diff --git a/bin/slubsipbuilder.pl b/bin/slubsipbuilder.pl
old mode 100644
new mode 100755
index de574ed0d573fae8e6010473aae62928395d97ed..932b8e3661bdb52e39eecbf93024d705ce87e006
--- a/bin/slubsipbuilder.pl
+++ b/bin/slubsipbuilder.pl
@@ -10,44 +10,54 @@
# REQUIREMENTS: ---
# BUGS: ---
# NOTES: related to official document
-# "SIP Spezifikation (v1.4.1)"
+# "SIP Spezifikation (v1.4.2)"
# AUTHOR: Andreas Romeyke (romeyke@slub-dresden.de)
# ORGANIZATION: SLUB
# VERSION: 1.1
-# CREATED: 10.05.2016
+# CREATED: 2019-07-23
#===============================================================================
+
+
use strict;
use warnings;
use Carp;
-use 5.20.0;
-use strict;
-use warnings;
-use Archive::Zip::SimpleZip qw($SimpleZipError);
-use Cwd;
-use DateTime::Format::ISO8601;
-use Digest::MD5 qw(md5);
-use File::Basename;
-use File::Copy qw(cp);
-use File::Find;
-use File::Path;
-use File::Slurp;
-use Getopt::Long;
-use LWP::UserAgent; # to get MARC data
-use MARC::Record;
-use Pod::Usage;
-use XML::LibXML;
-use XML::LibXSLT;
-use XML::XPath;
-use constant buffer => 100*1024*1024; # use 100MB as Buffer
-
-my $with_debug = 0;
-
-# this will patch the mods-xml as a workaround for bugs in LOCs xslt files
-sub patch_mods($) {
- my $modsobj = shift; # mods expected as XML Parser object
- # TODO: Bugfix for /mets:mets/mets:dmdSec[1]/mets:mdWrap[1]/mets:xmlData[1]/mods:modsCollection[1]/mods:mods[1]/mods:relatedItem[2]/mods:internetMediaType[1]
- my $xslt_patch_string =<<PATCH;
+use 5.28.0;
+package SLUB::LZA::SIPBuilder;
+ use DateTime::Format::ISO8601;
+ use File::Copy qw(cp);
+ use File::Find;
+ use Path::Tiny;
+ use LWP::UserAgent; # to get MARC data
+ use MARC::Record;
+ use XML::LibXML;
+ use XML::LibXSLT;
+ use XML::XPath;
+ use Carp;
+ my $marc_mods_url = 'http://www.loc.gov/standards/mods/v3/MARC21slim2MODS3-6.xsl';
+ my $marc_utils_url = 'http://www.loc.gov/standards/marcxml/xslt/MARC21slimUtils.xsl';
+ my $swb_url = 'https://sru.bsz-bw.de/swb';
+ my $searchkey = "pica.swn";
+ my $recordschema = "marcxmlvbos";
+ our $VERSION = '1.2';
+ our $with_debug=0;
+
+
+ # write data to file (UTF-8)
+ sub write_file($$) {
+ my $filename = $_[0];
+ my $value = $_[1];
+ open(my $fh, '>:encoding(UTF-8)', $filename) || (croak "Can't open '$filename', $!");
+ print $fh $value;
+ close($fh) || (croak "could not close file '$filename', $!");
+ return 1;
+ }
+
+ # this will patch the mods-xml as a workaround for bugs in LOCs xslt files
+ sub patch_mods($) {
+ my $modsobj = shift; # mods expected as XML Parser object
+ # TODO: Bugfix for /mets:mets/mets:dmdSec[1]/mets:mdWrap[1]/mets:xmlData[1]/mods:modsCollection[1]/mods:mods[1]/mods:relatedItem[2]/mods:internetMediaType[1]
+ my $xslt_patch_string = <<'PATCH';
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:xs="http://www.w3.org/2001/XMLSchema"
@@ -68,16 +78,16 @@ sub patch_mods($) {
</xsl:template>
</xsl:stylesheet>
PATCH
- my $xslt = XML::LibXSLT->new();
- my $xslt_patch = XML::LibXML->load_xml(string=>$xslt_patch_string, no_cdata=>1);
- my $stylesheet = $xslt->parse_stylesheet ( $xslt_patch);
- my $result = $stylesheet->transform( $modsobj );
- return $result;
-}
+ my $xslt = XML::LibXSLT->new();
+ my $xslt_patch = XML::LibXML->load_xml(string => $xslt_patch_string, no_cdata => 1);
+ my $stylesheet = $xslt->parse_stylesheet($xslt_patch);
+ my $result = $stylesheet->transform($modsobj);
+ return $result;
+ }
-sub patch_marc_response($) {
- my $marcobj = shift; # marcobj expected as XML Parser object
- my $xslt_patch_string =<<PATCH2;
+ sub patch_marc_response($) {
+ my $marcobj = shift; # marcobj expected as XML Parser object
+ my $xslt_patch_string = <<'PATCH2';
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns="http://www.loc.gov/MARC21/slim" xmlns:srw="http://www.loc.gov/zing/srw/"
@@ -102,155 +112,206 @@ sub patch_marc_response($) {
</xsl:template>
</xsl:stylesheet>
PATCH2
- my $xslt = XML::LibXSLT->new();
- my $xslt_patch = XML::LibXML->load_xml(string=>$xslt_patch_string, no_cdata=>1);
- my $stylesheet = $xslt->parse_stylesheet ( $xslt_patch);
- my $result = $stylesheet->transform( $marcobj );
- return $result;
-}
-
+ my $xslt = XML::LibXSLT->new();
+ my $xslt_patch = XML::LibXML->load_xml(string => $xslt_patch_string, no_cdata => 1);
+ my $stylesheet = $xslt->parse_stylesheet($xslt_patch);
+ my $result = $stylesheet->transform($marcobj);
+ return $result;
+ }
+ # check MARC21 utility xsl
+ sub check_marc21_utility {
+ my $xsl_dir = shift;
+ my $ua = shift;
+ my $marc_utils_basename = path($marc_utils_url)->basename;
+ my $marc_utils_path = path($xsl_dir)->child($marc_utils_basename);
+ if (!$marc_utils_path->is_file) {
+ say "Downloading MARC21 utility xsl '$marc_utils_url'";
+ my $result = $ua->get($marc_utils_url);
+ if ($result->is_error) {
+ croak "Failed to download '$marc_utils_url', " . $result->error_as_HTML;
+ }
+ say "Saving MARC21 utility xsl to file '$marc_utils_path'";
+ my $xsl = $result->decoded_content;
+ write_file($marc_utils_path, $xsl);
+ }
+ return $marc_utils_path;
+ }
-# the "old" approach does not handle umlauts or UTF8-chars above ASCII table
-# old: http://swb2.bsz-bw.de/sru/DB=2.1/username=/password=/?query=pica.ppn+%3D+"494384174"&startRecord=1&maximumRecords=10&recordSchema=marcxml
-# new: http://swb2.bsz-bw.de/sru/DB=2.1/username=/password=/?query=pica.ppn+%3D+"494384174"&startRecord=1&maximumRecords=10&recordSchema=marc21&recordPacking=xml&version=1.1
-# with stylesheet:
-# http://swb2.bsz-bw.de/sru/DB=2.1/username=/password=/?query=pica.ppn+%3D+%22494384174%22&version=1.1&operation=searchRetrieve&stylesheet=http%3A%2F%2Fswb2.bsz-bw.de%2Fsru%2FDB%3D2.1%2F%3Fxsl%3DsearchRetrieveResponse&recordSchema=marc21&maximumRecords=10&startRecord=1&recordPacking=xml&sortKeys=none&x-info-5-mg-requestGroupings=none
-sub get_mods_from ($$) { # $mods = ($url, $ppn)
