From 850e5888c924a7810b5d586bc305e5a14d0c7969 Mon Sep 17 00:00:00 2001
From: John Bauer <horatio@gmail.com>
Date: Wed, 13 Mar 2024 16:45:16 -0700
Subject: [PATCH] Add lemmas to a few of the MWTs that we combine for English. 
 A few others are still TODO, such as the n't 'll etc suite

---
 .../nlp/trees/ud/EnglishMWTCombiner.java       | 18 ++++++++++++++++--
 1 file changed, 16 insertions(+), 2 deletions(-)

diff --git a/src/edu/stanford/nlp/trees/ud/EnglishMWTCombiner.java b/src/edu/stanford/nlp/trees/ud/EnglishMWTCombiner.java
index e9fc8f6edd..ef2bb2c3b6 100644
--- a/src/edu/stanford/nlp/trees/ud/EnglishMWTCombiner.java
+++ b/src/edu/stanford/nlp/trees/ud/EnglishMWTCombiner.java
@@ -15,6 +15,8 @@ public SemanticGraph combineMWTs(SemanticGraph sg) {
 
     // combine using the CombineMWT operation, using the default concatenation for the MWT text
     String mwt = String.join(newline,
+                             // TODO: separate the contractions so we can adjust the lemmas?
+                             // In some other way fix those lemmas?
                              "<ssurgeon-pattern-list>",
                              "  <ssurgeon-pattern>",
                              "    <uid>1</uid>",
@@ -30,9 +32,19 @@ public SemanticGraph combineMWTs(SemanticGraph sg) {
                              "  </ssurgeon-pattern>",
                              "  <ssurgeon-pattern>",
                              "    <uid>3</uid>",
-                             "    <notes>Edit a node's MWT for wanna/gonna</notes>",
-                             "    <semgrex>" + XMLUtils.escapeXML("{word:/(?i)wan|gon/;after://}=first . {word:/(?i)na/}=second") + "</semgrex>",
+                             "    <notes>Edit a node's MWT for wanna</notes>",
+                             "    <semgrex>" + XMLUtils.escapeXML("{word:/(?i)wan/;after://}=first . {word:/(?i)na/}=second") + "</semgrex>",
                              "    <edit-list>CombineMWT -node first -node second</edit-list>",
+                             "    <edit-list>EditNode -node first -lemma want</edit-list>",
+                             "    <edit-list>EditNode -node second -lemma to</edit-list>",
+                             "  </ssurgeon-pattern>",
+                             "  <ssurgeon-pattern>",
+                             "    <uid>3b</uid>",
+                             "    <notes>Edit a node's MWT for gonna</notes>",
+                             "    <semgrex>" + XMLUtils.escapeXML("{word:/(?i)gon/;after://}=first . {word:/(?i)na/}=second") + "</semgrex>",
+                             "    <edit-list>CombineMWT -node first -node second</edit-list>",
+                             "    <edit-list>EditNode -node first -lemma go</edit-list>",
+                             "    <edit-list>EditNode -node second -lemma to</edit-list>",
                              "  </ssurgeon-pattern>",
                              "  <ssurgeon-pattern>",
                              "    <uid>4</uid>",
@@ -45,6 +57,8 @@ public SemanticGraph combineMWTs(SemanticGraph sg) {
                              "    <notes>Edit a node's MWT for 'tis and 'twas</notes>",
                              "    <semgrex>" + XMLUtils.escapeXML("{word:/'[tT]/}=first . {word:/(?i)is|was/}=second") + "</semgrex>",
                              "    <edit-list>CombineMWT -node first -node second</edit-list>",
+                             "    <edit-list>EditNode -node first -lemma it</edit-list>",
+                             "    <edit-list>EditNode -node second -lemma be</edit-list>",
                              "  </ssurgeon-pattern>",
                              "  <ssurgeon-pattern>",
                              "    <uid>6</uid>",