz-moved-to-github.git] / src / marcdisp.c
index 3079d81..8a25d60 100644 (file)
@@ -1,8 +1,6 @@
-/*
- * Copyright (C) 1995-2006, Index Data ApS
+/* This file is part of the YAZ toolkit.
+ * Copyright (C) 1995-2009 Index Data
  * See the file LICENSE for details.
- *
- * $Id: marcdisp.c,v 1.40 2006-12-18 10:33:22 adam Exp $
  */
 
 /**
 
 #include <stdarg.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <string.h>
 #include <ctype.h>
 #include <yaz/marcdisp.h>
 #include <yaz/wrbuf.h>
 #include <yaz/yaz-util.h>
 #include <yaz/nmem_xml.h>
+#include <yaz/snprintf.h>
 
 #if YAZ_HAVE_XML2
 #include <libxml/parser.h>
 #include <libxml/tree.h>
 #endif
 
+enum yaz_collection_state {
+    no_collection,
+    collection_first,
+    collection_second
+};
+   
 /** \brief node types for yaz_marc_node */
 enum YAZ_MARC_NODE_TYPE
 { 
@@ -84,6 +90,7 @@ struct yaz_marc_t_ {
     int xml;
     int debug;
     int write_using_libxml2;
+    enum yaz_collection_state enable_collection;
     yaz_iconv_t iconv_cd;
     char subfield_str[8];
     char endline_str[8];
@@ -99,6 +106,7 @@ yaz_marc_t yaz_marc_create(void)
     mt->xml = YAZ_MARC_LINE;
     mt->debug = 0;
     mt->write_using_libxml2 = 0;
+    mt->enable_collection = no_collection;
     mt->m_wr = wrbuf_alloc();
     mt->iconv_cd = 0;
     mt->leader_spec = 0;
@@ -115,7 +123,7 @@ void yaz_marc_destroy(yaz_marc_t mt)
     if (!mt)
         return ;
     nmem_destroy(mt->nmem);
-    wrbuf_free(mt->m_wr, 1);
+    wrbuf_destroy(mt->m_wr);
     xfree(mt->leader_spec);
     xfree(mt);
 }
@@ -125,13 +133,19 @@ NMEM yaz_marc_get_nmem(yaz_marc_t mt)
     return mt->nmem;
 }
 
+static void marc_iconv_reset(yaz_marc_t mt, WRBUF wr)
+{
+    wrbuf_iconv_reset(wr, mt->iconv_cd);
+}
+
 static int marc_exec_leader(const char *leader_spec, char *leader,
                             size_t size);
 
 
 static struct yaz_marc_node *yaz_marc_add_node(yaz_marc_t mt)
 {
-    struct yaz_marc_node *n = nmem_malloc(mt->nmem, sizeof(*n));
+    struct yaz_marc_node *n = (struct yaz_marc_node *)
+        nmem_malloc(mt->nmem, sizeof(*n));
     n->next = 0;
     *mt->nodes_pp = n;
     mt->nodes_pp = &n->next;
@@ -161,19 +175,9 @@ void yaz_marc_cprintf(yaz_marc_t mt, const char *fmt, ...)
 {
     va_list ap;
     char buf[200];
-    va_start(ap, fmt);
 
-#ifdef WIN32
-    _vsnprintf(buf, sizeof(buf)-1, fmt, ap);
-#else
-/* !WIN32 */
-#if HAVE_VSNPRINTF
-    vsnprintf(buf, sizeof(buf), fmt, ap);
-#else
-    vsprintf(buf, fmt, ap);
-#endif
-#endif
-/* WIN32 */
+    va_start(ap, fmt);
+    yaz_vsnprintf(buf, sizeof(buf)-1, fmt, ap);
     yaz_marc_add_comment(mt, buf);
     va_end (ap);
 }
@@ -260,7 +264,8 @@ void yaz_marc_add_subfield(yaz_marc_t mt,
 
