我有一个这种格式的词频列表:
3 yaz
1 yazlik
5 zemin
3 zemine
1 zeminde
2 zeminler
zemine 和 zeminde 是不同的字符串,但具有相同的根 zemin
我想像这样合并列表:
4 yaz
11 zemin
我如何使用 bash 或 python 做到这一点?
请您参考如下方法:
bash (4.0+) 的解决方案甚至可以处理未排序的列表:
$ cat script.sh
#!/bin/bash
declare -A roots # declare roots as an associative array (bash 4.0+)
while read n word; do
unset shortest longest
# check if the element (or its root) is already registered
for root in "${!roots[@]}"; do
if [[ "$root" =~ ^$word ]]; then
shortest=$word
longest=$root
elif [[ "$word" =~ ^$root ]]; then
shortest=$root
longest=$word
fi
done
# if registered, check if it must be replaced for a shorter one (its root)
if [ "$longest" ] && [ "${roots[$longest]}" ]; then
tmp_n=${roots["$longest"]}
unset roots["$longest"]
roots["$shortest"]=$tmp_n
fi
# register or update the element
let roots[${shortest:-$word}]+=$n
done < list
# print the result
for root in "${!roots[@]}"; do
echo "${roots[$root]} $root"
done
例子
$ cat list
3 yaz
1 yazlik
1 zeminde
5 zemin
3 zemine
2 zeminler
$ ./script.sh
4 yaz
11 zemin
