qurator-spk/ocrd-galley
1
0
Fork 0
mirror of https://github.com/qurator-spk/ocrd-galley.git synced 2025-07-01 01:19:52 +02:00
Code Issues Projects Releases Wiki Activity

🎨 ppn2ocr: Extract a function to make a workspace

Browse source
This commit is contained in:
Gerber, Mike 2020-05-22 16:53:20 +02:00
parent 18d4ab0ba1
commit 2a4b204fbe
1 changed files with 24 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

39
ppn2ocr
View file

@ -12,27 +12,36 @@ self_dir=`dirname $0`
self_dir=`realpath $self_dir`
workspace=$ppn
mkdir "$workspace"
cd "$workspace"
pwd
oai_identifier="oai%3Adigital.staatsbibliothek-berlin.de%3A$ppn"
oai_url="https://digital.staatsbibliothek-berlin.de/oai?verb=GetRecord&metadataPrefix=mets&identifier=$oai_identifier"
echo "$oai_url"
curl "$oai_url" | xmlstarlet sel -t -c '//*[local-name()="mets"]' > mets.xml
make_workspace () {
ppn=$1
workspace=$2
# Fix 'file:/' URLs to 'file:///'
sed -i 's#file:/\([^/]\)#file:///\1#' mets.xml
# Make workspace directory
mkdir "$workspace"
cd "$workspace"
# Patch mets.xml to use our NFS mount
sed -i 's#file:///goobi/tiff001/sbb/#file:///srv/digisam_images/sbb/#g' mets.xml
# Get METS from OAI-PMH
oai_identifier="oai%3Adigital.staatsbibliothek-berlin.de%3A$ppn"
oai_url="https://digital.staatsbibliothek-berlin.de/oai?verb=GetRecord&metadataPrefix=mets&identifier=$oai_identifier"
echo "$oai_url"
curl "$oai_url" | xmlstarlet sel -t -c '//*[local-name()="mets"]' > mets.xml
# Remove LOCAL file group as we do not have access to the files
ocrd workspace remove-group -rf --keep-files LOCAL
# Fix 'file:/' URLs to 'file:///'
sed -i 's#file:/\([^/]\)#file:///\1#' mets.xml
ocrd workspace validate mets.xml | grep -v "<notice>Won't download remote image"
# Patch mets.xml to use our NFS mount
sed -i 's#file:///goobi/tiff001/sbb/#file:///srv/digisam_images/sbb/#g' mets.xml
# Remove LOCAL file group as we do not have access to the files
ocrd workspace remove-group -rf --keep-files LOCAL
# Validate workspace
ocrd workspace validate mets.xml | grep -v "<notice>Won't download remote image"
}
make_workspace $ppn $ppn
$self_dir/run-docker-hub -I PRESENTATION --skip-validation