- my $url = shift;
- my $ppn = shift; # example: "457035137" for "Der Fichtelberg"
- #### where to find XSLT
- # my $marc_dc_url = 'http://www.loc.gov/standards/marcxml/xslt/MARC21slim2RDFDC.xsl';
- my $marc_mods_url = 'http://www.loc.gov/standards/mods/v3/MARC21slim2MODS3-6.xsl';
+ # check MARC21->MODS xsl
+ sub check_marc21_mods_xsl {
+ my $xsl_dir = shift;
+ my $ua = shift;
+ my $marc_mods_basename = path($marc_mods_url)->basename;
+ my $marc_mods_path = path($xsl_dir)->child($marc_mods_basename)->stringify;
+ my $marc_mods_patched_basename = path($marc_mods_url)->basename(".xsl") . ".patched.xsl";
+ my $marc_mods_patched_path = path($xsl_dir)->child($marc_mods_patched_basename);
+ if (! $marc_mods_patched_path->is_file) {
+ say "Downloading MARC21->MODS xsl '$marc_mods_url'";
+ my $result = $ua->get($marc_mods_url);
+ if ($result->is_error) {
+ croak "Failed to download '$marc_mods_url', " . $result->error_as_HTML;
+ }
+ say "Modifying MARC21->MODS xsl for offline use";
+ my $xsl = $result->decoded_content;
+ write_file($marc_mods_path, $xsl);
+ my $xsl_modified = $xsl;
+ my $marc_utils_path = check_marc21_utility( $xsl_dir, $ua);
+ $xsl_modified =~ s#$marc_utils_url#$marc_utils_path#g;
+ say "Saving MARC21->MODS xsl to file '$marc_mods_path'";
+ write_file($marc_mods_patched_path, $xsl_modified);
+ }
+ return $marc_mods_patched_path;
+ }
- my $ua = LWP::UserAgent->new;
- $ua->agent("MyApp/0.1 ");
- $ua->timeout(3600); #1h
- my $srubase=$url; # host
- my $srusearchkey="pica.ppn";
- my $sruvalue=$ppn;
- my $srumaxrecords=1;
- #my $sruschema="marcxml";
- my $sruschema="marc21";
- #my $sru = "${srubase}?query=${srusearchkey}+%3D+%22${sruvalue}%22&startRecord=1&maximumRecords=${srumaxrecords}&recordSchema=${sruschema}";
- my $sru = "${srubase}?query=${srusearchkey}+%3D+%22${sruvalue}%22&startRecord=1&maximumRecords=${srumaxrecords}&recordSchema=${sruschema}&recordPacking=xml&version=1.1&stylesheet=http%3A%2F%2Fswb2.bsz-bw.de%2Fsru%2FDB%3D2.1%2F%3Fxsl%3DsearchRetrieveResponse";
- #p ($sru); # debug output
- my $record = $ua->get($sru); # ask SWB for given PPN
- if ($record->is_success) {
- # parse ZiNG repsonse, extract MARC-data
- my $xp = XML::XPath->new( $record->decoded_content );
- my $parser = XML::LibXML->new();
- if ($with_debug) {
- say "write DEBUG_${ppn}_response.xml";
- write_file("DEBUG_${ppn}_response.xml", {binmode => ':utf8'}, $record->decoded_content);
+ sub check_xsl_directory {
+ # check xsl directory
+ my $xsl_dir = path(__FILE__)->parent->realpath->parent->child("xsl");
+ if (! $xsl_dir->is_dir) {
+ say "Rebuilding XSL directory '$xsl_dir'";
+ $xsl_dir->mkpath() || confess("could not mkdir '$xsl_dir', $!");
}
- my $marcblob = $parser->parse_string(
- $xp->findnodes_as_string('/*[local-name()="searchRetrieveResponse"]/*[local-name()="records"]/*[local-name()="record"]/*[local-name()="recordData"]/*')
- );
- my $marcblob_patched = patch_marc_response( $marcblob );
- if ($with_debug) {
- say "write DEBUG_${ppn}_marc_unpatched.xml";
- write_file("DEBUG_${ppn}_marc_unpatched.xml", {binmode => ':utf8'}, $marcblob);
- say "write DEBUG_${ppn}_marc.xml";
- write_file("DEBUG_${ppn}_marc.xml", {binmode => ':utf8'}, $marcblob_patched);
+ return $xsl_dir;
+ }
+
+ # specification SRU/SRW BSZ: https://wiki.k10plus.de/pages/viewpage.action?pageId=132874251
+ sub get_mods_from($$$$) {
+ # $mods = ($url, $ppn, $searchkey, $recordschema)
+ my $url = shift;
+ my $ppn = shift; # example: "457035137" for "Der Fichtelberg"
+ my $key = shift;
+ my $schema = shift;
+
+
+ #### where to find XSLT
+
+
+
+
+ my $ua = LWP::UserAgent->new;
+ $ua->agent("MyApp/0.1 ");
+ $ua->timeout(3600); #1h
+
+ my $xsl_dir = check_xsl_directory();
+ check_marc21_utility($xsl_dir, $ua);
+ check_marc21_mods_xsl($xsl_dir, $ua);
+
+
+ my $srubase = $url; # host
+ my $srusearchkey = $key; # SRU search key
+ my $sruvalue = $ppn;
+ my $srumaxrecords = 1;
+ my $srustartrecord = 1;
+ my $sruschema = $schema;
+ my $sru = "${srubase}?version=1.1&query=${srusearchkey}%3D${sruvalue}&operation=searchRetrieve&maximumRecords=${srumaxrecords}&startRecord=${srustartrecord}&recordSchema=${sruschema}";
+ if ($with_debug) {say "catalog-URL='$sru'";}
+ my $response = $ua->get($sru); # ask SWB for given PPN
+ if ($response->is_success) {
+ # parse ZiNG repsonse, extract MARC-data
+ my $xp = XML::XPath->new($response->decoded_content);
+ my $parser = XML::LibXML->new();
+ if ($with_debug) {
+ say "write DEBUG_${ppn}_response.xml";
+ write_file("DEBUG_${ppn}_response.xml", $response->decoded_content);
+ }
+ my $recordData = $xp->findnodes_as_string('/*[local-name()="searchRetrieveResponse"]/*[local-name()="records"]/*[local-name()="record"]/*[local-name()="recordData"]/*');
+ if (!$recordData) { croak("ERROR: Did not get any <recordData/> for PPN '$ppn' using '$sru'");}
+ my $marcblob = $parser->parse_string($recordData);
+
+
+ my $marcblob_patched = patch_marc_response($marcblob);
+ if ($with_debug) {
+ say "write DEBUG_${ppn}_marc_unpatched.xml";
+ write_file("DEBUG_${ppn}_marc_unpatched.xml", $marcblob);
+ say "write DEBUG_${ppn}_marc.xml";
+ write_file("DEBUG_${ppn}_marc.xml", $marcblob_patched);
+ }
+ my $marc_mods_patched_path = check_marc21_mods_xsl($xsl_dir, $ua);
+ my $xslt = XML::LibXSLT->new();
+ my $marcmods = XML::LibXML->load_xml(location => $marc_mods_patched_path, no_cdata => 1);
+ my $stylesheet = $xslt->parse_stylesheet($marcmods);
+ my $marc = $parser->parse_string($marcblob_patched);
+ my $result = $stylesheet->transform($marc);
+ if ($with_debug) {
+ say "write DEBUG_${ppn}_unpatched_mods.xml";
+ write_file("DEBUG_${ppn}_unpatched_mods.xml", $stylesheet->output_string($result));
+ }
+
+ $result = patch_mods($result);
+ my $result_string = $stylesheet->output_string($result);
+ return $result_string;
}
- my $xslt = XML::LibXSLT->new();