     if (mt->subfield_pp)
     {
-        struct yaz_marc_subfield *n = nmem_malloc(mt->nmem, sizeof(*n));
+        struct yaz_marc_subfield *n = (struct yaz_marc_subfield *)
+            nmem_malloc(mt->nmem, sizeof(*n));
         n->code_data = nmem_strdupn(mt->nmem, code_data, code_data_len);
         n->next = 0;
         /* mark subfield_pp to point to this one, so we append here next */
@@ -420,7 +425,7 @@ int yaz_marc_write_check(yaz_marc_t mt, WRBUF wr)
         case YAZ_MARC_COMMENT:
             wrbuf_iconv_write(wr, mt->iconv_cd, 
                               n->u.comment, strlen(n->u.comment));
-            wrbuf_puts(wr, ")\n");
+            wrbuf_puts(wr, "\n");
             break;
         default:
             break;
@@ -429,6 +434,17 @@ int yaz_marc_write_check(yaz_marc_t mt, WRBUF wr)
     return 0;
 }
 
+static size_t get_subfield_len(yaz_marc_t mt, const char *data,
+                               int identifier_length)
+{
+    /* if identifier length is 2 (most MARCs) or less (probably an error),
+       the code is a single character .. However we've
+       seen multibyte codes, so see how big it really is */
+    if (identifier_length > 2)
+        return identifier_length - 1;
+    else
+        return cdata_one_character(mt, data);
+}
 
 int yaz_marc_write_line(yaz_marc_t mt, WRBUF wr)
 {
@@ -458,13 +474,8 @@ int yaz_marc_write_line(yaz_marc_t mt, WRBUF wr)
                          n->u.datafield.indicator);
             for (s = n->u.datafield.subfields; s; s = s->next)
             {
-                /* if identifier length is 2 (most MARCs),
-                   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
-                    :
-                    cdata_one_character(mt, s->code_data);
+                size_t using_code_len = get_subfield_len(mt, s->code_data,
+                                                         identifier_length);
                 
                 wrbuf_puts (wr, mt->subfield_str); 
                 wrbuf_iconv_write(wr, mt->iconv_cd, s->code_data, 
@@ -472,8 +483,7 @@ int yaz_marc_write_line(yaz_marc_t mt, WRBUF wr)
                 wrbuf_iconv_puts(wr, mt->iconv_cd, " ");
                 wrbuf_iconv_puts(wr, mt->iconv_cd, 
                                  s->code_data + using_code_len);
-                wrbuf_iconv_puts(wr, mt->iconv_cd, " ");
-                wr->pos--;
+                marc_iconv_reset(mt, wr);
             }
             wrbuf_puts (wr, mt->endline_str);
             break;
@@ -481,14 +491,14 @@ int yaz_marc_write_line(yaz_marc_t mt, WRBUF wr)
             wrbuf_printf(wr, "%s", n->u.controlfield.tag);
             wrbuf_iconv_puts(wr, mt->iconv_cd, " ");
             wrbuf_iconv_puts(wr, mt->iconv_cd, n->u.controlfield.data);
-            wrbuf_iconv_puts(wr, mt->iconv_cd, " ");
-            wr->pos--;
+            marc_iconv_reset(mt, wr);
             wrbuf_puts (wr, mt->endline_str);
             break;
         case YAZ_MARC_COMMENT:
             wrbuf_puts(wr, "(");
             wrbuf_iconv_write(wr, mt->iconv_cd, 
                               n->u.comment, strlen(n->u.comment));
+            marc_iconv_reset(mt, wr);
             wrbuf_puts(wr, ")\n");
             break;
         case YAZ_MARC_LEADER:
@@ -499,6 +509,28 @@ int yaz_marc_write_line(yaz_marc_t mt, WRBUF wr)
     return 0;
 }
 
