tutorial

ODIN Union

# product of ODIN codesprint at CERN, 15-16 2013
# creators: Stefano Bortoli, Jan Dvorak, Alejandra Gonzalez-Beltran & Todd Vision

# Narrative description of steps current as of 15 Oct 2013 17:50, w/ code samples

# Step 1. Getting the data from DataCite
# http://search.datacite.org/api?q=nameIdentifier%3AORCID\%3A*&fl=doi,nameIdentifier,xml&fq=is_active:true&fq=has_metadata:true&wt=csv&rows=100000 
# in this directory as datacite1.csv
# the third column contains base64-encoded XML

#recover that XML:
mkdir data1
sed -e 1d -e 's/.*,//g' datacite1.csv | while read X ; do XX=`echo $X | shasum | awk '{print $1}'` ; echo $X | base64 -D >data1/$XX.xml ; echo $XX ; done

# Step 2. Convert DataCite XML to csv w/ custom xlst script
# process it to extract creators/contributors with ORCIDs
# runs isolate.xslt on all the files from data1/, making *.txt files in data2/
mkdir data2
ant

# build.xml is:
<project name="ODIN codesprint DataCite ORCID harmonization" default="xml-to-text-1">
  <target name="xml-to-text-1">
    <xslt style="isolate.xslt" basedir="data1" destdir="data2" extension=".txt"/>
  </target>
</project>

# isolate.xslt
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet version="2.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
    <xsl:output method="text" encoding="UTF-8"/>
    <xsl:template match="/">
        <xsl:for-each select="*[local-name()='resource']/*[local-name()='creators']/*[local-name()='creator' and *[local-name()='nameIdentifier']/@nameIdentifierScheme='ORCID']">
            <xsl:value-of select="*[local-name()='nameIdentifier' and @nameIdentifierScheme='ORCID']"/>
            <xsl:text>; </xsl:text>
            <xsl:value-of select="*[local-name()='creatorName']"/>
            <xsl:text>; </xsl:text>
            <xsl:value-of select="/*[local-name()='resource']/*[local-name()='identifier' and @identifierType='DOI']"/>
            <xsl:text>
</xsl:text>
        </xsl:for-each>
        <xsl:for-each select="*[local-name()='resource']/*[local-name()='contributors']/*[local-name()='contributor' and *[local-name()='nameIdentifier']/@nameIdentifierScheme='ORCID']">
            <xsl:value-of select="*[local-name()='nameIdentifier' and @nameIdentifierScheme='ORCID']"/>
            <xsl:text>; </xsl:text>
            <xsl:value-of select="*[local-name()='contributorName']"/>
            <xsl:text>; </xsl:text>
            <xsl:value-of select="/*[local-name()='resource']/*[local-name()='identifier' and @identifierType='DOI']"/>
            <xsl:text>
</xsl:text>
        </xsl:for-each>
    </xsl:template>

</xsl:stylesheet>

# step 3 Merge the files
cd data2
ls | xargs -n 1000 cat | sort >../data2.csv
# note that rwo lines at the beginning and several lines at the end of the file are junk row due to messy data

# Step 4. Extract creators, contributors, and any associated ORCIDs for them, with OpenRefine
# see attached OpenRefine files

# Step 5. Download CrossRef records with ORCIDs
# Search for specific ORCID, retrieving DOIs with any relationship to the ORCID given:
# http://search.crossref.org/dois?q=xxxx-xxxx-xxxx-xxxx (substitute specific ORCID for x)

# Step 6. Retrieve XML for each CrossRef DOI 
# No API key required, but queries may be throttled
# e.g. http://api.labs.crossref.org/doi:10.1890/12-2231.1.xml

# Step 7. Getting Data dump from ORCID
# URL masked