- my $marcmods = XML::LibXML->load_xml(location=>$marc_mods_url, no_cdata=>1);
- my $stylesheet = $xslt->parse_stylesheet ( $marcmods);
- my $marc = $parser->parse_string( $marcblob_patched );
- my $result = $stylesheet->transform( $marc);
- if ($with_debug) {
- say "write DEBUG_${ppn}_unpatched_mods.xml";
- write_file("DEBUG_${ppn}_unpatched_mods.xml", {binmode => ':utf8'}, $stylesheet->output_string( $result ));
+ else {
+ carp("Problem asking catalogue at $url using $ppn");
}
- $result = patch_mods( $result);
- my $result_string = $stylesheet->output_string( $result );
- return $result_string;
- } else {
- carp ("Problem asking catalogue at $url using $ppn");
+ return;
}
- return;
-}
-#===============================================================================
-
-my $directory;
-my $ppn;
-my $noppn;
-my $output;
-my $url;
-my $as_zip;
-my $external_id;
-my $external_workflow;
-my $external_isil="";
-my $external_value_descr;
-my $external_conservation_flag;
+sub create_filecopyhash {
+ my $directory = shift;
+ my $content = shift;
+ my %filecopyhash;
+ my $wanted=sub {
+ if (-d $_) {
+ # dir, do nothing
+ ();
+ } else {
+ my $file=$File::Find::name;
+ if ($file !~ m#^[-A-Za-z0-9_\.:\\/]+$#) {
+ confess("file '$file' does not match regex '^[-A-Za-z0-9_\.:\\/]+\$'");
+ }
+ my $source = $file;
+ $filecopyhash{$source}->{'source'}=$file;
+ $file=~s#^$directory/?##;
+ $filecopyhash{$source}{'relative'}="data/$file";
+ $filecopyhash{$source}{'target'}="$content/$file";
+ my $fh;
+ open($fh, "<", $source) or confess ("Can't open '$source', $!");
+ binmode($fh);
+ my $ctx = Digest::MD5->new;
+ $ctx->addfile(*$fh);
+ close ($fh);
+ my $md5 = $ctx->hexdigest;
+ $filecopyhash{$source}{'md5sum'}=$md5;
+ }
+ };
-our $VERSION = '1.0';
-GetOptions(
- "IE_directory=s" => \$directory,
- "ppn=s" => \$ppn,
- "noppn=s" => \$noppn,
- "SIP_output_path=s" => \$output,
- "as_zip" => \$as_zip,
- "url=s" => \$url,
- "external_id=s" => \$external_id,
- "external_workflow=s" => \$external_workflow,
- "external_ISIL=s" => \$external_isil,
- "external_value_descr=s" => \$external_value_descr,
- "external_conservation_flag" => \$external_conservation_flag,
- "debug" => \$with_debug,
-
- "help" => sub { pod2usage(1); exit(0); },
- ) or pod2usage(2);
-
-if (!defined $directory) { confess ("you need to specify an IE directory, which needs to be archived"); }
-if (!defined $ppn && !defined $noppn) { confess ("you need to specify a PPN, which exists in SWB catalogue"); }
-if (defined $ppn && defined $noppn) {confess ("you could only use --ppn=foo or --noppn=bar"); }
-if (!defined $output) { confess (" you need to specify an output path, where the SIP will be stored"); }
-if ($output !~ m#^/#) { confess("you need to specify an output path using absoluet paths, $!"); }
-if (!defined $url) { $url = "http://swb.bsz-bw.de/sru/DB=2.1/username=/password=/";}
-if (!defined $external_conservation_flag) { $external_conservation_flag="false"; } else { $external_conservation_flag="true"; }
-# additional checks
-if (! -d $directory) { confess("you need to specify an IE directory, which needs to be archived, $!"); }
-if ($directory !~ m#^/#) { confess("you need to specify an IE directory using absoluet paths, $!"); }
-#if (! -d $output) { confess("you need to specify an output path, where the SIP will be stored, $!"); }
-
-# get date
-my $export_to_archive_date = DateTime->now->iso8601();#
-my $file_date = $export_to_archive_date;
-$file_date =~ s/T/_/g;
-$file_date =~ s/:/-/g;
-# create output dir
-mkpath "$output" || confess("could not create SIP directory for '$output', $!");
-my $sip_root_dir = "PPN-${ppn}_${file_date}";
-my $content = "$output/$sip_root_dir/data";
-if (!defined $as_zip) {
- mkpath "$output/$sip_root_dir" || confess("could not create SIP directory for '$output/$sip_root_dir', $!");
- mkpath "$content" || confess("could not create SIP subdirectory for '$content', $!");
+ finddepth($wanted, $directory);
+ return \%filecopyhash;
}
-
-# prepare dmd-sec
-my $mods;
-if (defined $ppn) {
- $mods = get_mods_from($url, $ppn);
- if (1 == $with_debug) {
- write_file("DEBUG_${ppn}_mods.xml", {binmode => ':utf8'}, $mods);
+ sub prepare_dmd_section_with_ppn ($) {
+ my $ppn = shift;
+ my $mods = SLUB::LZA::SIPBuilder::get_mods_from($swb_url, $ppn, $searchkey, $recordschema);
+ if ($with_debug) {
+ SLUB::LZA::SIPBuilder::write_file("DEBUG_${ppn}_mods.xml", $mods);
+ }
+ # remove the <xml /> from beginning of the answer
+ $mods=~ s#<\?xml version="1.0" encoding="UTF-8"\?>#<!-- removed xml header from mods part -->#;
+ my $dmd =<<"DMD";
+<mets:dmdSec ID="DMDLOG_0000">
+ <!-- bibliographic metadata -->
+ <mets:mdWrap MDTYPE="MODS">
+ <mets:xmlData>
+ $mods
+ </mets:xmlData>
+ </mets:mdWrap>
+</mets:dmdSec>
+DMD
+ return $dmd;
}
-# remove the <xml /> from beginning of the answer
- $mods=~ s#<\?xml version="1.0" encoding="UTF-8"\?>#<!-- removed xml header from mods part -->#;
-} elsif (defined $noppn) {
- $mods =<<MODS;
-<mods version="3.5"
+
+ sub prepare_dmd_section_with_noppn ($) {
+ my $noppn = shift;
+ my $mods =<<"MODS";
+<mods version="3.6"
xmlns="http://www.loc.gov/mods/v3"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
- xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-5.xsd">
+ xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-6.xsd">
<identifier>$noppn</identifier>
</mods>
MODS
-}
-my $dmd =<<DMD;
+ my $dmd =<<"DMD";
<mets:dmdSec ID="DMDLOG_0000">
<!-- bibliographic metadata -->
<mets:mdWrap MDTYPE="MODS">
@@ -260,9 +321,17 @@ my $dmd =<<DMD;
</mets:mdWrap>
</mets:dmdSec>
DMD
+ return $dmd;
+ }
-# prepare amd-sec
-my $amd =<<AMD;
+ sub prepare_amd_section($$$$$$) {
+ my $export_to_archive_date = shift;
+ my $external_workflow = shift;
+ my $external_id = shift;
+ my $external_conservation_flag = shift;
+ my $external_isil = shift;
