pgestraier/trunk/pgest.c

/*
 * integrate Hyper Estraier into PostgreSQL
 *
 * Dobrica Pavlinusic <dpavlin@rot13.org> 2005-05-19
 *
 * TODO:
 * - all
 *
 * NOTES:
 * - clear structures with memset to support hash indexes (who whould like
 *   to create hash index on table returned from function?)
 * - number of returned rows is set by PostgreSQL evaluator, see:
 *   http://archives.postgresql.org/pgsql-hackers/2005-02/msg00546.php
 *
 * Based on:
 * - C example from PostgreSQL documentation (BSD licence)
 * - example002.c from Hyper Estraier (GPL)
 * - _textin/_textout from pgcurl.c (LGPL)
 *
 * This code is licenced under GPL
 */

#include "postgres.h"
#include "fmgr.h"
#include "funcapi.h"
#include "utils/builtins.h"
#include "utils/array.h"
#include "miscadmin.h"
#include <estraier.h>
#include <cabin.h>

#define _textin(str) DirectFunctionCall1(textin, CStringGetDatum(str))
#define _textout(str) DatumGetPointer(DirectFunctionCall1(textout, PointerGetDatum(str)))
#define GET_STR(textp) DatumGetCString(DirectFunctionCall1(textout, PointerGetDatum(textp)))
#define GET_TEXT(cstrp) DatumGetTextP(DirectFunctionCall1(textin, CStringGetDatum(cstrp)))


ESTDB *db;
ESTCOND *cond;
ESTDOC *doc;
const CBLIST *texts;
int ecode, *est_result, resnum, i, j;
int limit = 0;
int offset = 0;

/* define PostgreSQL v1 function */
PG_FUNCTION_INFO_V1(pgest);
Datum pgest(PG_FUNCTION_ARGS) {

        FuncCallContext *funcctx;
        int             call_cntr;
        int             max_calls;
        TupleDesc       tupdesc;
        TupleTableSlot  *slot;
        AttInMetadata   *attinmeta;
        char            *index_path;
        char            *query;

        /* stuff done only on the first call of the function */
        if (SRF_IS_FIRSTCALL()) {
                MemoryContext   oldcontext;

                /* take arguments from function */
                //index_path = _textout(PG_GETARG_TEXT_P(0));
                index_path = _textout(PG_GETARG_TEXT_P(0));
                query = _textout(PG_GETARG_TEXT_P(1));
                limit = PG_GETARG_INT32(2);
                offset = PG_GETARG_INT32(3);

                /* create a function context for cross-call persistence */
                funcctx = SRF_FIRSTCALL_INIT();

                /* switch to memory context appropriate for multiple function calls */
                oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);

                /* open the database */
                elog(DEBUG1, "pgest: est_db_open(%s)", index_path);
                
                if(!(db = est_db_open(index_path, ESTDBREADER, &ecode))){
                        elog(ERROR, "est_db_open: can't open %s [%d]: %s", index_path, ecode, est_err_msg(ecode));
                        SRF_RETURN_DONE(funcctx);
                }
                
                elog(INFO, "pgest: query=%s limit %d offset %d", query, limit, offset);
                
                /* create a search condition object */
                if (!(cond = est_cond_new())) {
                        elog(INFO, "pgest: est_cond_new failed");
                        SRF_RETURN_DONE(funcctx);
                }
                
                /* set the search phrase to the search condition object */
                est_cond_set_phrase(cond, query);

                /* get the result of search */
                est_result = est_db_search(db, cond, &resnum, NULL);
                
                /* total number of tuples to be returned */
                funcctx->max_calls = resnum;

                /* check if results exists */
                if ( 0 == funcctx->max_calls )
                        elog(INFO, "pgest: no results for: %s", query );

                elog(DEBUG1, "pgest: found %d hits for %s", funcctx->max_calls, query);

                /* Build a tuple description for a __pgest tuple */
                tupdesc = RelationNameGetTupleDesc("__pgest");

                /* allocate a slot for a tuple with this tupdesc */
                slot = TupleDescGetSlot(tupdesc);