+int yaz_marc_write_trailer(yaz_marc_t mt, WRBUF wr)
+{
+    if (mt->enable_collection == collection_second)
+    {
+        switch(mt->xml)
+        {
+        case YAZ_MARC_MARCXML:
+            wrbuf_printf(wr, "</collection>\n");
+            break;
+        case YAZ_MARC_XCHANGE:
+            wrbuf_printf(wr, "</collection>\n");
+            break;
+        }
+    }
+    return 0;
+}
+
+void yaz_marc_enable_collection(yaz_marc_t mt)
+{
+    mt->enable_collection = collection_first;
+}
+
 int yaz_marc_write_mode(yaz_marc_t mt, WRBUF wr)
 {
     switch(mt->xml)
@@ -544,8 +576,18 @@ static int yaz_marc_write_marcxml_ns1(yaz_marc_t mt, WRBUF wr,
         return -1;
     if (!atoi_n_check(leader+11, 1, &identifier_length))
         return -1;
-
-    wrbuf_printf(wr, "<record xmlns=\"%s\"", ns);
+    
+    if (mt->enable_collection != no_collection)
+    {
+        if (mt->enable_collection == collection_first)
+            wrbuf_printf(wr, "<collection xmlns=\"%s\">\n", ns);
+        mt->enable_collection = collection_second;
+        wrbuf_printf(wr, "<record");
+    }
+    else
+    {
+        wrbuf_printf(wr, "<record xmlns=\"%s\"", ns);
+    }
     if (format)
         wrbuf_printf(wr, " format=\"%.80s\"", format);
     if (type)
@@ -576,14 +618,8 @@ static int yaz_marc_write_marcxml_ns1(yaz_marc_t mt, WRBUF wr,
             wrbuf_printf(wr, ">\n");
             for (s = n->u.datafield.subfields; s; s = s->next)
             {
-                /* if identifier length is 2 (most MARCs),
-                   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
-                    :
-                    cdata_one_character(mt, s->code_data);
-                
+                size_t using_code_len = get_subfield_len(mt, s->code_data,
+                                                         identifier_length);
                 wrbuf_iconv_puts(wr, mt->iconv_cd, "    <subfield code=\"");
                 wrbuf_iconv_write_cdata(wr, mt->iconv_cd,
                                         s->code_data, using_code_len);
@@ -591,6 +627,7 @@ static int yaz_marc_write_marcxml_ns1(yaz_marc_t mt, WRBUF wr,
                 wrbuf_iconv_write_cdata(wr, mt->iconv_cd,
                                         s->code_data + using_code_len,
                                         strlen(s->code_data + using_code_len));
+                marc_iconv_reset(mt, wr);
                 wrbuf_iconv_puts(wr, mt->iconv_cd, "</subfield>");
                 wrbuf_puts(wr, "\n");
             }
@@ -601,7 +638,11 @@ static int yaz_marc_write_marcxml_ns1(yaz_marc_t mt, WRBUF wr,
             wrbuf_iconv_write_cdata(wr, mt->iconv_cd, n->u.controlfield.tag,
                                     strlen(n->u.controlfield.tag));
             wrbuf_iconv_puts(wr, mt->iconv_cd, "\">");
-            wrbuf_iconv_puts(wr, mt->iconv_cd, n->u.controlfield.data);
+            wrbuf_iconv_write_cdata(wr, mt->iconv_cd,
+                                    n->u.controlfield.data,
+                                    strlen(n->u.controlfield.data));
+
+            marc_iconv_reset(mt, wr);
             wrbuf_iconv_puts(wr, mt->iconv_cd, "</controlfield>");
             wrbuf_puts(wr, "\n");
             break;
@@ -629,6 +670,7 @@ static int yaz_marc_write_marcxml_ns(yaz_marc_t mt, WRBUF wr,
 {
     if (mt->write_using_libxml2)
     {
+#if YAZ_HAVE_XML2
         int ret;
         xmlNode *root_ptr;
 
