pdfu

creates a PDF from a URL to an EAD XML file

usage: pdfu [-h] [-t TEMPDIR] [-w] [--loglevel LOGLEVEL] url outfile

takes an EAD file and turn it into a PDF

positional arguments:
  url                   URL or path to source EAD XML file
  outfile               name for new PDF

optional arguments:
  -h, --help            show this help message and exit
  -t TEMPDIR, --tempdir TEMPDIR
  -w, --warnings        show python warnings supressed by default
  --loglevel LOGLEVEL

run_nightly_batch.py

run from crontab on cloud control node

usage: run_nightly_batch.py [-h] [--simpledb-domain SIMPLEDB_DOMAIN]
                         [--shadow-prefix SHADOW_PREFIX] [--generate-all]
                         [--ondemand]
                         eads bucket shadow

run the PDF batch

positional arguments:
  eads                  URL for crawler to start harvesting EAD XML won't
                        follow redirects
  bucket                s3://bucket[/optional/path] where the generated PDF
                        files go
  shadow                .tar.gz filename to store "shadow" file archive for
                        XTF

optional arguments:
  -h, --help            show this help message and exit
  --simpledb-domain SIMPLEDB_DOMAIN
                        "domain"/name of Amazon Simple DB database
  --shadow-prefix SHADOW_PREFIX
                        path the .tar.gz will unpack to
  --launch-only         launch worker for manual batch
  --generate-all        build all files
  --ondemand            use EC2 ondemand rather than EC2 spot market

crontab

# minute         0-59
#   hour           0-23
#     day of month   0-31
#       month          0-12 (or names, see below)
#         day of week    0-7 (0 or 7 is Sun, or use names)
# Eastern timzone

04 01 * * * ///appstrap/cronic/cronic ///code/pdfu/run_nightly_batch.py http:///oac-ead/prime2002/ s3:///pdf/ /usr/share/nginx/html/pdf-shadow.tar.gz 

install

Requires oac-ead-to-pdf and some python modules, which are installed into a virtualenv and a subdirectory by running.

./init.sh

re-run for OAC

--generate-all does not work for OAC...

to sort from smallest EAD file to the largest EAD file, you must run a command on the machine with the EAD files.

find prime2002/ -type f -printf "%s\t%p\n" | sort -n > sorted-prime.txt

edit with vi

:%s,prime2002/,,
:%s,.xml$,,

create the batch file

cat sorted-prime.txt | xargs -I {} echo "http://voro.cdlib.org/oac-ead/prime2002/{}.xml s3://pdf-generation/pdf/{}.pdf" > sbatch.txt

on remote host

xargs -a sbatch.txt -P 7 -n 2 ./pdfu/pdfu

it will die in the end, but we did the best we can