X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=util%2Fmarcdump.c;h=d441855ae8e5f41729ecab8c507f5d6aacde28f2;hb=05a79698b4956c63642ebd0b4deedf56098c255c;hp=43ce28fddfa9b4495eb19480d45940d6e743aa82;hpb=635dd986bff70edc8e8c39067bcc702e59215b56;p=yaz-moved-to-github.git diff --git a/util/marcdump.c b/util/marcdump.c index 43ce28f..d441855 100644 --- a/util/marcdump.c +++ b/util/marcdump.c @@ -1,20 +1,36 @@ /* - * Copyright (c) 1995-2002, Index Data + * Copyright (c) 1995-2004, Index Data * See the file LICENSE for details. - * Sebastian Hammer, Adam Dickmeiss * - * $Id: marcdump.c,v 1.15 2002-03-18 18:11:45 adam Exp $ + * $Id: marcdump.c,v 1.25 2004-08-07 08:18:20 adam Exp $ */ #if HAVE_CONFIG_H #include #endif +#if HAVE_XML2 +#include +#include + +#include +#include + +#endif + #include #include #include #include -#include +#include + +#if HAVE_LOCALE_H +#include +#endif +#if HAVE_LANGINFO_H +#include +#endif + #include #include #include @@ -29,12 +45,64 @@ static void usage(const char *prog) { - fprintf (stderr, "Usage: %s [-c cfile] [-x] [-v] file...\n", prog); + fprintf (stderr, "Usage: %s [-c cfile] [-f from] [-t to] [-x] [-O] [-X] [-I] [-v] file...\n", + prog); } +#if HAVE_XML2 +void print_xpath_nodes(xmlNodeSetPtr nodes, FILE* output) { + xmlNodePtr cur; + int size; + int i; + + assert(output); + size = (nodes) ? nodes->nodeNr : 0; + + fprintf(output, "Result (%d nodes):\n", size); + for(i = 0; i < size; ++i) { + assert(nodes->nodeTab[i]); + + if(nodes->nodeTab[i]->type == XML_NAMESPACE_DECL) + { + xmlNsPtr ns; + + ns = (xmlNsPtr)nodes->nodeTab[i]; + cur = (xmlNodePtr)ns->next; + if(cur->ns) { + fprintf(output, "= namespace \"%s\"=\"%s\" for node %s:%s\n", + ns->prefix, ns->href, cur->ns->href, cur->name); + } else { + fprintf(output, "= namespace \"%s\"=\"%s\" for node %s\n", + ns->prefix, ns->href, cur->name); + } + } + else if(nodes->nodeTab[i]->type == XML_ELEMENT_NODE) + { + cur = nodes->nodeTab[i]; + if(cur->ns) { + fprintf(output, "= element node \"%s:%s\"\n", + cur->ns->href, cur->name); + } + else + { + fprintf(output, "= element node \"%s\"\n", + cur->name); + } + } + else + { + cur = nodes->nodeTab[i]; + fprintf(output, "= node \"%s\": type %d\n", cur->name, cur->type); + } + } +} +#endif + int main (int argc, char **argv) { int r; + int libxml_dom_test = 0; + int print_offset = 0; char *arg; int verbose = 0; FILE *inf; @@ -43,75 +111,177 @@ int main (int argc, char **argv) int no = 0; int xml = 0; FILE *cfile = 0; + char *from = 0, *to = 0; + int num = 1; + +#if HAVE_LOCALE_H + setlocale(LC_CTYPE, ""); +#endif +#if HAVE_LANGINFO_H +#ifdef CODESET + to = nl_langinfo(CODESET); +#endif +#endif - while ((r = options("vc:x", argv, argc, &arg)) != -2) + while ((r = options("pvc:xOXIf:t:2", argv, argc, &arg)) != -2) { int count; no++; switch (r) { + case 'f': + from = arg; + break; + case 't': + to = arg; + break; case 'c': if (cfile) fclose (cfile); cfile = fopen (arg, "w"); break; case 'x': - xml = 1; + xml = YAZ_MARC_SIMPLEXML; break; + case 'O': + xml = YAZ_MARC_OAIMARC; + break; + case 'X': + xml = YAZ_MARC_MARCXML; + break; + case 'I': + xml = YAZ_MARC_ISO2709; + break; + case 'p': + print_offset = 1; + break; + case '2': + libxml_dom_test = 