From d1cd162f5776904e8972d2cfdf4b0d62afadd458 Mon Sep 17 00:00:00 2001
From: Fred Atherden <43879983+fred-atherden@users.noreply.github.com>
Date: Mon, 25 Nov 2024 15:29:39 +0000
Subject: [PATCH] RP XSL - add template to fix source elements

- Change to title case when source is in all caps
- If source starts with 'In ', then move this outside of source
---
 src/preprint-changes.xsl                      | 73 ++++++++++++++++++-
 test/tests/preprint-changes/all/input.xml     |  2 +
 test/tests/preprint-changes/all/output.xml    |  2 +
 .../ref-source-fixes/input.xml                |  8 ++
 .../ref-source-fixes/output.xml               |  8 ++
 test/xspec/preprint-changes.xspec             |  4 +
 6 files changed, 96 insertions(+), 1 deletion(-)
 create mode 100644 test/tests/preprint-changes/ref-source-fixes/input.xml
 create mode 100644 test/tests/preprint-changes/ref-source-fixes/output.xml

diff --git a/src/preprint-changes.xsl b/src/preprint-changes.xsl
index 09160594a7..f1d4e56d00 100644
--- a/src/preprint-changes.xsl
+++ b/src/preprint-changes.xsl
@@ -47,6 +47,40 @@
         <xsl:param name="string" as="xs:string"/>
         <xsl:value-of select="replace(replace(replace(translate(normalize-unicode($string,'NFD'),'ƀȼđɇǥħɨıɉꝁłøɍŧɏƶ','bcdeghiijklortyz'),'[\p{M}’]',''),'æ','ae'),'ß','ss')"/>
     </xsl:function>
+    
+    <xsl:function name="e:toTitleCase" as="xs:string">
+        <xsl:param name="input" as="xs:string"/>
+        <xsl:variable name="upper-exceptions" select="('PLOS','OSF','JOSS','JAMA','EMBO','RNA','BMJ','PRIDE','CRAN','EMDB','SSRN','EMPIAR','NEJM','FEMS','JMLR','APMIS','ISME','FASEB','PNAS','JBO','JBJS','AAPS','NIPS','BJOG','ISMRM')"/>
+        <xsl:variable name="case-exceptions" select="('eLife','iScience','eNeuro','arXiv','bioRxiv','medRxiv','ChemRxiv','PeerJ','PsyArXiv','PaleorXiv','AfricArXiv','EcoEvoRxiv')"/>
+        <xsl:variable name="lower-exceptions" select="('and', 'or', 'of', 'the', 'in', 'on', 'with', 'a', 'an')"/>
+        <xsl:variable name="words" select="tokenize($input,'\s+')" />
+        <xsl:variable name="processed-words">
+            <xsl:for-each select="$words">
+                <xsl:variable name="val" select="."/>
+                <xsl:choose>
+                    <xsl:when test="position()=1 and not(lower-case(.)=$case-exceptions) and not(upper-case(.)=$upper-exceptions)">
+                        <xsl:value-of select="concat(upper-case(substring(.,1,1)),lower-case(substring(.,2)))"/>
+                    </xsl:when>
+                    <xsl:when test="upper-case(.)=$upper-exceptions">
+                        <xsl:value-of select="upper-case(.)"/>
+                    </xsl:when>
+                    <xsl:when test="some $example in $case-exceptions satisfies (lower-case($example)=lower-case($val))">
+                        <xsl:value-of select="$case-exceptions[lower-case(.)=lower-case($val)]"/>
+                    </xsl:when>
+                    <xsl:when test="lower-case(.)=($lower-exceptions)">
+                        <xsl:value-of select="lower-case(.)"/>
+                    </xsl:when>
+                    <xsl:when test="string-length(.)=1">
+                        <xsl:value-of select="upper-case(.)"/>
+                    </xsl:when>
+                    <xsl:otherwise>
+                        <xsl:value-of select="concat(upper-case(substring(.,1,1)),lower-case(substring(.,2)))"/>
+                    </xsl:otherwise>
+                </xsl:choose>
+            </xsl:for-each>
+        </xsl:variable>
+         <xsl:sequence select="string-join($processed-words,' ')" />
+    </xsl:function>
 
      <xsl:template match="*|@*|text()|comment()|processing-instruction()">
         <xsl:copy>
@@ -799,7 +833,11 @@
         <xsl:variable name="mixed-citation-round-3">
             <xsl:apply-templates select="$mixed-citation-round-2" mode="mixed-citation-round-3"/>
          </xsl:variable>
