root/trunk/gsdl/packages/yaz/retrieval/d1_map.c @ 1343

Revision 1343, 9.1 KB (checked in by johnmcp, 20 years ago)

Added the YAZ toolkit source to the packages directory (for z39.50 stuff)

  • Property svn:keywords set to Author Date Id Revision
Line 
1/*
2 * Copyright (c) 1995-1999, Index Data.
3 * See the file LICENSE for details.
4 * Sebastian Hammer, Adam Dickmeiss
5 *
6 * $Log$
7 * Revision 1.1  2000/08/03 03:11:33  johnmcp
8 * Added the YAZ toolkit source to the packages directory (for z39.50 stuff)
9 *
10 * Revision 1.17  1999/11/30 13:47:12  adam
11 * Improved installation. Moved header files to include/yaz.
12 *
13 * Revision 1.16  1999/10/21 12:06:29  adam
14 * Retrieval module no longer uses ctype.h - functions.
15 *
16 * Revision 1.15  1999/08/27 09:40:32  adam
17 * Renamed logf function to yaz_log. Removed VC++ project files.
18 *
19 * Revision 1.14  1998/10/13 16:09:50  adam
20 * Added support for arbitrary OID's for tagsets, schemas and attribute sets.
21 * Added support for multiple attribute set references and tagset references
22 * from an abstract syntax file.
23 * Fixed many bad logs-calls in routines that read the various
24 * specifications regarding data1 (*.abs,*.att,...) and made the messages
25 * consistent whenever possible.
26 * Added extra 'lineno' argument to function readconf_line.
27 *
28 * Revision 1.13  1998/02/11 11:53:35  adam
29 * Changed code so that it compiles as C++.
30 *
31 * Revision 1.12  1997/11/18 09:51:09  adam
32 * Removed element num_children from data1_node. Minor changes in
33 * data1 to Explain.
34 *
35 * Revision 1.11  1997/09/17 12:10:36  adam
36 * YAZ version 1.4.
37 *
38 * Revision 1.10  1997/09/05 09:50:56  adam
39 * Removed global data1_tabpath - uses data1_get_tabpath() instead.
40 *
41 * Revision 1.9  1996/06/10 08:56:02  quinn
42 * Work on Summary.
43 *
44 * Revision 1.8  1996/05/01  12:45:31  quinn
45 * Support use of local tag names in abs file.
46 *
47 * Revision 1.7  1995/12/13  13:44:31  quinn
48 * Modified Data1-system to use nmem
49 *
50 * Revision 1.6  1995/12/12  16:37:08  quinn
51 * Added destroy element to data1_node.
52 *
53 * Revision 1.5  1995/12/12  14:11:31  quinn
54 * More work on the large-record problem.
55 *
56 * Revision 1.4  1995/12/11  15:22:37  quinn
57 * Added last_child field to the node.
58 * Rewrote schema-mapping.
59 *
60 * Revision 1.3  1995/11/01  16:34:56  quinn
61 * Making data1 look for tables in data1_tabpath
62 *
63 * Revision 1.2  1995/11/01  13:54:46  quinn
64 * Minor adjustments
65 *
66 * Revision 1.1  1995/11/01  11:56:08  quinn
67 * Added Retrieval (data management) functions en masse.
