SLR 分析器 C# 代码实现 - 解析文法并构建分析表

{/n //产生式结点类/n public class SLRNode/n {/n public string Left;/n public string Right;/n public SLRNode(string Left, string Right)/n {/n this.Left = Left;/n this.Right = Right;/n }/n }/n //项目集类/n public class SLRitemsets/n {/n public List Container/n = new List(100);/n //记录项目在项目集合中的序号/n }/n/n //DFA结点/n public struct DFA/n {/n public int from;/n public char symbol;/n public int to;/n public DFA(int from, char symbol, int to)/n {/n this.from = from;/n this.symbol = symbol;/n this.to = to;/n }/n }/n/n //分析表 结点/n public class Table/n {/n public bool error;//是否为ERROR/n public char type;//结点类型/n public int id;//数值/n public Table()/n {/n this.error = true;/n }/n public Table(char type, int id)/n {/n this.type = type;/n this.id = id;/n this.error = false;/n }/n }/n/n //分析句子/n public class Analyze/n {/n public List stack_state = new List(100);//记录状态栈/n public List stack_symbol = new List(100);//记录符号栈/n public List Input_str = new List(100);//记录输入串/n public List Tran_pro = new List(100);//记录所用产生式/n }/n/n public DFA[] dfa = new DFA[100];/n public int Pindex = 0; //dfa数组指针/n public Table[][] SLRAna;//分析表/n public Analyze Jz;/n public bool Success = false;/n public List SLRproNum = new List(50);//产生式 列表/n public List SLRobjNum = new List(50);//项目 列表/n public List proitemset = new List(100);//项目集合/n public List Gy_obj = new List(50);//归约项目序号集合/n public List Gy_itemset = new List(50);//含有归约项目的集合的序号 的集合/n public List Nchar = new List(50);//非终结符集合/n public List Echar = new List(50);//终结符集合/n/n Dictionary<char, HashSet> follow = new Dictionary<char, HashSet>();//非终结符的Follow集/n/n public string RStr = /'/';/n public string RStr_obitemset = /'/';//输出返回/n public string RStr_DFA = /'/';/n public string RStr_ANA = /'/';/n/n/n public void Buildprod(string str)/n {/n/n SLRNode SLr;/n int i = 0;/n string left = /'/';/n string right = /'/';/n left += 'S/'';/n right += str[0];/n SLr = new SLRNode(left, right);//拓广文法开始/n SLRproNum.Add(SLr);/n while (i < str.Length)/n {/n left = right = /'/';//还原/n int j = i;/n while (i < str.Length && str[i] != '//r' && str[i] != '//n')//换行符‘//r//n’/n {/n if (str[i] == ' ')/n {/n i++;/n continue;/n }/n if (str[i] == '|') // 遇到'|'可构造一条产生式/n {/n SLr = new SLRNode(left, right);/n SLRproNum.Add(SLr);/n right = /'/'; //产生式左边相同 右边重新积累/n i++; //跳过'|'/n continue;/n }/n if ((i - j) == 0)/n {/n if (!exist(Nchar, str[i]))//如果非终结符集合中不存在str[i],加入Nchar 产生式左边 只有非终结符 不必判断终结符/n Nchar.Add(str[i]);/n left += str[i++];/n }/n else if (i - j <= 2)/n i++;/n else/n {/n if (isFinalsymbol(str[i]) && !exist(Nchar, str[i]))//如果非终结符集合中不存在str[i],加入Nchar isfinalsymbol 非终结符返回T 终结符返回F/n Nchar.Add(str[i]);/n else if (!isFinalsymbol(str[i]) && !exist(Echar, str[i]))//产生式右边 需要判断终结符/n Echar.Add(str[i]);/n right += str[i++];/n }/n/n/n }//while/n/n i++;//跳过换行符/n if (left != /'/' && right != /'/')/n {/n SLr = new SLRNode(left, right);//构造每一行最后一个产生式，不存在'|'时就是该行产生式本身/n SLRproNum.Add(SLr);/n }/n }//while/n Echar.Add('#');/n/n //构造项目 对产生式集合LRproNum中的所有产生式都循环插'.'