Home >System Tutorial >LINUX >Understanding and Implementation of Finite State Machine FSM in Linux Programming
Finite State Machine (FSM for short) refers to a mathematical model consisting of a limited number of states and behaviors such as transitions and actions between these states. It has been widely used in the computer field. FSM is an efficient programming method used to implement the processing logic of a program within a logical unit. Especially in server programming, by performing corresponding processing based on different states or message types, the logic of the program can be made clearer and easier to understand. .
So, where are finite state machines usually used?
It can be applied to tokenizers (tokenizers) that process programming languages or natural languages, and implement grammar parsing and analysis through bottom-up grammar parsers (parser). In various communication protocols, the sender and receiver Processing messages by passing data between them, in game artificial intelligence, etc.
For the implementation of finite state machines, there are generally the following methods. I will introduce their advantages and disadvantages one by one.
Using if/else if statements to implement finite state machines is the simplest and easiest to understand method. Just use a large number of if/else if statements to determine the current status and perform corresponding logical processing.
The following is a simple state machine example. We use a large number of if/else if statements to implement it, perform corresponding operations according to different states, and implement state transitions.
enum { GET_UP, GO_TO_SCHOOL, HAVE_LUNCH, GO_HOME, DO_HOMEWORK, SLEEP, }; int main() { int state = GET_UP; //小明的一天 while ( 1 ) { if (state == GET_UP) { GetUp (); //具体调用的函数 state = GO_TO_SCHOOL; //状态的转移 } else if (state == GO_TO_SCHOOL) { Go2School (); state = HAVE_LUNCH; } else if (state == HAVE_LUNCH) { HaveLunch (); } ... else if (state == SLEEP) { Go2Bed (); state = GET_UP; } } return 0 ; }
After reading the above example, what do you think? Do you feel that although the program is simple and easy to understand, it uses a large number of if judgment statements, which makes the code very low-end and the code is bloated. This state machine has only a few states, and the code expansion is not obvious. However, if there are dozens of states that we need to process, the code of this state machine will be difficult to read.
The structure of FSM implemented using switch statements has become clearer, and its shortcomings are also obvious: although this design method is simple and processed through a lot of judgments, it is suitable for small-scale state switching processes, but if the scale Expansion is difficult to expand and maintain.
int state = GET_UP; //小明的一天 while ( 1 ) { switch (state) { case GET_UP: GetUp (); //具体调用的函数 state = GO_TO_SCHOOL; //状态的转移 break ; case GO_TO_SCHOOL: Go2School (); state = HAVE_LUNCH; break ; case HAVE_LUNCH: HaveLunch (); state = GO_HOME; break ; ... default : break ; } } return 0 ; }
The idea of using function pointers to implement FSM: establish the corresponding state table and action query table, locate the corresponding action processing function according to the state table, event, and action table, and then switch the state after the execution is completed.
当然使用函数指针实现的 FSM 的过程还是比较费时费力,但是这一切都是值得的,因为当你的程序规模大时候,基于这种表结构的状态机,维护程序起来也是得心应手。
下面给出一个使用函数指针实现的 FSM 的框架:
我们还是以 “小明的一天” 为例设计出该 FSM。
先给出该 FSM 的状态转移图:
下面讲解关键部分代码实现
首先我们定义出小明一天的活动状态:
//比如我们定义了小明一天的状态如下
enum { GET_UP, GO_TO_SCHOOL, HAVE_LUNCH, DO_HOMEWORK, SLEEP, };
我们也定义出会发生的事件
{ EVENT1 = 1 , EVENT2, EVENT3, };
定义状态表的数据结构
typedef struct FsmTable_s { int event ; //事件 int CurState ; //当前状态 void (*eventActFun)(); //函数指针 int NextState ; //下一个状态 } FsmTable_t ;
接下来定义出最重要 FSM 的状态表,我们整个 FSM 就是根据这个定义好的表来运转的。
FsmTable_t XiaoMingTable [] = { //{到来的事件,当前的状态,将要要执行的函数,下一个状态} { EVENT1, SLEEP, GetUp , GET_UP }, { EVENT2, GET_UP, Go2School , GO_TO_SCHOOL }, { EVENT3, GO_TO_SCHOOL, HaveLunch , HAVE_LUNCH }, { EVENT1, HAVE_LUNCH, DoHomework , DO_HOMEWORK }, { EVENT2, DO_HOMEWORK, Go2Bed , SLEEP }, //add your codes here };
状态机的注册、状态转移、事件处理的动作实现
/*状态机注册*/ void FSM_Regist( FSM_t * pFsm, FsmTable_t * pTable) { pFsm-> FsmTable = pTable; } /*状态迁移*/ void FSM_StateTransfer( FSM_t * pFsm, int state) { pFsm->curState = state; } /*事件处理*/ void FSM_EventHandle( FSM_t * pFsm, int event ) { FsmTable_t * pActTable = pFsm-> FsmTable ; void (*eventActFun)() = NULL; //函数指针初始化为空 int NextState ; int CurState = pFsm->curState; int flag = 0 ; //标识是否满足条件 int i; /*获取当前动作函数*/ for (i = 0 ; iif ( event == pActTable[i]. event && CurState == pActTable[i]. CurState ) { flag = 1 ; eventActFun = pActTable[i].eventActFun; NextState = pActTable[i]. NextState ; break ; } } if (flag) //如果满足条件了 { /*动作执行*/ if (eventActFun) { eventActFun(); } //跳转到下一个状态 FSM_StateTransfer(pFsm, NextState ); } else { // do nothing } }
主函数我们这样写,然后观察状态机的运转情况。
int main() { FSM_t fsm; InitFsm (&fsm); int event = EVENT1; //小明的一天,周而复始的一天又一天,进行着相同的活动 while ( 1 ) { printf( "event %d is coming...\n" , event ); FSM_EventHandle(&fsm, event ); printf( "fsm current state %d\n" , fsm.curState); test(& event ); sleep( 1 ); //休眠1秒,方便观察 } return 0 ; }
看一看该状态机跑起来的状态转移情况:
上面的图可以看出,当且仅当在指定的状态下来了指定的事件才会发生函数的执行以及状态的转移,否则不会发生状态的跳转。这种机制使得这个状态机不停地自动运转,有条不絮地完成任务。
与前两种方法相比,使用函数指针实现 FSM 能很好用于大规模的切换流程,只要我们实现搭好了 FSM 框架,以后进行扩展就很简单了(只要在状态表里加一行来写入新的状态处理就可以了)。
The above is the detailed content of Understanding and Implementation of Finite State Machine FSM in Linux Programming. For more information, please follow other related articles on the PHP Chinese website!