-        <xsl:apply-templates select="$mixed-citation-round-3" mode="mixed-citation-round-4"/>
+         <xsl:variable name="mixed-citation-round-4">
+            <xsl:apply-templates select="$mixed-citation-round-3" mode="mixed-citation-round-4"/>
+         </xsl:variable>
+        <xsl:apply-templates select="$mixed-citation-round-4" mode="mixed-citation-round-5"/>
+
     </xsl:template>
 
     <!-- Introduces author person-groups into refs when they are missing-->
@@ -943,6 +981,39 @@
          </xsl:copy>
      </xsl:template>
     
+    <!-- Fixes mistagged and all caps source elements -->
+     <xsl:template xml:id="ref-source-fixes" mode="mixed-citation-round-4" match="mixed-citation//source">
+         <xsl:choose>
+                <xsl:when test="matches(lower-case(.),'^in\s')">
+                    <xsl:choose>
+                        <xsl:when test="(ancestor::mixed-citation/@publication-type=('preprint','journal','other')) and upper-case(.)=.">
+                            <xsl:text>In </xsl:text>
+                            <xsl:copy>
+                                <xsl:value-of select="e:toTitleCase(substring-after(.,' '))"/>
+                            </xsl:copy>
+                        </xsl:when>
+                        <xsl:otherwise>
+                            <xsl:text>In </xsl:text>
+                            <xsl:copy>
+                                <xsl:value-of select="substring-after(.,' ')"/>
+                            </xsl:copy>
+                        </xsl:otherwise>
+                    </xsl:choose>
+                </xsl:when>
+                <xsl:when test="(ancestor::mixed-citation/@publication-type=('preprint','journal','other')) and upper-case(.)=.">
+                    <xsl:copy>
+                        <xsl:apply-templates select="@*"/>
+                        <xsl:value-of select="e:toTitleCase(.)"/>
+                    </xsl:copy>
+                </xsl:when>
+                <xsl:otherwise>
+                    <xsl:copy>
+                        <xsl:apply-templates select="*|@*|text()|comment()|processing-instruction()"/>
+                    </xsl:copy>
+                </xsl:otherwise>
+            </xsl:choose>
+     </xsl:template>
+    
     
     <!-- Add a CC-BY biorender statement for CC0 RPs -->
     <xsl:template xml:id="biorender-permissions" match="article[front//permissions/license/@xlink:href[contains(lower-case(.),'creativecommons.org/publicdomain/zero/')]]//*[caption[contains(lower-case(.),'biorender')]]">
diff --git a/test/tests/preprint-changes/all/input.xml b/test/tests/preprint-changes/all/input.xml
index c682f441d4..e24a944f95 100644
--- a/test/tests/preprint-changes/all/input.xml
+++ b/test/tests/preprint-changes/all/input.xml
@@ -113,6 +113,8 @@
 <ref id="c4"><label>4.</label><mixed-citation publication-type="book"><string-name><given-names>A.</given-names> <surname>Nagy</surname></string-name>, <string-name><given-names>M.</given-names> <surname>Gertsenstein</surname></string-name>, <string-name><given-names>K.</given-names> <surname>Vintersten</surname></string-name>, … and <string-name><given-names>R.</given-names> <surname>Behringer</surname></string-name>, <chapter-title>Manipulating the mouse embryo: a laboratory manual</chapter-title> (<publisher-name>Cold Spring Harbor Laboratory Press Cold Spring Harbor</publisher-name>, NY:, <year>2003</year>), vol. <volume>10</volume>.</mixed-citation></ref>
 <ref id="c5"><label>5.</label><mixed-citation publication-type="book"><string-name><given-names>J. M.</given-names> <surname>Glynn</surname></string-name>, <string-name><given-names>T. G.</given-names> <surname>Cotter</surname></string-name>, <string-name><given-names>D. R.</given-names> <surname>Green</surname></string-name> (<year>1992</year>) <article-title>Apoptosis induced by actinomycin D, camptothecin or aphidicolin can occur in all phases of the cell cycle</article-title>. (<publisher-loc>Portland</publisher-loc> <publisher-name>Press Ltd</publisher-name>.).</mixed-citation></ref>