                /* assign slot to function context */
                funcctx->slot = slot;

                /*
                 * generate attribute metadata needed later to produce tuples from raw
                 * C strings
                 */
                attinmeta = TupleDescGetAttInMetadata(tupdesc);
                funcctx->attinmeta = attinmeta;

                MemoryContextSwitchTo(oldcontext);

                elog(DEBUG1, "SRF_IS_FIRSTCALL done");
        }

        /* stuff done on every call of the function */
        funcctx = SRF_PERCALL_SETUP();

        call_cntr = funcctx->call_cntr;
        max_calls = funcctx->max_calls;
        slot = funcctx->slot;
        attinmeta = funcctx->attinmeta;

        if (limit && call_cntr > limit - 1) {
                elog(INFO, "call_cntr: %d limit: %d", call_cntr, limit);
                SRF_RETURN_DONE(funcctx);
        }

        if (call_cntr < max_calls) {
                char            **values;
                HeapTuple       tuple;
                Datum           result;

                elog(DEBUG1, "pgest: loop count %d", call_cntr);

                if (! est_result) {
                        elog(ERROR, "pgest: no estraier results");
                        SRF_RETURN_DONE(funcctx);
                }
                
                /*
                 * Prepare a values array for storage in our slot.
                 * This should be an array of C strings which will
                 * be processed later by the type input functions.
                 */

                if (doc = est_db_get_doc(db, est_result[call_cntr + offset], 0)) {
                
                        elog(DEBUG1, "URI: %s\n Title: %s\n",
                                est_doc_attr(doc, "@uri"),
                                est_doc_attr(doc, "@title")
                        );

                        values = (char **) palloc(4 * sizeof(char *));

//                      values[0] = (char *) palloc(strlen(_estval) * sizeof(char));

                        values[0] = attr2text(doc,"@id");
                        values[1] = attr2text(doc,"@uri");
                        values[2] = attr2text(doc,"@title");
                        values[3] = attr2text(doc,"@type");

                        /* destloy the document object */
                        elog(DEBUG2, "est_doc_delete");
                        est_doc_delete(doc);
                } else {
                        elog(INFO, "no result from estraier");
                        values[0] = NULL;
                        values[1] = NULL;
                        values[2] = NULL;
                        values[3] = NULL;
                }


                elog(DEBUG2, "build tuple");
                /* build a tuple */
                tuple = BuildTupleFromCStrings(attinmeta, values);

                elog(DEBUG2, "make tuple into datum");
                /* make the tuple into a datum */
                result = TupleGetDatum(slot, tuple);

                elog(DEBUG2, "cleanup");
                /* clean up ? */
/*
                pfree(values[0]);
                pfree(values[1]);
                pfree(values[2]);
                pfree(values[3]);
                pfree(values);
*/
                
                elog(DEBUG2, "cleanup over");
        
                SRF_RETURN_NEXT(funcctx, result);
        } else {
                elog(DEBUG1, "loop over");

                if(!est_db_close(db, &ecode)){
                        elog(INFO, "est_db_close error: %s", est_err_msg(ecode));
                }

                /* do when there is no more left */
                SRF_RETURN_DONE(funcctx);
        }
}

/* work in progress */
PG_FUNCTION_INFO_V1(pgest2);
Datum pgest2(PG_FUNCTION_ARGS)
{
        int             nrows = 3;
        int16           typlen;
        bool            typbyval;
        char            typalign;
        ReturnSetInfo   *rsinfo = (ReturnSetInfo *) fcinfo->resultinfo;
        AttInMetadata   *attinmeta;
        TupleDesc       tupdesc;
        Tuplestorestate *tupstore = NULL;
        HeapTuple       tuple;
        MemoryContext   per_query_ctx;
        MemoryContext   oldcontext;
        Datum           dvalue;
        char            **values;
        int             ncols;
        int             i, j;