+ my $external_value_descr = shift;
+ my $amd =<<"AMD";
<mets:amdSec ID="AMD">
<!-- SIP metadata for automated processing by submission application -->
<mets:techMD ID="ARCHIVE">
@@ -281,80 +350,149 @@ my $amd =<<AMD;
</mets:techMD>
</mets:amdSec>
AMD
-
-# create filecopyhash
-my %filecopyhash;
-my $wanted=sub {
- if (-d $_) {
- # dir, do nothing
- ()
- } else {
- my $file=$File::Find::name;
- if ($file !~ m#^[-A-Za-z0-9_\./]+$#) {
- confess("file '$file' does not match regex '^[-A-Za-z0-9_\./]+\$'");
- }
- my $source = $file;
- $filecopyhash{$source}->{'source'}=$file;
- $file=~s#^$directory/?##;
- $filecopyhash{$source}{'relative'}="data/$file";
- $filecopyhash{$source}{'target'}="$content/$file";
- my $fh;
- open($fh, "<", $source) or confess ("Can't open '$source' (current='",getcwd,"', $!\n");
- binmode($fh);
- my $ctx = Digest::MD5->new;
- $ctx->addfile(*$fh);
- close ($fh);
- my $md5 = $ctx->hexdigest;
- $filecopyhash{$source}{'md5sum'}=$md5;
+ return $amd;
}
-};
-
-finddepth($wanted, $directory);
-# create fileSec
-my $filesec=<<FILESEC1;
+ sub prepare_files_sections($) {
+ my $filecopyhash = shift;
+ my @fsec;
+ my $i=0;
+ foreach my $fkey (sort keys (%{$filecopyhash})) {
+ push @fsec, sprintf("<mets:file ID=\"FILE_%015u_LZA\" CHECKSUMTYPE=\"MD5\" CHECKSUM=\"%s\">", $i, $filecopyhash->{$fkey}->{"md5sum"});
+ push @fsec, sprintf("<mets:FLocat xmlns:xlink=\"http://www.w3.org/1999/xlink\" LOCTYPE=\"URL\" xlink:href=\"file://%s\"/>", $filecopyhash->{$fkey}->{"relative"});
+ push @fsec, "</mets:file>";
+ $i++;
+ }
+ my $files = join("\n", @fsec);
+ my $filesec=<<"FILESEC";
<mets:fileSec>
<mets:fileGrp USE="LZA">
-FILESEC1
-{
- my @fsec;
- my $i=0;
- foreach my $fkey (sort keys (%filecopyhash)) {
- push @fsec, sprintf("<mets:file ID=\"FILE_%015u_LZA\" CHECKSUMTYPE=\"MD5\" CHECKSUM=\"%s\">", $i, $filecopyhash{$fkey}->{"md5sum"});
- push @fsec, sprintf("<mets:FLocat xmlns:xlink=\"http://www.w3.org/1999/xlink\" LOCTYPE=\"URL\" xlink:href=\"file://%s\"/>", $filecopyhash{$fkey}->{"relative"});
- push @fsec, "</mets:file>";
- $i++;
- }
- $filesec = join("\n", $filesec, @fsec);
-}
-$filesec = $filesec . <<FILESEC2;
+ $files
</mets:fileGrp>
</mets:fileSec>
-FILESEC2
+FILESEC
+ return $filesec;
+ }
-# prepare structmap
-my $structmap =<<STRUCTMAP1;
+ sub prepare_struct_map($) {
+ my $filecopyhash = shift;
+ my @ssec;
+ my $i=0;
+ foreach my $fkey (sort keys (%{$filecopyhash})) {
+ push @ssec, sprintf("<mets:div ID=\"PHYS_%015u_LZA\" TYPE=\"fileorderSequence\">", $i);
+ push @ssec, sprintf("<mets:fptr FILEID=\"FILE_%015u_LZA\" />", $i);
+ push @ssec, "</mets:div>";
+ $i++;
+ }
+ my $structs = join("\n", @ssec);
+ my $structmap =<<"STRUCTMAP";
<mets:structMap TYPE="PHYSICAL">
<mets:div ID="PHYS_0000" TYPE="ieDir">
-STRUCTMAP1
-{
- my @ssec;
- my $i=0;
- foreach my $fkey (sort keys (%filecopyhash)) {
- push @ssec, sprintf("<mets:div ID=\"PHYS_%015u_LZA\" TYPE=\"fileorderSequence\">", $i);
- push @ssec, sprintf("<mets:fptr FILEID=\"FILE_%015u_LZA\" />", $i);
- push @ssec, "</mets:div>";
- $i++;
- }
- $structmap = join("\n", $structmap, @ssec);
-}
-$structmap = $structmap . <<STRUCTMAP2;
+ $structs
</mets:div>
</mets:structMap>
-STRUCTMAP2
+STRUCTMAP
+ return $structmap;
+ }
+
+ # end package
+
+package main;
+#===============================================================================
+
+BEGIN{
+ $INC{'SLUB/LZA/SIPBuilder.pm'} = 1; # needed because inlined module
+}
+return 1 if caller; # avoids main code running if module stuff is needed
+use SLUB::LZA::SIPBuilder;
+use Archive::Zip::SimpleZip qw($SimpleZipError);
+use Getopt::Long;
+use Path::Tiny;
+use Digest::MD5;
+use constant buffer => 100 * 1024 * 1024; # use 100MB as Buffer
+use File::Find;
+use File::Copy qw(cp);
+use Pod::Usage;
+
+my $directory;
+my $ppn;
+my $noppn;
+my $output;
+my $as_zip;
+my $external_id;
+my $external_workflow;
+my $external_isil="";
+my $external_value_descr;
+my $external_conservation_flag;
-# create sip.xml
-my $sip =<<METS;
+my $help;
+my $man;
+
+
+GetOptions(
+ "IE_directory=s" => \$directory, # required
+ "ppn=s" => \$ppn, # semi-optional (choice 1 of 2)
+ "noppn=s" => \$noppn, # semi-optional (choice 2 of 2)
+ "SIP_output_path=s" => \$output, # required
+ "as_zip" => \$as_zip, # optional, default: do not zip
+ "external_id=s" => \$external_id, # required
+ "external_workflow=s" => \$external_workflow, # required
+ "external_ISIL=s" => \$external_isil, # optional, default: no ISIL
+ "external_value_descr=s" => \$external_value_descr, # required
+ "external_conservation_flag" => \$external_conservation_flag, # optional, default: no special conservation
+ "debug" => \$SLUB::LZA::SIPBuilder::with_debug, # optional
+ "help|?" => \$help, # optional
+ "man" => \$man, # optional
+) or pod2usage(2);
+
+if ($help) { pod2usage(1); }
+if ($man) { pod2usage(-exitval => 0, -verbose => 2); }
+if (!defined $directory) { confess("you need to specify an IE directory, which needs to be archived"); }
+if ((defined $ppn) && (defined $noppn)) { confess("you can only specify either -ppn or -noppn"); }
+if ((!defined $ppn) && (!defined $noppn)) { confess("you need to specify a PPN with -ppn or use --noppn"); }
+if (!defined $output) { confess("you need to specify an output path, where the SIP will be stored"); }
+if (!defined $external_conservation_flag) { $external_conservation_flag="false"; } else { $external_conservation_flag="true"; }
+if (! -d $directory) { confess("you need to specify an IE directory, which needs to be archived, $!"); }
+$directory = path($directory)->realpath->stringify;
+path($output)->mkpath;
+$output = path($output)->realpath->stringify;
+if ($external_id !~ m#^[a-z0-9]+$#) { confess("you need to specify a valid external ID (^[a-z0-9]+\$)"); }
+if ($external_workflow !~ m#^[a-z0-9]+$#) { confess("you need to specify a valid external workflow (^[a-z0-9]+\$)"); }
