X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=ccl%2Fcclfind.c;h=8153a0e3a1e4c2060538b3b42194f4c987af2705;hb=087aa6548dbee815eed2e6bad4c5b105c35391cb;hp=4305df4f0be9840e1c0c59d67b9c6378b4afa375;hpb=657fb99115b87a5244e9a33bbe4ca3d9d18849c4;p=yaz-moved-to-github.git diff --git a/ccl/cclfind.c b/ccl/cclfind.c index 4305df4..8153a0e 100644 --- a/ccl/cclfind.c +++ b/ccl/cclfind.c @@ -1,12 +1,65 @@ +/* + * Copyright (c) 1995, the EUROPAGATE consortium (see below). + * + * The EUROPAGATE consortium members are: + * + * University College Dublin + * Danmarks Teknologiske Videnscenter + * An Chomhairle Leabharlanna + * Consejo Superior de Investigaciones Cientificas + * + * Permission to use, copy, modify, distribute, and sell this software and + * its documentation, in whole or in part, for any purpose, is hereby granted, + * provided that: + * + * 1. This copyright and permission notice appear in all copies of the + * software and its documentation. Notices of copyright or attribution + * which appear at the beginning of any file must remain unchanged. + * + * 2. The names of EUROPAGATE or the project partners may not be used to + * endorse or promote products derived from this software without specific + * prior written permission. + * + * 3. Users of this software (implementors and gateway operators) agree to + * inform the EUROPAGATE consortium of their use of the software. This + * information will be used to evaluate the EUROPAGATE project and the + * software, and to plan further developments. The consortium may use + * the information in later publications. + * + * 4. Users of this software agree to make their best efforts, when + * documenting their use of the software, to acknowledge the EUROPAGATE + * consortium, and the role played by the software in their work. + * + * THIS SOFTWARE IS PROVIDED "AS IS" AND WITHOUT WARRANTY OF ANY KIND, + * EXPRESS, IMPLIED, OR OTHERWISE, INCLUDING WITHOUT LIMITATION, ANY + * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. + * IN NO EVENT SHALL THE EUROPAGATE CONSORTIUM OR ITS MEMBERS BE LIABLE + * FOR ANY SPECIAL, INCIDENTAL, INDIRECT OR CONSEQUENTIAL DAMAGES OF + * ANY KIND, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA + * OR PROFITS, WHETHER OR NOT ADVISED OF THE POSSIBILITY OF DAMAGE, AND + * ON ANY THEORY OF LIABILITY, ARISING OUT OF OR IN CONNECTION WITH THE + * USE OR PERFORMANCE OF THIS SOFTWARE. + * + */ /* CCL find (to rpn conversion) * Europagate, 1995 * - * $Log: cclfind.c,v $ - * Revision 1.2 1995-09-27 15:02:44 quinn - * Modified function heads & prototypes. + * $Id: cclfind.c,v 1.33 2003-02-14 18:49:22 adam Exp $ * - * Revision 1.1 1995/04/10 10:28:19 quinn - * Added copy of CCL. + * Old Europagate log: + * + * Revision 1.16 1996/01/08 08:41:13 adam + * Removed unused function. + * + * Revision 1.15 1995/07/20 08:14:34 adam + * Qualifiers were observed too often. Instead tokens are treated as + * qualifiers only when separated by comma. + * + * Revision 1.14 1995/05/16 09:39:26 adam + * LICENSE. + * + * Revision 1.13 1995/04/17 09:31:42 adam + * Improved handling of qualifiers. Aliases or reserved words. * * Revision 1.12 1995/03/20 15:27:43 adam * Minor changes. @@ -45,71 +98,94 @@ * */ -#include #include -#include #include -#include +#include -static struct ccl_token *look_token; -static int ccl_error; -static CCL_bibset bibset; +/* returns type of current lookahead */ +#define KIND (cclp->look_token->kind) -#define KIND (look_token->kind) -#define ADVANCE look_token = look_token->next -#define ADVX(x) x=(x)->next +/* move one token forward */ +#define ADVANCE cclp->look_token = cclp->look_token->next -static struct ccl_rpn_attr *qual_val (struct ccl_rpn_attr *list, int type) +/* + * qual_val_type: test for existance of attribute type/value pair. + * qa: Attribute array + * type: Type of attribute to search for + * value: Value of attribute to seach for + * return: 1 if found; 0 otherwise. + */ +static int qual_val_type (struct ccl_rpn_attr **qa, int type, int value, + char **attset) { - while (list) - { - if (list->type == type) - return list; - list = list->next; - } - return NULL; -} + int i; + struct ccl_rpn_attr *q; -static int qual_val_type (struct ccl_rpn_attr *list, int type, int value) -{ - while (list) - { - if (list->type == type && list->value == value) - return 1; - list = list->next; - } + if (!qa) + return 0; + for (i = 0; (q=qa[i]); i++) + while (q) + { + if (q->type == type && q->value == value) + { + if (attset) + *attset = q->set; + return 1; + } + q = q->next; + } return 0; } +/* + * strxcat: concatenate strings. + * n: Null-terminated Destination string + * src: Source string to be appended (not null-terminated) + * len: Length of source string. + */ static void strxcat (char *n, const char *src, int len) { while (*n) - n++; + n++; while (--len >= 0) - *n++ = *src++; + *n++ = *src++; *n = '\0'; } +/* + * copy_token_name: Return copy of CCL token name + * tp: Pointer to token info. + * return: malloc(3) allocated copy of token name. + */ static char *copy_token_name (struct ccl_token *tp) { - char *str = malloc (tp->len + 1); - assert (str); + char *str = (char *)xmalloc (tp->len + 1); + ccl_assert (str); memcpy (str, tp->name, tp->len); str[tp->len] = '\0'; return str; } -static struct ccl_rpn_node *mk_node (enum rpn_node_kind kind) +/* + * mk_node: Create RPN node. + * kind: Type of node. + * return: pointer to allocated node. + */ +static struct ccl_rpn_node *mk_node (int kind) { struct ccl_rpn_node *p; - p = malloc (sizeof(*p)); - assert (p); + p = (struct ccl_rpn_node *)xmalloc (sizeof(*p)); + ccl_assert (p); p->kind = kind; return p; } -void MDF ccl_rpn_delete (struct ccl_rpn_node *rpn) +/* + * ccl_rpn_delete: Delete RPN tree. + * rpn: Pointer to tree. + */ +void ccl_rpn_delete (struct ccl_rpn_node *rpn) { struct ccl_rpn_attr *attr, *attr1; if (!rpn) @@ -123,472 +199,880 @@ void MDF ccl_rpn_delete (struct ccl_rpn_node *rpn) ccl_rpn_delete (rpn->u.p[1]); break; case CCL_RPN_TERM: - free (rpn->u.t.term); + xfree (rpn->u.t.term); for (attr = rpn->u.t.attr_list; attr; attr = attr1) { attr1 = attr->next; - free (attr); + if (attr->set) + xfree (attr->set); + xfree (attr); } break; case CCL_RPN_SET: - free (rpn->u.setname); + xfree (rpn->u.setname); break; case CCL_RPN_PROX: ccl_rpn_delete (rpn->u.p[0]); ccl_rpn_delete (rpn->u.p[1]); break; } - free (rpn); + xfree (rpn); +} + +static struct ccl_rpn_node *find_spec (CCL_parser cclp, + struct ccl_rpn_attr **qa); + +static int is_term_ok (int look, int *list) +{ + for (;*list >= 0; list++) + if (look == *list) + return 1; + return 0; } -static struct ccl_rpn_node *find_spec (struct ccl_rpn_attr **qa); -static struct ccl_rpn_node *search_terms (struct ccl_rpn_attr **qa); +static struct ccl_rpn_node *search_terms (CCL_parser cclp, + struct ccl_rpn_attr **qa); -static void add_attr (struct ccl_rpn_node *p, int type, int value) +/* + * add_attr: Add attribute (type/value) to RPN term node. + * p: RPN node of type term. + * type: Type of attribute + * value: Value of attribute + * set: Attribute set name + */ +static void add_attr (struct ccl_rpn_node *p, const char *set, + int type, int value) { struct ccl_rpn_attr *n; - n = malloc (sizeof(*n)); - assert (n); + n = (struct ccl_rpn_attr *)xmalloc (sizeof(*n)); + ccl_assert (n); + if (set) + { + n->set = (char*) xmalloc (strlen(set)+1); + strcpy (n->set, set); + } + else + n->set = 0; n->type = type; n->value = value; n->next = p->u.t.attr_list; p->u.t.attr_list = n; } -static struct ccl_rpn_node *search_term (struct ccl_rpn_attr **qa) +/* + * search_term: Parse CCL search term. + * cclp: CCL Parser + * qa: Qualifier attributes already applied. + * term_list: tokens we accept as terms in context + * multi: whether we accept "multiple" tokens + * return: pointer to node(s); NULL