1; + break; case 0: - inf = fopen (arg, "r"); + inf = fopen (arg, "rb"); count = 0; if (!inf) { fprintf (stderr, "%s: cannot open %s:%s\n", prog, arg, strerror (errno)); - exit (1); + exit(1); } if (cfile) fprintf (cfile, "char *marc_records[] = {\n"); - while (1) - { - WRBUF wr = wrbuf_alloc(); + if (1) + { + yaz_marc_t mt = yaz_marc_create(); + yaz_iconv_t cd = 0; - int len; - - r = fread (buf, 1, 5, inf); - if (r < 5) - break; - len = atoi_n(buf, 5); - if (len < 25 || len > 100000) - break; - len = len - 5; - r = fread (buf + 5, 1, len, inf); - if (r < len) - break; - r = yaz_marc_decode (buf, wr, verbose, -1, xml); - if (r <= 0) - break; - fwrite (wrbuf_buf(wr), wrbuf_len(wr), 1, stdout); - if (cfile) - { - char *p = buf; - int i; - if (count) - fprintf (cfile, ","); - fprintf (cfile, "{\n"); - for (i = 0; i < r; i++) + if (from && to) + { + cd = yaz_iconv_open(to, from); + if (!cd) + { + fprintf(stderr, "conversion from %s to %s " + "unsupported\n", from, to); + exit(2); + } + yaz_marc_iconv(mt, cd); + } + yaz_marc_xml(mt, xml); + yaz_marc_debug(mt, verbose); + while (1) + { + int len; + char *result; + int rlen; + + r = fread (buf, 1, 5, inf); + if (r < 5) + { + if (r && print_offset) + printf ("Extra %d bytes", r); + break; + } + if (print_offset) + { + long off = ftell(inf); + printf ("Record %d offset %ld\n", num, (long) off); + } + len = atoi_n(buf, 5); + if (len < 25 || len > 100000) + break; + len = len - 5; + r = fread (buf + 5, 1, len, inf); + if (r < len) + break; + r = yaz_marc_decode_buf (mt, buf, -1, &result, &rlen); + if (r <= 0) + break; + fwrite (result, rlen, 1, stdout); +#if HAVE_XML2 + if (libxml_dom_test) { - if ((i & 15) == 0) - fprintf (cfile, " \""); - fprintf (cfile, "\\x%02X", p[i] & 255); - - if (i < r - 1 && (i & 15) == 15) - fprintf (cfile, "\"\n"); - + xmlDocPtr doc = xmlParseMemory(result, rlen); + if (!doc) + fprintf(stderr, "xmLParseMemory failed\n"); + else + { + int i; + xmlXPathContextPtr xpathCtx; + xmlXPathObjectPtr xpathObj; + static const char *xpathExpr[] = { + "/record/datafield[@tag='245']/subfield[@code='a']", + "/record/datafield[@tag='100']/subfield", + "/record/datafield[@tag='245']/subfield[@code='a']", + "/record/datafield[@tag='650']/subfield", + "/record/datafield[@tag='650']", + 0}; + + xpathCtx = xmlXPathNewContext(doc); + + for (i = 0; xpathExpr[i]; i++) { + xpathObj = xmlXPathEvalExpression(xpathExpr[i], xpathCtx); + if(xpathObj == NULL) { + fprintf(stderr,"Error: unable to evaluate xpath expression \"%s\"\n", xpathExpr[i]); + } + else + { + print_xpath_nodes(xpathObj->nodesetval, stdout); + xmlXPathFreeObject(xpathObj); + } + } + xmlXPathFreeContext(xpathCtx); + xmlFreeDoc(doc); + } + } +#endif + if (cfile) + { + char *p = buf; + int i; + if (count) + fprintf (cfile, ","); + fprintf (cfile, "\n"); + for (i = 0; i < r; i++) + { + if ((i & 15) == 0) + fprintf (cfile, " \""); + fprintf (cfile, "\\x%02X", p[i] & 255); + + if (i < r - 1 && (i & 15) == 15) + fprintf (cfile, "\"\n"); + } - fprintf (cfile, "\"\n}"); - } - count++; + fprintf (cfile, "\"\n"); + } + num++; + } + count++; + if (cd) + yaz_iconv_close(cd); + yaz_marc_destroy(mt); } if (cfile) fprintf (cfile, "};\n"); + fclose(inf); break; case 'v': verbose++;