先前三个小节的内容介绍了PostgreSQL如何构建解析表达式的步骤,本节以int4pl为例介绍PostgreSQL如何解析表达式. SQL样例脚本. ``` testdb=# select 1+id,c2 from t_expr where id < 3; ``` ### 一、数据结构 **FmgrInfo** 在函数通过fmgr调用前,该结构体持有系统目录(字典)信息,用于检索相关信息. 如果相同的函数将被调用多次,检索只需要完成一次即可,该结构体会缓存多次使用. ``` /* * This struct holds the system-catalog information that must be looked up * before a function can be called through fmgr. If the same function is * to be called multiple times, the lookup need be done only once and the * info struct saved for re-use. * 在函数通过fmgr调用前,该结构体持有系统目录(字典)信息,用于检索相关信息. * 如果相同的函数将被调用多次,检索只需要完成一次即可,该结构体会缓存多次使用. * * Note that fn_expr really is parse-time-determined information about the * arguments, rather than about the function itself. But it's convenient * to store it here rather than in FunctionCallInfoData, where it might more * logically belong. * 注意,fn_expr实际上是关于参数的解析时确定的信息,而不是函数自身. * 但fn_expr在这里存储而不是FunctionCallInfoData中存储,因为从逻辑上来说,它就应该属于那. * * fn_extra is available for use by the called function; all other fields * should be treated as read-only after the struct is created. * fn_extra可用于被调用函数的使用;所有其他字段应该在结构体创建后被处理为只读. */ typedef struct FmgrInfo { //指向函数或者将被调用的处理器 PGFunction fn_addr; /* pointer to function or handler to be called */ //函数的oid Oid fn_oid; /* OID of function (NOT of handler, if any) */ //输入参数的个数,0..FUNC_MAX_ARGS short fn_nargs; /* number of input args (0..FUNC_MAX_ARGS) */ //函数是否严格(strict),输入NULL,输出NULL bool fn_strict; /* function is "strict" (NULL in => NULL out) */ //函数是否返回集合 bool fn_retset; /* function returns a set */ //如track_functions > this,则收集统计信息 unsigned char fn_stats; /* collect stats if track_functions > this */ //handler使用的额外空间 void *fn_extra; /* extra space for use by handler */ //存储fn_extra的内存上下文 MemoryContext fn_mcxt; /* memory context to store fn_extra in */ //表达式解析树,或者为NULL fmNodePtr fn_expr; /* expression parse tree for call, or NULL */ } FmgrInfo; typedef struct Node *fmNodePtr; ``` **FunctionCallInfoData** 该结构体存储了实际传递给fmgr-called函数的参数 ``` /* * This struct is the data actually passed to an fmgr-called function. * 该结构体存储了实际传递给fmgr-called函数的参数 * * The called function is expected to set isnull, and possibly resultinfo or * fields in whatever resultinfo points to. It should not change any other * fields. (In particular, scribbling on the argument arrays is a bad idea, * since some callers assume they can re-call with the same arguments.) * 被调用的函数期望设置isnull以及可能的resultinfo或者resultinfo指向的域字段. * 不应该改变其他字段. * (特别的,在参数数组上乱写是个坏主意,因为某些调用者假定它们可以使用相同的参数重复调用) */ typedef struct FunctionCallInfoData { //指向该调用的检索信息 FmgrInfo *flinfo; /* ptr to lookup info used for this call */ //调用上下文 fmNodePtr context; /* pass info about context of call */ //传递或返回关于结果的特别信息 fmNodePtr resultinfo; /* pass or return extra info about result */ //函数的collation Oid fncollation; /* collation for function to use */ #define FIELDNO_FUNCTIONCALLINFODATA_ISNULL 4 //如结果为NULL,则必须设置为T bool isnull; /* function must set true if result is NULL */ //实际传递的参数个数 short nargs; /* # arguments actually passed */ #define FIELDNO_FUNCTIONCALLINFODATA_ARG 6 //传递给函数的参数 Datum arg[FUNC_MAX_ARGS]; /* Arguments passed to function */ #define FIELDNO_FUNCTIONCALLINFODATA_ARGNULL 7 //如arg[i]为NULL,则对应的值为T bool argnull[FUNC_MAX_ARGS]; /* T if arg[i] is actually NULL */ } FunctionCallInfoData; /* * All functions that can be called directly by fmgr must have this signature. * (Other functions can be called by using a handler that does have this * signature.) * 所有函数可以通过fmgr直接调用,但必须持有签名. * (其他函数可通过使用handler的方式调用,也有此签名) */ typedef struct FunctionCallInfoData *FunctionCallInfo; ``` ### 二、源码解读 **ExecInterpExpr** ExecInterpExpr中与表达式求值相关的代码片段如下: ``` EEO_CASE(EEOP_FUNCEXPR) { FunctionCallInfo fcinfo = op->d.func.fcinfo_data; Datum d; fcinfo->isnull = false; d = op->d.func.fn_addr(fcinfo); *op->resvalue = d; *op->resnull = fcinfo->isnull; EEO_NEXT(); } ``` 如为函数表达式,则从ExecInitFunc初始化的步骤信息中获取统一的调用参数fcinfo,然后通过函数指针(用于封装)调用实际的函数进行表达式求值.通过统一的参数,统一的返回值,做到了实现的统一,体现了面向对象OO中多态的思想,这再次说明了OO是思想,用过程性语言一样可以实现. **int4pl** SQL样例脚本相应的实现函数是int4pl,其实现代码如下: ``` Datum int4pl(PG_FUNCTION_ARGS) { int32 arg1 = PG_GETARG_INT32(0); int32 arg2 = PG_GETARG_INT32(1); int32 result; if (unlikely(pg_add_s32_overflow(arg1, arg2, &result))) ereport(ERROR, (errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE), errmsg("integer out of range"))); PG_RETURN_INT32(result); } /* * If a + b overflows, return true, otherwise store the result of a + b into * *result. The content of *result is implementation defined in case of * overflow. */ static inline bool pg_add_s32_overflow(int32 a, int32 b, int32 *result) { #if defined(HAVE__BUILTIN_OP_OVERFLOW) return __builtin_add_overflow(a, b, result); #else int64 res = (int64) a + (int64) b; if (res > PG_INT32_MAX || res < PG_INT32_MIN) { *result = 0x5EED; /* to avoid spurious warnings */ return true; } *result = (int32) res; return false; #endif } ``` 函数实现相对比较简单,两个数简单相加,如溢出则返回T,否则返回F. ### 三、跟踪分析 N/A ### 四、参考资料 [PG Source Code](https://doxygen.postgresql.org)