on error. + */ +static struct ccl_rpn_node *search_term_x (CCL_parser cclp, + struct ccl_rpn_attr **qa, + int *term_list, int multi) { - struct ccl_rpn_node *p; - struct ccl_rpn_attr *attr; - struct ccl_token *lookahead = look_token; - int len = 0; - int no, i; - int left_trunc = 0; - int right_trunc = 0; - int mid_trunc = 0; - - if (KIND != CCL_TOK_TERM) - { - ccl_error = CCL_ERR_TERM_EXPECTED; - return NULL; - } - for (no = 0; lookahead->kind == CCL_TOK_TERM; no++) - { - for (i = 0; ilen; i++) - if (lookahead->name[i] == '?') - { - if (no == 0 && i == 0 && lookahead->len >= 1) - left_trunc = 1; - else if (lookahead->next->kind != CCL_TOK_TERM && - i == lookahead->len-1 && i >= 1) - right_trunc = 1; - else - mid_trunc = 1; - } - len += 1+lookahead->len; - lookahead = lookahead->next; - } - p = mk_node (CCL_RPN_TERM); - p->u.t.term = malloc (len); - assert (p->u.t.term); - p->u.t.attr_list = NULL; - p->u.t.term[0] = '\0'; - for (i = 0; ilook_token; + int and_list = 0; + int or_list = 0; + char *attset; + const char *truncation_aliases; + + truncation_aliases = + ccl_qual_search_special(cclp->bibset, "truncation"); + if (!truncation_aliases) + truncation_aliases = "?"; + + if (qual_val_type (qa, CCL_BIB1_STR, CCL_BIB1_STR_AND_LIST, 0)) + and_list = 1; + if (qual_val_type (qa, CCL_BIB1_STR, CCL_BIB1_STR_OR_LIST, 0)) + or_list = 1; + while (1) { - const char *src_str = look_token->name; - int src_len = look_token->len; - - if (i == 0 && left_trunc) + struct ccl_rpn_node *p; + size_t no, i; + int no_spaces = 0; + int left_trunc = 0; + int right_trunc = 0; + int mid_trunc = 0; + int relation_value = -1; + int position_value = -1; + int structure_value = -1; + int truncation_value = -1; + int completeness_value = -1; + int len = 0; + size_t max = 200; + if (and_list || or_list || !multi) + max = 1; + + /* ignore commas when dealing with and-lists .. */ + if (and_list && lookahead && lookahead->kind == CCL_TOK_COMMA) { - src_len--; - src_str++; + lookahead = lookahead->next; + ADVANCE; + continue; } - else if (i == no-1 && right_trunc) - src_len--; - if (i) - strcat (p->u.t.term, " "); - strxcat (p->u.t.term, src_str, src_len); - ADVANCE; - } - if (qa) - { - int i; - for (i=0; qa[i]; i++) + /* go through each TERM token. If no truncation attribute is yet + met, then look for left/right truncation markers (?) and + set left_trunc/right_trunc/mid_trunc accordingly */ + for (no = 0; no < max && is_term_ok(lookahead->kind, term_list); no++) { - struct ccl_rpn_attr *attr; + for (i = 0; ilen; i++) + if (lookahead->name[i] == ' ') + no_spaces++; + else if (strchr(truncation_aliases, lookahead->name[i])) + { + if (no == 0 && i == 0 && lookahead->len >= 1) + left_trunc = 1; + else if (!is_term_ok(lookahead->next->kind, term_list) && + i == lookahead->len-1 && i >= 1) + right_trunc = 1; + else + mid_trunc = 1; + } + len += 1+lookahead->len; + lookahead = lookahead->next; + } + + if (len == 0) + break; /* no more terms . stop . */ + + if (p_top) + { + if (or_list) + p = mk_node (CCL_RPN_OR); + else if (and_list) + p = mk_node (CCL_RPN_AND); + else + p = mk_node (CCL_RPN_AND); + p->u.p[0] = p_top; + p_top = p; + } + + /* create the term node, but wait a moment before adding the term */ + p = mk_node (CCL_RPN_TERM); + p->u.t.attr_list = NULL; + p->u.t.term = NULL; + + /* make the top node point to us.. */ + if (p_top) + p_top->u.p[1] = p; + else + p_top = p; + + + /* go through all attributes and add them to the attribute list */ + for (i=0; qa && qa[i]; i++) + { + struct ccl_rpn_attr *attr; + for (attr = qa[i]; attr; attr = attr->next) if (attr->value > 0) - add_attr (p, attr->type, attr->value); + { /* deal only with REAL attributes (positive) */ + switch (attr->type) + { + case CCL_BIB1_REL: + if (relation_value != -1) + continue; + relation_value = attr->value; + break; + case CCL_BIB1_POS: + if (position_value != -1) + continue; + position_value = attr->value; + break; + case CCL_BIB1_STR: + if (structure_value != -1) + continue; + structure_value = attr->value; + break; + case CCL_BIB1_TRU: + if (truncation_value != -1) + continue; + truncation_value = attr->value; + left_trunc = right_trunc = mid_trunc = 0; + break; + case CCL_BIB1_COM: + if (completeness_value != -1) + continue; + completeness_value = attr->value; + break; + } + add_attr (p, attr->set, attr->type, attr->value); + } } - attr = qa[0]; - } - else - attr = ccl_qual_search (bibset, "term", 4); - if (attr && qual_val_type (attr, CCL_BIB1_STR, CCL_BIB1_STR_WP)) - { - if (no == 1) - add_attr (p, CCL_BIB1_STR, 2); - else - add_attr (p, CCL_BIB1_STR, 1); - } - if (left_trunc && right_trunc) - { - if (attr && !qual_val_type (attr, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_BOTH)) + /* len now holds the number of characters in the RPN term */ + /* no holds the number of CCL tokens (1 or more) */ + + if (structure_value == -1 && + qual_val_type (qa, CCL_BIB1_STR, CCL_BIB1_STR_WP, &attset)) + { /* no structure attribute met. Apply either structure attribute + WORD or PHRASE depending on number of CCL tokens */ + if (no == 1 && no_spaces == 0) + add_attr (p, attset, CCL_BIB1_STR, 2); + else + add_attr (p, attset, CCL_BIB1_STR, 1); + } + + /* make the RPN token */ + p->u.t.term = (char *)xmalloc (len); + ccl_assert (p->u.t.term); + p->u.t.term[0] = '\0'; + for (i = 0; ilook_token->name; + int src_len = cclp->look_token->len; + + if (i == 0 && left_trunc) + { + src_len--; + src_str++; + } + else if (i == no-1 && right_trunc) + src_len--; + if (src_len) + { + int len = strlen(p->u.t.term); + if (len && + !strchr("-+", *src_str) && + !strchr("-+", p->u.t.term[len-1])) + { + strcat (p->u.t.term, " "); + } + } + strxcat (p->u.t.term, src_str, src_len); + ADVANCE; } - add_attr (p, CCL_BIB1_TRU, 3); - } - else if (right_trunc) - { - if (attr && !qual_val_type (attr, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_RIGHT)) + if (left_trunc && right_trunc) { - ccl_error = CCL_ERR_TRUNC_NOT_RIGHT; - if (qa) - free (qa); - ccl_rpn_delete (p); - return NULL; + if (!qual_val_type (qa, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_BOTH, + &attset)) + { + cclp->error_code = CCL_ERR_TRUNC_NOT_BOTH; + ccl_rpn_delete (p); + return NULL; + } + add_attr (p, attset, CCL_BIB1_TRU, 3); } - add_attr (p, CCL_BIB1_TRU, 1); - } - else if (left_trunc) - { - if (attr && !qual_val_type (attr, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_LEFT)) + else if (right_trunc) { - ccl_error = CCL_ERR_TRUNC_NOT_LEFT; - if (qa) - free (qa); - ccl_rpn_delete (p); - return NULL; + if (!qual_val_type (qa, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_RIGHT, + &attset)) + { + cclp->error_code = CCL_ERR_TRUNC_NOT_RIGHT; + ccl_rpn_delete (p); + return NULL; + } + add_attr (p, attset, CCL_BIB1_TRU, 1); } - add_attr (p, CCL_BIB1_TRU, 2); - } - else - { - if (attr && qual_val_type (attr, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_NONE)) - add_attr (p, CCL_BIB1_TRU, 100); + else if (left_trunc) + { + if (!qual_val_type (qa, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_LEFT, + &attset)) + { + cclp->error_code = CCL_ERR_TRUNC_NOT_LEFT; + ccl_rpn_delete (p); + return NULL; + } + add_attr (p, attset, CCL_BIB1_TRU, 2); + } + else + { + if (qual_val_type (qa, CCL_BIB1_TRU, CCL_BIB1_TRU_CAN_NONE, + &attset)) + add_attr (p, attset, CCL_BIB1_TRU, 100); + } + if (!multi) + break; } - return p; + if (!p_top) + cclp->error_code = CCL_ERR_TERM_EXPECTED; + return p_top; } -static struct ccl_rpn_node *qualifiers (struct ccl_token *la, - struct ccl_rpn_attr **qa) +static struct ccl_rpn_node *search_term (CCL_parser cclp, + struct ccl_rpn_attr **qa) { - struct ccl_token *lookahead = look_token; - struct ccl_rpn_attr **ap; - int no = 1; - int i, rel; - struct ccl_rpn_attr *attr; + static int list[] = {CCL_TOK_TERM, CCL_TOK_COMMA, -1}; + return