Fix incorrect submobject count of multi-part Tex/MathTex mobjects by stopping them from adding empty submobjects

manim/mobject/text/tex_mobject.py

@@ -344,10 +344,6 @@ def _break_up_by_substrings(self):
                 curr_index + num_submobs + len("".join(self.arg_separator.split()))
             )
             if num_submobs == 0:
-                # For cases like empty tex_strings, we want the corresponding
-                # part of the whole MathTex to be a VectorizedPoint
-                # positioned in the right part of the MathTex
-                sub_tex_mob.submobjects = [VectorizedPoint()]
                 last_submob_index = min(curr_index, len(self.submobjects) - 1)
                 sub_tex_mob.move_to(self.submobjects[last_submob_index], RIGHT)
             else:

tests/module/mobject/text/test_texmobject.py

@@ -2,6 +2,7 @@
 
 from pathlib import Path
 
+import numpy as np
 import pytest
 
 from manim import MathTex, SingleStringMathTex, Tex, TexTemplate, config, tempconfig
@@ -96,6 +97,23 @@ def test_tex_white_space_and_non_whitespace_args():
     assert len(tex[3]) == len("".join(str_part_4.split()))
 
 
+def test_multi_part_tex_with_empty_parts():
+    """Check that if a Tex or MathTex Mobject with multiple
+    string arguments is created where some of the parts render
+    as empty SVGs, then the number of family members with points
+    should still be the same as the snipped in one singular part.
+    """
+    tex_parts = ["(-1)", "^{", "0}"]
+    one_part_fomula = MathTex("".join(tex_parts))
+    multi_part_formula = MathTex(*tex_parts)
+
+    for one_part_glyph, multi_part_glyph in zip(
+        one_part_fomula.family_members_with_points(),
+        multi_part_formula.family_members_with_points(),
+    ):
+        np.testing.assert_allclose(one_part_glyph.points, multi_part_glyph.points)
+
+
 def test_tex_size():
     """Check that the size of a :class:`Tex` string is not changed."""
     text = Tex("what").center()