From 7e41e7cbdf4a4f09f54a2fecca8a8ef0d578a3a9 Mon Sep 17 00:00:00 2001
From: p-goulart <pedro.goulart@languagetool.com>
Date: Tue, 9 Apr 2024 16:23:40 +0200
Subject: [PATCH] [pt] Fix VIR_A_VERBO_VERBO

---
 .../AbstractAdvancedSynthesizerFilter.java    |  2 +-
 .../rules/pt/AdvancedSynthesizerFilter.java   | 25 +++++++++++++++++++
 .../org/languagetool/rules/pt/style.xml       | 13 +++++-----
 3 files changed, 32 insertions(+), 8 deletions(-)

diff --git a/languagetool-core/src/main/java/org/languagetool/rules/AbstractAdvancedSynthesizerFilter.java b/languagetool-core/src/main/java/org/languagetool/rules/AbstractAdvancedSynthesizerFilter.java
index 54eb376162f1..4f53a5ee3163 100644
--- a/languagetool-core/src/main/java/org/languagetool/rules/AbstractAdvancedSynthesizerFilter.java
+++ b/languagetool-core/src/main/java/org/languagetool/rules/AbstractAdvancedSynthesizerFilter.java
@@ -163,7 +163,7 @@ public RuleMatch acceptRuleMatch(RuleMatch match, Map<String, String> arguments,
     return match;
   }
 
-  private String getCompositePostag(String lemmaSelect, String postagSelect, String originalPostag,
+  public String getCompositePostag(String lemmaSelect, String postagSelect, String originalPostag,
       String desiredPostag, String postagReplace) {
     Pattern aPattern = Pattern.compile(lemmaSelect, Pattern.UNICODE_CASE);
     Pattern bPattern = Pattern.compile(postagSelect, Pattern.UNICODE_CASE);
diff --git a/languagetool-language-modules/pt/src/main/java/org/languagetool/rules/pt/AdvancedSynthesizerFilter.java b/languagetool-language-modules/pt/src/main/java/org/languagetool/rules/pt/AdvancedSynthesizerFilter.java
index 93d651bf3d7e..2975bb27ada1 100644
--- a/languagetool-language-modules/pt/src/main/java/org/languagetool/rules/pt/AdvancedSynthesizerFilter.java
+++ b/languagetool-language-modules/pt/src/main/java/org/languagetool/rules/pt/AdvancedSynthesizerFilter.java
@@ -19,10 +19,17 @@
 
 package org.languagetool.rules.pt;
 
+import org.languagetool.AnalyzedTokenReadings;
 import org.languagetool.rules.AbstractAdvancedSynthesizerFilter;
+import org.languagetool.rules.RuleMatch;
 import org.languagetool.synthesis.Synthesizer;
 import org.languagetool.synthesis.pt.PortugueseSynthesizer;
 
+import java.io.IOException;
+import java.util.List;
+import java.util.Map;
+import java.util.Objects;
+
 /*
  * Synthesize suggestions using the lemma from one token (lemma_from) 
  * and the POS tag from another one (postag_from).
@@ -46,4 +53,22 @@ protected boolean isSuggestionException(String token, String desiredPostag) {
     return false;
   }
 
+  protected String movePronounTag(String sourceTag, String destinationTag) {
+    String[] sourceTagParts = sourceTag.split(":");
+    String newTag = destinationTag;
+    if (sourceTagParts.length == 2) {
+      String[] destinationTagParts = destinationTag.split(":");
+      newTag = destinationTagParts[0] + ":" + sourceTagParts[1];
+    }
+    return newTag;
+  }
+
+  @Override
+  public String getCompositePostag(String lemmaSelect, String postagSelect, String originalPostag,
+                                   String desiredPostag, String postagReplace) {
+    if (Objects.equals(postagReplace, "keepPronoun")) {
+      return movePronounTag(originalPostag, desiredPostag);
+    }
+    return super.getCompositePostag(lemmaSelect, postagSelect, originalPostag, desiredPostag, postagReplace);
+  }
 }
\ No newline at end of file
diff --git a/languagetool-language-modules/pt/src/main/resources/org/languagetool/rules/pt/style.xml b/languagetool-language-modules/pt/src/main/resources/org/languagetool/rules/pt/style.xml
index 3a283bec05bf..84ce625daaf4 100644
--- a/languagetool-language-modules/pt/src/main/resources/org/languagetool/rules/pt/style.xml
+++ b/languagetool-language-modules/pt/src/main/resources/org/languagetool/rules/pt/style.xml
@@ -16098,21 +16098,16 @@ USA
         </rule>
 
 
-        <!-- VIR A VERBO verbo -->
         <rule id='VIR_A_VERBO_VERBO' name="[Simplificar] Vir + a + Verbo → Verbo" type='style'>
-            <!--IDEA shorten_it-->
-            <!--      Created by Marco A.G.Pinto, Portuguese rule 2021-08-22 (25-JUN-2021+)      -->
-            <!--
-      O nosso estudo terá grande aplicabilidade caso esta ameaça se venha a concretizar. → O nosso estudo terá grande aplicabilidade caso esta ameaça se concretize.
-      -->
             <pattern>
                 <token inflected='yes'>vir<exception>vi</exception></token><!-- please remove "vi" as a form of "vir" from dict -->
                 <token>a</token>
                 <token postag='VMN.+' postag_regexp='yes'/>
             </pattern>
-            <filter class="org.languagetool.rules.pt.AdvancedSynthesizerFilter" args="lemmaFrom:3 lemmaSelect:V.* postagFrom:1 postagSelect:V.*"/>
+            <filter class="org.languagetool.rules.pt.AdvancedSynthesizerFilter" args="lemmaFrom:3 lemmaSelect:V.+ postagFrom:1 postagSelect:V.+ postagReplace:keepPronoun"/>
             <message>&simplify_msg;</message>
             <example correction="concretize">O nosso estudo terá grande aplicabilidade caso esta ameaça se <marker>venha a concretizar</marker>.</example>
+            <example correction="especializar-se-ia">Acho que ele <marker>viria a especializar-se</marker> nisso.</example>
             <example>Eu vi a professora Gómez ontem.</example>
             <example>Quando eles vieram a receita quiseram fazer logo.</example>
         </rule>
@@ -18244,6 +18239,10 @@ USA
 
 
         <rulegroup id='VER_OBSERVAR_CONSTATAR' name="Ver → Observar/Constatar" type="style" tags="picky" tone_tags="academic" is_goal_specific="true">
+            <antipattern>
+                <token case_sensitive="yes">VI</token>
+                <example>Constantino VI no ano 790 da nossa era.</example>
+            </antipattern>
             <antipattern>
                 <token inflected='yes' skip='-1'>ver</token>
                 <token regexp='yes'>filmes?|vídeos?|séries?|mensagem|mensagens|cenas?|televisão|tv|lado</token>