我正在尝试将法语字符串（语言ID 1）合并到一列中。到目前为止，我能够在table1.title和table2.translated_topic中获得法语字符串，但我不确定如何连接它们。
Ver：Postgres 9.6.0
源表架构：
表1：knowledgebase_topics
id | title | language_id |
------------------------------------
64 | The Topic | 91 |
65 | The Topic 2 | 91 |
62 | Le fav sujet | 1 |
63 | Le fav sujet 2 | 1 |
61 | le bonjour | 1 |
表2：knowledgebase_topics_translations
id | translated_topic| knowledgebase_topic_id | language_id |
-------------------------------------------------------------
| Le sujet | 64 | 1 |
| Le sujet 2 | 65 | 1 |
| Fav The Topic | 62 | 91 |
| Fav The Topic 2 | 63 | 91 |
给出以下查询：
SELECT title, translated_topic, "kbt".language_id, "kbtt".language_id
FROM knowledgebase_topics as "kbt"
LEFT JOIN knowledgebase_topics_translations as "kbtt" on ("kbtt".knowledgebase_topic_id = "kbt".id)
INNER JOIN knowledgebase_topics_organizations as "kbto" on ("kbto".knowledgebase_topic_id = "kbt".id)
WHERE "kbto"."organization_id" = 1
AND to_tsvector("kbt".title) @@ to_tsquery('le')
OR to_tsvector("kbtt".translated_topic) @@ to_tsquery('le')
AND "kbt".language_id = 1
OR "kbtt".language_id = 1;
我得到以下结果：
title | translated_topic | language_id | language_id
----------------+------------------+-------------+-------------
The Topic | Le sujet | 91 | 1
The Topic 2 | Le sujet 2 | 91 | 1
Le fav sujet | Fav The Topic | 1 | 91
Le fav sujet 2 | Fav The Topic 2 | 1 | 91
le bonjour | | 1 |
所需结果： table1.title和table2.translated_topics已基于language_id == 1合并。两个表都有一个语言ID列。
title | language_id
----------------+--------------
Le sujet | 1
Le sujet 2 | 1
Le fav sujet | 1
Le fav sujet 2 | 1
le bonjour | 1
我该怎么做？
注意：我不只是想检查lang IDs = 1，例如
and "kbt".language_id = 1 AND (instead of OR) "kbtt".language_id = 1;
因为这导致2个语言ID 1的table 2缺少记录：
title | translated_topic | language_id | language_id
----------------+------------------+-------------+-------------
Le fav sujet | Fav The Topic | 1 | 91
Le fav sujet 2 | Fav The Topic 2 | 1 | 91
le bonjour | | 1 |
所以，我已经开始工作......但这是否具有效果？
SELECT title, "kbt".language_id
FROM knowledgebase_topics as "kbt"
INNER JOIN knowledgebase_topics_organizations as "kbto" on ("kbto".knowledgebase_topic_id = "kbt".id)
WHERE "kbto"."organization_id" = 1
AND to_tsvector("kbt".title) @@ to_tsquery('le')
AND "kbt".language_id = 1
UNION ALL
SELECT translated_topic, "kbtt".language_id
FROM knowledgebase_topics_translations as "kbtt"
INNER JOIN knowledgebase_topics_organizations as "kbto" on ("kbto".knowledgebase_topic_id = "kbtt".id)
WHERE "kbto"."organization_id" = 1
AND to_tsvector("kbtt".translated_topic) @@ to_tsquery('le')
AND "kbtt".language_id = 1;
提供输出：
title | language_id
----------------+-------------
le bonjour | 1
Le fav sujet | 1
Le fav sujet 2 | 1
Le sujet | 1
Le sujet 2 | 1
(5 rows)
答案 0 :(得分：2)
首先，观察我们如何用简洁的DDL最好地描述问题。最好在将来，你将学习如何写这样的问题..
CREATE TEMPORARY TABLE knowledgebase_topics AS
SELECT * FROM ( VALUES
(64,'The Topic',91),
(65,'The Topic 2',91),
(62,'Le fav sujet',1),
(63,'Le fav sujet 2',1),
(61,'le bonjour',1)
) AS t(knowledgebase_topic_id, title, language_id);
CREATE TEMPORARY TABLE knowledgebase_topics_translations AS
SELECT * FROM ( VALUES
('Le sujet' ,64,1 ),
('Le sujet 2' ,65,1 ),
('Fav The Topic' ,62,91 ),
('Fav The Topic 2',63,91 )
) AS t(translated_topic, knowledgebase_topic_id, language_id);
然后你只需要告诉我们你想要什么，我们就可以轻松地找到工作环境并回答你的问题。不需要英语！我们两个人都比较容易。
我们在UNION ALL中使用SELECT将其包装在language中，以便我们可以按ID排序，并轻松地在一个地方更改您要查找的SELECT title, language_id
FROM (
SELECT knowledgebase_topic_id, title, language_id
FROM knowledgebase_topics
UNION ALL
SELECT knowledgebase_topic_id, translated_topic, language_id
FROM knowledgebase_topics_translations
) AS t(id, title, language_id)
WHERE language_id = 1
ORDER BY id;
。
title │ language_id
────────────────┼─────────────
le bonjour │ 1
Le fav sujet │ 1
Le fav sujet 2 │ 1
Le sujet │ 1
Le sujet 2 │ 1
(5 rows)
MakeTheSprocFailWhenCalled();
try
{
ExecuteSomethingThatCallsASproc();
throw new Exception();
Assert.Fail("Should never have made it here");
}
catch(Exception e)
{
Assert.AreEqual(typeof(SqlException), e.GetType());
}
finally
{
UndoTheThingIDidInTheFirstLine();
}