+if (!$external_value_descr) { confess("you need to specify an external value description (reason for archiving)"); }
+
+#===============================================================================
+
+sub main {
+ # get date
+ my $export_to_archive_date = DateTime->now->iso8601();
+ my $file_date = $export_to_archive_date;
+ $file_date =~ s/T/_/g; # replace 'T' with '_'
+ $file_date =~ s/:/-/g; # replace ':' with '-'
+ # prepare dirs
+ my $sip_root_dir = (defined $ppn)? "PPN-${ppn}_${file_date}" : "ID-${noppn}_${file_date}";
+ my $content = path($output)->child($sip_root_dir)->child("data")->stringify;
+ if (!defined $as_zip) {
+ path($content)->mkpath;
+ }
+ my $filecopyhash = SLUB::LZA::SIPBuilder::create_filecopyhash($directory, $content);
+
+ # prepare dmd-sec
+ my $dmd = (defined $ppn)? SLUB::LZA::SIPBuilder::prepare_dmd_section_with_ppn( $ppn ) : SLUB::LZA::SIPBuilder::prepare_dmd_section_with_noppn( $noppn );
+ # prepare amd-sec
+ my $amd = SLUB::LZA::SIPBuilder::prepare_amd_section(
+ $export_to_archive_date,
+ $external_workflow,
+ $external_id,
+ $external_conservation_flag,
+ $external_isil,
+ $external_value_descr
+ );
+ # create fileSec
+ my $filesec = SLUB::LZA::SIPBuilder::prepare_files_sections($filecopyhash);
+ # prepare structmap
+ my $structmap = SLUB::LZA::SIPBuilder::prepare_struct_map($filecopyhash);
+ # create sip.xml
+ my $sip =<<"METS";
<?xml version="1.0" encoding="utf-8"?>
<mets:mets xmlns:mets="http://www.loc.gov/METS/"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
@@ -366,38 +504,48 @@ my $sip =<<METS;
</mets:mets>
METS
-# compress if needed
-if (!defined $as_zip) {
- write_file( "${output}/${sip_root_dir}/sip.xml",{binmode => ':utf8'}, $sip );
- # copy source to target
- foreach my $source (sort keys (%filecopyhash)) {
- my $target = $filecopyhash{$source}->{"target"};
- my $basename = dirname($target);
- #say "cp $source, $target ($basename)";
- if (! -d $basename) {
- mkpath $basename || confess ("could not mkdir '$basename', $!");
+ # write stuff out
+ if (!defined $as_zip) {
+ SLUB::LZA::SIPBuilder::write_file( path($output)->child($sip_root_dir)->child("sip.xml")->stringify, $sip );
+ # copy source to target
+ foreach my $source (sort keys (%{$filecopyhash})) {
+ my $target = path($filecopyhash->{$source}->{"target"})->stringify; # CHECK ON WINDOWS
+ my $basename = path($target)->parent->stringify;
+ if (! -d $basename) {
+ path($basename)->mkpath;
+ }
+ cp($source, $target, buffer) || confess ("could not copy from '$source' to '$target', $!");
}
- cp($source, $target, buffer) || confess ("could not copy from '$source' to '$target', $!");
- }
- say "SIP '$sip_root_dir' build successfully in '$output'";
-} else {
- # compress it
- my $zip_file_path = "$output/$sip_root_dir.zip";
- my $zip = Archive::Zip::SimpleZip->new( $zip_file_path, Zip64=>1 );
- $zip->addString($sip, Name=> "$sip_root_dir/sip.xml" );
- # copy source to target
- foreach my $source (sort keys (%filecopyhash)) {
- my $target = "$sip_root_dir/".$filecopyhash{$source}->{"relative"};
- my $basename = dirname($target);
- #say "cp $source, $target ($basename)";
- $zip->add( $source, Name=> $target) || confess ("could not zip copy from '$source' to '$target', $!");
- }
- unless ( $zip->close()) {
- confess "write error to '$zip_file_path', $SimpleZipError, $!";
+ say "SIP '$sip_root_dir' build successfully in '$output'";
+ } else {
+ # compress it
+ my $zip_file_path = path($output)->child("$sip_root_dir.zip")->stringify;
+ my $zip = Archive::Zip::SimpleZip->new( $zip_file_path, Zip64=>1 );
+ $zip->addString($sip, Name=>path($sip_root_dir)->child("sip.xml")->stringify);
+ # copy source to target
+ foreach my $source (sort keys (%{$filecopyhash})) {
+ my $target = path($sip_root_dir)->child($filecopyhash->{$source}->{"relative"})->stringify; # CHECK ON WINDOWS
+ my $basename = path($target)->parent->stringify;
+ $zip->add( $source, Name=> $target) || confess ("could not zip copy from '$source' to '$target', $!");
+ }
+ unless ( $zip->close()) {
+ confess "write error to '$zip_file_path', $SimpleZipError, $!";
+ }
+ say "SIP '$sip_root_dir' build successfully in '$zip_file_path'";
}
- say "SIP '$sip_root_dir' build successfully in '$zip_file_path'";
+ return;
}
+
+#===============================================================================
+
+main();
+
+#===============================================================================
+
+
+__END__
+
=pod
=head1 NAME
@@ -413,17 +561,16 @@ slubsipbuilder.pl [options]
-man full documentation
-IE_directory=<IE dir> existing IE directory (absolute path!)
- -ppn=<ppn>|-noppn=<noppn> PPN (swb catalogue) or any identifier (uses minimalistic MODS)
+ -ppn=<ppn>|-noppn=<noppn> SWB-PPN or any identifier (uses minimalistic MODS)
-SIP_output_path=<target dir> where to put the SIP dir (absolute path!)
-as_zip optional, if set a ZIP will be created
- -url=<SRU url> optional, URL of the SRU for PICA catalogues
-external_id=<id> mandatory, should be uniqe ID
-external_workflow=<workflow> mandatory, should be uniqe workflow name
-external_ISIL=<isil> optional, ISIL number of library
-external_value_descr=<text> mandatory, the reason why to archive
-external_conservation_flag optional, if set no other "original" still exists
-slubsipbuilder.pl --IE_directory=/processdir_from_goobi/10008 --ppn=457035137 --SIP_output_path=/tmp/mysip --external_id=10008 --external_workflow=goobitest --external_ISIL=de-14 --external_value_descr="Gesetzlicher Auftrag" --as_zip
+slubsipbuilder.pl --IE_directory=/export_dir_kitodo/10008 --ppn=457035137 --SIP_output_path=/tmp/mysip --external_id=10008 --external_workflow=kitodo --external_ISIL=DE-14 --external_value_descr="Gesetzlicher Auftrag"
=head1 OPTIONS
@@ -440,4 +587,3 @@ Print a brief help message and exits.