        /* check to see if caller supports us returning a tuplestore */
        if (!rsinfo || !(rsinfo->allowedModes & SFRM_Materialize))
                ereport(ERROR,
                                (errcode(ERRCODE_SYNTAX_ERROR),
                                 errmsg("materialize mode required, but it is not " \
                                                "allowed in this context")));

        /* get the requested return tuple description */
        tupdesc = rsinfo->expectedDesc;
        ncols = tupdesc->natts;

        /*
         * The requested tuple description better match up with the array
         * we were given.
         */
        /* OK, use it */
        attinmeta = TupleDescGetAttInMetadata(tupdesc);

        /* Now go to work */
        rsinfo->returnMode = SFRM_Materialize;

        per_query_ctx = fcinfo->flinfo->fn_mcxt;
        oldcontext = MemoryContextSwitchTo(per_query_ctx);

        /* initialize our tuplestore */
        tupstore = tuplestore_begin_heap(true, false, SortMem);

        values = (char **) palloc(ncols * sizeof(char *));

        for (i = 0; i < nrows; i++)
        {
                for (j = 0; j < ncols; j++)
                {
                        values[j] = DatumGetCString( "foo" );
                }
                /* construct the tuple */
                tuple = BuildTupleFromCStrings(attinmeta, values);

                /* now store it */
                tuplestore_puttuple(tupstore, tuple);
        }

        tuplestore_donestoring(tupstore);
        rsinfo->setResult = tupstore;

        /*
         * SFRM_Materialize mode expects us to return a NULL Datum. The actual
         * tuples are in our tuplestore and passed back through
         * rsinfo->setResult. rsinfo->setDesc is set to the tuple description
         * that we actually used to build our tuples with, so the caller can
         * verify we did what it was expecting.
         */
        rsinfo->setDesc = tupdesc;
        MemoryContextSwitchTo(oldcontext);

        return (Datum) 0;
}


/* make text var from attr */
char *attr2text(ESTDOC *doc, char *attr) {
        char *val;
        const char *attrval;
        int len;
        int attrlen;

        elog(DEBUG1, "doc: %08x, attr: %s", doc, attr);

        if ( (attrval = est_doc_attr(doc, attr)) && (attrlen = strlen(attrval)) ) {
                val = (char *) palloc(attrlen * sizeof(char));
        } else {
                return (Datum) NULL;
        }

        len = strlen(attrval);
        elog(DEBUG1, "attr2text(%s) = '%s' %d bytes", attr, attrval, len);

        len++;
        len *= sizeof(char);

        elog(DEBUG2, "palloc(%d)", len);

        val = palloc(len);

        memset(val, 0, len);
        strncpy(val, attrval, len);

        elog(DEBUG2, "val=%s", val);

        return val;
}

/* make integer variable from property */
/*
char *prop2int(SW_RESULT sw_res, char *propname) {
        char *val;
        unsigned long prop;
        int len;

        elog(DEBUG2, "prop2int(%s)", propname);

        prop = estResultPropertyULong( sw_res, propname );
        if (error_or_abort( est_handle )) return NULL;

        elog(DEBUG1, "prop2int(%s) = %lu", propname, prop);

        len = 128 * sizeof(char);
        elog(DEBUG2, "palloc(%d)", len);

        val = palloc(len);
        memset(val, 0, len);

        snprintf(val, len, "%lu", prop);

        elog(DEBUG2, "val=%s", val);

