From 8de37e96aafd688d99b1f20afb458c9daf6e9df2 Mon Sep 17 00:00:00 2001
From: taoufik <abacidtaoufik@gmail.com>
Date: Mon, 6 Jan 2020 11:39:51 +0100
Subject: [PATCH 1/4] Fix read_json category dtype

---
 pandas/io/json/_json.py             |  5 +++--
 pandas/tests/io/json/test_pandas.py | 12 ++++++++++++
 2 files changed, 15 insertions(+), 2 deletions(-)

diff --git a/pandas/io/json/_json.py b/pandas/io/json/_json.py
index 7f2aab569ab71..4270f312d56a3 100644
--- a/pandas/io/json/_json.py
+++ b/pandas/io/json/_json.py
@@ -12,7 +12,7 @@
 from pandas._typing import JSONSerializable
 from pandas.errors import AbstractMethodError
 
-from pandas.core.dtypes.common import ensure_str, is_period_dtype
+from pandas.core.dtypes.common import ensure_str, is_period_dtype, is_categorical_dtype
 
 from pandas import DataFrame, MultiIndex, Series, isna, to_datetime
 from pandas.core.construction import create_series_with_explicit_dtype
@@ -892,7 +892,8 @@ def _try_convert_data(self, name, data, use_dtypes=True, convert_dates=True):
                 )
                 if dtype is not None:
                     try:
-                        dtype = np.dtype(dtype)
+                        if not is_categorical_dtype(dtype):
+                            dtype = np.dtype(dtype)
                         return data.astype(dtype), True
                     except (TypeError, ValueError):
                         return data, False
diff --git a/pandas/tests/io/json/test_pandas.py b/pandas/tests/io/json/test_pandas.py
index 09d8a1d3f10ea..16429890e0606 100644
--- a/pandas/tests/io/json/test_pandas.py
+++ b/pandas/tests/io/json/test_pandas.py
@@ -1197,6 +1197,18 @@ def test_read_local_jsonl(self):
             expected = DataFrame([[1, 2], [1, 2]], columns=["a", "b"])
             tm.assert_frame_equal(result, expected)
 
+    def test_read_json_category_dtype(self):
+        json = (
+            '{"a": 0, "b": "A"}\n'
+            '{"a": 1, "b": "B"}\n'
+            '{"a": 2, "b": "B"}\n'
+            '{"a": 3, "b": "B"}\n'
+        )
+        json = StringIO(json)
+        result = read_json(json, lines=True, dtype={"a": "category"})
+        expected = DataFrame([[0, "foo"], [1, "bar"], [2, "foo"], [3, "bar"]])
+        tm.assert_frame_equal(result, expected)
+
     def test_read_jsonl_unicode_chars(self):
         # GH15132: non-ascii unicode characters
         # \u201d == RIGHT DOUBLE QUOTATION MARK

From 3f99f04e18d00f5a98678c674eb7f712ec1868c7 Mon Sep 17 00:00:00 2001
From: taoufik <abacidtaoufik@gmail.com>
Date: Mon, 6 Jan 2020 12:08:55 +0100
Subject: [PATCH 2/4] Use pandas_dtype instead of np.dtype

---
 pandas/io/json/_json.py | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/pandas/io/json/_json.py b/pandas/io/json/_json.py
index 4270f312d56a3..5c95ef48c8e9b 100644
--- a/pandas/io/json/_json.py
+++ b/pandas/io/json/_json.py
@@ -12,7 +12,12 @@
 from pandas._typing import JSONSerializable
 from pandas.errors import AbstractMethodError
 
-from pandas.core.dtypes.common import ensure_str, is_period_dtype, is_categorical_dtype
+from pandas.core.dtypes.common import (
+    ensure_str,
+    is_period_dtype,
+    is_categorical_dtype,
+    pandas_dtype,
+)
 
 from pandas import DataFrame, MultiIndex, Series, isna, to_datetime
 from pandas.core.construction import create_series_with_explicit_dtype
@@ -893,7 +898,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, convert_dates=True):
                 if dtype is not None:
                     try:
                         if not is_categorical_dtype(dtype):
-                            dtype = np.dtype(dtype)
+                            dtype = pandas_dtype(dtype)
                         return data.astype(dtype), True
                     except (TypeError, ValueError):
                         return data, False

From e92b4cb9cba7abbc7450f253d1f0e0f7fbbe719b Mon Sep 17 00:00:00 2001
From: taoufik <abacidtaoufik@gmail.com>
Date: Mon, 6 Jan 2020 12:23:15 +0100
Subject: [PATCH 3/4] Add tests

---
 pandas/tests/io/json/test_pandas.py | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/pandas/tests/io/json/test_pandas.py b/pandas/tests/io/json/test_pandas.py
index 16429890e0606..686a0ce99f9aa 100644
--- a/pandas/tests/io/json/test_pandas.py
+++ b/pandas/tests/io/json/test_pandas.py
@@ -8,6 +8,7 @@
 import pytest
 
 from pandas.compat import is_platform_32bit, is_platform_windows
+from pandas.core.dtypes.common import is_categorical
 import pandas.util._test_decorators as td
 
 import pandas as pd
@@ -1201,13 +1202,12 @@ def test_read_json_category_dtype(self):
         json = (
             '{"a": 0, "b": "A"}\n'
             '{"a": 1, "b": "B"}\n'
-            '{"a": 2, "b": "B"}\n'
+            '{"a": 2, "b": "A"}\n'
             '{"a": 3, "b": "B"}\n'
         )
         json = StringIO(json)
-        result = read_json(json, lines=True, dtype={"a": "category"})
-        expected = DataFrame([[0, "foo"], [1, "bar"], [2, "foo"], [3, "bar"]])
-        tm.assert_frame_equal(result, expected)
+        result = read_json(json, lines=True, dtype={"b": "category"})
+        assert is_categorical(result["b"])
 
     def test_read_jsonl_unicode_chars(self):
         # GH15132: non-ascii unicode characters

From de216026e8e24216a78cb6cbd96ff3fbb8ae10f8 Mon Sep 17 00:00:00 2001
From: taoufik <abacidtaoufik@gmail.com>
Date: Mon, 6 Jan 2020 13:30:39 +0100
Subject: [PATCH 4/4] Sort imports

---
 pandas/io/json/_json.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pandas/io/json/_json.py b/pandas/io/json/_json.py
index 5c95ef48c8e9b..8d44a54fbd988 100644
--- a/pandas/io/json/_json.py
+++ b/pandas/io/json/_json.py
@@ -13,10 +13,10 @@
 from pandas.errors import AbstractMethodError
 
 from pandas.core.dtypes.common import (
+    pandas_dtype,
     ensure_str,
     is_period_dtype,
     is_categorical_dtype,
-    pandas_dtype,
 )
 
 from pandas import DataFrame, MultiIndex, Series, isna, to_datetime