 <ref id="c6"><label>6.</label><mixed-citation publication-type="journal"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>Unveiling the sensory and interneuronal pathways of the neuroendocrine connectome in Drosophila</article-title>. <source>Elife</source> <volume>10</volume>:<fpage>e65745</fpage></mixed-citation></ref>
+<ref id="c7"><label>7.</label><mixed-citation publication-type="journal"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>Unveiling the sensory and interneuronal pathways of the neuroendocrine connectome in Drosophila</article-title>. <source>ELIFE</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
+<ref id="c8"><label>8.</label><mixed-citation publication-type="journal"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>..</article-title>. <source>PLANT PHYSIOLOGY</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
 </ref-list>
 </back>
 </article>
\ No newline at end of file
diff --git a/test/tests/preprint-changes/all/output.xml b/test/tests/preprint-changes/all/output.xml
index 72e7bcb480..bd8dea7f39 100644
--- a/test/tests/preprint-changes/all/output.xml
+++ b/test/tests/preprint-changes/all/output.xml
@@ -125,6 +125,8 @@
 <ref id="c4"><label>4.</label><mixed-citation publication-type="book"><person-group person-group-type="author"><string-name><given-names>A.</given-names> <surname>Nagy</surname></string-name>, <string-name><given-names>M.</given-names> <surname>Gertsenstein</surname></string-name>, <string-name><given-names>K.</given-names> <surname>Vintersten</surname></string-name>, <etal>…</etal> <string-name><given-names>R.</given-names> <surname>Behringer</surname></string-name></person-group>, <source>Manipulating the mouse embryo: a laboratory manual</source> (<publisher-name>Cold Spring Harbor Laboratory Press Cold Spring Harbor</publisher-name>, NY:, <year>2003</year>), vol. <volume>10</volume>.</mixed-citation></ref>
 <ref id="c5"><label>5.</label><mixed-citation publication-type="book"><person-group person-group-type="author"><string-name><given-names>J. M.</given-names> <surname>Glynn</surname></string-name>, <string-name><given-names>T. G.</given-names> <surname>Cotter</surname></string-name>, <string-name><given-names>D. R.</given-names> <surname>Green</surname></string-name></person-group> (<year>1992</year>) <source>Apoptosis induced by actinomycin D, camptothecin or aphidicolin can occur in all phases of the cell cycle</source>. (<publisher-loc>Portland</publisher-loc> <publisher-name>Press Ltd</publisher-name>.).</mixed-citation></ref>
 <ref id="c6"><label>6.</label><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>Unveiling the sensory and interneuronal pathways of the neuroendocrine connectome in Drosophila</article-title>. <source>Elife</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
+<ref id="c7"><label>7.</label><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>Unveiling the sensory and interneuronal pathways of the neuroendocrine connectome in Drosophila</article-title>. <source>eLife</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
+<ref id="c8"><label>8.</label><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>..</article-title>. <source>Plant Physiology</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
 </ref-list>
 </back>
 </article>
diff --git a/test/tests/preprint-changes/ref-source-fixes/input.xml b/test/tests/preprint-changes/ref-source-fixes/input.xml
new file mode 100644
index 0000000000..75c912690f
--- /dev/null
+++ b/test/tests/preprint-changes/ref-source-fixes/input.xml
@@ -0,0 +1,8 @@
+<!--Testing template with id: ref-source-fixes-->
+<ref-list>
+<ref id="c4"><label>4.</label><mixed-citation publication-type="other"><string-name><surname>Claus</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>..</article-title>. <source>AGAINST METHOD</source></mixed-citation></ref>
