我需要编写用于MySQL 5.x的一个聚合扩展功能（C实现）我已经走遍了文档（包括浏览SQL / udf_example.c），但我没有发现任何东西，是短暂的，给点意见，并表明我正是我需要做的。
I need to write an aggregate extension function (implemented in C) for mySQL 5.x. I have scoured the documentation (including browsing sql/udf_example.c) but I do not find anything that is brief, to the point and shows me just what I need to do.
这就是问题所在：
我有一个C函数，它接受这些美孚结构的数组，执行一个操作，阵列上，并返回一个double。
I have a C function that takes an array of these Foo structs, performs an operation, on the array, and returns a double.
结构FooBar的{ 字符*日期; 双周岁; 双怀特; 双倍工资; INT eye_color;};
struct FooBar { char * date; double age; double wight; double salary; int eye_color; };
/ *处理功能的 /双processFooBars（结构FooBar的的foobars，常量为size_t大小）;
/* Processing function / double processFooBars(struct FooBar foobars, const size_t size);
/ * MySQL表* /CREATE TABLE foo_bar这样的名称（the_date DATE，双岁，双重量，双薪，INT eye_color};
/* MySQL table */ CREATE TABLE foo_bar( the_date DATE, double age, double weight, double salary, int eye_color};
我希望能够创造这样一个聚合函数：（我也许使用PostgreSQL语法）
I want to be able to create an aggregate function thus: (I maybe using PostgreSQL syntax)
CREATE凝聚函数proc_foobar RETURNS REAL SONAME myshlib.so ALIAS my_wrapper_func
CREATE AGGREGATE FUNCTION proc_foobar RETURNS REAL soname myshlib.so ALIAS my_wrapper_func
然后我就可以在MySQL查询使用这样的：
I can then use it in a MySQL Query thus:
选择proc_foobar（）作为喜爱程度从foo_bar这样WHERE the_date'1-JAN-09和1日 - 12月09'之间。
SELECT proc_foobar() as likeability FROM foo_bar WHERE the_date BETWEEN '1-Jan-09' and '1-Dec-09'
这是什么查询应该那么做将是从表foo_bar这样获取所有符合条件的记录，将它们传递到周围processFooBar我的包装功能，然后将提取FooBar的结构从收到的记录，然后将它们传递给C函数该做的工作并返回值。
What this query should then do would be to fetch all the the matching records from the table foo_bar, pass them to my wrapper function around processFooBar, which will then extract FooBar structs from the records received and then pass them to the C function that does the work and returns the value.
它的简单使用（伪）code解释：
Its simpler to explain using (pseudo)code:
#ifdefined __cplusplus extern "C" { #endif /* this is the wrapper function that mySQL calls and passes the records to */ double my_wrapper_func(/*Matching rows sent by mySQL + other info .. ?*/) { /* create FooBar Array from received record */ struct FooBar ** the_array = ExtractArrayFromRowset(/*some params*/); double result = processFooBar(the_array, ARRAY_SIZE_MACRO(*the_array)); /* free resources */ FreeFooBarArray(the_array); RETURN_DOUBLE(result); /* or similar macro to return a double in MySQL */ } #ifdefined __cplusplus }; #endif任何人都可以提供一个小片段（或直接我一个片断），显示我怎么可以写my_wrapper_func - 更重要的一点还是如上所述，作为我如何实现写集合函数所需的功能在C / C ++的扩展功能。
Could anyone provide a little snippet (or direct me to a snippet) that shows me how I can write the my_wrapper_func - or more to the point how I can implement the required functionality of writing an aggregate function as described above, as an extension function in C/C++.
推荐答案不回答你的问题，但在MySQL UDF文章是pretty好：
Doesn't answer your question but article on MySQL udf is pretty good:
HTTP：//www.$c$cproject/ KB /数据库/ MySQL的％5FUDFs.aspx
更多推荐
MySQL的总UDF（用户定义函数）用C
发布评论