        return val;
}
*/
1	dpavlin	1	/*
2			* integrate Hyper Estraier into PostgreSQL
3			*
4			* Dobrica Pavlinusic <dpavlin@rot13.org> 2005-05-19
5			*
6			* TODO:
7			* - all
8			*
9			* NOTES:
10			* - clear structures with memset to support hash indexes (who whould like
11			* to create hash index on table returned from function?)
12			* - number of returned rows is set by PostgreSQL evaluator, see:
13			* http://archives.postgresql.org/pgsql-hackers/2005-02/msg00546.php
14			*
15			* Based on:
16			* - C example from PostgreSQL documentation (BSD licence)
17			* - example002.c from Hyper Estraier (GPL)
18			* - _textin/_textout from pgcurl.c (LGPL)
19			*
20			* This code is licenced under GPL
21			*/
22
23			#include "postgres.h"
24			#include "fmgr.h"
25			#include "funcapi.h"
26			#include "utils/builtins.h"
27			#include "utils/array.h"
28			#include "miscadmin.h"
29			#include <estraier.h>
30			#include <cabin.h>
31
32			#define _textin(str) DirectFunctionCall1(textin, CStringGetDatum(str))
33			#define _textout(str) DatumGetPointer(DirectFunctionCall1(textout, PointerGetDatum(str)))
34			#define GET_STR(textp) DatumGetCString(DirectFunctionCall1(textout, PointerGetDatum(textp)))
35			#define GET_TEXT(cstrp) DatumGetTextP(DirectFunctionCall1(textin, CStringGetDatum(cstrp)))
36
37
38			ESTDB *db;
39			ESTCOND *cond;
40			ESTDOC *doc;
41			const CBLIST *texts;
42			int ecode, *est_result, resnum, i, j;
43	dpavlin	3	int limit = 0;
44			int offset = 0;
45	dpavlin	1
46			/* define PostgreSQL v1 function */
47			PG_FUNCTION_INFO_V1(pgest);
48			Datum pgest(PG_FUNCTION_ARGS) {
49
50			FuncCallContext *funcctx;
51			int call_cntr;
52			int max_calls;
53			TupleDesc tupdesc;
54			TupleTableSlot *slot;
55			AttInMetadata *attinmeta;
56			char *index_path;
57			char *query;
58
59			/* stuff done only on the first call of the function */
60			if (SRF_IS_FIRSTCALL()) {
61			MemoryContext oldcontext;
62
63			/* take arguments from function */
64			//index_path = _textout(PG_GETARG_TEXT_P(0));
65			index_path = _textout(PG_GETARG_TEXT_P(0));
66			query = _textout(PG_GETARG_TEXT_P(1));
67	dpavlin	3	limit = PG_GETARG_INT32(2);
68			offset = PG_GETARG_INT32(3);
69	dpavlin	1
70			/* create a function context for cross-call persistence */
71			funcctx = SRF_FIRSTCALL_INIT();
72
73			/* switch to memory context appropriate for multiple function calls */
74			oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
75
76			/* open the database */
77			elog(DEBUG1, "pgest: est_db_open(%s)", index_path);
78
79			if(!(db = est_db_open(index_path, ESTDBREADER, &ecode))){
80			elog(ERROR, "est_db_open: can't open %s [%d]: %s", index_path, ecode, est_err_msg(ecode));
81			SRF_RETURN_DONE(funcctx);
82			}
83
84	dpavlin	3	elog(INFO, "pgest: query=%s limit %d offset %d", query, limit, offset);
85	dpavlin	1
86			/* create a search condition object */
87			if (!(cond = est_cond_new())) {
88			elog(INFO, "pgest: est_cond_new failed");
89			SRF_RETURN_DONE(funcctx);
90			}
91
92			/* set the search phrase to the search condition object */
93			est_cond_set_phrase(cond, query);
94
95			/* get the result of search */
96			est_result = est_db_search(db, cond, &resnum, NULL);
97
98			/* total number of tuples to be returned */
99			funcctx->max_calls = resnum;
100
101			/* check if results exists */
102			if ( 0 == funcctx->max_calls )
103			elog(INFO, "pgest: no results for: %s", query );
104
105	dpavlin	2	elog(DEBUG1, "pgest: found %d hits for %s", funcctx->max_calls, query);
