From 328f359135172f077dec24636e6e6884aa8068f8 Mon Sep 17 00:00:00 2001
From: William Ayd <will_ayd@innobi.io>
Date: Sun, 21 Jan 2024 13:43:58 -0500
Subject: [PATCH] JSON read support (#236)

---
 pantab/src/pantab.cpp       |  5 +++--
 pantab/tests/test_reader.py | 44 +++++++++++++++++++++++++++++++++++++
 2 files changed, 47 insertions(+), 2 deletions(-)

diff --git a/pantab/src/pantab.cpp b/pantab/src/pantab.cpp
index 3c6a8c07..68ee83fd 100644
--- a/pantab/src/pantab.cpp
+++ b/pantab/src/pantab.cpp
@@ -666,8 +666,9 @@ static auto arrowTypeFromHyper(const hyperapi::SqlType &sqltype)
         case hyperapi::TypeTag::Oid : return NANOARROW_TYPE_UINT32;
         case hyperapi::TypeTag::Double : return NANOARROW_TYPE_DOUBLE;
         case hyperapi::TypeTag::Bytes : return NANOARROW_TYPE_LARGE_BINARY;
-        case hyperapi::TypeTag::Varchar : case hyperapi::TypeTag::Char :
-            case hyperapi::TypeTag::Text : return NANOARROW_TYPE_LARGE_STRING;
+        case hyperapi::TypeTag::Varchar : case hyperapi::TypeTag::
+        Char : case hyperapi::TypeTag::Text : case hyperapi::TypeTag::
+        Json : return NANOARROW_TYPE_LARGE_STRING;
         case hyperapi::TypeTag::Bool : return NANOARROW_TYPE_BOOL;
         case hyperapi::TypeTag::Date : return NANOARROW_TYPE_DATE32;
         case hyperapi::TypeTag::Timestamp : case hyperapi::TypeTag::
diff --git a/pantab/tests/test_reader.py b/pantab/tests/test_reader.py
index 542395bf..03007e2b 100644
--- a/pantab/tests/test_reader.py
+++ b/pantab/tests/test_reader.py
@@ -1,3 +1,5 @@
+import json
+
 import pandas as pd
 import pandas.testing as tm
 import pytest
@@ -158,3 +160,45 @@ def test_read_oid(tmp_hyper):
 
     result = pantab.frame_from_hyper(tmp_hyper, table=table_name)
     tm.assert_frame_equal(result, expected)
+
+
+def test_read_json(tmp_hyper):
+    # Hyper just uses string to serialize/de-serialize, but we don't have an API
+    # yet for users to control writing JSON. So just testing the read until then
+    column_name = "JSON Column"
+    table_name = tab_api.TableName("public", "table")
+    table = tab_api.TableDefinition(
+        table_name=table_name,
+        columns=[
+            tab_api.TableDefinition.Column(
+                name=column_name,
+                type=tab_api.SqlType.json(),
+                nullability=tab_api.NOT_NULLABLE,
+            )
+        ],
+    )
+
+    with tab_api.HyperProcess(
+        telemetry=tab_api.Telemetry.DO_NOT_SEND_USAGE_DATA_TO_TABLEAU
+    ) as hyper:
+        with tab_api.Connection(
+            endpoint=hyper.endpoint,
+            database=tmp_hyper,
+            create_mode=tab_api.CreateMode.CREATE_AND_REPLACE,
+        ) as connection:
+            connection.catalog.create_table(table_definition=table)
+
+            with tab_api.Inserter(connection, table) as inserter:
+                inserter.add_rows(
+                    [[json.dumps({"foo": 42})], [json.dumps({"bar": -42})]]
+                )
+                inserter.execute()
+
+    expected = pd.DataFrame(
+        [[json.dumps({"foo": 42})], [json.dumps({"bar": -42})]],
+        columns=[column_name],
+        dtype="large_string[pyarrow]",
+    )
+
+    result = pantab.frame_from_hyper(tmp_hyper, table=table_name)
+    tm.assert_frame_equal(result, expected)