2. 程式人生 > 其它 >PostgreSQL模糊查詢

PostgreSQL模糊查詢

阿新 • • 發佈：2021-12-05

1、建立測試表、插入資料

postgres=# create table test_like(id int, name varchar);
CREATE TABLE
postgres=# insert into test_like select generate_series(1,1000000),md5(random()::varchar);
INSERT 0 1000000
postgres=# analyze test_like ;
ANALYZE

2、未建立索引的執行計劃

postgres=# explain select * from test_like where name like '111%' ;
                             QUERY PLAN
--------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Seq Scan on test_like  (cost=0.00..7950.56 rows=33 width=37)
         Filter: ((name)::text ~~ '111%'::text)
(4 rows)

3、普通btree索引不走索引

postgres=# create index idx_test_like_name on test_like (name);
CREATE INDEX
postgres=# analyze test_like ;
ANALYZE
postgres=# explain select * from test_like where name like '111%' ;
                             QUERY PLAN
--------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Seq Scan on test_like  (cost=0.00..7950.56 rows=33 width=37)
         Filter: ((name)::text ~~ '111%'::text)
(4 rows)

4、前模糊匹配查詢

1）collate "C"

postgres=# create index idx_test_like_name_c on test_like (name collate "C");
CREATE INDEX
postgres=# analyze test_like ;
ANALYZE
postgres=# explain select * from test_like where name like '111%' ;
                                         QUERY PLAN
--------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=2.99..108.88 rows=33 width=37)
         Filter: ((name)::text ~~ '111%'::text)
         ->  Bitmap Index Scan on idx_test_like_name_c  (cost=0.00..2.98 rows=56 width=0)
               Index Cond: (((name)::text >= '111'::text) AND ((name)::text < '112'::text))
(6 rows)

2）操作符類varchar_pattern_ops方式

postgres=# create index idx_test_like_name_ops on test_like (name varchar_pattern_ops);
CREATE INDEX
postgres=# analyze test_like ;
ANALYZE
postgres=# explain select * from test_like where name like '111%' ;
                                           QUERY PLAN
------------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=2.97..105.20 rows=33 width=37)
         Filter: ((name)::text ~~ '111%'::text)
         ->  Bitmap Index Scan on idx_test_like_name_ops  (cost=0.00..2.96 rows=54 width=0)
               Index Cond: (((name)::text ~>=~ '111'::text) AND ((name)::text ~<~ '112'::text))
(6 rows)

但後匹配、中間匹配不支援：

postgres=# explain select * from test_like where name like '%111%1' ;
                             QUERY PLAN
--------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Seq Scan on test_like  (cost=0.00..7950.56 rows=33 width=37)
         Filter: ((name)::text ~~ '%111%1'::text)
(4 rows)

postgres=# explain select * from test_like where name like '%111' ;
                             QUERY PLAN
--------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Seq Scan on test_like  (cost=0.00..7950.56 rows=33 width=37)
         Filter: ((name)::text ~~ '%111'::text)
(4 rows)

postgres=# explain select * from test_like where name like '111%' ;
                                           QUERY PLAN
------------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=2.98..107.04 rows=33 width=37)
         Filter: ((name)::text ~~ '111%'::text)
         ->  Bitmap Index Scan on idx_test_like_name_ops  (cost=0.00..2.97 rows=55 width=0)
               Index Cond: (((name)::text ~>=~ '111'::text) AND ((name)::text ~<~ '112'::text))
(6 rows)

postgres=# explain select * from test_like where name like '%111%' ;
                             QUERY PLAN
--------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Seq Scan on test_like  (cost=0.00..7950.56 rows=33 width=37)
         Filter: ((name)::text ~~ '%111%'::text)
(4 rows)

5、方式二：使用pg_trim外掛

The pg_trgm module provides functions and operators for determining the similarity of alphanumeric text based on trigram matching, as well as index operator classes that support fast searching for similar strings.

A trigram is a group of three consecutive characters taken from a string. We can measure the similarity of two strings by counting the number of trigrams they share. This simple idea turns out to be very effective for measuring the similarity of words in many natural languages.

