Upgrade with USAS structures

clarin-eric · Oct 11, 2023 · 2103dae · 2103dae
1 parent 2b4d6aa
commit 2103dae
Show file tree

Hide file tree

Showing 7 changed files with 65 additions and 39 deletions.
diff --git a/Schema/ParlaMint-TEI.ana.rnc b/Schema/ParlaMint-TEI.ana.rnc
@@ -26,23 +26,26 @@ include "ParlaMint-TEI.rnc" {
 
   ## The only element that contains analysed text is the segment. It can,
   ##       however, contain transcription commentary.
-  seg =
-    element seg {
-      global.atts
-      >> a:documentation [ "Optional @corresp for MTed corpora." ],
-      corresp.att,
-      (comment | pb | sentence)+
-    }
+  seg = element seg { global.atts, (comment | pb | sentence)+ }
 }
 
 ## A sentence.
 sentence =
   element s {
-    global.atts
-    >> a:documentation [ "Optional @corresp for MTed corpora." ],
-    corresp.att,
-    (comment | pb | ner | word | punct)+,
-    syntax?
+    global.atts, (comment | pb | ner | phr | word | punct)+, syntax?
+  }
+
+## Multi-Word Unit
+phr =
+  element phr {
+    global.atts,
+    attribute function { xsd:token },
+    attribute ana { xsd:anyURI }
+    >> a:documentation [
+         "Currently we only support phrases with semantic markers"
+       ],
+    attribute type { "sem" },
+    mwe_tokens
   }
 
 ## Named entity element
@@ -70,6 +73,9 @@ ner.atts =
   global.atts,
   attribute ana { xsd:anyURI }
 
+## Annotations that can appear in a MWE.
+mwe_tokens = (word | punct)+
+
 ## Annotations that can appear in a NE.
 ner_tokens = (word | punct | comment | pb | ner)+
 
@@ -123,7 +129,12 @@ token.atts =
   attribute msd { non-empty.string },
   attribute ana { anyURIs }?,
   attribute pos { non-empty.string }?,
-  attribute function { normalized-space.string }?
+  attribute function {
+    # This is different from TEI, where it is a name token (which can't contain comma!?)
+
+    # ref name="normalized-space.string"/
+    xsd:token
+  }?
 
 ## Obligatory attributes of word.
 word.atts =

diff --git a/Schema/ParlaMint-TEI.ana.rng b/Schema/ParlaMint-TEI.ana.rng
@@ -36,8 +36,6 @@
       however, contain transcription commentary.</a:documentation>
       <element name="seg">
         <ref name="global.atts"/>
-	<a:documentation>Optional @corresp for MTed corpora.</a:documentation>
-	<ref name="corresp.att"/>
         <oneOrMore>
           <choice>
             <ref name="comment"/>
@@ -54,13 +52,12 @@
     <a:documentation>A sentence.</a:documentation>
     <element name="s">
       <ref name="global.atts"/>
-      <a:documentation>Optional @corresp for MTed corpora.</a:documentation>
-      <ref name="corresp.att"/>
       <oneOrMore>
         <choice>
           <ref name="comment"/>
           <ref name="pb"/>
           <ref name="ner"/>
+          <ref name="phr"/>
           <ref name="word"/>
           <ref name="punct"/>
         </choice>
@@ -71,6 +68,24 @@
     </element>
   </define>
 
+  <define name="phr">
+    <a:documentation>Multi-Word Unit</a:documentation>
+    <element name="phr">
+      <ref name="global.atts"/>
+      <attribute name="function">
+        <data type="token"/>
+      </attribute>
+      <attribute name="ana">
+        <data type="anyURI"/>
+      </attribute>
+      <a:documentation>Currently we only support phrases with semantic markers</a:documentation>
+      <attribute name="type">
+        <value>sem</value>
+      </attribute>
+      <ref name="mwe_tokens"/>
+    </element>
+  </define>
+
   <define name="ner">
     <a:documentation>Named entity element</a:documentation>
     <choice>
@@ -139,6 +154,16 @@
     </attribute>
   </define>
 
+  <define name="mwe_tokens">
+    <a:documentation>Annotations that can appear in a MWE.</a:documentation>
+    <oneOrMore>
+      <choice>
+        <ref name="word"/>
+        <ref name="punct"/>
+      </choice>
+    </oneOrMore>
+  </define>
+
   <define name="ner_tokens">
     <a:documentation>Annotations that can appear in a NE.</a:documentation>
     <oneOrMore>
@@ -261,7 +286,9 @@
     </optional>
     <optional>
       <attribute name="function">
-	<ref name="normalized-space.string"/>
+	<!-- This is different from TEI, where it is a name token (which can't contain comma!?) -->
+	<!--ref name="normalized-space.string"/-->
+        <data type="token"/>
       </attribute>
     </optional>
   </define>

diff --git a/Schema/ParlaMint-TEI.rnc b/Schema/ParlaMint-TEI.rnc
@@ -174,7 +174,7 @@ title =
   }
 