B<This program> will process the given IE directory, add bibliographic metadata from catalogue with given PICA number and check and create a SIP directory ready for SLUBarchiv
=cut
-# vim: set tabstop=4
diff --git a/t/slubsipbuilder.t b/t/slubsipbuilder.t
new file mode 100644
index 0000000000000000000000000000000000000000..30945bcd0abe33b68370ab5210f351f0f4a5525f
--- /dev/null
+++ b/t/slubsipbuilder.t
@@ -0,0 +1,326 @@
+#!/usr/bin/perl -w
+use strict;
+use warnings;
+use diagnostics;
+
+use Test::More tests => 8;
+use Test::Exception;
+use Test::File;
+use Path::Tiny;
+
+### prepare
+BEGIN {
+ use Path::Tiny;
+ push @INC, Path::Tiny::path(__FILE__)->parent->parent->path("bin")->absolute->stringify;
+ require "slubsipbuilder.pl";
+ $INC{'SLUB/LZA/SIPBuilder.pm'} = 1; # needed because inlined module
+}
+my $unpatched_mods=<<'UNPATCHED_MODS';
+<?xml version="1.0" encoding="UTF-8"?>
+<mods xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://www.loc.gov/mods/v3" version="3.6" xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-6.xsd"><titleInfo><nonSort xml:space="preserve">Der </nonSort><title>Fichtelberg</title><subTitle>Berg der unbekannten Rekorrde</subTitle></titleInfo><name type="personal"><namePart>Schneider, Dirk</namePart><role><roleTerm type="text">FilmemacherIn</roleTerm></role><role><roleTerm authority="marcrelator" type="code">fmk</roleTerm></role><nameIdentifier>(DE-627)1235502279 (DE-576)165502274</nameIdentifier></name><typeOfResource>moving image</typeOfResource><genre authority="rdacontent">zweidimensionales bewegtes Bild</genre><genre authority="gnd-content">Film</genre><originInfo><place><placeTerm type="code" authority="marccountry">xx</placeTerm></place><dateIssued encoding="marc">2014</dateIssued><issuance>monographic</issuance></originInfo><originInfo eventType="publication"><place><placeTerm type="text">[Leipzig]</placeTerm></place><publisher>top ten tv</publisher><dateIssued>[2014]</dateIssued></originInfo><language><languageTerm authority="iso639-2b" type="code">ger</languageTerm></language><physicalDescription><form authority="marccategory">electronic resource</form><form authority="marcsmd">remote</form><extent>1 Online-Ressource (1 Videodatei, 29:49) farbig</extent><form type="media" authority="rdamedia">Computermedien</form><form type="carrier" authority="rdacarrier">Online-Ressource</form></physicalDescription><targetAudience authority="marctarget">juvenile</targetAudience><note type="statement of responsibility" altRepGroup="00">ein Film von Dirk Schneider</note><note>Dokumentarfilm. Deutschland. 2014</note><relatedItem type="series"><titleInfo><title>MDR</title></titleInfo></relatedItem><relatedItem type="series"><titleInfo><title>Der Osten - entdecke wo du lebst</title></titleInfo></relatedItem><identifier type="oclc">946544758</identifier><recordInfo><descriptionStandard>rda</descriptionStandard><recordContentSource authority="marcorg">DE-576</recordContentSource><recordCreationDate encoding="marc">160304</recordCreationDate><recordChangeDate encoding="iso8601">20160510144338.0</recordChangeDate><recordIdentifier source="DE-576">457035137</recordIdentifier><recordOrigin>Converted from MARCXML to MODS version 3.6 using MARC21slim2MODS3-6.xsl
+ (Revision 1.119 2018/06/21)</recordOrigin><languageOfCataloging><languageTerm authority="iso639-2b" type="code">ger</languageTerm></languageOfCataloging></recordInfo></mods>
+UNPATCHED_MODS
+my $patched_mods=<<'PATCHED_MODS';
+<?xml version="1.0" encoding="UTF-8"?>
+<mods xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://www.loc.gov/mods/v3" version="3.6" xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-6.xsd"><titleInfo><nonSort xml:space="preserve">Der </nonSort><title>Fichtelberg</title><subTitle>Berg der unbekannten Rekorrde</subTitle></titleInfo><name type="personal"><namePart>Schneider, Dirk</namePart><role><roleTerm type="text">FilmemacherIn</roleTerm></role><role><roleTerm authority="marcrelator" type="code">fmk</roleTerm></role><nameIdentifier>(DE-627)1235502279 (DE-576)165502274</nameIdentifier></name><typeOfResource>moving image</typeOfResource><genre authority="rdacontent">zweidimensionales bewegtes Bild</genre><genre authority="gnd-content">Film</genre><originInfo><place><placeTerm type="code" authority="marccountry">xx</placeTerm></place><dateIssued encoding="marc">2014</dateIssued><issuance>monographic</issuance></originInfo><originInfo eventType="publication"><place><placeTerm type="text">[Leipzig]</placeTerm></place><publisher>top ten tv</publisher><dateIssued>[2014]</dateIssued></originInfo><language><languageTerm authority="iso639-2b" type="code">ger</languageTerm></language><physicalDescription><form authority="marccategory">electronic resource</form><form authority="marcsmd">remote</form><extent>1 Online-Ressource (1 Videodatei, 29:49) farbig</extent><form type="media" authority="rdamedia">Computermedien</form><form type="carrier" authority="rdacarrier">Online-Ressource</form></physicalDescription><targetAudience authority="marctarget">juvenile</targetAudience><note type="statement of responsibility" altRepGroup="00">ein Film von Dirk Schneider</note><note>Dokumentarfilm. Deutschland. 2014</note><relatedItem type="series"><titleInfo><title>MDR</title></titleInfo></relatedItem><relatedItem type="series"><titleInfo><title>Der Osten - entdecke wo du lebst</title></titleInfo></relatedItem><identifier type="oclc">946544758</identifier><recordInfo><descriptionStandard>rda</descriptionStandard><recordContentSource authority="marcorg">DE-576</recordContentSource><recordCreationDate encoding="marc">160304</recordCreationDate><recordChangeDate encoding="iso8601">20160510144338.0</recordChangeDate><recordIdentifier source="DE-576">457035137</recordIdentifier><recordOrigin>Converted from MARCXML to MODS version 3.6 using MARC21slim2MODS3-6.xsl
+ (Revision 1.119 2018/06/21)</recordOrigin><languageOfCataloging><languageTerm authority="iso639-2b" type="code">ger</languageTerm></languageOfCataloging></recordInfo></mods>
+PATCHED_MODS
+my $unpatched_mods_obj = XML::LibXML->load_xml(string => $unpatched_mods);
+
+my $unpatched_marcblob=<<'UNPATCHED_MARCBLOB';
+<?xml version="1.0"?>
+<record xmlns="http://www.loc.gov/MARC21/slim">
+ <leader> cgm a22 4500</leader>
+ <controlfield tag="001">457035137</controlfield>
+ <controlfield tag="003">DE-576</controlfield>
+ <controlfield tag="005">20160510144338.0</controlfield>
+ <controlfield tag="006">m o | | </controlfield>
+ <controlfield tag="007">cr uuu---uuuuu</controlfield>
+ <controlfield tag="007">vu uuuuuu</controlfield>
+ <controlfield tag="008">160304s2014 xx ger c</controlfield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(DE-627)1655506501</subfield>