Note: pg_trgm ignores non-word characters (non-alphanumerics) when extracting trigrams from a string. Each word is considered to have two spaces prefixed and one space suffixed when determining the set of trigrams contained in the string. For example, the set of trigrams in the string "cat" is " c", " ca", "cat", and "at ". The set of trigrams in the string "foo|bar" is " f", " fo", "foo", "oo ", " b", " ba", "bar", and "ar ".

postgres=# create extension pg_trgm;
CREATE EXTENSION
postgres=# create index idx_test_like_name_trgm on test_like using gin (name gin_trgm_ops);
CREATE INDEX
postgres=# analyze test_like ;
ANALYZE
postgres=# explain select * from test_like where name like '111%' ;
                                          QUERY PLAN
----------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=18.26..81.59 rows=33 width=37)
         Recheck Cond: ((name)::text ~~ '111%'::text)
         ->  Bitmap Index Scan on idx_test_like_name_trgm  (cost=0.00..18.25 rows=33 width=0)
               Index Cond: ((name)::text ~~ '111%'::text)
(6 rows)

postgres=# explain select * from test_like where name like '%111%' ;
                                           QUERY PLAN
------------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=62.16..3816.37 rows=6731 width=37)
         Recheck Cond: ((name)::text ~~ '%111%'::text)
         ->  Bitmap Index Scan on idx_test_like_name_trgm  (cost=0.00..60.48 rows=6731 width=0)
               Index Cond: ((name)::text ~~ '%111%'::text)
(6 rows)

postgres=# explain select * from test_like where name like '%111' ;
                                          QUERY PLAN
----------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=12.26..75.59 rows=33 width=37)
         Recheck Cond: ((name)::text ~~ '%111'::text)
         ->  Bitmap Index Scan on idx_test_like_name_trgm  (cost=0.00..12.25 rows=33 width=0)
               Index Cond: ((name)::text ~~ '%111'::text)
(6 rows)

postgres=# explain select * from test_like where name like '%111%1' ;
                                         QUERY PLAN
---------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=6.26..69.59 rows=33 width=37)
         Recheck Cond: ((name)::text ~~ '%111%1'::text)
         ->  Bitmap Index Scan on idx_test_like_name_trgm  (cost=0.00..6.25 rows=33 width=0)
               Index Cond: ((name)::text ~~ '%111%1'::text)
(6 rows)

缺陷 ：就像上面描述的那樣，pg_trgm把字串切成多個三元組，然後做匹配，當匹配關鍵詞少於三個的時候，只有字首匹配會走索引：

postgres=# explain select * from test_like where name like '%11%' ;
                              QUERY PLAN
-----------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Seq Scan on test_like  (cost=0.00..7950.56 rows=40384 width=37)
         Filter: ((name)::text ~~ '%11%'::text)
(4 rows)

postgres=# explain select * from test_like where name like '%111%' ;
                                           QUERY PLAN
------------------------------------------------------------------------------------------------
 Remote Fast Query Execution  (cost=0.00..0.00 rows=0 width=0)
   Node/s: dn001, dn002, dn003
   ->  Bitmap Heap Scan on test_like  (cost=34.08..2908.28 rows=3365 width=37)
         Recheck Cond: ((name)::text ~~ '%111%'::text)
         ->  Bitmap Index Scan on idx_test_like_name_trgm  (cost=0.00..33.24 rows=3365 width=0)
               Index Cond: ((name)::text ~~ '%111%'::text)
(6 rows)
嚴以律己、寬以待人

相關推薦

PostgreSQL模糊查詢

1、建立測試表、插入資料 postgres=# create table test_like(id int, name varchar); CREATE TABLE

MongoDB多條件模糊查詢示例程式碼

前言 模糊查詢是資料庫的基本操作之一，實現對給定的字串是否與指定的模式進行匹配。如果字元完全匹配，可以用=等號表示，如果部分匹配可認為是一種模糊查詢。在關係型資料中，通過SQL使用like ‘%fens%\'的語法。那

pgsql查詢優化之模糊查詢例項詳解

前言 一直以來，對於搜尋時模糊匹配的優化一直是個讓人頭疼的問題，好在強大pgsql提供了優化方案，下面就來簡單談一談如何通過索引來優化模糊匹配

MongoDB模糊查詢操作案例詳解(類關係型資料庫的 like 和 not like)

