From 5d83f2360a1162052c4fc35e4c6d4c4c82133d0a Mon Sep 17 00:00:00 2001
From: mromanello <matteo.romanello@gmail.com>
Date: Mon, 30 Aug 2010 10:51:12 +0100
Subject: [PATCH] Initialized repo and imported files

---
 biblio_script.sh |  53 +++++++++++++++
 parscit2mods.xsl | 174 +++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 227 insertions(+)
 create mode 100755 biblio_script.sh
 create mode 100644 parscit2mods.xsl

diff --git a/biblio_script.sh b/biblio_script.sh
new file mode 100755
index 0000000..9eeda96
--- /dev/null
+++ b/biblio_script.sh
@@ -0,0 +1,53 @@
+#!/usr/bin/env python
+# Author: Matteo Romanello, <matteo.romanello@gmail.com>
+
+import os,sys
+
+# paths to executables 
+PARSCIT_PATH="/Applications/ParsCit/bin/"
+BIBUTILS_PATH="/Applications/bibutils_4.8/"
+SAXON_PATH="/56k/Downloads/saxonhe9-2-1-2j/saxon9he.jar"
+
+# paths to resources
+XSLT_TRANFORM_PATH="/56k/Documents/Research/training_parscit/xslt/parscit2mods.xsl"
+
+def parscit_to_mods(parscit_out):
+	saxon_cmd="java -jar %s -xsl:%s -s:%s" %(SAXON_PATH,XSLT_TRANFORM_PATH,parscit_out)
+	out=os.popen(saxon_cmd).readlines()
+	print "Transforming Parscit's output into mods xml..."
+	return out
+	
+def mods_to_bibtex(mods_xml):
+	bibutils_cmd="%sxml2bib %s"%(BIBUTILS_PATH,mods_xml)
+	out=os.popen(bibutils_cmd).readlines()
+	return out
+
+if(len(sys.argv)>1):
+	inp_file=sys.argv[1] #I should check that this file exists
+	out_dir=sys.argv[2] #I should check that this directory exists
+	print "Extracting references from the input file..."
+	parscit_out = os.popen("%sparseRefStrings.pl %s" %(PARSCIT_PATH,inp_file)).readlines()
+	parscit_xml='%sparscit_temp.xml'%out_dir
+	file = open(parscit_xml,'w')
+	for line in parscit_out:
+		file.write(line)
+	file.close()
+	
+	# transform parscit's output into mods 3.x
+	parscit_mods='%sparscit_mods.xml'%out_dir
+	file = open(parscit_mods,'w')
+	for line in parscit_to_mods(parscit_xml):
+		file.write(line)
+	file.close()
+	
+	# transform mods intermediate xml into bibtex
+	parscit_bibtex='%sparscit.bib'%out_dir
+	print "Transforming intermediate mods xml into Bibtex..."