search_term_x(cclp, qa, list, 0); +} - if (qa) - { - ccl_error = CCL_ERR_DOUBLE_QUAL; - return NULL; - } - for (lookahead = look_token; lookahead != la; lookahead=lookahead->next) - no++; - ap = malloc (no * sizeof(*ap)); - assert (ap); - for (i=0; look_token != la; i++) - { - ap[i] = ccl_qual_search (bibset, look_token->name, look_token->len); - if (!ap[i]) - { - ccl_error = CCL_ERR_UNKNOWN_QUAL; - free (ap); - return NULL; - } - ADVANCE; - if (KIND == CCL_TOK_COMMA) - ADVANCE; - } - ap[i] = NULL; - if (! (attr = qual_val (ap[0], CCL_BIB1_REL)) || - attr->value != CCL_BIB1_REL_ORDER) +static struct ccl_rpn_node *qualifiers2 (CCL_parser cclp, + struct ccl_rpn_attr **ap) +{ + char *attset; + int rel; + + if (!qual_val_type(ap, CCL_BIB1_REL, CCL_BIB1_REL_ORDER, &attset)) { /* unordered relation */ struct ccl_rpn_node *p; if (KIND != CCL_TOK_EQ) { - ccl_error = CCL_ERR_EQ_EXPECTED; - free (ap); + cclp->error_code = CCL_ERR_EQ_EXPECTED; return NULL; } ADVANCE; if (KIND == CCL_TOK_LP) { ADVANCE; - if (!(p = find_spec (ap))) + if (!(p = find_spec (cclp, ap))) { - free (ap); return NULL; } if (KIND != CCL_TOK_RP) { - ccl_error = CCL_ERR_RP_EXPECTED; + cclp->error_code = CCL_ERR_RP_EXPECTED; ccl_rpn_delete (p); - free (ap); return NULL; } ADVANCE; } else - p = search_terms (ap); - free (ap); + p = search_terms (cclp, ap); return p; } + /* ordered relation ... */ rel = 0; - if (look_token->len == 1) + if (cclp->look_token->len == 1) { - if (look_token->name[0] == '<') + if (cclp->look_token->name[0] == '<') rel = 1; - else if (look_token->name[0] == '=') + else if (cclp->look_token->name[0] == '=') rel = 3; - else if (look_token->name[0] == '>') + else if (cclp->look_token->name[0] == '>') rel = 5; } - else if (look_token->len == 2) + else if (cclp->look_token->len == 2) { - if (!memcmp (look_token->name, "<=", 2)) + if (!memcmp (cclp->look_token->name, "<=", 2)) rel = 2; - else if (!memcmp (look_token->name, ">=", 2)) + else if (!memcmp (cclp->look_token->name, ">=", 2)) rel = 4; - else if (!memcmp (look_token->name, "<>", 2)) + else if (!memcmp (cclp->look_token->name, "<>", 2)) rel = 6; } if (!rel) - ccl_error = CCL_ERR_BAD_RELATION; + cclp->error_code = CCL_ERR_BAD_RELATION; else { struct ccl_rpn_node *p; - + ADVANCE; /* skip relation */ - if (KIND == CCL_TOK_TERM) + if (KIND == CCL_TOK_TERM && + cclp->look_token->next && cclp->look_token->next->len == 1 && + cclp->look_token->next->name[0] == '-') { struct ccl_rpn_node *p1; - p1 = search_term (ap); - if (KIND == CCL_TOK_MINUS) + if (!(p1 = search_term (cclp, ap))) + return NULL; + ADVANCE; /* skip '-' */ + if (KIND == CCL_TOK_TERM) /* = term - term ? */ { - ADVANCE; /* skip '-' */ - if (KIND == CCL_TOK_TERM) /* = term - term ? */ + struct ccl_rpn_node *p2; + + if (!(p2 = search_term (cclp, ap))) { - struct ccl_rpn_node *p2; - - p2 = search_term (ap); - p = mk_node (CCL_RPN_AND); - p->u.p[0] = p1; - add_attr (p1, CCL_BIB1_REL, 4); - p->u.p[1] = p2; - add_attr (p2, CCL_BIB1_REL, 2); - free (ap); - return p; - } - else /* = term - */ - { - add_attr (p1, CCL_BIB1_REL, 4); - free (ap); - return p1; + ccl_rpn_delete (p1); + return NULL; } + p = mk_node (CCL_RPN_AND); + p->u.p[0] = p1; + add_attr (p1, attset, CCL_BIB1_REL, 4); + p->u.p[1] = p2; + add_attr (p2, attset, CCL_BIB1_REL, 2); + return p; } - else + else /* = term - */ { - add_attr (p1, CCL_BIB1_REL, rel); - free (ap); + add_attr (p1, attset, CCL_BIB1_REL, 4); return p1; } } - else if (KIND == CCL_TOK_MINUS) /* = - term ? */ + else if (cclp->look_token->len == 1 && + cclp->look_token->name[0] == '-') /* = - term ? */ + { + ADVANCE; + if (!(p = search_term (cclp, ap))) + return NULL; + add_attr (p, attset, CCL_BIB1_REL, 2); + return p; + } + else if (KIND == CCL_TOK_LP) { ADVANCE; - p = search_term (ap); - add_attr (p, CCL_BIB1_REL, 2); - free (ap); + if (!