/n SLRNode Lobj;/n for (i = 0; i < SLRproNum.Count; i++)/n {/n left = /'/';/n right = /'/';/n for (int j = 0; j <= SLRproNum[i].Right.Length; j++)//j可以等于length 项目共length+1个/n {/n left = SLRproNum[i].Left;/n right = CreObj(SLRproNum[i].Right, j);//在第j个位置插入'.'/n if (j == SLRproNum[i].Right.Length && SLRobjNum.Count != 1)/n {/n //在产生式最后的位置插入. 即为归约项目 项目集中1号位置为接受项目/n Gy_obj.Add(SLRobjNum.Count);//归约项目在项目集中的序号 不用+1 本身就是从0开始的/n }/n Lobj = new SLRNode(left, right);/n SLRobjNum.Add(Lobj);/n left = /'/';//还原/n right = /'/';/n }/n }/n Creteitemsets();//项目集/n RStr_obitemset += /'//r//n项目集构建：//r//n/';/n for (int j = 0; j < proitemset.Count; j++)/n {/n RStr_obitemset += 'I' + j.ToString() + ':' + /'//r//n/';/n for (i = 0; i < proitemset[j].Container.Count; i++)/n {/n RStr_obitemset += SLRobjNum[proitemset[j].Container[i]].Left.ToString() + /'->/' + SLRobjNum[proitemset[j].Container[i]].Right.ToString() + /'//r//n/';/n }/n }/n //return RStr_obitemset;/n/n }/n/n/n/n/n //求项目集/n public void Creteitemsets()/n {/n List lr_item = new List(100);//记录项目的序号/n lr_item.Add(0);/n lr_item = Closure(lr_item);//构造初始项目集 求闭包/n/n SLRitemsets LR_C = new SLRitemsets();/n LR_C.Container = lr_item;//集合----项目集序号的集合/n proitemset.Add(LR_C);//集合的集合----存放项目集序号集合 的集合/n/n/n for (int i = 0; i < proitemset.Count; i++)//整体集合中 第i个项目集/n {/n proitemset[i].Container.Sort();//排序由小到大 后面用于判断是否存在的比较/n int[] flag = new int[proitemset[i].Container.Count];/n for (int fi = 0; fi < proitemset[i].Container.Count; fi++)//标志位，用来判断该序号是否已经构造/n {/n flag[fi] = 0;/n }/n/n for (int j = 0; j < proitemset[i].Container.Count; j++)//第i个项目集的第j个项目/n {/n if (flag[j] == 1)//如果已经访问过 就不再构造 找下一个项目/n continue;/n int index = proitemset[i].Container[j];/n for (int pi = 0; pi < SLRobjNum[index].Right.Length - 1; pi++)//length-1是避免匹配到.在最后的规约项目/n {/n if (SLRobjNum[index].Right[pi] == '.')/n {/n/n List lr2_club = new List(100);//记录项目的序号/n char symbol = SLRobjNum[index].Right[pi + 1];//记录.a转移状态a.的符号a/n lr2_club.Add((index + 1));//如果遇到.a形式的项目序号为index 那么项目a.的序号为index+1/n for (int m1 = j + 1; m1 < proitemset[i].Container.Count; m1++)/n {/n //在第i个项目集中找到了可以移动的.：.a 重新遍历第i个项目集j项目之后的 找到同样可以移动a的项目集/n int index2 = proitemset[i].Container[m1];/n for (int m2 = 0; m2 < SLRobjNum[index2].Right.Length - 1; m2++)/n {/n if (SLRobjNum[index2].Right[m2] == '.' && SLRobjNum[index2].Right[m2 + 1] == symbol)/n {/n flag[m1] = 1;//标记位置为1 已经访问 之后不再访问/n lr2_club.Add(index2 + 1);/n }/n }/n }/n lr2_club = Closure(lr2_club);//求闭包/n int value = isexist(lr2_club);/n if (value == -1)//-1表示不存在相同的/n {/n for (int m3 = 0; m3 < Gy_obj.Count; m3++)/n {/n if (isnexist(lr2_club, Gy_obj[m3]))/n {/n Gy_itemset.Add(proitemset.Count);/n }/n }/n SLRitemsets LR_C2 = new SLRitemsets();/n dfa[Pindex++] = new DFA(i, symbol, proitemset.Count);//count不用加1 本身从0开始/n LR_C2.Container = lr2_club;/n proitemset.Add(LR_C2);/n }/n else/n {/n dfa[Pindex++] = new DFA(i, symbol, value);/n }/n break;/n }/n }/n }//end-forj/n }//end-fori/n/n }//end-Cre_club/n/n public void SLRAnaly()/n {/n Table tnode = new Table();/n/n SLRAna = new Table[proitemset.Count][];/n for (int i = 0; i < proitemset.Count; i++)/n SLRAna[i] = new Table[Echar.Count + Nchar.Count];/n/n for (int i = 0; i < proitemset.Count; i++)//初始化 赋予ERROR属性/n for (int j = 0; j < Echar.Count + Nchar.Count; j++)//为终结符加r状态 /n SLRAna[i][j] = tnode;/n/n tnode = new Table('A', 0);/n SLRAna[1][FindID(Echar, '#')] = tnode;//项目集1必定是接受项目 构建[1][#]:acc的情况 先直接赋值好 dfa里没有/n/n for (int i = 0; i < Gy_itemset.Count; i++)/n {/n SLRNode item = SLRobjNum[proitemset[Gy_itemset[i]].Container[0]];/n char left = item.Left[0];/n List follow = GetFollow(left);/n foreach (char c in follow)/n {/n int CID = FindID(Echar, c);/n SLRAna[Gy_itemset[i]][CID] = new Table('r', Find_pro(item)); /n }/n /n }/n/n for (int i = 0; i < Pindex; i++)/n {/n if (isFinalsymbol(dfa[i].symbol))//symbol为非终结符 添加状态N/n {/n int CID = FindID(Nchar, dfa[i].symbol);/n SLRAna[dfa[i].from][CID + Echar.Count] = new Table('N', dfa[i].to);/n }/n else //不是归约项目 添加状态S/n {/n int CID = FindID(Echar, dfa[i].symbol);/n SLRAna[dfa[i].from][CID] = new Table('S', dfa[i].to);/n }/n }/n }/n/n public void BuildFollow()/n {/n foreach (char c in Nchar)/n {/n follow.Add(c, new HashSet());/n }/n/n follow['S'].Add('#'); // 开始符号的 Follow 集中加入 #/n/n bool changed = true;/n while (changed)/n {/n changed = false;/n foreach (SLRNode node in SLRproNum)/n {/n string right = node.Right;/n HashSet set = follow[node.Left[0]];/n for (int i = right.Length - 1; i >= 0; i--)/n {/n char c = right[i];/n if (isFinalsymbol(c))/n {/n set.Clear();/n set.Add(c);/n continue;/n }/n HashSet follow_c = follow[c];/n if (i == right.Length - 1)/n {/n if (follow_c.Except(set).Count() > 0)/n {/n set.UnionWith(follow_c);/n changed = true;/n }/n }/n else/n {/n string beta = right.Substring(i + 1);/n HashSet first_beta = First(beta);/n if (first_beta.Contains('#'))/n {/n first_beta.ExceptWith(new HashSet { '#' });/n if (follow_c.Except(set).Count() > 0 || first_beta.Except(set).Count() > 0)/n {/n set.UnionWith(follow_c);/n set.UnionWith(first_beta);/n changed = true;/n }/n }/n else/n {/n if (follow_c.Except(set).Count() > 0 || first_beta.Except(set).Count() > 0)/n {/n set.UnionWith(first_beta);/n changed = true;/n }/n }/n }/n }/n }/n }/n }/n/n public HashSet First(string str)/n {/n HashSet set = new HashSet();/n if (str.Length == 0)/n {/n set.Add('#');/n return set;/n }/n if (isFinalsymbol(str[0]))/n {/n set.Add(str[0]);/n return set;/n }/n HashSet first = First(str.Substring(1));/n if (first.Contains('#'))/n {/n set.UnionWith(first);/n set.ExceptWith(new HashSet { '#' });/n set.UnionWith(First(str.Substring(1)));/n }/n else/n {/n set.UnionWith(first);/n }/n return set;/n }/n/n public bool isFinalsymbol(char c)/n {/n return (c >= 'a' && c <= 'z') || (c >= '0' && c <= '9') || c == '+' || c == '-' || c == '*' || c == '/' || c == '(' || c == ')' || c == '=';/n }/n/