├── .classpath ├── .gitattributes ├── .gitignore ├── .project ├── README.md ├── bin ├── analyize_error │ └── Error.class └── parse │ ├── Exp.class │ ├── Exp_B.class │ ├── Exp_Operate.class │ ├── Exp_calcul.class │ ├── Exp_comp.class │ ├── Exp_cond.class │ ├── Exp_func.class │ ├── Exp_funcname.class │ ├── Exp_ife.class │ ├── Exp_mark.class │ ├── Exp_type.class │ ├── Exp_whi.class │ ├── HaveToken.class │ ├── Main.class │ ├── Parse_Semantics.class │ ├── Parse_Token.class │ ├── Symbol_Table.class │ ├── Table.class │ ├── Token_List.class │ ├── Variable_Table.class │ ├── type_cal.class │ ├── type_com.class │ ├── type_op.class │ └── type_ret.class ├── src ├── analyize_error │ └── Error.java └── parse │ ├── Exps.java │ ├── HaveToken.java │ ├── Main.java │ ├── Parse_Semantics.java │ ├── Parse_Token.java │ ├── Symbol_Table.java │ ├── Token_List.java │ └── Variable_Table.java ├── test.t └── token.py /.classpath: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | -------------------------------------------------------------------------------- /.gitattributes: -------------------------------------------------------------------------------- 1 | # Auto detect text files and perform LF normalization 2 | * text=auto 3 | 4 | # Custom for Visual Studio 5 | *.cs diff=csharp 6 | 7 | # Standard to msysgit 8 | *.doc diff=astextplain 9 | *.DOC diff=astextplain 10 | *.docx diff=astextplain 11 | *.DOCX diff=astextplain 12 | *.dot diff=astextplain 13 | *.DOT diff=astextplain 14 | *.pdf diff=astextplain 15 | *.PDF diff=astextplain 16 | *.rtf diff=astextplain 17 | *.RTF diff=astextplain 18 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # Windows image file caches 2 | Thumbs.db 3 | ehthumbs.db 4 | 5 | # Folder config file 6 | Desktop.ini 7 | 8 | # Recycle Bin used on file shares 9 | $RECYCLE.BIN/ 10 | 11 | # Windows Installer files 12 | *.cab 13 | *.msi 14 | *.msm 15 | *.msp 16 | 17 | # Windows shortcuts 18 | *.lnk 19 | 20 | # ========================= 21 | # Operating System Files 22 | # ========================= 23 | 24 | # OSX 25 | # ========================= 26 | 27 | .DS_Store 28 | .AppleDouble 29 | .LSOverride 30 | 31 | # Thumbnails 32 | ._* 33 | 34 | # Files that might appear on external disk 35 | .Spotlight-V100 36 | .Trashes 37 | 38 | # Directories potentially created on remote AFP share 39 | .AppleDB 40 | .AppleDesktop 41 | Network Trash Folder 42 | Temporary Items 43 | .apdisk 44 | -------------------------------------------------------------------------------- /.project: -------------------------------------------------------------------------------- 1 | 2 | 3 | compile 4 | 5 | 6 | 7 | 8 | 9 | org.eclipse.jdt.core.javabuilder 10 | 11 | 12 | 13 | 14 | 15 | org.eclipse.jdt.core.javanature 16 | 17 | 18 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # compiler_java_python 2 | 简易编译器实现，最终翻译成伪汇编代码 3 | 4 | 项目太老，原先在eclipse上实现的，现在已经迁移到AndroidStudio，把分支切换到dev_android打开即可 5 | 6 | ## 编译器主要编译t语言（暂且这么称呼） 7 | 8 | 9 | ## t语言关键字： 10 | - func 用于定义函数 11 | - main 主函数名，文件中必须包含主函数 12 | - while 循环关键字 13 | - if 条件分支关键字 14 | - else 必须与if 一起才能识别 15 | - num 数字类型 16 | - string 字符串类型 17 | - bool 布尔值类型 18 | 19 | ## t语言的比较符号： 20 | - > 21 | - < 22 | - >= 23 | - <= 24 | - == 25 | 26 | 示例： 27 | 28 | 29 | func main(){ 30 | 31 | num a,b; 32 | 33 | a=1; 34 | 35 | b=0; 36 | 37 | while(b<5){ 38 | 39 | num c,k; 40 | 41 | b=b+1; 42 | 43 | if(b<3){ 44 | 45 | c=b+b*a; 46 | 47 | }else{ 48 | 49 | c=b; 50 | 51 | } 52 | } 53 | 54 | num d,c; 55 | 56 | c=1; 57 | 58 | a=b\*a\*c+b*4; 59 | 60 | } 61 | 62 | 输出伪汇编代码（三地址码的四元式）格式为：<操作符，操作数1，操作数2，结果存放> 63 | 以上示例的输出结果： 64 | - DEFINE , num , null , a 65 | - DEFINE , num , null , b 66 | - SEND , 1 , null , a 67 | - SEND , 0 , null , b 68 | - LABEL , null , null , Label1 69 | - CJMP_L , b , 5 , Label2 70 | - DEFINE , num , null , c 71 | - DEFINE , num , null , k 72 | - ADD , b , 1 , $ 73 | - SEND , $ , null , b 74 | - CJMP_L , b , 3 , Label3 75 | - MUL , b , a , $ 76 | - ADD , b , $ , $ 77 | - SEND , $ , null , c 78 | - LABEL , null , null , Label3 79 | - SEND , b , null , c 80 | - JMP , null , null , Label1 81 | - LABEL , null , null , Label2 82 | - DEFINE , num , null , d 83 | - DEFINE , num , null , c 84 | - SEND , 1 , null , c 85 | - MUL , b , a , $ 86 | - MUL , $ , c , $ 87 | - MUL , b , 4 , $ 88 | - ADD , $ , $ , $ 89 | - SEND , $ , null , a 90 | 91 | 92 | 93 | 94 | ##个别操作符的含义： 95 | - DEFINE: 定义变量 96 | - SEND : 赋值 97 | - JMP: 无条件跳转 98 | - CJMP_E: 等于时执行下面的语句，否则跳转到某个label 99 | - CJMP_B : 大于时........ 100 | - CJMP_B_E : 大于等于的时候...... 101 | - LABEL: 定义一个标签（用于跳转） 102 | - $ : 用于存放左边操作数运算的结果，或者代表上面的运算结果，主要作用是把被分解的表达式连接起来 103 | - null 表示该操作数不进行操作，仅仅起占位的作用 104 | 105 | 106 | 107 | ## 补充说明： 108 | - 为了简化编译工作 109 | - t语言只支持一个mian(),函数，所以不分析函数之间的跳转 110 | - t语言只支持变量之间的加减乘除，自动识别运算优先级，不允许使用（） 111 | - t语言以只支持while这一种循环，而且不支持break跳出循环。 112 | 113 | 由于第一次编写编译器，所以很多工作可能不是很好，但是对于理解编译原理，还是有很大帮助的，有意向的同学欢迎交流.... 114 | -------------------------------------------------------------------------------- /bin/analyize_error/Error.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/analyize_error/Error.class -------------------------------------------------------------------------------- /bin/parse/Exp.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp.class -------------------------------------------------------------------------------- /bin/parse/Exp_B.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_B.class -------------------------------------------------------------------------------- /bin/parse/Exp_Operate.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_Operate.class -------------------------------------------------------------------------------- /bin/parse/Exp_calcul.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_calcul.class -------------------------------------------------------------------------------- /bin/parse/Exp_comp.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_comp.class -------------------------------------------------------------------------------- /bin/parse/Exp_cond.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_cond.class -------------------------------------------------------------------------------- /bin/parse/Exp_func.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_func.class -------------------------------------------------------------------------------- /bin/parse/Exp_funcname.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_funcname.class -------------------------------------------------------------------------------- /bin/parse/Exp_ife.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_ife.class -------------------------------------------------------------------------------- /bin/parse/Exp_mark.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_mark.class -------------------------------------------------------------------------------- /bin/parse/Exp_type.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_type.class -------------------------------------------------------------------------------- /bin/parse/Exp_whi.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Exp_whi.class -------------------------------------------------------------------------------- /bin/parse/HaveToken.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/HaveToken.class -------------------------------------------------------------------------------- /bin/parse/Main.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Main.class -------------------------------------------------------------------------------- /bin/parse/Parse_Semantics.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Parse_Semantics.class -------------------------------------------------------------------------------- /bin/parse/Parse_Token.