68 *
69 *
70 */
71
72#include <stdio.h>
73#include <stdlib.h>
74#include <string.h>
75
76#include <yaz/oid.h>
77#include <yaz/log.h>
78#include <yaz/readconf.h>
79#include <yaz/tpath.h>
80#include <yaz/data1.h>
81#include <yaz/d1_map.h>
82
83data1_maptab *data1_read_maptab (data1_handle dh, const char *file)
84{
85    NMEM mem = data1_nmem_get (dh);
86    data1_maptab *res = (data1_maptab *)nmem_malloc(mem, sizeof(*res));
87    FILE *f;
88    int lineno = 0;
89    int argc;
90    char *argv[50], line[512];
91    data1_mapunit **mapp;
92    int local_numeric = 0;
93
94    if (!(f = yaz_path_fopen(data1_get_tabpath(dh), file, "r")))
95    {
96    yaz_log(LOG_WARN|LOG_ERRNO, "%s", file);
97    return 0;
98    }
99
100    res->name = 0;
101    res->target_absyn_ref = VAL_NONE;
102    res->map = 0;
103    mapp = &res->map;
104    res->next = 0;
105
106    while ((argc = readconf_line(f, &lineno, line, 512, argv, 50)))
107    if (!strcmp(argv[0], "targetref"))
108    {
109        if (argc != 2)
110        {
111        yaz_log(LOG_WARN, "%s:%d: Bad # args for targetref",
112            file, lineno);
113        continue;
114        }
115        if ((res->target_absyn_ref = oid_getvalbyname(argv[1]))
116        == VAL_NONE)
117        {
118        yaz_log(LOG_WARN, "%s:%d: Unknown reference '%s'",
119            file, lineno, argv[1]);
120        continue;
121        }
122    }
123    else if (!strcmp(argv[0], "targetname"))
124    {
125        if (argc != 2)
126        {
127        yaz_log(LOG_WARN, "%s:%d: Bad # args for targetname",
128            file, lineno);
129        continue;
130        }
131        res->target_absyn_name =
132        (char *)nmem_malloc(mem, strlen(argv[1])+1);
133        strcpy(res->target_absyn_name, argv[1]);
134    }
135    else if (!yaz_matchstr(argv[0], "localnumeric"))
136        local_numeric = 1;
137    else if (!strcmp(argv[0], "name"))
138    {
139        if (argc != 2)
140        {
141        yaz_log(LOG_WARN, "%s:%d: Bad # args for name", file, lineno);
142        continue;
143        }
144        res->name = (char *)nmem_malloc(mem, strlen(argv[1])+1);
145        strcpy(res->name, argv[1]);
146    }
147    else if (!strcmp(argv[0], "map"))
148    {
149        data1_maptag **mtp;
150        char *ep, *path = argv[2];
151
152        if (argc < 3)
153        {
154        yaz_log(LOG_WARN, "%s:%d: Bad # of args for map",
155            file, lineno);
156        continue;
157        }
158        *mapp = (data1_mapunit *)nmem_malloc(mem, sizeof(**mapp));
159        (*mapp)->next = 0;
160        if (argc > 3 && !data1_matchstr(argv[3], "nodata"))
161        (*mapp)->no_data = 1;
162        else
163        (*mapp)->no_data = 0;
164        (*mapp)->source_element_name =
165        (char *)nmem_malloc(mem, strlen(argv[1])+1);
166        strcpy((*mapp)->source_element_name, argv[1]);
167        mtp = &(*mapp)->target_path;
168        if (*path == '/')
169        path++;
170        for (ep = strchr(path, '/'); path; (void)((path = ep) &&
171        (ep = strchr(path, '/'))))
172        {
173        int type, np;
174        char valstr[512], parm[512];
175
176        if (ep)
177            ep++;
178        if ((np = sscanf(path, "(%d,%[^)]):%[^/]", &type, valstr,
179            parm)) < 2)
180        {
181            yaz_log(LOG_WARN, "%s:%d: Syntax error in map "
182                "directive: %s", file, lineno, argv[2]);
183            fclose(f);
184            return 0;
185        }
186        *mtp = (data1_maptag *)nmem_malloc(mem, sizeof(**mtp));
187        (*mtp)->next = 0;
188        (*mtp)->type = type;
189        if (np > 2 && !data1_matchstr(parm, "new"))
190            (*mtp)->new_field = 1;
191        else
192            (*mtp)->new_field = 0;
193        if ((type != 3 || local_numeric) && d1_isdigit(*valstr))
194                {
195            (*mtp)->which = D1_MAPTAG_numeric;
196            (*mtp)->value.numeric = atoi(valstr);
197        }
198        else
199        {
200            (*mtp)->which = D1_MAPTAG_string;
201            (*mtp)->value.string =
202            (char *)nmem_malloc(mem, strlen(valstr)+1);
203            strcpy((*mtp)->value.string, valstr);
204        }
205        mtp = &(*mtp)->next;
206        }
207        mapp = &(*mapp)->next;
208    }
209    else
210        yaz_log(LOG_WARN, "%s:%d: Unknown directive '%s'",
211            file, lineno, argv[0]);
212
213    fclose(f);
214    return res;
215}
216
217/*
218 * Locate node with given elementname.
219 * NOTE: This is stupid - we don't find repeats this way.
220 */
221static data1_node *find_node(data1_node *p, char *elementname)
222{
223    data1_node *c, *r;
224
225    for (c = p->child; c; c = c->next)
226    if (c->which == DATA1N_tag && c->u.tag.element &&
227        !data1_matchstr(c->u.tag.element->name, elementname))
228        return c;
229    else if ((r = find_node(c, elementname)))
230        return r;
231    return 0;
232}
233
234/*
235 * See if the node n is equivalent to the tag t.
236 */
237static int tagmatch(data1_node *n, data1_maptag *t)
238{
239    if (n->which != DATA1N_tag)
240    return 0;
241    if (n->u.tag.element)
242    {
243    if (n->u.tag.element->tag->tagset)
244    {
245        if (n->u.tag.element->tag->tagset->type != t->type)
246        return 0;
247    }
248    else if (t->type != 3)
249        return 0;
250    if (n->u.tag.element->tag->which == DATA1T_numeric)
251    {
252        if (t->which != D1_MAPTAG_numeric)
253        return 0;
254        if (n->u.tag.element->tag->value.numeric != t->value.numeric)
255        return 0;
256    }
257    else
258    {
259        if (t->which != D1_MAPTAG_string)
260        return 0;
261        if (data1_matchstr(n->u.tag.element->tag->value.string,
262        t->value.string))
263        return 0;
264    }
265    }
266    else /* local tag */
267    {
268    char str[10];
269
270    if (t->type != 3)
271        return 0;
272    if (t->which == D1_MAPTAG_numeric)
273        sprintf(str, "%d", t->value.numeric);
274    else
275        strcpy(str, t->value.string);
276    if (data1_matchstr(n->u.tag.tag, str))
277        return 0;
278    }
279    return 1;
280}
281
282static int map_children(data1_handle dh, data1_node *n, data1_maptab *map,
283            data1_node *res, NMEM mem)
284{
285    data1_node *c;
286    data1_mapunit *m;
287    /*
288     * locate each source element in turn.
289     */
290    for (c = n->child; c; c = c->next)
291    if (c->which == DATA1N_tag && c->u.tag.element)
292    {
293        for (m = map->map; m; m = m->next)
294        {
295        if (!data1_matchstr(m->source_element_name,
296            c->u.tag.element->name))
297        {
298            data1_node *pn = res;
299            data1_node *cur = pn->last_child;
300            data1_maptag *mt;
301
302            /*
303             * process the target path specification.
304             */
305            for (mt = m->target_path; mt; mt = mt->next)
306            {
307            if (!cur || mt->new_field || !tagmatch(cur, mt))
308            {
309                cur = data1_mk_node (dh, mem);
310                cur->which = DATA1N_tag;
311                cur->u.tag.element = 0;
312                cur->u.tag.tag = mt->value.string;
313                cur->u.tag.node_selected = 0;
314                cur->parent = pn;
315                cur->root = pn->root;
316                if (!pn->child)
317                pn->child = cur;
318                if (pn->last_child)
319                pn->last_child->next = cur;
320                pn->last_child = cur;
321            }
322           
323            if (mt->next)
324                pn = cur;
325            else if (!m->no_data)
326            {
327                cur->child = c->child;
328                cur->last_child = c->last_child;
329                c->child = 0;
330                c->last_child = 0;
331            }
332            }
333            break;
334        }
335        }
336        if (map_children(dh, c, map, res, mem) < 0)
337        return -1;
338    }
339    return 0;
340}
341
342/*
343 * Create a (possibly lossy) copy of the given record based on the
344 * table. The new copy will refer back to the data of the original record,
345 * which should not be discarded during the lifetime of the copy.
346 */
347data1_node *data1_map_record (data1_handle dh, data1_node *n,
348                  data1_maptab *map, NMEM m)
349{
350    data1_node *res = data1_mk_node(dh, m);
351
352    res->which = DATA1N_root;
353    res->u.root.type = map->target_absyn_name;
354    if (!(res->u.root.absyn = data1_get_absyn(dh, map->target_absyn_name)))
355    {
356    yaz_log(LOG_WARN, "%s: Failed to load target absyn '%s'",
357        map->name, map->target_absyn_name);
358    }
359    res->parent = 0;
360    res->root = res;
361
362    if (map_children(dh, n, map, res, m) < 0)
363    {
364    data1_free_tree(dh, res);
365    return 0;
366    }
367    return res;
368}
369
Note: See TracBrowser for help on using the browser.