Trigger (C) question - Mailing list pgsql-general

From Mitch Vincent
Subject Trigger (C) question
Date
Msg-id 02d501bfe114$6b351660$0300000a@doot.org
Whole thread Raw
Responses Re: Trigger (C) question  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general
This is a bit off topic but email to the author of the trigger bounced..

In the fti.c (full text index) trigger I see..

sprintf(query, "INSERT INTO %s (string, id) VALUES ($1, $2)", indexname);

below.. I need to add something else here.. Basically I'm eliminating the
need for the text table, I'm going to integrate the fti indexed table into
an existing one I have but I need to insert another ID into the resumes_fti
(my indexed fti table)..

The above makes the query string, as I can see but what in the world is the
$1 $2 and where did they come from? Since it looks like that's how the
values of string and id are put into the query, I need to know so I can add
one more in there..

Below is the whole function I'm referring to in fti.c

Thanks!!!


if (isinsert)
 {
  char    *substring,
       *column;
  void    *pplan;
  Oid     *argtypes;
  Datum  values[2];
  int   colnum;
  struct varlena *data;
  EPlan    *plan;

  sprintf(query, "I%s$%s", args[0], args[1]);
  plan = find_plan(query, &InsertPlans, &nInsertPlans);

  /* no plan yet, so allocate mem for argtypes */
  if (plan->nplans <= 0)
  {
   argtypes = (Oid *) palloc(2 * sizeof(Oid));

   argtypes[0] = VARCHAROID; /* create table t_name (string
           * varchar, */
   argtypes[1] = OIDOID;  /* id   oid);    */

   /* prepare plan to gain speed */
   sprintf(query, "INSERT INTO %s (string, id) VALUES ($1, $2)",
     indexname);
   pplan = SPI_prepare(query, 2, argtypes);
   if (!pplan)
    elog(ERROR, "Full Text Indexing: SPI_prepare returned NULL "
      "in insert");

   pplan = SPI_saveplan(pplan);
   if (pplan == NULL)
    elog(ERROR, "Full Text Indexing: SPI_saveplan returned NULL"
      " in insert");

   plan->splan = (void **) malloc(sizeof(void *));
   *(plan->splan) = pplan;
   plan->nplans = 1;
  }





pgsql-general by date:

Previous
From: Travis Bauer
Date:
Subject: Re: Restricting queries by the presence of a WHERE clause
Next
From: mikeo
Date:
Subject: disk backups