重新格式化段落
大部分功能强大的文本编辑器都提供重新格式化段落的命令；供用户切分段落，使文本行数不要超出我们看到的屏幕范围。这样我们就引入了 fmt 命令，虽然一些 fmt 的实现有较多的选项可用，但其实只用:-s 仅切割较长的行，但不会将短行结合成较长的行，而 -w n 则设置输出行宽度为 n 个字符 (默认通常是 75 个)。
fmt 命令的语法:
fmt [option] [file-list]
fmt 通过将所有非空白的长度设置为几乎相同，来进行简单的文本格式化。
-s截断长行，但不合并-t除每个段落的第 1 行外都缩进-u改变格式化，使字之间出现一个空格，句子之间出现两个空格-w n将输出的行宽改为 n 个字符。不带该选项时，fmt 输出的行宽度为 75 个字符
例如，我有一个文件 demo，内容为:
A long time ago, there was a huge apple tree.A little boy loved to come and play around it every day. He climbed to the tree top, ate the apples, took a nap under the shadow… He loved the tree and the tree loved to play with him.
使用命令 fmt -s demo，输出为:
A long time ago, there was a huge apple tree.A little boy lovedto come and play around it every day. He climbed to the tree top, atethe apples, took a nap under the shadow… He loved the tree and thetree loved to play with him.
该命令的含义是节段 2 长行。
使用 fmt -t demo 命令的意思是说排除首行的缩进，结果为:
A long time ago, there was a huge apple tree. A little boy loved to come and play around it every day. He climbed to the tree top, ate the apples, took a nap under the shadow… He loved the tree and the tree loved to play with him.
使用 fmt -u demo 命令的意思是说格式化单词和句子的间隔。输出为:
A long time ago, there was a huge apple tree. A little boy loved to comeand play around it every day. He climbed to the tree top, ate the apples,took a nap under the shadow… He loved the tree and the tree loved toplay with him.
显然 A little boy 前面的多个空格变成了两个。
使用命令 fmt -w 40 demo 意思是说指定行的宽度，这里的行宽为 40 个字符。所以输出为:
A long time ago, there was a hugeapple tree. A little boyloved to come and play around itevery day. He climbed to the tree top,ate the apples, took a nap under theshadow… He loved the tree and thetree loved to play with him.
仅作切割的选项： -s，在你想将长的行绕回，短的行保持不动时很好用，这么做也能使结果与原始版本间的差异达到最小，例如:
fmt -s -w 10 《《 EOFone two three four fivesixseveneight输出为:one twothreefour fivesixseveneight
fmt 的小案例:
下面以拼音字典为例：
字典文件：/usr/dict/words 或者 /usr/share/dict/words。
sed -n -e 9991,10010p /usr/share/dict/words | fmtsed -n -e 9991,10010p /usr/share/dict/words | fmt -w 30
观察上面两行命令的输出。
复习一下 sed 命令:
sed 是一个很好的文件处理工具，本身是一个管道命令，主要是以行为单位进行处理，可以将数据行进行替换、删除、新增、选取等特定工作。假设我们有一个文件 demo
sed '1d' demo 删除第一行sed -n '1p' 显示第一行sed -n '/root/p' demo 查询包括关键字 root 所在所有行sed '1a hahaha' demo 第一行后增加字符串 hahahased '1,3a hahaha' demo 第一行到第三行后增加字符串 hahahased '1c hihihi' demo 第一行代替为 hihihised '1,2c hihihi' demo 第一行和第二行替换为 hihihi
替换一行中的某部分
格式：sed 's/ 要替换的字符串 / 新的字符串 /g' （要替换的字符串可以用正则表达式），案例:sed 's/root/hahaha/g' 替换 root 为 hahahased 's/root//g' 删除 rootsed -i '$a bye' ab #在文件 ab 中最后一行直接输入 "bye"
原文来源：http://wiki.jikexueyuan.com/project/learn-shell/twenty-one-day.html