让Greenplum 支持反转索引

GP的反转索引可以通过函数reverse来实现，但是这个函数在GP的版本中没有，所以需要port过来。
可以在9.5的代码中找到
src/backend/utils/adt/varlena.c

vi reverse.c

#include <string.h>
#include "postgres.h"
#include "fmgr.h"

PG_MODULE_MAGIC;

PG_FUNCTION_INFO_V1(text_reverse);


/*
 * Return reversed string
 */
Datum
text_reverse(PG_FUNCTION_ARGS)
{
        text       *str = PG_GETARG_TEXT_PP(0);
        const char *p = VARDATA_ANY(str);
        int                     len = VARSIZE_ANY_EXHDR(str);
        const char *endp = p + len;
        text       *result;
        char       *dst;

        result = palloc(len + VARHDRSZ);
        dst = (char *) VARDATA(result) + len;
        SET_VARSIZE(result, len + VARHDRSZ);

        if (pg_database_encoding_max_length() > 1)
        {
                /* multibyte version */
                while (p < endp)
                {
                        int                     sz;

                        sz = pg_mblen(p);
                        dst -= sz;
                        memcpy(dst, p, sz);
                        p += sz;
                }
        }
        else
        {
                /* single byte version */
                while (p < endp)
                        *(--dst) = *p++;
        }

        PG_RETURN_TEXT_P(result);
}

编译

gcc -O3 -Wall -Wextra -Werror -I /home/digoal/gpsrc/src/include -g -fPIC -c ./reverse.c -o reverse.o
gcc -O3 -Wall -Wextra -Werror -I /home/digoal/gpsrc/src/include -g -shared reverse.o -o libreverse.so

cp libreverse.so /home/digoal/gphome/lib

拷贝到所有节点

gpscp -f ./host /home/digoal/gphome/lib/libreverse.so =:/home/digoal/gphome/lib/

创建函数并测试可用性

postgres=# create or replace function reverse(text) returns text as '/home/digoal/gphome/lib/libreverse.so', 'text_reverse' language C STRICT immutable;
CREATE FUNCTION
postgres=# select reverse('abc');
 reverse 
---------
 cba
(1 row)

postgres=# select reverse('a f d12');
 reverse 
---------
 21d f a
(1 row)

postgres=# select reverse(null);
 reverse 
---------

(1 row)

创建反转索引测试

postgres=# create table t(id int, info text);
NOTICE:  Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'id' as the Greenplum Database data distribution key for this table.
HINT:  The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
CREATE TABLE
postgres=# create index idx on t(reverse(info));
CREATE INDEX
postgres=# insert into t select id,md5(random()::text) from generate_series(1,1000000) t(id);
INSERT 0 1000000

postgres=# select id,info,reverse(info) from t limit 10;
  id  |               info               |             reverse              
------+----------------------------------+----------------------------------
  197 | fefb23cb1705e6faa74601e6cd0dc8d7 | 7d8cd0dc6e10647aaf6e5071bc32bfef
  314 | 64c3d79458fc0ba2413e5493582830dd | dd0382853945e3142ab0cf85497d3c46
  426 | e0486de86c2c6bd72912fbc33d059de8 | 8ed950d33cbf21927db6c2c68ed6840e
  715 | c5087148a8086e2201a63b158268adbb | bbda862851b36a1022e6808a8417805c
  768 | 50ebdba7ff260d5adc11495313817221 | 12271831359411cda5d062ff7abdbe05
  944 | 8da13db138858ec1f78193f5c16cc310 | 013cc61c5f39187f1ce858831bd31ad8
 1057 | cf029096c2c66714861d0adba9ab49d8 | 8d94ba9abda0d16841766c2c690920fc
 1233 | ae6eb1bdf32b15c73c1a04adf54b9881 | 1889b45fda40a1c37c51b23fdb1be6ea
 1286 | 9943f89159055e0450765ab0968a1ad8 | 8da1a8690ba5670540e55095198f3499
 1575 | b8f0337315d238070984b9883a965c57 | 75c569a3889b489070832d5137330f8b
(10 rows)

postgres=# select * from t where reverse(info)>='7d8cd0dc6e10647aaf6e507' and reverse(info)<'7d8cd0dc6e10647aaf6e508';
 id  |               info               
-----+----------------------------------
 197 | fefb23cb1705e6faa74601e6cd0dc8d7
(1 row)

postgres=# explain analyze select * from t where reverse(info)>='7d8cd0dc6e10647aaf6e507' and reverse(info)<'7d8cd0dc6e10647aaf6e508';
                                                           QUERY PLAN                                                           
--------------------------------------------------------------------------------------------------------------------------------
 Gather Motion 240:1  (slice1; segments: 240)  (cost=11012.90..13972.90 rows=40001 width=37)
   Rows out:  1 rows at destination with 9.113 ms to first row, 43 ms to end, start offset by 1.317 ms.
   ->  Bitmap Heap Scan on t  (cost=11012.90..13972.90 rows=167 width=37)
         Recheck Cond: reverse(info) >= '7d8cd0dc6e10647aaf6e507'::text AND reverse(info) < '7d8cd0dc6e10647aaf6e508'::text
         Rows out:  1 rows (seg46) with 0.156 ms to first row, 0.178 ms to end, start offset by 9.850 ms.
         ->  Bitmap Index Scan on idx  (cost=0.00..11002.90 rows=167 width=0)
               Index Cond: reverse(info) >= '7d8cd0dc6e10647aaf6e507'::text AND reverse(info) < '7d8cd0dc6e10647aaf6e508'::text
               Bitmaps out:  Avg 1.0 x 240 workers.  Max 1 (seg0) with 0.021 ms to end, start offset by 8.845 ms.
               Work_mem used:  9K bytes.
 Slice statistics:
   (slice0)    Executor memory: 475K bytes.
   (slice1)    Executor memory: 321K bytes avg x 240 workers, 329K bytes max (seg46).  Work_mem: 9K bytes max.
 Statement statistics:
   Memory used: 128000K bytes
 Total runtime: 71.958 ms
(15 rows)

适用场景：
1. 带后缀的检索。