X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=data1%2Fd1_absyn.c;h=4a008df8e18637b6cfa6579ad605a9b467676d24;hb=6acd2e140d59433a002f2f1890cb4a6e25895c68;hp=45620d5791ecdb03d61ef00d3f9043ed518b23ef;hpb=4eb3b54bb2ca9af74f39f000d3d40dba99ded887;p=idzebra-moved-to-github.git diff --git a/data1/d1_absyn.c b/data1/d1_absyn.c index 45620d5..4a008df 100644 --- a/data1/d1_absyn.c +++ b/data1/d1_absyn.c @@ -1,6 +1,6 @@ -/* $Id: d1_absyn.c,v 1.16 2004-12-13 20:51:27 adam Exp $ - Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004 - Index Data Aps +/* $Id: d1_absyn.c,v 1.23 2006-02-07 00:17:44 adam Exp $ + Copyright (C) 1995-2005 + Index Data ApS This file is part of the Zebra server. @@ -439,65 +439,74 @@ void fix_element_ref (data1_handle dh, data1_absyn *absyn, data1_element *e) */ -const char * mk_xpath_regexp (data1_handle dh, char *expr) +static const char * mk_xpath_regexp (data1_handle dh, const char *expr) { - char *p = expr; - char *pp; - char *s; + const char *p = expr; int abs = 1; - int i; - int j; - int e=0; - int is_predicate = 0; - - static char *stack[32]; - static char res[1024]; - char *r = ""; + int e = 0; + char *stack[32]; + char *res_p, *res = 0; + size_t res_size = 1; - if (*p != '/') { return (""); } + if (*p != '/') + return (""); p++; - if (*p == '/') { abs=0; p++; } - - while (*p) { - i=0; - while (*p && !strchr("/",*p)) { - i++; p++; - } - stack[e] = (char *) nmem_malloc (data1_nmem_get (dh), i+1); + if (*p == '/') + { + abs =0; + p++; + } + while (*p) + { + int is_predicate = 0; + char *s; + int i, j; + for (i = 0; *p && !strchr("/",*p); i++, p++) + ; + res_size += (i+3); /* we'll add / between later .. */ + stack[e] = (char *) nmem_malloc(data1_nmem_get(dh), i+1); s = stack[e]; - for (j=0; j< i; j++) { - pp = p-i+j; - if (*pp == '[') { - is_predicate=1; - } - else if (*pp == ']') { - is_predicate=0; - } - else { - if (!is_predicate) { - if (*pp == '*') - *s++ = '.'; - *s++ = *pp; + for (j = 0; j < i; j++) + { + const char *pp = p-i+j; + if (*pp == '[') + is_predicate=1; + else if (*pp == ']') + is_predicate=0; + else + { + if (!is_predicate) { + if (*pp == '*') + *s++ = '.'; + *s++ = *pp; + } } - } } *s = 0; e++; - if (*p) {p++;} + if (*p) + p++; } - e--; p = &res[0]; i=0; - sprintf (p, "^"); p++; - while (e >= 0) { - /* !!! res size is not checked !!! */ - sprintf (p, "%s/",stack[e]); - p += strlen(stack[e]) + 1; - e--; + res_p = res = nmem_malloc(data1_nmem_get(dh), res_size + 10); + + if (stack[e-1][0] == '@') /* path/@attr spec (leaf is attribute) */ + strcpy(res_p, "/"); + else + strcpy(res_p, "[^@]*/"); /* path .. (index all cdata below it) */ + res_p = res_p + strlen(res_p); + while (--e >= 0) { + sprintf(res_p, "%s/", stack[e]); + res_p += strlen(stack[e]) + 1; } - if (!abs) { sprintf (p, ".*"); p+=2; } - sprintf (p, "$"); p++; - r = nmem_strdup (data1_nmem_get (dh), res); - yaz_log(YLOG_DEBUG,"Got regexp: %s",r); - return (r); + if (!abs) + { + sprintf(res_p, ".*"); + res_p += 2; + } + sprintf (res_p, "$"); + res_p++; + yaz_log(YLOG_DEBUG, "Got regexp: %s", res); + return res; } /* *ostrich* @@ -561,12 +570,17 @@ static int parse_termlists (data1_handle dh, data1_termlist ***tpp, nmem_malloc(data1_nmem_get(dh), sizeof(**tp)); (*tp)->next = 0; +#if NATTR + (*tp)->index_name = nmem_strdup(data1_nmem_get(dh), element_name); + if (*attname == '!' && xpelement) + (*tp)->index_name = 0; +#else if (!xpelement) { if (*attname == '!') strcpy(attname, element_name); } - if (!((*tp)->att = data1_getattbyname(dh, res->attset, - attname))) { + if (!((*tp)->att = data1_getattbyname(dh, res->attset, attname))) + { if ((!xpelement) || (*attname != '!')) { yaz_log(YLOG_WARN, "%s:%d: Couldn't find att '%s' in attset", @@ -576,7 +590,7 @@ static int parse_termlists (data1_handle dh, data1_termlist ***tpp, (*tp)->att = 0; } } - +#endif if (r == 2 && (source = strchr(structure, ':'))) *source++ = '\0'; /* cut off structure .. */ else @@ -596,6 +610,33 @@ static int parse_termlists (data1_handle dh, data1_termlist ***tpp, return 0; } +/* quinn + * Converts a 'melm' field[$subfield] pattern to a simple xpath + */ +static int melm2xpath(char *melm, char *buf) +{ + char *dollar; + char *field = melm; + char *subfield; + char *fieldtype; + if ((dollar = strchr(melm, '$'))) { + *dollar = '\0'; + subfield = ++dollar; + } else + subfield = ""; + if (field[0] == '0' && field[1] == '0') + fieldtype = "controlfield"; + else + fieldtype = "datafield"; + sprintf(buf, "/*/%s[@tag=\"%s\"]", fieldtype, field); + if (*subfield) + sprintf(buf + strlen(buf), "/subfield[@code=\"%s\"]", subfield); + else if (field[0] != '0' || field[1] != '0') + strcat(buf, "/subfield"); + yaz_log(YLOG_DEBUG, "Created xpath: '%s'", buf); + return 0; +} + const char *data1_systag_lookup(data1_absyn *absyn, const char *tag, const char *default_value) { @@ -843,20 +884,28 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, pop, 2003-01-17 */ - else if (!strcmp(cmd, "xelm")) { + else if (!strcmp(cmd, "xelm") || !strcmp(cmd, "melm")) { int i; char *p, *xpath_expr, *termlists; const char *regexp; struct DFA *dfa = dfa = dfa_init(); data1_termlist **tp; + char melm_xpath[128]; if (argc < 3) { yaz_log(YLOG_WARN, "%s:%d: Bad # of args to xelm", file, lineno); continue; } - xpath_expr = argv[1]; + + if (!strcmp(cmd, "melm")) { + if (melm2xpath(argv[1], melm_xpath) < 0) + continue; + xpath_expr = melm_xpath; + } else { + xpath_expr = argv[1]; + } termlists = argv[2]; regexp = mk_xpath_regexp(dh, xpath_expr); i = dfa_parse (dfa, ®exp); @@ -1002,6 +1051,9 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, } else if (!strcmp(cmd, "attset")) { +#if NATTR + yaz_log(YLOG_WARN, "%s:%d: attset obsolete", file, lineno); +#else char *name; data1_attset *attset; @@ -1023,6 +1075,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, (*attset_childp)->child = attset; (*attset_childp)->next = 0; attset_childp = &(*attset_childp)->next; +#endif } else if (!strcmp(cmd, "tagset")) {