@@ -648,6 +690,9 @@ static int yaz_marc_write_marcxml_ns(yaz_marc_t mt, WRBUF wr,
             xmlFreeDoc(doc);
         }
         return ret;
+#else
+        return -1;
+#endif
     }
     else
         return yaz_marc_write_marcxml_ns1(mt, wr, ns, format, type);
@@ -655,6 +700,8 @@ static int yaz_marc_write_marcxml_ns(yaz_marc_t mt, WRBUF wr,
 
 int yaz_marc_write_marcxml(yaz_marc_t mt, WRBUF wr)
 {
+    /* set leader 09 to 'a' for UNICODE */
+    /* http://www.loc.gov/marc/bibliographic/ecbdldrd.html#mrcblea */
     if (!mt->leader_spec)
         yaz_marc_modify_leader(mt, 9, "a");
     return yaz_marc_write_marcxml_ns(mt, wr, "http://www.loc.gov/MARC21/slim",
@@ -666,22 +713,23 @@ int yaz_marc_write_marcxchange(yaz_marc_t mt, WRBUF wr,
                                const char *type)
 {
     return yaz_marc_write_marcxml_ns(mt, wr,
-                                     "http://www.bs.dk/standards/MarcXchange",
+                                     "info:lc/xmlns/marcxchange-v1",
                                      0, 0);
 }
 
 
+#if YAZ_HAVE_XML2
 int yaz_marc_write_xml(yaz_marc_t mt, xmlNode **root_ptr,
                        const char *ns, 
                        const char *format,
                        const char *type)
 {
-#if YAZ_HAVE_XML2
     struct yaz_marc_node *n;
     int identifier_length;
     const char *leader = 0;
     xmlNode *record_ptr;
     xmlNsPtr ns_record;
+    WRBUF wr_cdata = 0;
 
     for (n = mt->nodes; n; n = n->next)
         if (n->which == YAZ_MARC_LEADER)
@@ -695,10 +743,13 @@ int yaz_marc_write_xml(yaz_marc_t mt, xmlNode **root_ptr,
     if (!atoi_n_check(leader+11, 1, &identifier_length))
         return -1;
 
+    wr_cdata = wrbuf_alloc();
+
     record_ptr = xmlNewNode(0, BAD_CAST "record");
     *root_ptr = record_ptr;
 
     ns_record = xmlNewNs(record_ptr, BAD_CAST ns, 0);
+    xmlSetNs(record_ptr, ns_record);
 
     if (format)
         xmlNewProp(record_ptr, BAD_CAST "format", BAD_CAST format);
@@ -730,35 +781,32 @@ int yaz_marc_write_xml(yaz_marc_t mt, xmlNode **root_ptr,
             }
             for (s = n->u.datafield.subfields; s; s = s->next)
             {
-                char code_val[8];
-
                 xmlNode *ptr_subfield;
-                /* if identifier length is 2 (most MARCs),
-                   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
-                    :
-                    cdata_one_character(mt, s->code_data);
-
-                if (using_code_len >= sizeof(code_val)-1)
-                    continue;
-
+                size_t using_code_len = get_subfield_len(mt, s->code_data,
+                                                         identifier_length);
+                wrbuf_rewind(wr_cdata);
+                wrbuf_iconv_puts(wr_cdata, mt->iconv_cd,
+                                 s->code_data + using_code_len);
+                marc_iconv_reset(mt, wr_cdata);
                 ptr_subfield = xmlNewTextChild(
                     ptr, ns_record, 
-                    BAD_CAST "subfield", 
-                    BAD_CAST (s->code_data + using_code_len));
-                
-                memcpy(code_val, s->code_data, using_code_len);
-                code_val[using_code_len] = '\0';
+                    BAD_CAST "subfield",  BAD_CAST wrbuf_cstr(wr_cdata));
 
