From 063d2f77f8991ad30760564cf39c4f02ac91f38e Mon Sep 17 00:00:00 2001
From: xxyzz <gitpull@protonmail.com>
Date: Fri, 20 Oct 2023 09:54:55 +0800
Subject: [PATCH] Append default override page JSON file to the "--override"
 option

Chinese and German Wiktionary extractor requires some pages to be
overriden.
---
 overrides/.gitignore                          |  0
 .../wiktextract/data/overrides}/de.json       |  0
 .../wiktextract/data/overrides}/zh.json       |  0
 src/wiktextract/wiktwords.py                  | 22 ++++++++++++++++---
 4 files changed, 19 insertions(+), 3 deletions(-)
 delete mode 100644 overrides/.gitignore
 rename {overrides => src/wiktextract/data/overrides}/de.json (100%)
 rename {overrides => src/wiktextract/data/overrides}/zh.json (100%)

diff --git a/overrides/.gitignore b/overrides/.gitignore
deleted file mode 100644
index e69de29b..00000000
diff --git a/overrides/de.json b/src/wiktextract/data/overrides/de.json
similarity index 100%
rename from overrides/de.json
rename to src/wiktextract/data/overrides/de.json
diff --git a/overrides/zh.json b/src/wiktextract/data/overrides/zh.json
similarity index 100%
rename from overrides/zh.json
rename to src/wiktextract/data/overrides/zh.json
diff --git a/src/wiktextract/wiktwords.py b/src/wiktextract/wiktwords.py
index 936a51fe..de4c9a36 100755
--- a/src/wiktextract/wiktwords.py
+++ b/src/wiktextract/wiktwords.py
@@ -19,6 +19,11 @@
 from pathlib import Path
 from typing import TextIO
 
+if sys.version_info < (3, 10):
+    from importlib_resources import files
+else:
+    from importlib.resources import files
+
 from wikitextprocessor import Wtp
 from wikitextprocessor.dumpparser import analyze_and_overwrite_pages
 
@@ -246,8 +251,7 @@ def main():
         "--override",
         type=str,
         action="append",
-        help="Override module(s) by one in file or files in directory "
-        "(for debugging)",
+        help="Path of JSON file contains override page data",
     )
     parser.add_argument(
         "--use-thesaurus",
@@ -439,8 +443,20 @@ def main():
         pr = cProfile.Profile()
         pr.enable()
 
+    skip_extract_dump = wxr.wtp.saved_page_nums() > 0
+    default_override_json_path = (
+        files("wiktextract")
+        / "data"
+        / "overrides"
+        / f"{args.dump_file_language_code}.json"
+    )
+    if default_override_json_path.exists() and not skip_extract_dump:
+        if args.override is None:
+            args.override = [default_override_json_path]
+        elif default_override_json_path not in args.override:
+            args.override.append(default_override_json_path)
+
     try:
-        skip_extract_dump = wxr.wtp.saved_page_nums() > 0
         if args.path is not None:
             namespace_ids = {
                 wxr.wtp.NAMESPACE_DATA.get(name, {}).get("id")