BUG: fix duplicate entries in LaTeX List of Tables when using longtable environments

doc/source/whatsnew/v1.2.0.rst

@@ -342,7 +342,7 @@ Other
 ^^^^^
 - Bug in :meth:`DataFrame.replace` and :meth:`Series.replace` incorrectly raising ``AssertionError`` instead of ``ValueError`` when invalid parameter combinations are passed (:issue:`36045`)
 - Bug in :meth:`DataFrame.replace` and :meth:`Series.replace` with numeric values and string ``to_replace`` (:issue:`34789`)
--
+- Bug in :func:`LongTableBuilder.middle_separator` was duplicating LaTeX longtable entires in the List of Tables of a LaTeX document (:issue:`34360`)
 
 .. ---------------------------------------------------------------------------
 

pandas/io/formats/latex.py

@@ -476,7 +476,16 @@ def _caption_and_label(self) -> str:
     @property
     def middle_separator(self) -> str:
         iterator = self._create_row_iterator(over="header")
+
+        # the content between \endfirsthead and \endhead commands
+        # mitigates repeated List of Tables entries in the final LaTeX
+        # document when dealing with longtable environments; GH #34360
         elements = [
+            "\\midrule",
+            "\\endfirsthead",
+            f"\\caption[]{{{self.caption}}} \\\\" if self.caption else "",
+            self.top_separator,
+            self.header,
             "\\midrule",
             "\\endhead",
             "\\midrule",

pandas/tests/io/formats/test_to_latex.py

@@ -411,6 +411,11 @@ def test_to_latex_longtable(self):
         df = DataFrame({"a": [1, 2], "b": ["b1", "b2"]})
         withindex_result = df.to_latex(longtable=True)
         withindex_expected = r"""\begin{longtable}{lrl}
+\toprule
+{} &  a &   b \\
+\midrule
+\endfirsthead
+
 \toprule
 {} &  a &   b \\
 \midrule
@@ -430,6 +435,11 @@ def test_to_latex_longtable(self):
 
         withoutindex_result = df.to_latex(index=False, longtable=True)
         withoutindex_expected = r"""\begin{longtable}{rl}
+\toprule
+ a &  b \\
+\midrule
+\endfirsthead
+
 \toprule
  a &  b \\
 \midrule
@@ -525,6 +535,9 @@ def test_to_latex_longtable_caption_label(self):
 
         df = DataFrame({"a": [1, 2], "b": ["b1", "b2"]})
 
+        # test when no caption and no label is provided
+        # is performed by test_to_latex_longtable()
+
         # test when only the caption is provided
         result_c = df.to_latex(longtable=True, caption=the_caption)
 
@@ -533,6 +546,11 @@ def test_to_latex_longtable_caption_label(self):
 \toprule
 {} &  a &   b \\
 \midrule
+\endfirsthead
+\caption[]{a table in a \texttt{longtable} environment} \\
+\toprule
+{} &  a &   b \\
+\midrule
 \endhead
 \midrule
 \multicolumn{3}{r}{{Continued on next page}} \\
@@ -552,6 +570,11 @@ def test_to_latex_longtable_caption_label(self):
 
         expected_l = r"""\begin{longtable}{lrl}
 \label{tab:longtable}\\
+\toprule
+{} &  a &   b \\
+\midrule
+\endfirsthead
+
 \toprule
 {} &  a &   b \\
 \midrule
@@ -578,6 +601,11 @@ def test_to_latex_longtable_caption_label(self):
 \toprule
 {} &  a &   b \\
 \midrule
+\endfirsthead
+\caption[]{a table in a \texttt{longtable} environment} \\
+\toprule
+{} &  a &   b \\
+\midrule
 \endhead
 \midrule
 \multicolumn{3}{r}{{Continued on next page}} \\
@@ -623,6 +651,11 @@ def test_to_latex_longtable_position(self):
         result_p = df.to_latex(longtable=True, position=the_position)
 
         expected_p = r"""\begin{longtable}[t]{lrl}
+\toprule
+{} &  a &   b \\
+\midrule
+\endfirsthead
+
 \toprule
 {} &  a &   b \\
 \midrule