Add support for inlineTags extraAttributes.

This allows you to use XPath to add attributes to inline tags, in addition to the one they have in the document. See how-to-configure-indexing.md details.
INL · Jun 4, 2024 · 53b9f13 · 53b9f13
1 parent d8801f5
commit 53b9f13
Show file tree

Hide file tree

Showing 5 changed files with 102 additions and 1 deletion.
diff --git a/engine/src/main/java/nl/inl/blacklab/indexers/config/ConfigInlineTag.java b/engine/src/main/java/nl/inl/blacklab/indexers/config/ConfigInlineTag.java
@@ -10,6 +10,39 @@
  */
 public class ConfigInlineTag {
 
+    /** Configuration for extra attributes to index using XPath */
+    public static class ConfigExtraAttribute {
+        /** Attribute name */
+        private String name;
+        /** XPath to get attribute's value */
+        private String valuePath;
+
+        public ConfigExtraAttribute() {
+
+        }
+
+        public ConfigExtraAttribute(String name, String valuePath) {
+            this.name = name;
+            this.valuePath = valuePath;
+        }
+
+        public String getName() {
+            return name;
+        }
+
+        public void setName(String name) {
+            this.name = name;
+        }
+
+        public String getValuePath() {
+            return valuePath;
+        }
+
+        public void setValuePath(String valuePath) {
+            this.valuePath = valuePath;
+        }
+    }
+
     /** XPath to the inline tag, relative to the container element */
     private String path;
 
@@ -34,6 +67,9 @@ public class ConfigInlineTag {
     /** If set: ignore excludeAttributes and don't index attributes not in this list. */
     private List<String> includeAttributes = null;
 
+    /** Extra attributes to index with the tag via an XPath expression */
+    private List<ConfigExtraAttribute> extraAttributes = Collections.emptyList();
+
     public ConfigInlineTag() {
     }
 
@@ -94,4 +130,16 @@ public void setIncludeAttributes(List<String> includeAttributes) {
     public List<String> getIncludeAttributes() {
         return includeAttributes;
     }
+
+    public void setExtraAttributes(List<ConfigExtraAttribute> extraAttributes) {
+        this.extraAttributes = extraAttributes;
+    }
+
+    public List<ConfigExtraAttribute> getExtraAttributes() {
+        return extraAttributes;
+    }
+
+    public boolean hasDetailedAttributeConfig() {
+        return includeAttributes != null || !excludeAttributes.isEmpty() || !extraAttributes.isEmpty();
+    }
 }
diff --git a/engine/src/main/java/nl/inl/blacklab/indexers/config/DocIndexerSaxon.java b/engine/src/main/java/nl/inl/blacklab/indexers/config/DocIndexerSaxon.java
@@ -364,6 +364,12 @@ private void handleInlineOpenTag(ConfigAnnotatedField annotatedField, Map<Span,
                     }
                 }
             }
+            // Index any extra attributes using the provided XPath expressions.
+            for (ConfigInlineTag.ConfigExtraAttribute extraAttribute: currentInline.config.getExtraAttributes()) {
+                String value = xpathValue(extraAttribute.getValuePath(), nodeInfo);
+                if (value != null)
+                    atts.put(extraAttribute.getName(), value);
+            }
             inlineTag(nodeInfo.getDisplayName(), true, atts);
 
             // Add tag to the list of tags to close at the correct position.

diff --git a/engine/src/main/java/nl/inl/blacklab/indexers/config/DocIndexerVTD.java b/engine/src/main/java/nl/inl/blacklab/indexers/config/DocIndexerVTD.java
@@ -319,6 +319,9 @@ private void collectPunctuation(ConfigAnnotatedField annotatedField, List<Inline
     private void collectInlineTags(ConfigAnnotatedField annotatedField, List<InlineObject> inlineObjects) {
         int i = 0;
         for (ConfigInlineTag inlineTag : annotatedField.getInlineTags()) {
+            if (inlineTag.hasDetailedAttributeConfig())
+                warn("Detailed inline tag attribute configuration not supported in VTD indexer. Ignoring for tag: " +
+                        inlineTag.getPath() + " (for support, add 'processor: saxon' to .blf.yaml file)");
             // Collect the occurrences of this inline tag
             String tokenIdXPath = inlineTag.getTokenIdPath();
             // We want to capture token ids for this inline tag. Create the AutoPilot.

diff --git a/engine/src/main/java/nl/inl/blacklab/indexers/config/InputFormatReader.java b/engine/src/main/java/nl/inl/blacklab/indexers/config/InputFormatReader.java
@@ -568,6 +568,9 @@ private void readInlineTags(Entry<String, JsonNode> itsEntry, ConfigAnnotatedFie
                     readStringList(e, exclAttr);
                     t.setExcludeAttributes(exclAttr);
                     break;
+                case "extraAttributes":
+                    t.setExtraAttributes(readExtraAttributes(e));
+                    break;
                 default:
                     throw new InvalidInputFormatConfig("Unknown key " + e.getKey() + " in inline tag " + t.getPath());
                 }
@@ -576,6 +579,40 @@ private void readInlineTags(Entry<String, JsonNode> itsEntry, ConfigAnnotatedFie
         }
     }
 
+    /**
+     * Extra attributes to index by XPath for inline tag
+     *
+     * @return
+     */
+    private List<ConfigInlineTag.ConfigExtraAttribute> readExtraAttributes(Entry<String, JsonNode> e) {
+        List<ConfigInlineTag.ConfigExtraAttribute> extraAttr = new ArrayList<>();
+        Iterator<JsonNode> itExtraAttr = array(e).elements();
+        while (itExtraAttr.hasNext()) {
+            JsonNode ea = itExtraAttr.next();
+            ConfigInlineTag.ConfigExtraAttribute a = new ConfigInlineTag.ConfigExtraAttribute();
+            Iterator<Entry<String, JsonNode>> itExtraAttrEntry = obj(ea, "extra attribute").fields();
+            while (itExtraAttrEntry.hasNext()) {
+                Entry<String, JsonNode> eea = itExtraAttrEntry.next();
+                switch (eea.getKey()) {
+                case "name":
+                    a.setName(str(eea));
+                    break;
+                case "value":
+                    a.setValuePath(fixedStringToXpath(str(eea)));
+                    break;
+                case "valuePath":
+                    a.setValuePath(str(eea));
+                    break;
+                default:
+                    throw new InvalidInputFormatConfig(
+                            "Unknown key " + eea.getKey() + " in extra attribute " + a.getName());
+                }
+            }
+            extraAttr.add(a);
+        }
+        return extraAttr;
+    }
+
     private void readMetadata(Entry<String, JsonNode> mdEntry, ConfigInputFormat cfg) {
         JsonNode node = mdEntry.getValue();
         if (node instanceof ObjectNode) {

diff --git a/site/docs/guide/how-to-configure-indexing.md b/site/docs/guide/how-to-configure-indexing.md
@@ -219,7 +219,14 @@ inlineTags:
         - "xml:id" 
     - path: .//p
       includeAttributes:
-        - "type"  # Only index the type attribute
+        - "type"  # Only index this tag's type attribute
+        
+      # Any extra attributes to index with this tag using XPath
+      extraAttributes:
+        # if e.g. input is <p xml:id="par-12">...</p> , index  number="12"
+        - name: "number"
+          valuePath: "substring-after(@xml:id, 'par-')"
+          
     - path: .//ne
       displayAs: named-entity    # what CSS class to use (when using autogenerated XSLT)    
 ```