+<ref id="c5"><label>5.</label><mixed-citation publication-type="journal"><string-name><surname>Claus</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>..</article-title>. <source>PLOS ONE</source></mixed-citation></ref>
+<ref id="c6"><label>6.</label><mixed-citation publication-type="preprint"><string-name><surname>Claus</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>..</article-title>. <source>BIORXIV</source></mixed-citation></ref>
+<ref id="c7"><label>7.</label><mixed-citation publication-type="journal"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>Unveiling the sensory and interneuronal pathways of the neuroendocrine connectome in Drosophila</article-title>. <source>ELIFE</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
+<ref id="c8"><label>8.</label><mixed-citation publication-type="journal"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal> (<year>2021</year>) <article-title>..</article-title>. <source>PLANT PHYSIOLOGY</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
+</ref-list>
\ No newline at end of file
diff --git a/test/tests/preprint-changes/ref-source-fixes/output.xml b/test/tests/preprint-changes/ref-source-fixes/output.xml
new file mode 100644
index 0000000000..b772e84796
--- /dev/null
+++ b/test/tests/preprint-changes/ref-source-fixes/output.xml
@@ -0,0 +1,8 @@
+<!--Testing template with id: ref-source-fixes-->
+<ref-list>
+<ref id="c4"><label>4.</label><mixed-citation publication-type="other"><person-group person-group-type="author"><string-name><surname>Claus</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>..</article-title>. <source>Against Method</source></mixed-citation></ref>
+<ref id="c5"><label>5.</label><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Claus</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>..</article-title>. <source>PLOS One</source></mixed-citation></ref>
+<ref id="c6"><label>6.</label><mixed-citation publication-type="preprint"><person-group person-group-type="author"><string-name><surname>Claus</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>..</article-title>. <source>bioRxiv</source></mixed-citation></ref>
+<ref id="c7"><label>7.</label><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>Unveiling the sensory and interneuronal pathways of the neuroendocrine connectome in Drosophila</article-title>. <source>eLife</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
+<ref id="c8"><label>8.</label><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Huckesfeld</surname> <given-names>S</given-names></string-name>, <etal>et al</etal></person-group> (<year>2021</year>) <article-title>..</article-title>. <source>Plant Physiology</source> <volume>10</volume>:<elocation-id>e65745</elocation-id></mixed-citation></ref>
+</ref-list>
\ No newline at end of file
diff --git a/test/xspec/preprint-changes.xspec b/test/xspec/preprint-changes.xspec
index 98082f6b80..8571de6e08 100644
--- a/test/xspec/preprint-changes.xspec
+++ b/test/xspec/preprint-changes.xspec
@@ -96,6 +96,10 @@
       <x:context href="../tests/preprint-changes/find-and-tag-preprint-refs/input.xml"/>
       <x:expect label="Testing the template: find-and-tag-preprint-refs" href="../tests/preprint-changes/find-and-tag-preprint-refs/output.xml"/>
     </x:scenario>
+    <x:scenario label="ref-source-fixes">
+      <x:context href="../tests/preprint-changes/ref-source-fixes/input.xml"/>
+      <x:expect label="Testing the template: ref-source-fixes" href="../tests/preprint-changes/ref-source-fixes/output.xml"/>
+    </x:scenario>
     <x:scenario label="biorender-permissions">
       <x:context href="../tests/preprint-changes/biorender-permissions/input.xml"/>
       <x:expect label="Testing the template: biorender-permissions" href="../tests/preprint-changes/biorender-permissions/output.xml"/>