(p = find_spec (cclp, ap))) + return NULL; + if (KIND != CCL_TOK_RP) + { + cclp->error_code = CCL_ERR_RP_EXPECTED; + ccl_rpn_delete (p); + return NULL; + } + ADVANCE; return p; } - ccl_error = CCL_ERR_TERM_EXPECTED; + else + { + if (!(p = search_terms (cclp, ap))) + return NULL; + add_attr (p, attset, CCL_BIB1_REL, rel); + return p; + } + cclp->error_code = CCL_ERR_TERM_EXPECTED; } - free (ap); return NULL; } -static struct ccl_rpn_node *search_terms (struct ccl_rpn_attr **qa) +/* + * qualifiers1: Parse CCL qualifiers and search terms. + * cclp: CCL Parser + * la: Token pointer to RELATION token. + * qa: Qualifier attributes already applied. + * return: pointer to node(s); NULL on error. + */ +static struct ccl_rpn_node *qualifiers1 (CCL_parser cclp, struct ccl_token *la, + struct ccl_rpn_attr **qa) +{ + struct ccl_token *lookahead = cclp->look_token; + struct ccl_token *look_start = cclp->look_token; + struct ccl_rpn_attr **ap; + struct ccl_rpn_node *node = 0; + const char *field_str; + int no = 0; + int seq = 0; + int i; + int mode_merge = 1; +#if 0 + if (qa) + { + cclp->error_code = CCL_ERR_DOUBLE_QUAL; + return NULL; + } +#endif + for (lookahead = cclp->look_token; lookahead != la; + lookahead=lookahead->next) + no++; + if (qa) + for (i=0; qa[i]; i++) + no++; + ap = (struct ccl_rpn_attr **)xmalloc ((no ? (no+1) : 2) * sizeof(*ap)); + ccl_assert (ap); + + field_str = ccl_qual_search_special(cclp->bibset, "field"); + if (field_str) + { + if (!strcmp (field_str, "or")) + mode_merge = 0; + else if (!strcmp (field_str, "merge")) + mode_merge = 1; + } + if (!mode_merge) + { + /* consider each field separately and OR */ + lookahead = look_start; + while (lookahead != la) + { + ap[1] = 0; + seq = 0; + while ((ap[0] = ccl_qual_search (cclp, lookahead->name, + lookahead->len, seq)) != 0) + { + struct ccl_rpn_node *node_sub; + cclp->look_token = la; + + node_sub = qualifiers2(cclp, ap); + if (!node_sub) + { + ccl_rpn_delete (node); + xfree (ap); + return 0; + } + if (node) + { + struct ccl_rpn_node *node_this = mk_node(CCL_RPN_OR); + node_this->u.p[0] = node; + node_this->u.p[1] = node_sub; + node = node_this; + } + else + node = node_sub; + seq++; + } + if (seq == 0) + { + cclp->look_token = lookahead; + cclp->error_code = CCL_ERR_UNKNOWN_QUAL; + xfree (ap); + return NULL; + } + lookahead = lookahead->next; + if (lookahead->kind == CCL_TOK_COMMA) + lookahead = lookahead->next; + } + } + else + { + /* merge attributes from ALL fields - including inherited ones */ + while (1) + { + struct ccl_rpn_node *node_sub; + int found = 0; + lookahead = look_start; + for (i = 0; lookahead != la; i++) + { + ap[i] = ccl_qual_search (cclp, lookahead->name, + lookahead->len, seq); + if (ap[i]) + found++; + if (!ap[i] && seq > 0) + ap[i] = ccl_qual_search (cclp, lookahead->name, + lookahead->len, 0); + if (!ap[i]) + { + cclp->look_token = lookahead; + cclp->error_code = CCL_ERR_UNKNOWN_QUAL; + xfree (ap); + return NULL; + } + lookahead = lookahead->next; + if (lookahead->kind == CCL_TOK_COMMA) + lookahead = lookahead->next; + } + if (qa) + { + struct ccl_rpn_attr **qa0 = qa; + + while (*qa0) + ap[i++] = *qa0++; + } + ap[i] = NULL; + + if (!found) + break; + + cclp->look_token = lookahead; + + node_sub = qualifiers2(cclp, ap); + if (!node_sub) + { + ccl_rpn_delete (node); + break; + } + if (node) + { + struct ccl_rpn_node *node_this = mk_node(CCL_RPN_OR); + node_this->u.p[0] = node; + node_this->u.p[1] = node_sub; + node = node_this; + } + else + node = node_sub; + seq++; + } + } + xfree (ap); + return node; +} + + +/* + * search_terms: Parse CCL search terms - including proximity. + * cclp: CCL Parser + * qa: Qualifier attributes already applied. + * return: pointer to node(s); NULL on error. + */ +static struct ccl_rpn_node *search_terms (CCL_parser cclp, + struct ccl_rpn_attr **qa) { + static int list[] = { + CCL_TOK_TERM, CCL_TOK_COMMA,CCL_TOK_EQ, CCL_TOK_REL, CCL_TOK_SET, -1}; struct ccl_rpn_node *p1, *p2, *pn; - p1 = search_term (qa); + p1 = search_term_x (cclp, qa, list, 1); if (!p1) return NULL; while (1) { - if (KIND == CCL_TOK_PROX) - { - ADVANCE; - p2 = search_term (qa); + if (KIND == CCL_TOK_PROX) + { + struct ccl_rpn_node *p_prox = 0; + /* ! word order specified */ + /* % word order not specified */ + p_prox = mk_node(CCL_RPN_TERM); + p_prox->u.t.term = (char *) xmalloc(cclp->look_token->len); + memcpy(p_prox->u.t.term, cclp->look_token->name, + cclp->look_token->len); + p_prox->u.t.term[cclp->look_token->len] = 0; + p_prox->u.t.attr_list = 0; + + ADVANCE; + p2 = search_term_x (cclp, qa, list, 1); if (!p2) { ccl_rpn_delete (p1); return NULL; } - pn = mk_node (CCL_RPN_PROX); - pn->u.p[0] = p1; - pn->u.p[1] = p2; - p1 = pn; - } - else if (KIND == CCL_TOK_TERM) - { - p2 = search_term (qa); + pn = mk_node (CCL_RPN_PROX); + pn->u.p[0] = p1; + pn->u.p[1] = p2; + pn->u.p[2] = p_prox; + p1 = pn; + } + else if (is_term_ok(KIND, list)) + { + p2 = search_term_x (cclp, qa, list, 1); if (!p2) { ccl_rpn_delete (p1); return NULL; } - pn = mk_node (CCL_RPN_PROX); - pn->u.p[0] = p1; - pn->u.p[1] = p2; - p1 = pn; - } - else - break; + pn = mk_node (CCL_RPN_PROX); + pn->u.p[0] = p1; + pn->u.p[1] = p2; + pn->u.p[2] = 0; + p1 = pn; + } + else + break; } return p1; } -static struct ccl_rpn_node *search_elements (struct ccl_rpn_attr **qa) +/* + * search_elements: Parse CCL search elements + * cclp: CCL Parser + * qa: Qualifier attributes already applied. + * return: pointer to node(s); NULL on error. + */ +static struct ccl_rpn_node *search_elements (CCL_parser cclp, + struct ccl_rpn_attr **qa) { struct ccl_rpn_node *p1; struct ccl_token *lookahead; if (KIND == CCL_TOK_LP) { - ADVANCE; - p1 = find_spec (qa); + ADVANCE; + p1 = find_spec (cclp, qa); if (!p1) return NULL; if (KIND != CCL_TOK_RP) { - ccl_error = CCL_ERR_RP_EXPECTED; + cclp->error_code = CCL_ERR_RP_EXPECTED; ccl_rpn_delete (p1); return NULL; } - ADVANCE; - return p1; + ADVANCE; + return p1; } else if (KIND == CCL_TOK_SET) { - ADVANCE; + ADVANCE; if (KIND == CCL_TOK_EQ) ADVANCE; if (KIND != CCL_TOK_TERM) { - ccl_error = CCL_ERR_SETNAME_EXPECTED; + cclp->error_code = CCL_ERR_SETNAME_EXPECTED; return NULL; } - p1 = mk_node (CCL_RPN_SET); - p1->u.setname = copy_token_name (look_token); - ADVANCE; - return p1; + p1 = mk_node (CCL_RPN_SET); + p1->u.setname = copy_token_name (cclp->look_token); + ADVANCE; + return p1; } - lookahead = look_token; + lookahead = cclp->look_token; - while (lookahead->kind==CCL_TOK_TERM || lookahead->kind==CCL_TOK_COMMA) - lookahead = lookahead->next; - if (lookahead->kind == CCL_TOK_REL || lookahead->kind == CCL_TOK_EQ) - return qualifiers (lookahead, qa); - return search_terms (qa); + while (lookahead->kind==CCL_TOK_TERM) + { + lookahead = lookahead->next; + if (lookahead->kind == CCL_TOK_REL || lookahead->kind == CCL_TOK_EQ) + return qualifiers1 (cclp, lookahead, qa); + if (lookahead->kind != CCL_TOK_COMMA) + break; + lookahead = lookahead->next; + } + if (qa) + return search_terms (cclp, qa); + else + { + struct ccl_rpn_attr *qa[2]; + struct ccl_rpn_node *node = 0; + int seq; + lookahead = cclp->look_token; + + qa[1] = 0; + for(seq = 0; ;seq++) + { + struct ccl_rpn_node *node_sub; + qa[0] = ccl_qual_search(cclp, "term", 4, seq); + if (!qa[0]) + break; + + cclp->look_token = lookahead; + + node_sub = search_terms (cclp, qa); + if (!node_sub) + { + ccl_rpn_delete (node); + return 0; + } + if (node) + { + struct ccl_rpn_node *node_this = mk_node(CCL_RPN_OR); + node_this->u.p[0] = node; + node_this->u.p[1] = node_sub; + node_this->u.p[2] = 0; + node = node_this; + } + else + node = node_sub; + } + if (!node) + node = search_terms (cclp, 0); + return node; + } } -static struct ccl_rpn_node *find_spec (struct ccl_rpn_attr **qa) +/* + * find_spec: Parse CCL find specification + * cclp: CCL Parser + * qa: Qualifier attributes already applied. + * return: pointer to node(s); NULL on error. + */ +static struct ccl_rpn_node *find_spec (CCL_parser cclp, + struct ccl_rpn_attr **qa) { struct ccl_rpn_node *p1, *p2, *pn; - if (!