1.作用與語法描述 作用： 正則表示式是使用指定字串來描述、匹配一系列符合某個句法規則的字串。許多程式設計語言都支援利用正則表示式進行字串操作。MongoDB 使用 $regex 操作符來設定匹配字串的正則表示式。

mongo中模糊查詢的綜合應用

前言 以下是一個mongo查詢的綜合應用，即介紹一個生產中實際應用的模糊查詢，當然其實也很簡單，主要用到mongo中的模糊查詢和$or查詢，以及並的關係，下面是一個mongo中的一條記錄

Golang Mongodb模糊查詢的使用示例

前言 在日常使用的Mongodb中，有一項功能叫做模糊查詢（使用正則匹配），例如:

Python操作mongodb資料庫進行模糊查詢操作示例

本文例項講述了Python操作mongodb資料庫進行模糊查詢操作。分享給大家供大家參考，具體如下：

oracle sql語言模糊查詢--萬用字元like的使用教程詳解

oracle在Where子句中，可以對datetime、char、varchar欄位型別的列用Like子句配合萬用字元選取那些“很像...”的資料記錄，以下是可使用的萬用字元：

mybatis 模糊查詢的實現方法

mybatis 模糊查詢的實現方法 mybatis的逆向助手確實好用，可以省去很多編寫常規sql語句的時間，但是它沒辦法自動生成模糊查詢語句，但開發中模糊查詢是必不可少的，所以，需要手動對mapper編寫模糊查詢功能。

mysql中使用instr進行模糊查詢方法介紹

在mysql中使用內部函式instr，可代替傳統的like方式查詢，並且速度更快。 instr 函式，第一個引數是欄位，第二個引數是要查詢的串，返回串的位置，第一個是1，如果沒找到就是0.

詳解基於MVC的資料查詢模組進行模糊查詢

完成一個簡單的基於MVC的資料查詢模組，要求能夠按照name進行模糊查詢。 Index.jsp：

資料庫學習之MySQL (七）——模糊查詢 萬用字元 like ‘between and’ in ‘is null’ 安全等於

MySQL學習專欄 正在持續更新中：） 文章目錄likebetween andinis null 和 is not null安全等於 <=>

MySQL模糊查詢語句整理集合

SQL模糊查詢語句 一般模糊語句語法如下： SELECT 欄位 FROM 表 WHERE 某欄位 Like 條件

Java如何使用elasticsearch進行模糊查詢

這篇文章主要介紹了Java如何使用elasticsearch進行模糊查詢,文中通過示例程式碼介紹的非常詳細，對大家的學習或者工作具有一定的參考學習價值,需要的朋友可以參考下

Mybatis模糊查詢及自動對映實現詳解

這篇文章主要介紹了Mybatis模糊查詢及自動對映實現詳解,文中通過示例程式碼介紹的非常詳細，對大家的學習或者工作具有一定的參考學習價值,需要的朋友可以參考下

Mysql模糊查詢之LIKE CONCAT(&#039;%&#039;,#{name},&#039;%&#039;)

眾所周知,SQL語句中的like模糊查詢select * from table where name like‘%張蛋%’, 實際開發中經常用

python中資料庫like模糊查詢方式

在Python中%是一個格式化字元，所以如果需要使用%則需要寫成%%。 將在Python中執行的sql語句改為：

Python Django2 model 查詢介紹(條件、範圍、模糊查詢)

條件查詢 範圍查詢 模糊查詢 條件查詢 all() 返回全部結果集 filter(**kwargs) 返回滿足引數定義的結果集 例如

20 LIKE 模糊查詢

前面介紹瞭如何查詢資料庫單條和多條資料，本小節介紹如何使用關鍵字模糊查詢符合要求的結果集，模糊查詢在實際業務中主要用於搜尋關鍵字查詢需要的資訊。

HBase根據rowKey模糊查詢

根據rowKey作為過濾條件查詢HBase，需要用到過濾器RowFilter。 RowFilter屬於比較過濾器的一種，比較過濾器繼承自CompareFilter。建立一個比較過濾器需要傳入兩個引數，分別是比較運算子和比較器。