是时候在模板中使用JSONb datatype作为输入,而PostgreSQL查询作为模板系统... Mustache将是完美的,如果有一些针对PLpgSQL的Mustache实现(或适用于{{ 3}})......似乎没有人。
但是有C:如何使用/适应PLv8?
在像SELECT tplEngine_plv8(input_jsonb,tpl_mustashe) as text_result FROM t
这样的上下文中多次调用小胡子的最佳方式(性能)是什么?
测试:good source-code for Javascript-mustache
CREATE TABLE story (
id int NOT NULL PRIMARY KEY,
title text NOT NULL,
UNIQUE(title)
);
CREATE TABLE story_character (
id_story int REFERENCES story(id) NOT NULL,
name text NOT NULL,
UNIQUE(id_story,name)
);
INSERT INTO story (id,title) VALUES (1,'African jungle story'),
(2,'Story of India jungle');
INSERT INTO story_character(id_story,name) VALUES
(1,'Tarzan'), (1,'Jane'), (2,'Mowgli'), (2,'Baloo');
CREATE VIEW t AS
select id_story, jsonb_build_object('title',title,
'names', jsonb_agg( jsonb_build_object('name', name) )
) AS j
from story s INNER JOIN story_character c ON s.id=c.id_story
group by id_story,title;
所以使用VIEW我们有胡子模板的名称和标题,
SELECT tplEngine_plv8(
j,
'<br/>* <b>{{title}}</b> with: {{#names}} <i>{{name}}</i>; {{/names}}'
) as result
FROM t;
并将其与JSFiddle结果进行比较。
使用EXPLAIN ANALYZE
...也许在测试表中添加几百个随机值。并且,测试也调用策略:一次一个或通过数组。
CREATE FUNCTION mustache_engine(
p_input JSONB,
p_template TEXT
) RETURNS TEXT language plv8 as $$
// copy https://rawgit.com/janl/mustache.js/master/mustache.js
// and somethings more
$$;
或
CREATE FUNCTION mustache_engine(
p_inputs JSONB[], -- many inputs
p_templates TEXT -- one template
) RETURNS TEXT[] -- many resuts
language plv8 as $$ ... $$;
CREATE FUNCTION mustache_engine( -- many input-template pairs
p_inputs JSONB[],
p_templates TEXT[]
) RETURNS TEXT[] -- many resuts
language plv8 as $$ ... $$;
答案 0 :(得分:1)
create or replace function mustache(template text, view json, partials json) returns text
language plv8 as
<< copy https://rawgit.com/janl/mustache.js/master/mustache.js >>
var mustache=this.Mustache;
return mustache.render(template, view, partials)
$$