+ </datafield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(DE-576)457035137</subfield>
+ </datafield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(DE-599)BSZ457035137</subfield>
+ </datafield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(OCoLC)946544758</subfield>
+ </datafield>
+ <datafield tag="040" ind1=" " ind2=" ">
+ <subfield code="a">DE-576</subfield>
+ <subfield code="b">ger</subfield>
+ <subfield code="c">DE-576</subfield>
+ <subfield code="e">rda</subfield>
+ </datafield>
+ <datafield tag="041" ind1=" " ind2=" ">
+ <subfield code="a">ger</subfield>
+ </datafield>
+ <datafield tag="245" ind1="1" ind2="4">
+ <subfield code="a">Der Fichtelberg</subfield>
+ <subfield code="b">Berg der unbekannten Rekorrde</subfield>
+ <subfield code="c">ein Film von Dirk Schneider</subfield>
+ </datafield>
+ <datafield tag="264" ind1=" " ind2="1">
+ <subfield code="a">[Leipzig]</subfield>
+ <subfield code="b">top ten tv</subfield>
+ <subfield code="c">[2014]</subfield>
+ </datafield>
+ <datafield tag="264" ind1=" " ind2="4">
+ <subfield code="c">© 2014</subfield>
+ </datafield>
+ <datafield tag="300" ind1=" " ind2=" ">
+ <subfield code="a">1 Online-Ressource (1 Videodatei, 29:49)</subfield>
+ <subfield code="b">farbig</subfield>
+ </datafield>
+ <datafield tag="336" ind1=" " ind2=" ">
+ <subfield code="a">zweidimensionales bewegtes Bild</subfield>
+ <subfield code="b">tdi</subfield>
+ <subfield code="2">rdacontent</subfield>
+ </datafield>
+ <datafield tag="337" ind1=" " ind2=" ">
+ <subfield code="a">Computermedien</subfield>
+ <subfield code="b">c</subfield>
+ <subfield code="2">rdamedia</subfield>
+ </datafield>
+ <datafield tag="338" ind1=" " ind2=" ">
+ <subfield code="a">Online-Ressource</subfield>
+ <subfield code="b">cr</subfield>
+ <subfield code="2">rdacarrier</subfield>
+ </datafield>
+ <datafield tag="490" ind1="0" ind2=" ">
+ <subfield code="a">MDR</subfield>
+ </datafield>
+ <datafield tag="490" ind1="0" ind2=" ">
+ <subfield code="a">Der Osten - entdecke wo du lebst</subfield>
+ </datafield>
+ <datafield tag="500" ind1=" " ind2=" ">
+ <subfield code="a">Dokumentarfilm. Deutschland. 2014</subfield>
+ </datafield>
+ <datafield tag="591" ind1=" " ind2=" ">
+ <subfield code="a">Fernsehmitschnitt (SWB)</subfield>
+ </datafield>
+ <datafield tag="655" ind1=" " ind2="7">
+ <subfield code="a">Film</subfield>
+ <subfield code="0">(DE-588)4017102-4</subfield>
+ <subfield code="0">(DE-627)104559683</subfield>
+ <subfield code="0">(DE-576)208918531</subfield>
+ <subfield code="2">gnd-content</subfield>
+ </datafield>
+ <datafield tag="700" ind1="1" ind2=" ">
+ <subfield code="a">Schneider, Dirk</subfield>
+ <subfield code="e">FilmemacherIn</subfield>
+ <subfield code="0">(DE-627)1235502279</subfield>
+ <subfield code="0">(DE-576)165502274</subfield>
+ <subfield code="4">fmk</subfield>
+ </datafield>
+ <datafield tag="935" ind1=" " ind2=" ">
+ <subfield code="c">vide</subfield>
+ </datafield>
+ <datafield tag="937" ind1=" " ind2=" ">
+ <subfield code="a">Dokumentarfilm</subfield>
+ <subfield code="b">Deutschland</subfield>
+ <subfield code="c">2014</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">000 xxxxxcx a22 zn 4500</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">001 901795887</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">003 DE-576</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">004 457035137</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">005 20160510125331</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">008 160304||||||||||||||||ger|||||||</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">040 </subfield>
+ <subfield code="a">DE-14</subfield>
+ <subfield code="c">DE-576</subfield>
+ <subfield code="d">DE-14</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">852 </subfield>
+ <subfield code="z">Fernsehmitschnitt: MDR, 04.02.2014. - Beilage</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">852 </subfield>
+ <subfield code="a">DE-14</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">852 1</subfield>
+ <subfield code="9">00</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">866 </subfield>
+ <subfield code="x">ddsu,pn</subfield>
+ </datafield>
+</record>
+UNPATCHED_MARCBLOB
+
+my $patched_marcblob =<<'PATCHED_MARCBLOB';
+<?xml version="1.0"?>
+<record xmlns="http://www.loc.gov/MARC21/slim">
+ <leader> cgm a22 4500</leader>
+ <controlfield tag="001">457035137</controlfield>
+ <controlfield tag="003">DE-576</controlfield>
+ <controlfield tag="005">20160510144338.0</controlfield>
+ <controlfield tag="006">m o | | </controlfield>
+ <controlfield tag="007">cr uuu---uuuuu</controlfield>
+ <controlfield tag="007">vu uuuuuu</controlfield>
+ <controlfield tag="008">160304s2014 xx ger c</controlfield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(DE-627)1655506501</subfield>
+ </datafield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(DE-576)457035137</subfield>
+ </datafield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(DE-599)BSZ457035137</subfield>
+ </datafield>
+ <datafield tag="035" ind1=" " ind2=" ">
+ <subfield code="a">(OCoLC)946544758</subfield>
+ </datafield>
+ <datafield tag="040" ind1=" " ind2=" ">
+ <subfield code="a">DE-576</subfield>
+ <subfield code="b">ger</subfield>
+ <subfield code="c">DE-576</subfield>
+ <subfield code="e">rda</subfield>
+ </datafield>
+ <datafield tag="041" ind1=" " ind2=" ">
+ <subfield code="a">ger</subfield>
+ </datafield>
+ <datafield tag="245" ind1="1" ind2="4">
+ <subfield code="a">Der Fichtelberg</subfield>
+ <subfield code="b">Berg der unbekannten Rekorrde</subfield>
+ <subfield code="c">ein Film von Dirk Schneider</subfield>
+ </datafield>
+ <datafield tag="264" ind1=" " ind2="1">
+ <subfield code="a">[Leipzig]</subfield>
+ <subfield code="b">top ten tv</subfield>
+ <subfield code="c">[2014]</subfield>
+ </datafield>
+ <datafield tag="264" ind1=" " ind2="4">
+ <subfield code="c">© 2014</subfield>
+ </datafield>
+ <datafield tag="300" ind1=" " ind2=" ">
+ <subfield code="a">1 Online-Ressource (1 Videodatei, 29:49)</subfield>
+ <subfield code="b">farbig</subfield>
+ </datafield>
+ <datafield tag="336" ind1=" " ind2=" ">
+ <subfield code="a">zweidimensionales bewegtes Bild</subfield>
+ <subfield code="b">tdi</subfield>
+ <subfield code="2">rdacontent</subfield>
+ </datafield>
+ <datafield tag="337" ind1=" " ind2=" ">
+ <subfield code="a">Computermedien</subfield>
+ <subfield code="b">c</subfield>
+ <subfield code="2">rdamedia</subfield>
+ </datafield>
+ <datafield tag="338" ind1=" " ind2=" ">
+ <subfield code="a">Online-Ressource</subfield>
