Merge branch 'main' of https://github.com/wibarab/corpus-data

wibarab · Nov 6, 2023 · 3f73d02 · 3f73d02
2 parents 6522608 + 24e96b0
commit 3f73d02
Show file tree

Hide file tree

Showing 93 changed files with 22,922 additions and 42,999 deletions.
diff --git a/080_scripts_generic/080_01_ELAN2TEI/ELAN2TEI.ipynb b/080_scripts_generic/080_01_ELAN2TEI/ELAN2TEI.ipynb
diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/params.xsl b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/params.xsl
@@ -1,7 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<xsl:stylesheet xmlns:xs="http://www.w3.org/2001/XMLSchema"
+<xsl:stylesheet xmlns:tei="http://www.tei-c.org/ns/1.0"
                 xmlns:xd="http://www.oxygenxml.com/ns/doc/xsl"
-                xmlns:tei="http://www.tei-c.org/ns/1.0"
+                xmlns:xs="http://www.w3.org/2001/XMLSchema"
                 xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                 version="2.0"
                 xml:id="params">

diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/postTokenization/1.xsl b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/postTokenization/1.xsl
@@ -1,8 +1,8 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<xsl:stylesheet xmlns:xs="http://www.w3.org/2001/XMLSchema"
-                xmlns:xtoks="http://acdh.oeaw.ac.at/xtoks"
+<xsl:stylesheet xmlns:tei="http://www.tei-c.org/ns/1.0"
+                xmlns:xs="http://www.w3.org/2001/XMLSchema"
                 xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
-                xmlns:tei="http://www.tei-c.org/ns/1.0"
+                xmlns:xtoks="http://acdh.oeaw.ac.at/xtoks"
                 exclude-result-prefixes="#all"
                 version="2.0"
                 xml:base="postTokenization.xsl">

diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/profile.xml b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/profile.xml
@@ -54,6 +54,12 @@
         <doc-attribute name="topic">
             <expression>//tei:textClass/tei:catRef/@target</expression>
         </doc-attribute>
+        <doc-attribute name="researcher">
+            <expression>//tei:recording//tei:persName/@ref</expression>
+        </doc-attribute>
+        <doc-attribute name="campaign">
+            <expression>//tei:recording/tei:p/tei:ref</expression>
+        </doc-attribute>
         <doc-attribute name="geo">
             <expression>//tei:settingDesc/tei:place</expression>
         </doc-attribute>

diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_addP.xsl b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_addP.xsl
@@ -1,7 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<xsl:stylesheet xmlns:xs="http://www.w3.org/2001/XMLSchema"
+<xsl:stylesheet xmlns:tei="http://www.tei-c.org/ns/1.0"
                 xmlns:xd="http://www.oxygenxml.com/ns/doc/xsl"
-                xmlns:tei="http://www.tei-c.org/ns/1.0"
+                xmlns:xs="http://www.w3.org/2001/XMLSchema"
                 xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                 version="2.0"
                 exclude-result-prefixes="#all">

diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_toks.xsl b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_toks.xsl
@@ -1,7 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<xsl:stylesheet xmlns:xs="http://www.w3.org/2001/XMLSchema"
+<xsl:stylesheet xmlns:tei="http://www.tei-c.org/ns/1.0"
                 xmlns:xd="http://www.oxygenxml.com/ns/doc/xsl"
-                xmlns:tei="http://www.tei-c.org/ns/1.0"
+                xmlns:xs="http://www.w3.org/2001/XMLSchema"
                 xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                 version="2.0"
                 exclude-result-prefixes="#all">

diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_vert2txt.xsl b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_vert2txt.xsl
@@ -1,9 +1,9 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<xsl:stylesheet xmlns:xs="http://www.w3.org/2001/XMLSchema"
+<xsl:stylesheet xmlns:tei="http://www.tei-c.org/ns/1.0"
                 xmlns:xd="http://www.oxygenxml.com/ns/doc/xsl"
-                xmlns:xtoks="http://acdh.oeaw.ac.at/xtoks"
-                xmlns:tei="http://www.tei-c.org/ns/1.0"
+                xmlns:xs="http://www.w3.org/2001/XMLSchema"
                 xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+                xmlns:xtoks="http://acdh.oeaw.ac.at/xtoks"
                 version="2.0"
                 exclude-result-prefixes="#all">
    <xsl:output method="text"/>

diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_xtoks2vert.xsl b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/wrapper_xtoks2vert.xsl
@@ -1,9 +1,9 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<xsl:stylesheet xmlns:xs="http://www.w3.org/2001/XMLSchema"
+<xsl:stylesheet xmlns:tei="http://www.tei-c.org/ns/1.0"
                 xmlns:xd="http://www.oxygenxml.com/ns/doc/xsl"
-                xmlns:xtoks="http://acdh.oeaw.ac.at/xtoks"
-                xmlns:tei="http://www.tei-c.org/ns/1.0"
+                xmlns:xs="http://www.w3.org/2001/XMLSchema"
                 xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+                xmlns:xtoks="http://acdh.oeaw.ac.at/xtoks"
                 version="2.0"
                 exclude-result-prefixes="#all">
    <xsl:include href="params.xsl"/>
@@ -79,6 +79,16 @@
                      name="topic"
                      select="string-join(//tei:textClass/tei:catRef/@target/normalize-space(), ' ')"/>
    </xsl:template>
+   <xsl:template match="//tei:recording//tei:persName/@ref" mode="doc-attributes">
+      <xsl:attribute namespace="http://acdh.oeaw.ac.at/xtoks"
+                     name="researcher"
+                     select="string-join(//tei:recording//tei:persName/@ref/normalize-space(), ' ')"/>
+   </xsl:template>
+   <xsl:template match="//tei:recording/tei:p/tei:ref" mode="doc-attributes">
+      <xsl:attribute namespace="http://acdh.oeaw.ac.at/xtoks"
+                     name="campaign"
+                     select="string-join(//tei:recording/tei:p/tei:ref/normalize-space(), ' ')"/>
+   </xsl:template>
    <xsl:template match="//tei:settingDesc/tei:place" mode="doc-attributes">
       <xsl:attribute namespace="http://acdh.oeaw.ac.at/xtoks"
                      name="geo"

diff --git a/080_scripts_generic/080_01_ELAN2TEI/tokenizer/xsl/make_xsl.xsl b/080_scripts_generic/080_01_ELAN2TEI/tokenizer/xsl/make_xsl.xsl
@@ -263,6 +263,7 @@
 
         <!-- Create the tokenizer wrapper stylesheet -->
         <xsl:result-document href="{$output-path}/wrapper_toks.xsl" indent="yes">
+            <xsl:message select="$output-path"></xsl:message>
             <xsl:element name="xsl:stylesheet" namespace="http://www.w3.org/1999/XSL/Transform">
                 <xsl:namespace name="xs">http://www.w3.org/2001/XMLSchema</xsl:namespace>
                 <xsl:namespace name="xd">http://www.oxygenxml.com/ns/doc/xsl</xsl:namespace>

diff --git a/082_scripts_xsl/postprocessTEICorpoOutput.xsl b/082_scripts_xsl/postprocessTEICorpoOutput.xsl
@@ -11,8 +11,11 @@
         </xsl:copy>
     </xsl:template>
     <xsl:template match="tei:spanGrp"/>
-   <xsl:template match="@who">
+    <xsl:template match="@xml:id[matches(.,'^\d')]">
+        <xsl:attribute name="{name(.)}" select="concat(substring(parent::*/local-name(.),1,1),.)"/>
+    </xsl:template>
+    <!--   <xsl:template match="@who">
        <xsl:attribute name="who">WER-BIN-ICH</xsl:attribute>
-   </xsl:template>     
+   </xsl:template>     -->
 
 </xsl:stylesheet>