diff --git a/pypandoc/__init__.py b/pypandoc/__init__.py
index 9cc682f..f0d6a27 100644
--- a/pypandoc/__init__.py
+++ b/pypandoc/__init__.py
@@ -138,15 +138,20 @@ def convert_file(source_file:Union[list, str, Path, Generator], to:str, format:U
     :raises OSError: if pandoc is not found; make sure it has been installed and is available at
             path.
     """
-    # check if we have a working directory
-    # if we don't, we use the current working directory
-    if cworkdir is None:
-        cworkdir = os.getcwd()
-
     # TODO: remove 'encoding' parameter and warning
     if encoding != "utf-8":
         logger.warning("The 'encoding' parameter will be removed in version 1.13. Just remove the parameter, because currently the method does not use it.")
 
+    # This if block effectively adds support for pathlib.Path objects
+    # and generators produced by pathlib.Path().glob().
+    if not isinstance(source_file, str):
+        try:
+            source_file = list(map(str, source_file))
+        except TypeError:
+            source_file = str(source_file)
+
+    if not _identify_path(source_file):
+        raise RuntimeError("source_file is not a valid path")
     if _is_network_path(source_file): # if the source_file is an url
         format = _identify_format_from_path(source_file, format)
         return _convert_input(source_file, format, 'path', to, extra_args=extra_args,
@@ -154,45 +159,13 @@ def convert_file(source_file:Union[list, str, Path, Generator], to:str, format:U
                           verify_format=verify_format, sandbox=sandbox,
                           cworkdir=cworkdir)
 
-    # convert the source file to a path object internally
-    if isinstance(source_file, str):
-        source_file = Path(source_file)
-    elif isinstance(source_file, list):
-        source_file = [Path(x) for x in source_file]
-    elif isinstance(source_file, Generator):
-        source_file = [Path(x) for x in source_file]
-
-
-    # we are basically interested to figure out if its an absolute path or not
-    # if it's not, we want to prefix the working directory
-    # if it's a list, we want to prefix the working directory to each item if it's not an absolute path
-    # if it is, just use the absolute path
-    if isinstance(source_file, list):
-        source_file = [x if x.is_absolute() else Path(cworkdir, x) for x in source_file]
-    elif isinstance(source_file, Generator):
-        source_file = (x if x.is_absolute() else Path(cworkdir, x) for x in source_file)
-    # check ifjust a single path was given
-    elif isinstance(source_file, Path):
-        source_file = source_file if source_file.is_absolute() else Path(cworkdir, source_file)
-
-
     discovered_source_files = []
-    # if we have a list of files, we need to glob them
-    # if we have a single file, we need to glob it
-    # remember that we already converted the source_file to a path object
-    # so for glob.glob use both the dir and file name
-    if isinstance(source_file, list):
-        for single_source in source_file:
-            discovered_source_files.extend(glob.glob(str(single_source)))
-        if discovered_source_files == []:
-            discovered_source_files = source_file
-    else:
-        discovered_source_files.extend(glob.glob(str(source_file)))
-        if discovered_source_files == []:
-            discovered_source_files = [source_file]
+    if isinstance(source_file, str):
+        discovered_source_files += glob.glob(source_file)
+    if isinstance(source_file, list): # a list of possibly file or file patterns. Expand all with glob
+        for filepath in source_file:
+            discovered_source_files.extend(glob.glob(filepath))
 
-    if not _identify_path(discovered_source_files):
-        raise RuntimeError("source_file is not a valid path")
     format = _identify_format_from_path(discovered_source_files[0], format)
     if len(discovered_source_files) == 1:
         discovered_source_files = discovered_source_files[0]