+ <subfield code="b">cr</subfield>
+ <subfield code="2">rdacarrier</subfield>
+ </datafield>
+ <datafield tag="490" ind1="0" ind2=" ">
+ <subfield code="a">MDR</subfield>
+ </datafield>
+ <datafield tag="490" ind1="0" ind2=" ">
+ <subfield code="a">Der Osten - entdecke wo du lebst</subfield>
+ </datafield>
+ <datafield tag="500" ind1=" " ind2=" ">
+ <subfield code="a">Dokumentarfilm. Deutschland. 2014</subfield>
+ </datafield>
+ <datafield tag="591" ind1=" " ind2=" ">
+ <subfield code="a">Fernsehmitschnitt (SWB)</subfield>
+ </datafield>
+ <datafield tag="655" ind1=" " ind2="7">
+ <subfield code="a">Film</subfield>
+ <subfield code="0">(DE-588)4017102-4</subfield>
+ <subfield code="0">(DE-627)104559683</subfield>
+ <subfield code="0">(DE-576)208918531</subfield>
+ <subfield code="2">gnd-content</subfield>
+ </datafield>
+ <datafield tag="700" ind1="1" ind2=" ">
+ <subfield code="a">Schneider, Dirk</subfield>
+ <subfield code="e">FilmemacherIn</subfield>
+ <subfield code="0">(DE-627)1235502279</subfield>
+ <subfield code="0">(DE-576)165502274</subfield>
+ <subfield code="4">fmk</subfield>
+ </datafield>
+ <datafield tag="935" ind1=" " ind2=" ">
+ <subfield code="c">vide</subfield>
+ </datafield>
+ <datafield tag="937" ind1=" " ind2=" ">
+ <subfield code="a">Dokumentarfilm</subfield>
+ <subfield code="b">Deutschland</subfield>
+ <subfield code="c">2014</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">000 xxxxxcx a22 zn 4500</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">001 901795887</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">003 DE-576</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">004 457035137</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">005 20160510125331</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">008 160304||||||||||||||||ger|||||||</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">040 </subfield>
+ <subfield code="a">DE-14</subfield>
+ <subfield code="c">DE-576</subfield>
+ <subfield code="d">DE-14</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">852 </subfield>
+ <subfield code="z">Fernsehmitschnitt: MDR, 04.02.2014. - Beilage</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">852 </subfield>
+ <subfield code="a">DE-14</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">852 1</subfield>
+ <subfield code="9">00</subfield>
+ </datafield>
+ <datafield tag="LOK" ind1=" " ind2=" ">
+ <subfield code="0">866 </subfield>
+ <subfield code="x">ddsu,pn</subfield>
+ </datafield>
+</record>
+PATCHED_MARCBLOB
+my $unpatched_marcblob_obj = XML::LibXML->load_xml(string => $unpatched_marcblob);
+
+### tests
+BEGIN { use_ok("SLUB::LZA::SIPBuilder"); }
+like(SLUB::LZA::SIPBuilder::get_mods_from("https://sru.bsz-bw.de/swb", "457035137", "pica.swn", "marcxmlvbos"), qr//, "get_mods_from()");
+is(SLUB::LZA::SIPBuilder::patch_mods($unpatched_mods_obj), $patched_mods, "patch_mods()" );
+is(SLUB::LZA::SIPBuilder::patch_marc_response($unpatched_marcblob_obj), $patched_marcblob, "patch_marc_response()");
+# ensure no dir exists, then run test
+my $xsl_path = path(__FILE__)->parent->parent->child('xsl');
+if ($xsl_path->is_dir) { $xsl_path->remove_tree; }
+is(SLUB::LZA::SIPBuilder::check_xsl_directory(), $xsl_path->absolute, "check_xsl_directory(), return value if not exist");
+ok($xsl_path->is_dir, "check_xsl_directory(), created if not exist");
+is(SLUB::LZA::SIPBuilder::check_xsl_directory(), $xsl_path->absolute, "check_xsl_directory(), return value if exist");
+ok($xsl_path->is_dir, "check_xsl_directory(), untouched if exist");
+
+1;
diff --git a/xsd/archive.xsd b/xsd/archive.xsd
index 4bc783c9dc25239bbc43f6d0aa786202571eb7b3..15d034a11877d937a19c3d5eea17397bfecf0d07 100644
--- a/xsd/archive.xsd
+++ b/xsd/archive.xsd
@@ -1,25 +1,30 @@
<?xml version="1.0" encoding="UTF-8"?>
-<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" elementFormDefault="qualified" targetNamespace="http://slub-dresden.de" xmlns:archive="http://slub-dresden.de" xmlns:v3="http://www.loc.gov/mods/v3" xmlns:mets="http://www.loc.gov/METS/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink">
- <xs:import namespace="http://www.loc.gov/METS/" schemaLocation="sip.xsd"/>
- <xs:import namespace="http://www.loc.gov/mods/v3" schemaLocation="v3.xsd"/>
- <xs:import namespace="http://www.w3.org/1999/xlink" schemaLocation="xlink.xsd"/>
- <xs:import namespace="http://www.w3.org/2001/XMLSchema-instance" schemaLocation="xsi.xsd"/>
- <xs:element name="record">
- <xs:complexType>
- <xs:sequence>
- <xs:element ref="archive:exportToArchiveDate"/>
- <xs:element ref="archive:externalId"/>
- <xs:element ref="archive:externalWorkflow"/>
- <xs:element ref="archive:hasConservationReason"/>
- <xs:element ref="archive:externalIsilId"/>
- <xs:element ref="archive:archivalValueDescription"/>
- </xs:sequence>
- </xs:complexType>
- </xs:element>
- <xs:element name="exportToArchiveDate" type="xs:dateTime"/>
- <xs:element name="externalId" type="xs:string"/>
- <xs:element name="externalWorkflow" type="xs:string"/>
- <xs:element name="hasConservationReason" type="xs:boolean"/>
- <xs:element name="externalIsilId" type="xs:string"/>
- <xs:element name="archivalValueDescription" type="xs:string"/>
+<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" elementFormDefault="qualified" attributeFormDefault="unqualified" targetNamespace="http://slub-dresden.de/slubarchiv" xmlns:archive="http://slub-dresden.de/slubarchiv">
+ <xs:element name="record" type="archive:record"/>
+ <xs:complexType name="record">
+ <xs:all>
+ <xs:element name="archivalValueDescription" type="archive:NonEmptyString" minOccurs="1" maxOccurs="1"/>
+ <xs:element name="exportToArchiveDate" type="xs:dateTime" minOccurs="1" maxOccurs="1"/>
+ <xs:element name="externalId" type="archive:LzaIdSubstring" minOccurs="1" maxOccurs="1"/>
+ <xs:element name="externalIsilId" type="archive:NonEmptyString" minOccurs="0" maxOccurs="1"/>
+ <xs:element name="externalWorkflow" type="archive:LzaIdSubstring" minOccurs="1" maxOccurs="1"/>
+ <xs:element name="hasConservationReason" type="xs:boolean" minOccurs="1" maxOccurs="1"/>
+ </xs:all>
+ <xs:attribute name="version" type="archive:SipVersionString" use="required"/>
+ </xs:complexType>
+ <xs:simpleType name="NonEmptyString">
+ <xs:restriction base="xs:string">
+ <xs:pattern value="[\s\S]*[^ ][\s\S]*"/>
+ </xs:restriction>
+ </xs:simpleType>
+ <xs:simpleType name="LzaIdSubstring">
+ <xs:restriction base="xs:string">
+ <xs:pattern value="[a-z0-9]+"/>
+ </xs:restriction>
+ </xs:simpleType>
+ <xs:simpleType name="SipVersionString">
+ <xs:restriction base="xs:string">
+ <xs:enumeration value="v2017.1"/>
+ </xs:restriction>
+ </xs:simpleType>
</xs:schema>