seb

Hi,

I'm struggling with a task that seems perfect for awk, as is usually the
case with BibTeX files. The file has thousands of entries like this:


@ARTICLE{913,
author = {Doe, J.},
title = {title},
journal = {some journal},
year = {1492},
volume = {30},
pages = {40-50},
key = {913},
pdf = {filename.pdf}
}


Many entries already have that last 'pdf' line, but a lot of entries need
to have that line added. The file names are the base names (without the
directory) of files available in some directory. The idea is therefore to
find which files in the directory with the pdf files do not have a 'pdf'
line entry somewhere on the BibTeX file.

I know how I can get a list of the file names already in the BibTeX file,
but am not sure how to proceed from there. I would highly appreciate some
suggestions on how to approach this. Thanks in advance.


Cheers,

--
Seb

bob harris

cd /to/directory/interest
ls | awk '
NR == FNR && $1 == "pdf" {
#
# reading BibTeX and found a pdf entry
#
gsub(/{}/,"",$3)
files[$3] = 1 # build hash of pdf entries
}
NR == FNR {
#
# Reading stdin of directory names
#
if ( !($1 in files) ) {
print # no BibTeX pdf entry for this file
}
}
' BibTeX.file - # dash says read stdin _AFTER_ BibTeX

Bob Harris

ed morton

ITYM:

cd /to/directory/interest
ls | awk '
NR == FNR {
if ($1 == "pdf") {
# reading BibTeX and found a pdf entry
gsub(/[{}]/,"",$3)
files[$3] = 1 # build hash of pdf entries
}
next
}
# Reading stdin of directory names
!($0 in files) # no BibTeX pdf entry for this file
' BibTeX.file - # dash says read stdin _AFTER_ BibTeX

It still won't work for files that contain newlines in them but
hopefully you don't have that situation.

Regards,

Ed.

seb

Indeed, I don't. This dealt with the task completely. It's amazing how
quickly my awk skills get rusty if I don't use it for a while. Thanks to
both of you for getting me up to speed again.


Cheers,

--
Seb