Raise FileNotFoundError in read_json if input looks like file path but file is missing (pandas-dev#46718)

* raise FileNotFoundError in _get_data_from_filepath()

* update tests test_read_non_existent + test_read_expands_user_home_dir

* add changelog entry in doc/source/whatsnew/v1.5.0.rst

* use pandas.io.common._compression_to_extension instead of hard-coded extensions

* move changelog entry from IO to other API changes

* fix ImportError from _compression_to_extension -> _extension_to_compression rename

* add test read_json very long file path

* remove extra period in extension checking

Co-authored-by: Matthew Roeschke <[email protected]>

Author: 2 people

doc/source/whatsnew/v1.5.0.rst

@@ -435,6 +435,7 @@ Other API changes
   <https://developers.googleblog.com/2022/02/making-oauth-flows-safer.html?m=1#disallowed-oob>`_.
   The ``auth_local_webserver = False`` option is planned to stop working in
   October 2022. (:issue:`46312`)
+- :func:`read_json` now raises ``FileNotFoundError`` (previously ``ValueError``) when input is a string ending in ``.json``, ``.json.gz``, ``.json.bz2``, etc. but no such file exists. (:issue:`29102`)
 -
 
 .. ---------------------------------------------------------------------------

pandas/io/json/_json.py

@@ -52,6 +52,7 @@
 
 from pandas.io.common import (
     IOHandles,
+    _extension_to_compression,
     file_exists,
     get_handle,
     is_fsspec_url,
@@ -698,6 +699,9 @@ def _get_data_from_filepath(self, filepath_or_buffer):
 
         This method turns (1) into (2) to simplify the rest of the processing.
         It returns input types (2) and (3) unchanged.
+
+        It raises FileNotFoundError if the input is a string ending in
+        one of .json, .json.gz, .json.bz2, etc. but no such file exists.
         """
         # if it is a string but the file does not exist, it might be a JSON string
         filepath_or_buffer = stringify_path(filepath_or_buffer)
@@ -716,6 +720,14 @@ def _get_data_from_filepath(self, filepath_or_buffer):
                 errors=self.encoding_errors,
             )
             filepath_or_buffer = self.handles.handle
+        elif (
+            isinstance(filepath_or_buffer, str)
+            and filepath_or_buffer.lower().endswith(
+                (".json",) + tuple(f".json{c}" for c in _extension_to_compression)
+            )
+            and not file_exists(filepath_or_buffer)
+        ):
+            raise FileNotFoundError(f"File {filepath_or_buffer} does not exist")
 
         return filepath_or_buffer
 

pandas/tests/io/json/test_pandas.py

@@ -1566,6 +1566,20 @@ def test_read_json_with_url_value(self, url):
         expected = DataFrame({"url": [url]})
         tm.assert_frame_equal(result, expected)
 
+    @pytest.mark.parametrize(
+        "compression",
+        ["", ".gz", ".bz2", ".tar"],
+    )
+    def test_read_json_with_very_long_file_path(self, compression):
+        # GH 46718
+        long_json_path = f'{"a" * 1000}.json{compression}'
+        with pytest.raises(
+            FileNotFoundError, match=f"File {long_json_path} does not exist"
+        ):
+            # path too long for Windows is handled in file_exists() but raises in
+            # _get_data_from_filepath()
+            read_json(long_json_path)
+
     @pytest.mark.parametrize(
         "date_format,key", [("epoch", 86400000), ("iso", "P1DT0H0M0S")]
     )

pandas/tests/io/test_common.py

@@ -187,7 +187,7 @@ def test_iterator(self):
             (pd.read_hdf, "tables", FileNotFoundError, "h5"),
             (pd.read_stata, "os", FileNotFoundError, "dta"),
             (pd.read_sas, "os", FileNotFoundError, "sas7bdat"),
-            (pd.read_json, "os", ValueError, "json"),
+            (pd.read_json, "os", FileNotFoundError, "json"),
             (pd.read_pickle, "os", FileNotFoundError, "pickle"),
         ],
     )
@@ -253,7 +253,7 @@ def test_write_missing_parent_directory(self, method, module, error_class, fn_ex
             (pd.read_hdf, "tables", FileNotFoundError, "h5"),
             (pd.read_stata, "os", FileNotFoundError, "dta"),
             (pd.read_sas, "os", FileNotFoundError, "sas7bdat"),
-            (pd.read_json, "os", ValueError, "json"),
+            (pd.read_json, "os", FileNotFoundError, "json"),
             (pd.read_pickle, "os", FileNotFoundError, "pickle"),
         ],
     )