(p1 = search_elements (qa))) + if (!(p1 = search_elements (cclp, qa))) return NULL; while (1) { - switch (KIND) - { - case CCL_TOK_AND: - ADVANCE; - p2 = search_elements (qa); + switch (KIND) + { + case CCL_TOK_AND: + ADVANCE; + p2 = search_elements (cclp, qa); if (!p2) { ccl_rpn_delete (p1); return NULL; } - pn = mk_node (CCL_RPN_AND); - pn->u.p[0] = p1; - pn->u.p[1] = p2; - p1 = pn; - continue; - case CCL_TOK_OR: - ADVANCE; - p2 = search_elements (qa); + pn = mk_node (CCL_RPN_AND); + pn->u.p[0] = p1; + pn->u.p[1] = p2; + pn->u.p[2] = 0; + p1 = pn; + continue; + case CCL_TOK_OR: + ADVANCE; + p2 = search_elements (cclp, qa); if (!p2) { ccl_rpn_delete (p1); return NULL; } - pn = mk_node (CCL_RPN_OR); - pn->u.p[0] = p1; - pn->u.p[1] = p2; - p1 = pn; - continue; - case CCL_TOK_NOT: - ADVANCE; - p2 = search_elements (qa); + pn = mk_node (CCL_RPN_OR); + pn->u.p[0] = p1; + pn->u.p[1] = p2; + pn->u.p[2] = 0; + p1 = pn; + continue; + case CCL_TOK_NOT: + ADVANCE; + p2 = search_elements (cclp, qa); if (!p2) { ccl_rpn_delete (p1); return NULL; } - pn = mk_node (CCL_RPN_NOT); - pn->u.p[0] = p1; - pn->u.p[1] = p2; - p1 = pn; - continue; - } - break; + pn = mk_node (CCL_RPN_NOT); + pn->u.p[0] = p1; + pn->u.p[1] = p2; + pn->u.p[2] = 0; + p1 = pn; + continue; + } + break; } return p1; } -struct ccl_rpn_node MDF *ccl_find (CCL_bibset abibset, struct ccl_token *list, - int *error, const char **pos) +struct ccl_rpn_node *ccl_parser_find (CCL_parser cclp, struct ccl_token *list) { struct ccl_rpn_node *p; - look_token = list; - bibset = abibset; - p = find_spec (NULL); + + + cclp->look_token = list; + p = find_spec (cclp, NULL); if (p && KIND != CCL_TOK_EOL) { if (KIND == CCL_TOK_RP) - ccl_error = CCL_ERR_BAD_RP; + cclp->error_code = CCL_ERR_BAD_RP; else - ccl_error = CCL_ERR_OP_EXPECTED; + cclp->error_code = CCL_ERR_OP_EXPECTED; ccl_rpn_delete (p); p = NULL; } - *pos = look_token->name; + cclp->error_pos = cclp->look_token->name; if (p) - *error = CCL_ERR_OK; + cclp->error_code = CCL_ERR_OK; else - *error = ccl_error; + cclp->error_code = cclp->error_code; return p; } -struct ccl_rpn_node MDF *ccl_find_str (CCL_bibset bibset, const char *str, +/* + * ccl_find: Parse CCL find - token representation + * bibset: Bibset to be used for the parsing + * list: List of tokens + * error: Pointer to integer. Holds error no. on completion. + * pos: Pointer to char position. Holds approximate error position. + * return: RPN tree on successful completion; NULL otherwise. + */ +struct ccl_rpn_node *ccl_find (CCL_bibset bibset, struct ccl_token *list, + int *error, const char **pos) +{ + struct ccl_rpn_node *p; + CCL_parser cclp = ccl_parser_create (); + + cclp->bibset = bibset; + + p = ccl_parser_find (cclp, list); + + *error = cclp->error_code; + *pos = cclp->error_pos; + + ccl_parser_destroy (cclp); + + return p; +} + +/* + * ccl_find_str: Parse CCL find - string representation + * bibset: Bibset to be used for the parsing + * str: String to be parsed + * error: Pointer to integer. Holds error no. on completion. + * pos: Pointer to char position. Holds approximate error position. + * return: RPN tree on successful completion; NULL otherwise. + */ +struct ccl_rpn_node *ccl_find_str (CCL_bibset bibset, const char *str, int *error, int *pos) { + CCL_parser cclp = ccl_parser_create (); struct ccl_token *list; - struct ccl_rpn_node *rpn; - const char *char_pos; + struct ccl_rpn_node *p; + + cclp->bibset = bibset; - list = ccl_tokenize (str); - rpn = ccl_find (bibset, list, error, &char_pos); + list = ccl_parser_tokenize (cclp, str); + p = ccl_parser_find (cclp, list); + + *error = cclp->error_code; if (*error) - *pos = char_pos - str; - return rpn; + *pos = cclp->error_pos - str; + ccl_parser_destroy (cclp); + ccl_token_del (list); + return p; }