-                xmlNewProp(ptr_subfield, BAD_CAST "code", BAD_CAST code_val);
+                wrbuf_rewind(wr_cdata);
+                wrbuf_iconv_write(wr_cdata, mt->iconv_cd,
+                                  s->code_data, using_code_len);
+                xmlNewProp(ptr_subfield, BAD_CAST "code",
+                           BAD_CAST wrbuf_cstr(wr_cdata));
             }
             break;
         case YAZ_MARC_CONTROLFIELD:
+            wrbuf_rewind(wr_cdata);
+            wrbuf_iconv_puts(wr_cdata, mt->iconv_cd, n->u.controlfield.data);
+            marc_iconv_reset(mt, wr_cdata);
+            
             ptr = xmlNewTextChild(record_ptr, ns_record,
                                   BAD_CAST "controlfield",
-                                  BAD_CAST n->u.controlfield.data);
+                                  BAD_CAST wrbuf_cstr(wr_cdata));
             
             xmlNewProp(ptr, BAD_CAST "tag", BAD_CAST n->u.controlfield.tag);
             break;
@@ -772,11 +820,10 @@ int yaz_marc_write_xml(yaz_marc_t mt, xmlNode **root_ptr,
             break;
         }
     }
+    wrbuf_destroy(wr_cdata);
     return 0;
-#else
-    return -1;
-#endif
 }
+#endif
 
 int yaz_marc_write_iso2709(yaz_marc_t mt, WRBUF wr)
 {
@@ -826,9 +873,11 @@ int yaz_marc_write_iso2709(yaz_marc_t mt, WRBUF wr)
                 /* write dummy IDFS + content */
                 wrbuf_iconv_putchar(wr_data_tmp, mt->iconv_cd, ' ');
                 wrbuf_iconv_puts(wr_data_tmp, mt->iconv_cd, s->code_data);
+                marc_iconv_reset(mt, wr_data_tmp);
             }
             /* write dummy FS (makes MARC-8 to become ASCII) */
             wrbuf_iconv_putchar(wr_data_tmp, mt->iconv_cd, ' ');
+            marc_iconv_reset(mt, wr_data_tmp);
             data_length += wrbuf_len(wr_data_tmp);
             break;
         case YAZ_MARC_CONTROLFIELD:
@@ -837,7 +886,9 @@ int yaz_marc_write_iso2709(yaz_marc_t mt, WRBUF wr)
             wrbuf_rewind(wr_data_tmp);
             wrbuf_iconv_puts(wr_data_tmp, mt->iconv_cd, 
                              n->u.controlfield.data);
+            marc_iconv_reset(mt, wr_data_tmp);
             wrbuf_iconv_putchar(wr_data_tmp, mt->iconv_cd, ' ');/* field sep */
+            marc_iconv_reset(mt, wr_data_tmp);
             data_length += wrbuf_len(wr_data_tmp);
             break;
         case YAZ_MARC_COMMENT:
@@ -871,9 +922,9 @@ int yaz_marc_write_iso2709(yaz_marc_t mt, WRBUF wr)
     
     wrbuf_write(wr, wrbuf_buf(wr_head), 24);
     wrbuf_write(wr, wrbuf_buf(wr_dir), wrbuf_len(wr_dir));
-    wrbuf_free(wr_head, 1);
-    wrbuf_free(wr_dir, 1);
-    wrbuf_free(wr_data_tmp, 1);
+    wrbuf_destroy(wr_head);
+    wrbuf_destroy(wr_dir);
+    wrbuf_destroy(wr_data_tmp);
 
     for (n = mt->nodes; n; n = n->next)
     {
@@ -888,17 +939,13 @@ int yaz_marc_write_iso2709(yaz_marc_t mt, WRBUF wr)
             {
                 wrbuf_putc(wr, ISO2709_IDFS);
                 wrbuf_iconv_puts(wr, mt->iconv_cd, s->code_data);
-                /* write dummy blank - makes MARC-8 to become ASCII */
-                wrbuf_iconv_putchar(wr, mt->iconv_cd, ' ');
-                wr->pos--;
+                marc_iconv_reset(mt, wr);
             }
             wrbuf_putc(wr, ISO2709_FS);
             break;
         case YAZ_MARC_CONTROLFIELD:
             wrbuf_iconv_puts(wr, mt->iconv_cd, n->u.controlfield.data);
-            /* write dummy blank - makes MARC-8 to become ASCII */
-            wrbuf_iconv_putchar(wr, mt->iconv_cd, ' ');
-            wr->pos--;
+            marc_iconv_reset(mt, wr);
             wrbuf_putc(wr, ISO2709_FS);
             break;
         case YAZ_MARC_COMMENT:
@@ -924,14 +971,14 @@ int yaz_marc_decode_wrbuf(yaz_marc_t mt, const char *buf, int bsize, WRBUF wr)
 }
 
