File: sed.info,  Node: Multiline techniques,  Next: Branching and flow control,  Prev: Hold and Pattern Buffers,  Up: advanced sed

6.3 Multiline techniques - using D,G,H,N,P to process multiple lines
====================================================================

Multiple lines can be processed as one buffer using the
‘D’,‘G’,‘H’,‘N’,‘P’.  They are similar to their lowercase counterparts
(‘d’,‘g’, ‘h’,‘n’,‘p’), except that these commands append or subtract
data while respecting embedded newlines - allowing adding and removing
lines from the pattern and hold spaces.

   They operate as follows:
‘D’
     _deletes_ line from the pattern space until the first newline, and
     restarts the cycle.

‘G’
     _appends_ line from the hold space to the pattern space, with a
     newline before it.

‘H’
     _appends_ line from the pattern space to the hold space, with a
     newline before it.

‘N’
     _appends_ line from the input file to the pattern space.

‘P’
     _prints_ line from the pattern space until the first newline.

   The following example illustrates the operation of ‘N’ and ‘D’
commands:

     $ seq 6 | sed -n 'N;l;D'
     1\n2$
     2\n3$
     3\n4$
     4\n5$
     5\n6$

  1. ‘sed’ starts by reading the first line into the pattern space (i.e.
     ‘1’).
  2. At the beginning of every cycle, the ‘N’ command appends a newline
     and the next line to the pattern space (i.e.  ‘1’, ‘\n’, ‘2’ in the
     first cycle).
  3. The ‘l’ command prints the content of the pattern space
     unambiguously.
  4. The ‘D’ command then removes the content of pattern space up to the
     first newline (leaving ‘2’ at the end of the first cycle).
  5. At the next cycle the ‘N’ command appends a newline and the next
     input line to the pattern space (e.g.  ‘2’, ‘\n’, ‘3’).

   A common technique to process blocks of text such as paragraphs
(instead of line-by-line) is using the following construct:

     sed '/./{H;$!d} ; x ; s/REGEXP/REPLACEMENT/'

  1. The first expression, ‘/./{H;$!d}’ operates on all non-empty lines,
     and adds the current line (in the pattern space) to the hold space.
     On all lines except the last, the pattern space is deleted and the
     cycle is restarted.

  2. The other expressions ‘x’ and ‘s’ are executed only on empty lines
     (i.e.  paragraph separators).  The ‘x’ command fetches the
     accumulated lines from the hold space back to the pattern space.
     The ‘s///’ command then operates on all the text in the paragraph
     (including the embedded newlines).

   The following example demonstrates this technique:
     $ cat input.txt
     a a a aa aaa
     aaaa aaaa aa
     aaaa aaa aaa

     bbbb bbb bbb
     bb bb bbb bb
     bbbbbbbb bbb

     ccc ccc cccc
     cccc ccccc c
     cc cc cc cc

     $ sed '/./{H;$!d} ; x ; s/^/\nSTART-->/ ; s/$/\n<--END/' input.txt

     START-->
     a a a aa aaa
     aaaa aaaa aa
     aaaa aaa aaa
     <--END

     START-->
     bbbb bbb bbb
     bb bb bbb bb
     bbbbbbbb bbb
     <--END

     START-->
     ccc ccc cccc
     cccc ccccc c
     cc cc cc cc
     <--END

   For more annotated examples, *note Text search across multiple
lines:: and *note Line length adjustment::.