+	file = open(parscit_bibtex,'w')
+	for line in mods_to_bibtex(parscit_mods):
+		file.write(line)
+	file.close()
+		
+	
+else:
+	print"Usage: <inputFile> <outDir>"
\ No newline at end of file
diff --git a/parscit2mods.xsl b/parscit2mods.xsl
new file mode 100644
index 0000000..7101de8
--- /dev/null
+++ b/parscit2mods.xsl
@@ -0,0 +1,174 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="2.0" xmlns:mods="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-3-2.xsd"
+    xmlns:xd="http://www.pnp-software.com/XSLTdoc">
+    <xd:author>Matteo Romanello</xd:author>
+    <xsl:output method="xml" indent="yes"/>
+    <xsl:template match="citationList">
+        <xsl:element name="modsCollection" namespace="http://www.loc.gov/mods/v3">
+            <xsl:attribute name="xsi:schemaLocation" namespace="http://www.w3.org/2001/XMLSchema-instance">http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-2.xsd</xsl:attribute>
+            <xsl:comment> ### JOURNAL ARTICLES ### </xsl:comment>
+            <xsl:apply-templates select="citation[journal]"/>
+            <xsl:comment> ### BOOKS ### </xsl:comment>
+            <xsl:apply-templates select="citation[title and not(booktitle) and not(pages) and not(journal)]"/>
+        </xsl:element>
+    </xsl:template>
+    <xd:doc> Here is where most of the inetersting stuff happen. </xd:doc>
+    <xsl:template match="citation">
+        <xsl:element name="mods" namespace="http://www.loc.gov/mods/v3">
+            <xsl:attribute name="ID">
+                <!-- add compatibility check for saxon or not -->
+                <xsl:value-of select="generate-id()"/>
+            </xsl:attribute>
+            <xsl:apply-templates select="title"/>
+            <!-- heuristic to determine the kind of resource -->
+            <xsl:choose>
+                <!-- CASE 1: Paper in a Journal-->
+                <xsl:when test="./journal">
+                    <xsl:apply-templates select="authors" mode="journal_article"/>
+                    <xsl:element name="relatedItem" namespace="http://www.loc.gov/mods/v3">
+                        <xsl:attribute name="type">host</xsl:attribute>
+                        <xsl:apply-templates select="journal" mode="journal_article"/>
+                        <xsl:element name="originInfo" namespace="http://www.loc.gov/mods/v3">
+                            <xsl:element name="issuance" namespace="http://www.loc.gov/mods/v3">continuing</xsl:element>
+                        </xsl:element>
+                        <xsl:element name="part" namespace="http://www.loc.gov/mods/v3">
+                            <xsl:apply-templates select="*[name()!='authors'][name()!='journal']" mode="journal_article"/>
+                        </xsl:element>
+                        <xsl:element name="genre" namespace="http://www.loc.gov/mods/v3">
+                            <xsl:attribute name="authority">marc</xsl:attribute>
+                            <xsl:text>journal</xsl:text>
+                        </xsl:element>
+                        <xsl:element name="genre" namespace="http://www.loc.gov/mods/v3">
+                            <xsl:text>academic journal</xsl:text>
+                        </xsl:element>
+                    </xsl:element>
+                </xsl:when>
+                <!-- CASE 2: Book -->
+                <xsl:when test=".[title and not(booktitle) and not(pages) and not(journal)]">
+                    <xsl:apply-templates select="authors" mode="book"/>
+                    <xsl:element name="originInfo" namespace="http://www.loc.gov/mods/v3">
+                        <xsl:apply-templates select="location | date | publisher" mode="book"/>
+                        <xsl:element name="issuance" namespace="http://www.loc.gov/mods/v3">monographic</xsl:element>
+                    </xsl:element>
+                </xsl:when>
+                <xsl:otherwise>
+                    <xsl:apply-templates select="*[name()!='authors']"/>
+                </xsl:otherwise>
+            </xsl:choose>
+            <!--<xsl:element name="typeOfResource" namespace="http://www.loc.gov/mods/v3">text</xsl:element>-->
+            <xsl:element name="identifier" namespace="http://www.loc.gov/mods/v3">
+                <xsl:attribute name="type">citekey</xsl:attribute>
+                <xsl:value-of select="generate-id()"/>
+            </xsl:element>
+        </xsl:element>
+    </xsl:template>
+    <xd:doc>Auhtors of a journal article</xd:doc>
+    <xsl:template match="authors" mode="journal_article">
+        <xsl:apply-templates select="author">
+            <xsl:with-param name="mode">journal_article</xsl:with-param>
+        </xsl:apply-templates>
+    </xsl:template>
+    <xd:doc>Auhtors of a book</xd:doc>
+    <xsl:template match="authors" mode="book">
