生成候选词不大于5的缩写词库
程序是用SAS写的，本来也打算用python写的，碰到困难，就用SAS了（SAS也困难重重，写了一天，才搞定，╮(╯_╰)╭）
首先获取数据，本程序利用了宏变量，宏程序；宏变量写在程序开头，方便修改
libname aa "d:\english_word";
/*是否合格的定义：大于候选个数限制且长度大于已输入单词，的为不合格*/
%let wait_max=5;/*候选个数限制*/
%let infile="d:\word.txt";
%let outfile="d:\output.txt";
/*读取txt*/
data aa.allword_temp;
infile &infile.;
attrib word length=$20.;
input word $;
/*input word $20.;*/
/*是错误写法，$20.指连续的20位，效果是每隔一行读一次，语法错误会导致各种奇怪的错误*/
run;
/*去重*/
proc sql noprint;
create table aa.allword as
select distinct word from aa.allword_temp;
quit;
;
接下来的工作就麻烦了。
/*0-3位缩写不合格的*/
proc sql noprint;
create table aa.temp2 as
select distinct word,substr(word,1,3) as suox from aa.allword
group by suox
having count(word)>&wait_max. and length(word)>3;
/*0-3位缩写合格的*/
create table aa.yes2 as
select distinct word,substr(word,1,3) as suox from aa.allword
group by suox
having count(word)<= &wait_max. or length(word)<=3;
quit;
;
测试一下，缩写成3位,检查，能正确运行之后，就可以写宏程序（弄一个循环）
我想弄一个sql里面的循环的，发现sql里面没有循环结构，反正我搞不懂
/*3位缩写，不合格的*/
proc sql noprint;
create table aa.temp3 as
select word,substr(word,1,3) as suox from aa.temp2
group by suox
having count(word)>&wait_max. and length(word)>3
;
/*3位缩写，合格的*/
/*不宜从allword里选,而是从上一步不合格（0-3位缩写)的数中选*/
create table aa.yes3 as
select word,substr(word,1,3) as suox from aa.temp2
group by suox
having count(word)<=&wait_max. or length(word)<=3
;
quit;
下面就是完整代码链接：
