如何合并sed命令实现匹配行及后续指定行打印，及文件末尾同类操作？

Great question! Let's tackle both parts of your problem—combining sed commands to get the start of the file plus lines after a match, and the reverse (lines before a match to the end of the file).

Printing from the start of the file to a pattern, plus N lines after #

Yes, you absolutely can do this in a single sed command—no need to pipe two separate commands together. Let's use your example where we want to match dolor and include 2 lines after it.

First, let's assume your input file (example.txt) has one word per line like this:

lorem
ipsum
dolor
sit
amet
consectetur
adipiscing
elit

The goal is to print up to dolor plus the next 2 lines, resulting in:

lorem
ipsum
dolor
sit
amet

Here's a clean sed command to do this:

sed -n '1,/dolor/p; /dolor/{n;p;n;q}' example.txt

Let's break this down:

• -n: Tells sed only to print lines we explicitly specify (avoids duplicate output)
• 1,/dolor/p: Prints every line from the start of the file up to (and including) the line matching dolor
• /dolor/{n;p;n;q}: When we hit the dolor line:
  • n: Reads the next line into the pattern space
  • p: Prints that line
  • n: Reads the second line after dolor
  • p: Prints that line
  • q: Quits sed immediately so we don't print any lines beyond this point

If you need a more flexible version (say, for any number of lines after the match, not just 2), you can use a loop with labels to count lines dynamically:

sed '/dolor/{:loop; N; s/\n/ /g; /\n.*\n/!b loop; q}' example.txt

• :loop: Defines a reusable label
• N: Adds the next line to the pattern space
• /\n.*\n/!b loop: If we don't have 2 newline characters (meaning we haven't added 2 lines after the match yet), jump back to the loop label
• q: Quit once we've collected the match plus 2 lines

Doing the reverse: Printing from N lines before a pattern to the end of the file #

For the end-of-file equivalent—printing from a few lines before the match to the end of the file—we can use a similar approach, or leverage tac (which reverses the file) to reuse our earlier logic.

Using your example, if we want to print 2 lines before dolor to the end, the output would be:

ipsum
dolor
sit
amet
consectetur
adipiscing
elit

Here's how to do it with tac + sed:

tac example.txt | sed '/dolor/{:loop; N; 2ba; q}' | tac

• tac example.txt: Reverses the file so dolor appears near the top
• sed '/dolor/{:loop; N; 2ba; q}': Grabs the dolor line plus the next 2 lines (which are the original file's lines before dolor)
• tac: Reverses the result back to normal order

If you want a pure sed solution without tac, you can use the hold space to track the last N lines until you hit the match:

sed -e '1{N;N}' -e '/dolor/!{D;N;b}' example.txt

• 1{N;N}: Starts by reading the first 3 lines (the match line plus 2 before it)
• /dolor/!{D;N;b}: If the current lines don't include dolor, delete the oldest line (D), read a new line (N), and repeat the loop
• Once dolor is found, sed stops looping and prints all remaining lines normally

内容的提问来源于stack exchange，提问作者jmlemetayer