 ## A "meeting" (e.g. session) of the parliament.
-meeting = element meeting { global.atts, ana.att, corresp.att, text }
+meeting = element meeting { global.atts, ana.att, text }
 
 ## Publisher of the corpus.
 publisher =
@@ -218,9 +218,7 @@ text-body =
 ## The definition of the div element.
 \div =
   element div {
-    global.atts
-    >> a:documentation [ "Optional @corresp for MTed corpora." ],
-    corresp.att,
+    global.atts,
 
     ## A standard div with utterances.
     ((attribute type { "debateSection" },
@@ -249,18 +247,14 @@ u =
          "Utterance has obligatory @ana, for the role of the speaker."
        ],
     ana.att,
-    (attribute source { anyURIs }?)
-    >> a:documentation [ "Optional @corresp for MTed corpora." ],
-    corresp.att,
+    attribute source { anyURIs }?,
     (seg | comment | pb)+
   }
 
 ## A segment (i.e paragraph) inside a speech.
 seg =
   element seg {
-    global.atts
-    >> a:documentation [ "Optional @corresp for MTed corpora." ],
-    corresp.att,
+    global.atts,
     (comment
      | pb
      | text

diff --git a/Schema/ParlaMint-TEI.rng b/Schema/ParlaMint-TEI.rng
@@ -318,7 +318,6 @@
     <element name="meeting">
       <ref name="global.atts"/>
       <ref name="ana.att"/>
-      <ref name="corresp.att"/>
       <text/>
     </element>
   </define>
@@ -408,8 +407,6 @@
     <a:documentation>The definition of the div element.</a:documentation>
     <element name="div">
       <ref name="global.atts"/>
-      <a:documentation>Optional @corresp for MTed corpora.</a:documentation>
-      <ref name="corresp.att"/>
       <choice>
         <a:documentation>A standard div with utterances.</a:documentation>
         <group>
@@ -479,8 +476,6 @@
           <ref name="anyURIs"/>
         </attribute>
       </optional>
-      <a:documentation>Optional @corresp for MTed corpora.</a:documentation>
-      <ref name="corresp.att"/>
       <oneOrMore>
         <choice>
           <ref name="seg"/>
@@ -495,8 +490,6 @@
     <a:documentation>A segment (i.e paragraph) inside a speech.</a:documentation>
     <element name="seg">
       <ref name="global.atts"/>
-      <a:documentation>Optional @corresp for MTed corpora.</a:documentation>
-      <ref name="corresp.att"/>
       <oneOrMore>
         <choice>
           <ref name="comment"/>

diff --git a/Schema/ParlaMint.rnc b/Schema/ParlaMint.rnc
@@ -267,7 +267,7 @@ head =
   }
 
 ## A page break, possibly with its source URI.
-pb = element pb { global.atts, corresp.att, source.att, empty }
+pb = element pb { global.atts, source.att, empty }
 
 ## A series of paragraphs. Paragraphs are only used in the teiHeader.
 paras = element p { global.atts, annotated.text }+
@@ -339,6 +339,7 @@ corresp.att = attribute corresp { anyURIs }?
 global.atts =
   id.att?,
   lang.att,
+  corresp.att,
   attribute n { xsd:string }?
 
 ## The identifier attribute, giving the unique ID of the

diff --git a/Schema/ParlaMint.rng b/Schema/ParlaMint.rng
@@ -549,7 +549,6 @@
     <a:documentation>A page break, possibly with its source URI.</a:documentation>
     <element name="pb">
       <ref name="global.atts"/>
-      <ref name="corresp.att"/>
       <ref name="source.att"/>
       <empty/>
     </element>
@@ -733,6 +732,7 @@
       <ref name="id.att"/>
     </optional>
     <ref name="lang.att"/>
+    <ref name="corresp.att"/>
     <optional>
       <attribute name="n">
         <data type="string"/>

diff --git a/Schema/compile.log b/Schema/compile.log
@@ -7,5 +7,5 @@ java -jar /usr/share/java/trang.jar ParlaMint-listOrg.rng ParlaMint-listOrg.rnc
 java -jar /usr/share/java/trang.jar ParlaMint-listPerson.rng ParlaMint-listPerson.rnc
 java -jar /usr/share/java/trang.jar ParlaMint-taxonomy.rng ParlaMint-taxonomy.rnc
 make[1]: Leaving directory '/home/project/corpora/Parla/ParlaMint/ParlaMint/Schema'
-4.92user 0.61system 0:02.46elapsed 224%CPU (0avgtext+0avgdata 67164maxresident)k
-1872inputs+824outputs (0major+71458minor)pagefaults 0swaps
+5.33user 0.56system 0:02.59elapsed 227%CPU (0avgtext+0avgdata 67200maxresident)k
+1600inputs+872outputs (0major+71468minor)pagefaults 0swaps