better unalign

unfoldingWord-dev · Jan 20, 2022 · 1799567 · 1799567
1 parent aaade76
commit 1799567
Showing 1 changed file with 15 additions and 11 deletions.
diff --git a/usfm/usfm_utils.py b/usfm/usfm_utils.py
@@ -3,15 +3,19 @@
 from __future__ import unicode_literals
 import re
 
-def usfm3_to_usfm2(usfm):
+
+def unalign_usfm(aligned_usfm):
     """
-    Converts a USFM 3 string to a USFM 2 compatible string
-    :param usfm3:
-    :return: the USFM 2 version of the string
+    Converts an aligned USFM string to an unaligned USFM compatible string
+    :param aligned_usfm:
+    :return: the unaligned USFM of the string
     """
-    # Kind of usfm3 to usfm2
-    usfm = re.sub(r'\\zaln-s[^\*]*\*', r'', usfm, flags=re.UNICODE | re.MULTILINE)
+    # Remove all tags used for alignments and words
+    usfm = re.sub(r'\\ts(-s)*\s*\\\*\s*', r'', aligned_usfm, flags=re.UNICODE | re.MULTILINE)
+    usfm = re.sub(r'\\zaln-s[^*]*?\*', r'', usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r'\\zaln-e\\\*', r'', usfm, flags=re.UNICODE | re.MULTILINE)
+    usfm = re.sub(r'\\k-s.*?\\\*', r'', usfm, flags=re.UNICODE | re.MULTILINE)
+    usfm = re.sub(r'\\k-e\\\*', r'', usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r'\\w ([^|]+)\|.*?\\w\*', r'\1', usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r'^\n', '', usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r'^([^\\].*)\n(?=[^\\])', r'\1 ', usfm, flags=re.UNICODE | re.MULTILINE)
@@ -22,14 +26,14 @@ def usfm3_to_usfm2(usfm):
     usfm = re.sub(r"\s*' s(?!\w)", "'s", usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r'\\s5', '', usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r'\\fqa([^*]+)\\fqa(?![*])', r'\\fqa\1\\fqa*', usfm, flags=re.UNICODE | re.MULTILINE)
-    
+
     # Pair up quotes by chapter
-    chapters = re.compile(r'\\c').split(usfm)
+    chapters = re.compile(r'\\c ').split(usfm)
     usfm = chapters[0]
     for chapter in chapters[1:]:
-        chapter = re.sub(r'\s*"\s*([^"]+)\s*"\s*', r' "\1" ', chapter, flags=re.UNICODE | re.MULTILINE | re.DOTALL)
-        usfm += '\c'+chapter
-    usfm = re.sub(r'\\(\w+\**)([^\w* \n])', r'\\\1 \2', usfm, flags=re.UNICODE | re.MULTILINE) # \\q1" => \q1 "
+        chapter = re.sub(r'[ \t]*"([^"]+)"[ \t]*', r' "\1" ', chapter, flags=re.UNICODE | re.MULTILINE | re.DOTALL)
+        usfm += '\\c {0}'.format(chapter)
+    usfm = re.sub(r'\\(\w+\**)([^\w* \n])', r'\\\1 \2', usfm, flags=re.UNICODE | re.MULTILINE)  # \\q1" => \q1 "
     usfm = re.sub(r" ' ", r" '", usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r' +([:;.?,!\]})-])', r'\1', usfm, flags=re.UNICODE | re.MULTILINE)
     usfm = re.sub(r'([{(\[-]) +', r'\1', usfm, flags=re.UNICODE | re.MULTILINE)