From daa714a9b481631ba10f8f592bedc2e175d6334b Mon Sep 17 00:00:00 2001
From: Jamie Hardt <jamiehardt@me.com>
Date: Thu, 31 Oct 2024 20:08:10 -0700
Subject: [PATCH] Removing "a" as a synonym from all cats in English

---
 json/en.json         | 3 ---
 tools/ucsxls2json.py | 9 +++++++++
 2 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/json/en.json b/json/en.json
index da7e8a0..756e6dc 100644
--- a/json/en.json
+++ b/json/en.json
@@ -12550,7 +12550,6 @@
   "CatShort": "GAME",
   "Explanations": "Game sounds not fitting another category in this list. Cards would go here. Jenga.",
   "Synonyms": [
-   "a",
    "bag",
    "ball",
    "balloon",
@@ -22234,7 +22233,6 @@
   "CatShort": "VEH",
   "Explanations": "Very old vehicles, such a Model A, Model T. Mainly cars, but possible other very old vehicles, trucks, tractors.",
   "Synonyms": [
-   "a",
    "antique",
    "austin",
    "auto",
@@ -23550,7 +23548,6 @@
   "CatShort": "VOX",
   "Explanations": "A single voice singing, a child singing. For a crowd, use CROWDS-SINGING",
   "Synonyms": [
-   "a",
    "cappella",
    "caroling",
    "chanting",
diff --git a/tools/ucsxls2json.py b/tools/ucsxls2json.py
index cad6dcc..b2f985c 100644
--- a/tools/ucsxls2json.py
+++ b/tools/ucsxls2json.py
@@ -85,6 +85,15 @@ for lang in langs:
                     split_pattern = r'\W+'
                 syn_list = re.split(split_pattern, syns_raw)
                 category['Synonyms'] = [s.lower() for s in syn_list]
+
+                # if the sub-category name is not in the synonym list, include it
+                # if category['SubCategory'].lower() not in category['Synonyms']:
+                    # category['Synonyms'].append(category['SubCategory'].lower())
+                
+                if lang == 'en':
+                    for i, syn in enumerate(category['Synonyms']):
+                        if syn == "a":
+                            del category['Synonyms'][i]
         
         schedule.append(category)