+        <xsl:apply-templates select="author">
+            <xsl:with-param name="mode">book</xsl:with-param>
+        </xsl:apply-templates>
+    </xsl:template>
+    <xd:doc>Handles the creation of name elements in mods format. <xd:param type="string">The current mode.</xd:param>
+    </xd:doc>
+    <xsl:template match="author">
+        <xsl:param name="mode"/>
+        <xsl:element name="name" namespace="http://www.loc.gov/mods/v3">
+            <xsl:attribute name="type">personal</xsl:attribute>
+            <xsl:for-each select="tokenize(.,' ')">
+                <xsl:element name="namePart" namespace="http://www.loc.gov/mods/v3">
+                    <xsl:choose>
+                        <xsl:when test="string-length(.)=1">
+                            <xsl:attribute name="type">given</xsl:attribute>
+                        </xsl:when>
+                        <xsl:otherwise>
+                            <xsl:attribute name="type">family</xsl:attribute>
+                        </xsl:otherwise>
+                    </xsl:choose>
+                    <xsl:value-of select="."/>
+                </xsl:element>
+            </xsl:for-each>
+            <xsl:element name="role" namespace="http://www.loc.gov/mods/v3">
+                <xsl:element name="roleTerm" namespace="http://www.loc.gov/mods/v3">
+                    <xsl:attribute name="authority">marcrelator</xsl:attribute>
+                    <xsl:attribute name="type">text</xsl:attribute>
+                    <xsl:choose>
+                        <xsl:when test="$mode='journal_article'">
+                            <xsl:text>author</xsl:text>
+                        </xsl:when>
+                        <xsl:when test="$mode='book'">
+                            <xsl:text>creator</xsl:text>
+                        </xsl:when>
+                    </xsl:choose>
+                </xsl:element>
+            </xsl:element>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template name="title">
+        <xsl:element name="titleInfo" namespace="http://www.loc.gov/mods/v3">
+            <xsl:element name="title" namespace="http://www.loc.gov/mods/v3">
+                <xsl:value-of select="."/>
+            </xsl:element>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template match="title">
+        <xsl:call-template name="title"/>
+    </xsl:template>
+    <xsl:template match="journal" mode="journal_article">
+        <xsl:call-template name="title"/>
+    </xsl:template>
+    <xsl:template match="pages" mode="journal_article">
+        <xsl:element name="extent" namespace="http://www.loc.gov/mods/v3">
+            <xsl:attribute name="unit">page</xsl:attribute>
+            <xsl:element name="start" namespace="http://www.loc.gov/mods/v3">
+                <xsl:value-of select="tokenize(.,'--')[1]"/>
+            </xsl:element>
+            <xsl:element name="end" namespace="http://www.loc.gov/mods/v3">
+                <xsl:value-of select="tokenize(.,'--')[2]"/>
+            </xsl:element>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template match="volume" mode="journal_article">
+        <xsl:element name="detail" namespace="http://www.loc.gov/mods/v3">
+            <xsl:attribute name="type">volume</xsl:attribute>
+            <xsl:element name="number" namespace="http://www.loc.gov/mods/v3">
+                <xsl:value-of select="."/>
+            </xsl:element>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template mode="journal_article" match="date">
+        <xsl:element name="date" namespace="http://www.loc.gov/mods/v3">
+            <xsl:value-of select="."/>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template match="location" mode="book">
+        <xsl:element name="place" namespace="http://www.loc.gov/mods/v3">
+            <xsl:element name="placeTerm" namespace="http://www.loc.gov/mods/v3">
+                <xsl:attribute name="type">text</xsl:attribute>
+                <xsl:value-of select="."/>
+            </xsl:element>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template match="date" mode="book">
+        <xsl:element name="dateIssued" namespace="http://www.loc.gov/mods/v3">
+            <xsl:value-of select="."/>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template match="publisher" mode="book">
+        <xsl:element name="publisher" namespace="http://www.loc.gov/mods/v3">
+            <xsl:value-of select="."/>
+        </xsl:element>
+    </xsl:template>
+    <xsl:template match="notes" mode="journal_article">
+        <xsl:comment>
+            <xsl:value-of select="."/>
+        </xsl:comment>
+    </xsl:template>
+    <xsl:template mode="journal_article" match="location"/>
+    <xsl:template mode="journal_article" match="title"/>
+</xsl:stylesheet>