qurator-spk/ocrd-galley
1
0
Fork 0
mirror of https://github.com/qurator-spk/ocrd-galley.git synced 2025-10-31 11:14:12 +01:00
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` self_dir=`realpath $self_dir`
workspace=$ppn make_workspace () {
mkdir "$workspace" ppn=$1
cd "$workspace" workspace=$2
pwd
oai_identifier="oai%3Adigital.staatsbibliothek-berlin.de%3A$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
# Fix 'file:/' URLs to 'file:///' # Make workspace directory
sed -i 's#file:/\([^/]\)#file:///\1#' mets.xml mkdir "$workspace"
cd "$workspace"
# Patch mets.xml to use our NFS mount # Get METS from OAI-PMH
sed -i 's#file:///goobi/tiff001/sbb/#file:///srv/digisam_images/sbb/#g' mets.xml 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 # Fix 'file:/' URLs to 'file:///'
ocrd workspace remove-group -rf --keep-files LOCAL 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 $self_dir/run-docker-hub -I PRESENTATION --skip-validation