Properly preserve specified markers (matching fix in machine)

Author: Enkidu93

machine/corpora/update_usfm_parser_handler.py

@@ -323,20 +323,21 @@ def _replace_with_new_tokens(self, state: UsfmParserState, closed: bool = True)
         )
 
         use_new_tokens = (
-            (
+            not self._is_in_preserved_paragraph(marker)
+            and (
                 self._text_behavior == UpdateUsfmTextBehavior.STRIP_EXISTING
-                and not self._is_in_preserved_paragraph(marker)
+                or (
+                    self._has_new_text()
+                    and (not existing_text or self._text_behavior != UpdateUsfmTextBehavior.PREFER_EXISTING)
+                )
             )
-            or (
-                self._has_new_text()
-                and (not existing_text or self._text_behavior != UpdateUsfmTextBehavior.PREFER_EXISTING)
-            )
-        ) and (
-            not in_embed
-            or (
-                self._is_in_note_text()
-                and not in_nested_embed
-                and self._embed_behavior == UpdateUsfmMarkerBehavior.PRESERVE
+            and (
+                not in_embed
+                or (
+                    self._is_in_note_text()
+                    and not in_nested_embed
+                    and self._embed_behavior == UpdateUsfmMarkerBehavior.PRESERVE
+                )
             )
         )
 
@@ -346,7 +347,9 @@ def _replace_with_new_tokens(self, state: UsfmParserState, closed: bool = True)
             else:
                 self._add_new_tokens()
 
-        if existing_text and self._text_behavior == UpdateUsfmTextBehavior.PREFER_EXISTING:
+        if existing_text and (
+            self._text_behavior == UpdateUsfmTextBehavior.PREFER_EXISTING or self._is_in_preserved_paragraph(marker)
+        ):
             if in_embed:
                 self._clear_new_embed_tokens()
             else:

tests/corpora/test_update_usfm_parser_handler.py

@@ -101,6 +101,54 @@ def test_get_usfm_strip_all_text() -> None:
     assess(target, result)
 
 
+def test_get_usfm_strip_paragraphs_preserve_paragraph_styles():
+    rows = [
+        (scr_ref("MAT 1:0/1:rem"), "New remark"),
+        (scr_ref("MAT 1:0/3:ip"), "Another new remark"),
+        (scr_ref("MAT 1:1"), "Update 1"),
+    ]
+    usfm = r"""\id MAT
+\c 1
+\rem Update remark
+\r reference
+\ip This is another remark, but with a different marker
+\v 1 This is a verse
+"""
+
+    target = update_usfm(
+        rows,
+        usfm,
+        text_behavior=UpdateUsfmTextBehavior.STRIP_EXISTING,
+        paragraph_behavior=UpdateUsfmMarkerBehavior.STRIP,
+    )
+    result = r"""\id MAT
+\c 1
+\rem Update remark
+\r reference
+\ip Another new remark
+\v 1 Update 1
+"""
+
+    assess(target, result)
+
+    targetDiffParagraph = update_usfm(
+        rows,
+        usfm,
+        text_behavior=UpdateUsfmTextBehavior.STRIP_EXISTING,
+        paragraph_behavior=UpdateUsfmMarkerBehavior.STRIP,
+        preserve_paragraph_styles=["ip"],
+    )
+    resultDiffParagraph = r"""\id MAT
+\c 1
+\rem New remark
+\r
+\ip This is another remark, but with a different marker
+\v 1 Update 1
+"""
+
+    assess(targetDiffParagraph, resultDiffParagraph)
+
+
 def test_preserve_paragraphs():
     rows = [
         (