 int yaz_marc_decode_buf (yaz_marc_t mt, const char *buf, int bsize,
-                         char **result, int *rsize)
+                         const char **result, size_t *rsize)
 {
     int r;
 
     wrbuf_rewind(mt->m_wr);
     r = yaz_marc_decode_wrbuf(mt, buf, bsize, mt->m_wr);
     if (result)
-        *result = wrbuf_buf(mt->m_wr);
+        *result = wrbuf_cstr(mt->m_wr);
     if (rsize)
         *rsize = wrbuf_len(mt->m_wr);
     return r;
@@ -954,6 +1001,11 @@ void yaz_marc_iconv(yaz_marc_t mt, yaz_iconv_t cd)
     mt->iconv_cd = cd;
 }
 
+yaz_iconv_t yaz_marc_get_iconv(yaz_marc_t mt)
+{
+    return mt->iconv_cd;
+}
+
 void yaz_marc_modify_leader(yaz_marc_t mt, size_t off, const char *str)
 {
     struct yaz_marc_node *n;
@@ -967,53 +1019,6 @@ void yaz_marc_modify_leader(yaz_marc_t mt, size_t off, const char *str)
         }
 }
 
-/* deprecated */
-int yaz_marc_decode(const char *buf, WRBUF wr, int debug, int bsize, int xml)
-{
-    yaz_marc_t mt = yaz_marc_create();
-    int r;
-
-    mt->debug = debug;
-    mt->xml = xml;
-    r = yaz_marc_decode_wrbuf(mt, buf, bsize, wr);
-    yaz_marc_destroy(mt);
-    return r;
-}
-
-/* deprecated */
-int marc_display_wrbuf (const char *buf, WRBUF wr, int debug, int bsize)
-{
-    return yaz_marc_decode(buf, wr, debug, bsize, 0);
-}
-
-/* deprecated */
-int marc_display_exl (const char *buf, FILE *outf, int debug, int bsize)
-{
-    yaz_marc_t mt = yaz_marc_create();
-    int r;
-
-    mt->debug = debug;
-    r = yaz_marc_decode_wrbuf (mt, buf, bsize, mt->m_wr);
-    if (!outf)
-        outf = stdout;
-    if (r > 0)
-        fwrite (wrbuf_buf(mt->m_wr), 1, wrbuf_len(mt->m_wr), outf);
-    yaz_marc_destroy(mt);
-    return r;
-}
-
-/* deprecated */
-int marc_display_ex (const char *buf, FILE *outf, int debug)
-{
-    return marc_display_exl (buf, outf, debug, -1);
-}
-
-/* deprecated */
-int marc_display (const char *buf, FILE *outf)
-{
-    return marc_display_ex (buf, outf, 0);
-}
-
 int yaz_marc_leader_spec(yaz_marc_t mt, const char *leader_spec)
 {
     xfree(mt->leader_spec);
@@ -1040,7 +1045,7 @@ static int marc_exec_leader(const char *leader_spec, char *leader, size_t size)
         no = sscanf(cp, "%d=%20[^,]%n", &pos, val, &no_read);
         if (no < 2 || no_read < 3)
             return -1;
-        if (pos < 0 || pos >= size)
+        if (pos < 0 || (size_t) pos >= size)
             return -1;
 
         if (*val == '\'')
@@ -1093,6 +1098,7 @@ void yaz_marc_write_using_libxml2(yaz_marc_t mt, int enable)
 /*
  * Local variables:
  * c-basic-offset: 4
+ * c-file-style: "Stroustrup"
  * indent-tabs-mode: nil
  * End:
  * vim: shiftwidth=4 tabstop=8 expandtab