106	dpavlin	1
107			/* Build a tuple description for a __pgest tuple */
108			tupdesc = RelationNameGetTupleDesc("__pgest");
109
110			/* allocate a slot for a tuple with this tupdesc */
111			slot = TupleDescGetSlot(tupdesc);
112
113			/* assign slot to function context */
114			funcctx->slot = slot;
115
116			/*
117			* generate attribute metadata needed later to produce tuples from raw
118			* C strings
119			*/
120			attinmeta = TupleDescGetAttInMetadata(tupdesc);
121			funcctx->attinmeta = attinmeta;
122
123			MemoryContextSwitchTo(oldcontext);
124
125	dpavlin	2	elog(DEBUG1, "SRF_IS_FIRSTCALL done");
126	dpavlin	1	}
127
128			/* stuff done on every call of the function */
129			funcctx = SRF_PERCALL_SETUP();
130
131			call_cntr = funcctx->call_cntr;
132			max_calls = funcctx->max_calls;
133			slot = funcctx->slot;
134			attinmeta = funcctx->attinmeta;
135	dpavlin	3
136			if (limit && call_cntr > limit - 1) {
137			elog(INFO, "call_cntr: %d limit: %d", call_cntr, limit);
138			SRF_RETURN_DONE(funcctx);
139			}
140
141	dpavlin	1	if (call_cntr < max_calls) {
142			char **values;
143			HeapTuple tuple;
144			Datum result;
145
146	dpavlin	2	elog(DEBUG1, "pgest: loop count %d", call_cntr);
147	dpavlin	1
148			if (! est_result) {
149			elog(ERROR, "pgest: no estraier results");
150			SRF_RETURN_DONE(funcctx);
151			}
152
153			/*
154			* Prepare a values array for storage in our slot.
155			* This should be an array of C strings which will
156			* be processed later by the type input functions.
157			*/
158
159	dpavlin	3	if (doc = est_db_get_doc(db, est_result[call_cntr + offset], 0)) {
160	dpavlin	1
161	dpavlin	2	elog(DEBUG1, "URI: %s\n Title: %s\n",
162	dpavlin	1	est_doc_attr(doc, "@uri"),
163			est_doc_attr(doc, "@title")
164			);
165
166			values = (char *) palloc(4 sizeof(char *));
167
168			// values[0] = (char ) palloc(strlen(_estval) sizeof(char));
169
170			values[0] = attr2text(doc,"@id");
171			values[1] = attr2text(doc,"@uri");
172			values[2] = attr2text(doc,"@title");
173			values[3] = attr2text(doc,"@type");
174
175			/* destloy the document object */
176			elog(DEBUG2, "est_doc_delete");
177			est_doc_delete(doc);
178			} else {
179			elog(INFO, "no result from estraier");
180			values[0] = NULL;
181			values[1] = NULL;
182			values[2] = NULL;
183			values[3] = NULL;
184			}
185
186
187			elog(DEBUG2, "build tuple");
188			/* build a tuple */
189			tuple = BuildTupleFromCStrings(attinmeta, values);
190
191			elog(DEBUG2, "make tuple into datum");
192			/* make the tuple into a datum */
193			result = TupleGetDatum(slot, tuple);
194
195			elog(DEBUG2, "cleanup");
196			/* clean up ? */
197			/*
198			pfree(values[0]);
199			pfree(values[1]);
200			pfree(values[2]);
201			pfree(values[3]);
202			pfree(values);
203			*/
204
205			elog(DEBUG2, "cleanup over");
206
207			SRF_RETURN_NEXT(funcctx, result);
208			} else {
209	dpavlin	2	elog(DEBUG1, "loop over");
210	dpavlin	1
211			if(!est_db_close(db, &ecode)){
212			elog(INFO, "est_db_close error: %s", est_err_msg(ecode));
213			}
214
215			/* do when there is no more left */
216			SRF_RETURN_DONE(funcctx);
217			}
218			}
219
220			/* work in progress */
221			PG_FUNCTION_INFO_V1(pgest2);
222			Datum pgest2(PG_FUNCTION_ARGS)
223			{
224			int nrows = 3;
225			int16 typlen;
226			bool typbyval;
227			char typalign;
228			ReturnSetInfo rsinfo = (ReturnSetInfo ) fcinfo->resultinfo;
229			AttInMetadata *attinmeta;
230			TupleDesc tupdesc;