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Parse_Token.class -------------------------------------------------------------------------------- /bin/parse/Symbol_Table.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Symbol_Table.class -------------------------------------------------------------------------------- /bin/parse/Table.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Table.class -------------------------------------------------------------------------------- /bin/parse/Token_List.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Token_List.class -------------------------------------------------------------------------------- /bin/parse/Variable_Table.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/Variable_Table.class -------------------------------------------------------------------------------- /bin/parse/type_cal.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/type_cal.class -------------------------------------------------------------------------------- /bin/parse/type_com.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/type_com.class -------------------------------------------------------------------------------- /bin/parse/type_op.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/type_op.class -------------------------------------------------------------------------------- /bin/parse/type_ret.class: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/bin/parse/type_ret.class -------------------------------------------------------------------------------- /src/analyize_error/Error.java: -------------------------------------------------------------------------------- 1 | package analyize_error; 2 | 3 | public class Error { 4 | public Error(){ 5 | 6 | } 7 | public void error_print(int id,String content,String line,String row){ 8 | String error_content=null; 9 | switch(id){ 10 | case 0:error_content="expression was in wrong place. ";break; 11 | case 1:error_content="expression was not wrote rightly. ";break; 12 | case 2:error_content="misss "+content+" .";break; 13 | case 3:error_content="variable "+content+" was not defined. ";break; 14 | case 4:error_content="the type was wrong (type only num,string,bool).";break; 15 | case 5: error_content="the variable"+content+" can't be used. ";break; 16 | case 6:error_content="the mark "+content+" was missing. ";break; 17 | case 7 :error_content="the key word "+content+" was wrong.";break; 18 | case 8: error_content="the type "+content+" can not match.";break; 19 | case 9 :error_content="the type "+content+" can not be calculatered. ";break; 20 | default:error_content="something wrong--";break; 21 | } 22 | print(error_content,line,row); 23 | 24 | } 25 | public void print(String content,String line,String row){ 26 | System.out.println(content+"in line:"+line+" row:"+row); 27 | } 28 | 29 | } 30 | -------------------------------------------------------------------------------- /src/parse/Exps.java: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/src/parse/Exps.java -------------------------------------------------------------------------------- /src/parse/HaveToken.java: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/src/parse/HaveToken.java -------------------------------------------------------------------------------- /src/parse/Main.java: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/src/parse/Main.java -------------------------------------------------------------------------------- /src/parse/Parse_Semantics.java: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/src/parse/Parse_Semantics.java -------------------------------------------------------------------------------- /src/parse/Parse_Token.java: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/src/parse/Parse_Token.java -------------------------------------------------------------------------------- /src/parse/Symbol_Table.java: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/src/parse/Symbol_Table.java -------------------------------------------------------------------------------- /src/parse/Token_List.java: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ladingwu/compiler_java_python/268442050d7080cf7af74107982a67fb7ab8b7fd/src/parse/Token_List.java -------------------------------------------------------------------------------- /src/parse/Variable_Table.java: -------------------------------------------------------------------------------- 1 | package parse; 2 | 3 | public class Variable_Table { 4 | 5 | } 6 | -------------------------------------------------------------------------------- /test.t: -------------------------------------------------------------------------------- 1 | func main(){ 2 | num a,b; 3 | a=1; 4 | b=0; 5 | while(b<5){ 6 | num c,k; 7 | b=b+1; 8 | if(b<3){ 9 | c=b+b*a; 10 | }else{ 11 | c=b; 12 | } 13 | } 14 | num d,c; 15 | c=1; 16 | a=b*a*c+b*4; 17 | 18 | } -------------------------------------------------------------------------------- /token.py: -------------------------------------------------------------------------------- 1 | import os,sys 2 | msg_table=['func','main','if','while', 3 | 'else','num','string','bool','(',')', 4 | ';',',','{','}','[',']','=','+','-','*' 5 | ,'/','%','&','|','>','<','==','>=','<=','"','"','id'] 6 | def read_token(file): 7 | with open(file,'r')as f: 8 | return f.readlines() 9 | 10 | def token_recogn(lines): 11 | msg_line=0 12 | #k='qwertyuioplkjhgfdsazxvcbnm_' 13 | #s=k.upper() 14 | q1='>=' 15 | q2='<=' 16 | q3='==' 17 | m=',;+=*/-(){}""><' 18 | wrong_mark='；，。；“”【】' 19 | for line in lines: 20 | words='' 21 | msg_line+=1 22 | msg_row=0 23 | str_i=0 24 | msg='' 25 | for word in line: 26 | if word ==' ' or word =='\t' or word=='\n': 27 | if(word=='\t'): 28 | msg_row+=4 29 | if words != '': 30 | deal_words(words, msg_line, msg_row) 31 | words='' 32 | elif word in m : 33 | if(msg_row+10: 37 | 38 | deal_words(words,msg_line,msg_row-1) 39 | deal_words(msg+word, msg_line,msg_row) 40 | msg='' 41 | else: 42 | deal_words(msg+word,msg_line,msg_row) 43 | msg='' 44 | words='' 45 | elif word in wrong_mark: 46 | if len(words)>0: 47 | 48 | deal_words(words,msg_line,msg_row-1) 49 | deal_words(word, msg_line,msg_row) 50 | 51 | else: 52 | deal_words(word,msg_line,msg_row) 53 | else: 54 | words=words+word 55 | msg_row+=1 56 | def check_id(s): 57 | m='qwertyuiopasdfghjklzxcvbnm_' 58 | n=m.upper() 59 | k='0987654321' 60 | flag=True 61 | #print(n) 62 | j=0 63 | for i in s: 64 | if j==0: 65 | j+=1 66 | if i in k: 67 | flag=False 68 | break 69 | 70 | if i not in m and i not in n and i not in k: 71 | 72 | flag=False 73 | break 74 | return flag 75 | def check_num(s): 76 | k="0987654321" 77 | flag=True 78 | for i in s: 79 | if i not in k: 80 | flag=False 81 | break 82 | return flag 83 | last_word='' #为了读取字符串 84 | count=0 85 | def deal_words(words,msg_line,msg_row): 86 | i=0 87 | # print(words) 88 | global msg_table,last_word,count 89 | length=len(msg_table) 90 | #print('-----------------------------------',words) 91 | if words=='"': 92 | count+=1 93 | if (count%2)!=0 and last_word=='"': 94 | last_word='' 95 | # 表示字符常量 96 | print('id_string'+':'+words+':'+str(msg_line)+':'+str(msg_row)) 97 | return 98 | 99 | for t in msg_table: 100 | if i==length-1: 101 | 102 | if check_id(words): 103 | print('id'+':'+words+':'+str(msg_line)+':'+str(msg_row)) 104 | break 105 | elif check_num(words): 106 | # 表示常数变量 107 | print('id_num'+':'+words+':'+str(msg_line)+':'+str(msg_row)) 108 | break 109 | else: 110 | error(words,msg_line,msg_row) 111 | break 112 | if words==t: 113 | if i<8: 114 | 115 | print(words+':'+words+':'+str(msg_line)+':'+str(msg_row)) 116 | elif i