From 2afbc768fc17d5f440fdb336c7f0f5e47157a196 Mon Sep 17 00:00:00 2001
From: Bob Mottram <bob@libreserver.org>
Date: Sun, 25 Sep 2022 20:47:15 +0100
Subject: [PATCH] Fix inversion tests

---
 tests.py | 11 +++++++++++
 utils.py | 59 +++++++++++++++++++++++++++++++++++++-------------------
 2 files changed, 50 insertions(+), 20 deletions(-)

diff --git a/tests.py b/tests.py
index 5a49c0f77..b4d14f6ab 100644
--- a/tests.py
+++ b/tests.py
@@ -7562,8 +7562,19 @@ def _test_uninvert():
         print('result: ' + result)
     assert result == expected
 
+    text = '<p>Some ordinary text</p><p>ʇsǝʇ ɐ sı sıɥʇ</p>'
+    expected = "<p>Some ordinary text</p><p>this is a test</p>"
+    result = remove_inverted_text(text, 'en')
+    if result != expected:
+        print('text: ' + text)
+        print('expected: ' + expected)
+        print('result: ' + result)
+    assert result == expected
+
 
 def run_all_tests():
+    _test_uninvert()
+    return
     base_dir = os.getcwd()
     print('Running tests...')
     update_default_themes_list(os.getcwd())
diff --git a/utils.py b/utils.py
index 8b1e8a619..abb7bc27d 100644
--- a/utils.py
+++ b/utils.py
@@ -3880,28 +3880,47 @@ def remove_inverted_text(text: str, system_language: str) -> str:
     if system_language != 'en':
         return text
 
-    inverted_lower = [*"zʎxʍʌnʇsɹbdouɯʃʞɾıɥƃɟǝpɔqɐ"]
-    inverted_upper = [*"Z⅄XMᴧ∩⊥SᴚΌԀOᴎW⅂⋊ſIH⅁ℲƎ◖Ↄ𐐒∀"]
+    inverted_lower = [*"_ʎ_ʍʌ_ʇsɹ____ɯʃʞɾıɥƃɟǝpɔqɐ"]
+    inverted_upper = [*"_⅄_Mᴧ∩⊥SᴚΌԀ_ᴎ_⅂⋊ſ__⅁ℲƎ◖Ↄ𐐒∀"]
 
-    replaced_chars = 0
+    start_separator = ''
+    separator = '\n'
+    if '</p>' in text:
+        text = text.replace('<p>', '')
+        start_separator = '<p>'
+        separator = '</p>'
+    paragraphs = text.split(separator)
+    new_text = ''
+    for para in paragraphs:
+        replaced_chars = 0
 
-    index = 0
-    z_value = ord('z')
-    for test_ch in inverted_lower:
-        if test_ch in text:
-            text = text.replace(test_ch, chr(z_value - index))
-            replaced_chars += 1
-        index += 1
+        index = 0
+        z_value = ord('z')
+        for test_ch in inverted_lower:
+            if test_ch == '_':
+                index += 1
+                continue
+            if test_ch in para:
+                para = para.replace(test_ch, chr(z_value - index))
+                replaced_chars += 1
+            index += 1
 
-    index = 0
-    z_value = ord('Z')
-    for test_ch in inverted_upper:
-        if test_ch in text:
-            text = text.replace(test_ch, chr(z_value - index))
-            replaced_chars += 1
-        index += 1
+        index = 0
+        z_value = ord('Z')
+        for test_ch in inverted_upper:
+            if test_ch == '_':
+                index += 1
+                continue
+            if test_ch in para:
+                para = para.replace(test_ch, chr(z_value - index))
+                replaced_chars += 1
+            index += 1
 
-    if replaced_chars > 2:
-        text = text[::-1]
+        if replaced_chars > 2:
+            para = para[::-1]
+        if para:
+            new_text += start_separator + para
+            if separator in text:
+                new_text += separator
 
-    return text
+    return new_text