mjhjmtu

I have a file whose content is

(bookmarks
 ("Cover"
 "#01.djvu" )
 ("Title page"
 "#all_24223_to_00243.cpc0002.djvu" )
 ("Preface"
 "#all_24223_to_00243.cpc0004.djvu" )
 ...

I want to change its content to be

(bookmarks
 ("Cover"
 "#2" )
 ("Title page"
 "#3" )
 ("Preface"
 "#5" )
...

by keeping the number just before .djvu, removing leading zeros, and adding one to it. I wonder how you would do that using awk?

Thanks.

That's more a job for perl:

perl -pe 's/"#K.*?(d+).djvu(?=")/$1+1/ge' <file

With variable:

INCR=1 perl -pe 's/"#K.*?(d+).djvu(?=")/$1+$ENVINCR/ge' <file

Or:

perl -spe 's/"#K.*?(d+).djvu(?=")/$1+$incr/ge' -- -incr=1 <file

Here's a GNU awk solution:

awk '/^ *(/print!/^ *(/split($1,aa,"[0-9]+",bb);printf ""#%s" )n", bb[length(bb)]+1'

or the identical, but spread over a few lines for readability:

awk '/^ *(/ print 
 !/^ *(/ split( $1, aa, "[0-9]+", bb )
 printf ""#%s" )n", bb[length(bb)]+1 '

1. /^ * and !/^ *(/ are two address rules, covering lines beginning with optional spaces and an open parenthesis ... and lines that don't.




2. split( $1, aa, "[0-9]+", bb ) For lines that don't, split the line into two arrays. aa is the line content delimited by the regex "[0-9]+", and bb are the delimiters that matched the regex. The final element of bb is what interests you.




3. printf ""#%s" )n" formats the output line, waiting for a single variable...




4. bb[length(bb)]+1 one plus the value of the last element of bb.

gawk '
 sub(/#.*.djvu/, "#" $1 + 1 ".djvu")
 print
' FPAT='[0-9]+.djvu' input.txt

The idea is following:

• extract the .djvu extension with leading numbers from the djvu filename, using the [0-9]+.djvu pattern (FPAT). Example: original filename is #all_24223_to_00243.cpc0002.djvu, the extracted part will be 0002.djvu.


• substitute the previous djvu filename #.*.djvu to the extracted one, increasing it by 1 previously. Example: take whole line $0 and substitute #all_24223_to_00243.cpc0002.djvu inside it, to the 0002.djvu + 1 (this expression results to the plain number 3, because of how the string to number conversion works in the gawk). Add the # sign and .djvu extension to it. Result: #3.djvu.

This solution will work only for lines with one djvu filename, as in your sample input.

Input

(bookmarks
 ("Cover"
 "#01.djvu" )
 ("Title page"
 "#all_24223_to_00243.cpc0002.djvu" )
 ("Preface"
 "#all_24223_to_00243.cpc0004.djvu" )

Output

(bookmarks
 ("Cover"
 "#2.djvu" )
 ("Title page"
 "#3.djvu" )
 ("Preface"
 "#5.djvu" )