[pyarrow] Add skip_empty=True option to merging.

law/contrib/pyarrow/util.py

@@ -26,6 +26,7 @@ def merge_parquet_files(
     callback: Callable[[int], Any] | None = None,
     writer_opts: dict[str, Any] | None = None,
     copy_single: bool = False,
+    skip_empty: bool = True,
 ) -> str:
     """
     Merges parquet files in *src_paths* into a new file at *dst_path*. Intermediate directories are
@@ -35,7 +36,8 @@ def merge_parquet_files(
     *callback* can refer to a callable accepting a single integer argument representing the index of
     the file after it was merged. *writer_opts* can be a dictionary of keyword arguments that are
     passed to the *ParquetWriter* instance. When *src_paths* contains only a single file and
-    *copy_single* is *True*, the file is copied to *dst_path* and no merging takes place.
+    *copy_single* is *True*, the file is copied to *dst_path* and no merging takes place. Files
+    containing empty tables are skipped unless *skip_empty* is *False*.
 
     The absolute, expanded *dst_path* is returned.
     """
@@ -81,7 +83,9 @@ def merge_parquet_files(
 
         # write the remaining ones
         for i, path in enumerate(src_paths[1:], 1):
-            writer.write_table(pq.read_table(path))
+            _table = pq.read_table(path)
+            if not skip_empty or _table.num_rows > 0:
+                writer.write_table(_table)
             callback(i)
 
     return dst_path
@@ -94,8 +98,7 @@ def merge_parquet_task(
     local: bool = False,
     cwd: str | pathlib.Path | LocalDirectoryTarget | None = None,
     force: bool = True,
-    writer_opts: dict[str, Any] | None = None,
-    copy_single: bool = False,
+    **kwargs: Any,
 ) -> None:
     """
     This method is intended to be used by tasks that are supposed to merge parquet files, e.g. when
@@ -108,8 +111,8 @@ def merge_parquet_task(
     :py:meth:`law.Task.publish_step` methods. When *force* is *True*, any existing output file is
     overwritten.
 
-    *writer_opts* and *copy_single* are forwarded to :py:func:`merge_parquet_files` which is used
-    internally for the actual merging.
+    All additional *kwargs* are forwarded to :py:func:`merge_parquet_files` which is used internally
+    for the actual merging.
     """
     abspath = lambda p: os.path.abspath(os.path.expandvars(os.path.expanduser(get_path(p))))
 
@@ -130,12 +133,7 @@ def merge(inputs, output):
                 output.remove()
 
             # merge
-            merge_parquet_files(
-                [inp.abspath for inp in inputs],
-                output.abspath,
-                writer_opts=writer_opts,
-                copy_single=copy_single,
-            )
+            merge_parquet_files([inp.abspath for inp in inputs], output.abspath, **kwargs)
 
         stat = output.exists(stat=True)
         if not stat: