Question

我希望逐行阅读connectedclients.now，在阅读每个内容时执行某些操作，并在完成工作时删除该行。

实际尝试过：

ClientWork.sh：

unset n
while read -r user work codename; do
  echo $user $work $codename
  : $[n++]
done <connectedclients.now
sed "1 $n d" connectedclients.now

来自stackexchange

的原始代码

将sed: -e expression #1, char 5: unexpected，＆＃39;`作为一个问题。有什么想法来解决它吗？

Answer 1

你的算术错误：$[n++]，而是尝试：

((n++))

检查http://mywiki.wooledge.org/ArithmeticExpression

如果你想逐行删除：

n=0
while read -r user work codename; do
    echo "$user $work $codename"
    ((n++))
    sed -i "$n d" connectedclients.now
done < connectedclients.now

Answer 2

如果你真的需要在处理每一行时删除行，我首先阅读整个文件：

file=connectedclients.now
echo "before processing, $(wc -l < "$file") lines"

mapfile -t all_lines < "$file"
for line in "${all_lines[@]}"; do
    read -r user work codename <<<"$line"
    echo $user $work $codename
    sed -i 1d "$file"
done

echo "after processing, $(wc -l < "$file") lines"

逐行读取文件并删除后删除

2 个答案: