Merge branch 'master' of github.com:proycon/foliatools

proycon · Apr 28, 2021 · dc1cb5f · dc1cb5f
2 parents c532991 + 998dc47
commit dc1cb5f
Show file tree

Hide file tree

Showing 4 changed files with 27 additions and 6 deletions.
diff --git a/foliatools/__init__.py b/foliatools/__init__.py
@@ -1,3 +1,3 @@
 """FoLiA-tools contains various Python-based command line tools for working with FoLiA XML (Format for Linguistic Annotation)"""
 
-VERSION = "2.5.0"
+VERSION = "2.5.1"
diff --git a/foliatools/tei2folia.py b/foliatools/tei2folia.py
@@ -72,6 +72,8 @@ def convert(filename, transformer, parser=None, **kwargs):
         if data.find(b"xmlns=\"http://www.tei-c.org/ns/1.0\"") == -1:
             data = data.replace(b"<TEI ",b"<TEI xmlns=\"http://www.tei-c.org/ns/1.0\" ")
             data = data.replace(b"<TEI.2 ",b"<TEI.2 xmlns=\"http://www.tei-c.org/ns/1.0\" ")
+            data = data.replace(b"<teiTrim ",b"<teiTrim xmlns=\"http://www.tei-c.org/ns/1.0\" ")
+            data = data.replace(b"<teiTrim>",b"<teiTrim xmlns=\"http://www.tei-c.org/ns/1.0\">")
         parsedsource = lxml.etree.fromstring(data, parser)
         del data
     else:

diff --git a/foliatools/tei2folia.xsl b/foliatools/tei2folia.xsl
@@ -62,11 +62,11 @@ Heavily adapted by Maarten van Gompel (Radboud University)
 
 <!-- *************************************************** DOCUMENT & METADATA ************************************************** -->
 
-<xsl:template match="TEI|TEI.2">
+<xsl:template match="TEI|TEI.2|teiTrim">
     <xsl:message terminate="yes">ERROR: TEI document lacks proper XML namespace declarations! Run tei2folia --forcenamespace to try to recover automatically or add xmlns="http://www.tei-c.org/ns/1.0" manually.</xsl:message>
 </xsl:template>
 
-<xsl:template match="tei:TEI|tei:TEI.2">
+<xsl:template match="tei:TEI|tei:TEI.2|tei:teiTrim">
 <FoLiA xmlns:xlink="http://www.w3.org/1999/xlink" xmlns="http://ilk.uvt.nl/folia" version="2.3.0" generator="tei2folia.xsl">
   <xsl:attribute name="xml:id"><xsl:value-of select="$docid"/></xsl:attribute>
   <metadata type="native">
@@ -108,7 +108,17 @@ Heavily adapted by Maarten van Gompel (Radboud University)
   </metadata>
   <text>
     <xsl:attribute name="xml:id"><xsl:value-of select="$docid"/>.text</xsl:attribute>
-    <xsl:apply-templates select="//tei:text/*" mode="structure"/>
+    <xsl:choose>
+    <xsl:when test="./tei:body">
+        <!-- We see this behaviour in some files, no tei:text but directly a body, not sure if it's valid TEI but we can deal with it  -->
+        <div class="body">
+            <xsl:apply-templates select="//tei:body/*" mode="structure"/>
+        </div>
+    </xsl:when>
+    <xsl:otherwise>
+        <xsl:apply-templates select="/text/tei:text/*" mode="structure"/>
+    </xsl:otherwise>
+    </xsl:choose>
   </text>
 </FoLiA>
 </xsl:template>
@@ -439,7 +449,16 @@ Heavily adapted by Maarten van Gompel (Radboud University)
     </cell>
 </xsl:template>
 
-<xsl:template match="tei:p|tei:speaker|tei:trailer|tei:closer|tei:opener|tei:lxx|tei:byline|tei:salute" mode="structure">
+<xsl:template match="tei:p|tei:speaker|tei:trailer|tei:closer|tei:opener|tei:lxx|tei:byline|tei:salute|tei:para|tei:remark" mode="structure">
+    <xsl:call-template name="p"/>
+</xsl:template>
+
+<!-- These are not real TEI elements but they occur in some dialects and we squash them to paragraphs -->
+<xsl:template match="tei:para|tei:remark|tei:folio" mode="structure">
+    <xsl:if test="$quiet = 'false'">
+    <xsl:message terminate="no">NOTICE: Found an unofficial TEI element that doesn't exist in the P5 specification <xsl:value-of select="name(.)"/> (in <xsl:value-of select="name(parent::node())" />), we're converting these to a FoLiA paragraph</xsl:message>
+    </xsl:if>
+    <comment>[tei2folia WARNING] Unofficial TEI element: <xsl:value-of select="name(.)"/> (in <xsl:value-of select="name(parent::node())" />) was converted to a paragraph</comment>
     <xsl:call-template name="p"/>
 </xsl:template>
 

diff --git a/setup.py b/setup.py
@@ -11,7 +11,7 @@ def read(fname):
 
 setup(
     name = "FoLiA-tools",
-    version = "2.5.0", #also change in __init__.py
+    version = "2.5.1", #also change in __init__.py
     author = "Maarten van Gompel",
     author_email = "proycon@anaproy.nl",
     description = ("FoLiA-tools contains various Python-based command line tools for working with FoLiA XML (Format for Linguistic Annotation)"),