Use the same annotator fudging logic in the server as well as the mai…

…n program
stanfordnlp · Mar 20, 2022 · b0d1e46 · b0d1e46
1 parent d694e20
commit b0d1e46
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 7 deletions.
diff --git a/src/edu/stanford/nlp/pipeline/StanfordCoreNLP.java b/src/edu/stanford/nlp/pipeline/StanfordCoreNLP.java
@@ -255,13 +255,7 @@ public StanfordCoreNLP(Properties props, boolean enforceRequirements, AnnotatorP
       this.properties.setProperty("annotators", newAnnotators);
     }
 
-    // if cleanxml is requested and tokenize is here,
-    // make it part of tokenize rather than its own annotator
-    unifyTokenizeProperty(this.properties, STANFORD_CLEAN_XML, STANFORD_TOKENIZE + "." + STANFORD_CLEAN_XML);
-    // ssplit is always part of tokenize now
-    unifyTokenizeProperty(this.properties, STANFORD_SSPLIT, null);
-    // cdc_tokenize is also absorbed into tokenize
-    replaceAnnotator(this.properties, STANFORD_CDC_TOKENIZE, STANFORD_TOKENIZE);
+    normalizeAnnotators(this.properties);
 
     // cdm [2017]: constructAnnotatorPool (PropertiesUtils.getSignature) requires non-null Properties, so after properties setup
     this.pool = annotatorPool != null ? annotatorPool : constructAnnotatorPool(props, getAnnotatorImplementations());
@@ -311,6 +305,19 @@ public StanfordCoreNLP(Properties props, boolean enforceRequirements, AnnotatorP
     this.pipelineSetupTime = tim.report();
   }
 
+  /**
+   * update the annotators, hopefully in a backwards compatible manner
+   */
+  static void normalizeAnnotators(Properties properties) {
+    // if cleanxml is requested and tokenize is here,
+    // make it part of tokenize rather than its own annotator
+    unifyTokenizeProperty(properties, STANFORD_CLEAN_XML, STANFORD_TOKENIZE + "." + STANFORD_CLEAN_XML);
+    // ssplit is always part of tokenize now
+    unifyTokenizeProperty(properties, STANFORD_SSPLIT, null);
+    // cdc_tokenize is also absorbed into tokenize
+    replaceAnnotator(properties, STANFORD_CDC_TOKENIZE, STANFORD_TOKENIZE);
+  }
+
   /**
    * The cdc_tokenize annotator is now part of tokenize
    */

diff --git a/src/edu/stanford/nlp/pipeline/StanfordCoreNLPServer.java b/src/edu/stanford/nlp/pipeline/StanfordCoreNLPServer.java
@@ -472,6 +472,7 @@ private Properties getProperties(HttpExchange httpExchange) throws UnsupportedEn
     urlProperties.forEach(props::setProperty);
 
     // Get the annotators
+    StanfordCoreNLP.normalizeAnnotators(props);
     String annotators = props.getProperty("annotators");
     // If the properties contains a custom annotator, then do not enforceRequirements.
     if (annotators != null && !PropertiesUtils.hasPropertyPrefix(props, CUSTOM_ANNOTATOR_PREFIX) && PropertiesUtils.getBool(props, "enforceRequirements", true)) {