231			Tuplestorestate *tupstore = NULL;
232			HeapTuple tuple;
233			MemoryContext per_query_ctx;
234			MemoryContext oldcontext;
235			Datum dvalue;
236			char **values;
237			int ncols;
238			int i, j;
239
240			/* check to see if caller supports us returning a tuplestore */
241			if (!rsinfo \|\| !(rsinfo->allowedModes & SFRM_Materialize))
242			ereport(ERROR,
243			(errcode(ERRCODE_SYNTAX_ERROR),
244			errmsg("materialize mode required, but it is not " \
245			"allowed in this context")));
246
247			/* get the requested return tuple description */
248			tupdesc = rsinfo->expectedDesc;
249			ncols = tupdesc->natts;
250
251			/*
252			* The requested tuple description better match up with the array
253			* we were given.
254			*/
255			/* OK, use it */
256			attinmeta = TupleDescGetAttInMetadata(tupdesc);
257
258			/* Now go to work */
259			rsinfo->returnMode = SFRM_Materialize;
260
261			per_query_ctx = fcinfo->flinfo->fn_mcxt;
262			oldcontext = MemoryContextSwitchTo(per_query_ctx);
263
264			/* initialize our tuplestore */
265			tupstore = tuplestore_begin_heap(true, false, SortMem);
266
267			values = (char *) palloc(ncols sizeof(char *));
268
269			for (i = 0; i < nrows; i++)
270			{
271			for (j = 0; j < ncols; j++)
272			{
273			values[j] = DatumGetCString( "foo" );
274			}
275			/* construct the tuple */
276			tuple = BuildTupleFromCStrings(attinmeta, values);
277
278			/* now store it */
279			tuplestore_puttuple(tupstore, tuple);
280			}
281
282			tuplestore_donestoring(tupstore);
283			rsinfo->setResult = tupstore;
284
285			/*
286			* SFRM_Materialize mode expects us to return a NULL Datum. The actual
287			* tuples are in our tuplestore and passed back through
288			* rsinfo->setResult. rsinfo->setDesc is set to the tuple description
289			* that we actually used to build our tuples with, so the caller can
290			* verify we did what it was expecting.
291			*/
292			rsinfo->setDesc = tupdesc;
293			MemoryContextSwitchTo(oldcontext);
294
295			return (Datum) 0;
296			}
297
298
299			/* make text var from attr */
300			char attr2text(ESTDOC doc, char *attr) {
301			char *val;
302			const char *attrval;
303			int len;
304	dpavlin	4	int attrlen;
305	dpavlin	1
306	dpavlin	2	elog(DEBUG1, "doc: %08x, attr: %s", doc, attr);
307	dpavlin	1
308	dpavlin	4	if ( (attrval = est_doc_attr(doc, attr)) && (attrlen = strlen(attrval)) ) {
309			val = (char ) palloc(attrlen sizeof(char));
310	dpavlin	1	} else {
311			return (Datum) NULL;
312			}
313
314			len = strlen(attrval);
315	dpavlin	2	elog(DEBUG1, "attr2text(%s) = '%s' %d bytes", attr, attrval, len);
316	dpavlin	1
317			len++;
318			len *= sizeof(char);
319
320			elog(DEBUG2, "palloc(%d)", len);
321
322			val = palloc(len);
323
324			memset(val, 0, len);
325			strncpy(val, attrval, len);
326
327			elog(DEBUG2, "val=%s", val);
328
329			return val;
330			}
331
332			/* make integer variable from property */
333			/*
334			char prop2int(SW_RESULT sw_res, char propname) {
335			char *val;
336			unsigned long prop;
337			int len;
338
339			elog(DEBUG2, "prop2int(%s)", propname);
340
341			prop = estResultPropertyULong( sw_res, propname );
342			if (error_or_abort( est_handle )) return NULL;
343
344			elog(DEBUG1, "prop2int(%s) = %lu", propname, prop);
345
346			len = 128 * sizeof(char);
347			elog(DEBUG2, "palloc(%d)", len);
348
349			val = palloc(len);
350			memset(val, 0, len);
351
352			snprintf(val, len, "%lu", prop);
353
354			elog(DEBUG2, "val=%s", val);
355
356			return val;
357			}
358			*/