From 1bda3bdf9fa107bbb85d6f1ef8a3fa1c6f3ec859 Mon Sep 17 00:00:00 2001
From: Phillip Cloud <417981+cpcloud@users.noreply.github.com>
Date: Sun, 20 Aug 2023 07:15:21 -0400
Subject: [PATCH] feat(polars): support reading ndjson

---
 ibis/backends/polars/__init__.py     | 33 +++++++++++++++++++++++++++-
 ibis/backends/snowflake/__init__.py  |  6 ++---
 ibis/backends/tests/test_register.py |  2 ++
 3 files changed, 37 insertions(+), 4 deletions(-)

diff --git a/ibis/backends/polars/__init__.py b/ibis/backends/polars/__init__.py
index c13c10c5bb24..fbda39b7e72c 100644
--- a/ibis/backends/polars/__init__.py
+++ b/ibis/backends/polars/__init__.py
@@ -144,7 +144,7 @@ def sql(
     def read_csv(
         self, path: str | Path, table_name: str | None = None, **kwargs: Any
     ) -> ir.Table:
-        """Register a CSV file as a table in the current database.
+        """Register a CSV file as a table.
 
         Parameters
         ----------
@@ -172,6 +172,37 @@ def read_csv(
             self._add_table(table_name, pl.read_csv(path, **kwargs))
         return self.table(table_name)
 
+    def read_json(
+        self, path: str | Path, table_name: str | None = None, **kwargs: Any
+    ) -> ir.Table:
+        """Register a JSON file as a table.
+
+        Parameters
+        ----------
+        path
+            A string or Path to a JSON file; globs are supported
+        table_name
+            An optional name to use for the created table. This defaults to
+            a sequentially generated name.
+        **kwargs
+            Additional keyword arguments passed to Polars loading function.
+            See https://pola-rs.github.io/polars/py-polars/html/reference/api/polars.scan_ndjson.html
+            for more information.
+
+        Returns
+        -------
+        ir.Table
+            The just-registered table
+        """
+        path = normalize_filename(path)
+        table_name = table_name or gen_name("read_json")
+        try:
+            self._add_table(table_name, pl.scan_ndjson(path, **kwargs))
+        except pl.exceptions.ComputeError:
+            # handles compressed json files
+            self._add_table(table_name, pl.read_ndjson(path, **kwargs))
+        return self.table(table_name)
+
     def read_delta(
         self, path: str | Path, table_name: str | None = None, **kwargs: Any
     ) -> ir.Table:
diff --git a/ibis/backends/snowflake/__init__.py b/ibis/backends/snowflake/__init__.py
index d5ad72ba3c66..fbdce8b8b7e7 100644
--- a/ibis/backends/snowflake/__init__.py
+++ b/ibis/backends/snowflake/__init__.py
@@ -615,7 +615,7 @@ def read_csv(
         Parameters
         ----------
         path
-            Path to the CSV file
+            A string or Path to a CSV file; globs are supported
         table_name
             Optional name for the table; if not passed, a random name will be generated
         kwargs
@@ -711,7 +711,7 @@ def read_json(
         Parameters
         ----------
         path
-            File or list of files
+            A string or Path to a JSON file; globs are supported
         table_name
             Optional table name
         kwargs
@@ -786,7 +786,7 @@ def read_parquet(
         Parameters
         ----------
         path
-            Path to a Parquet file
+            A string or Path to a Parquet file; globs are supported
         table_name
             Optional table name
         kwargs
diff --git a/ibis/backends/tests/test_register.py b/ibis/backends/tests/test_register.py
index be695b6550c0..3732609e6cb1 100644
--- a/ibis/backends/tests/test_register.py
+++ b/ibis/backends/tests/test_register.py
@@ -509,7 +509,9 @@ def test_read_csv_glob(con, tmp_path, ft_data):
 @pytest.mark.notyet(
     [
         "bigquery",
+        "clickhouse",
         "dask",
+        "datafusion",
         "impala",
         "mssql",
         "mysql",