From 4b03fd01002e6bc486715a0ae693dfaa8fcdcd58 Mon Sep 17 00:00:00 2001
From: kukushking <3997468+kukushking@users.noreply.github.com>
Date: Wed, 30 Jun 2021 15:30:09 +0100
Subject: [PATCH 1/2] Add dataset headers in wr.s3.to_csv

---
 awswrangler/s3/_write_text.py |  4 ++--
 tests/test_s3_text.py         | 11 +++++++++++
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/awswrangler/s3/_write_text.py b/awswrangler/s3/_write_text.py
index 6061798b5..37e2db248 100644
--- a/awswrangler/s3/_write_text.py
+++ b/awswrangler/s3/_write_text.py
@@ -456,7 +456,7 @@ def to_csv(  # pylint: disable=too-many-arguments,too-many-locals,too-many-state
         if database and table:
             quoting: Optional[int] = csv.QUOTE_NONE
             escapechar: Optional[str] = "\\"
-            header: Union[bool, List[str]] = False
+            header: Union[bool, List[str]] = pandas_kwargs.get("header", False)
             date_format: Optional[str] = "%Y-%m-%d %H:%M:%S.%f"
             pd_kwargs: Dict[str, Any] = {}
             compression: Optional[str] = pandas_kwargs.get("compression", None)
@@ -529,7 +529,7 @@ def to_csv(  # pylint: disable=too-many-arguments,too-many-locals,too-many-state
                     catalog_table_input=catalog_table_input,
                     catalog_id=catalog_id,
                     compression=pandas_kwargs.get("compression"),
-                    skip_header_line_count=None,
+                    skip_header_line_count=True if header else None,
                     serde_library=serde_library,
                     serde_parameters=serde_parameters,
                 )
diff --git a/tests/test_s3_text.py b/tests/test_s3_text.py
index 185d48d0d..370ce2f63 100644
--- a/tests/test_s3_text.py
+++ b/tests/test_s3_text.py
@@ -119,6 +119,17 @@ def test_csv(path):
         wr.s3.read_csv(path=paths, iterator=True)
 
 
+@pytest.mark.parametrize("header", [True, ["identifier"]])
+def test_csv_dataset_header(path, header):
+    df0 = pd.DataFrame({"id": [1, 2, 3]})
+    path0 = f"{path}test_csv_dataset0.csv"
+    wr.s3.to_csv(df=df0, path=path0, dataset=True, index=False, header=header)
+    df1 = wr.s3.read_csv(path=path0)
+    if isinstance(header, list):
+        df0.columns = header
+    assert df0.equals(df1)
+
+
 def test_json(path):
     df0 = pd.DataFrame({"id": [1, 2, 3]})
     path0 = f"{path}test_json0.json"

From c0424971788a0c986be59b6c2c80911d9f06b223 Mon Sep 17 00:00:00 2001
From: kukushking <3997468+kukushking@users.noreply.github.com>
Date: Thu, 1 Jul 2021 11:46:53 +0100
Subject: [PATCH 2/2] Add test case

---
 tests/test_s3_text.py | 41 ++++++++++++++++++++++++++++++++++++++---
 1 file changed, 38 insertions(+), 3 deletions(-)

diff --git a/tests/test_s3_text.py b/tests/test_s3_text.py
index 370ce2f63..e730caa93 100644
--- a/tests/test_s3_text.py
+++ b/tests/test_s3_text.py
@@ -120,16 +120,51 @@ def test_csv(path):
 
 
 @pytest.mark.parametrize("header", [True, ["identifier"]])
-def test_csv_dataset_header(path, header):
-    df0 = pd.DataFrame({"id": [1, 2, 3]})
+def test_csv_dataset_header(path, header, glue_database, glue_table):
     path0 = f"{path}test_csv_dataset0.csv"
-    wr.s3.to_csv(df=df0, path=path0, dataset=True, index=False, header=header)
+    df0 = pd.DataFrame({"id": [1, 2, 3]})
+    wr.s3.to_csv(
+        df=df0,
+        path=path0,
+        dataset=True,
+        database=glue_database,
+        table=glue_table,
+        index=False,
+        header=header,
+    )
     df1 = wr.s3.read_csv(path=path0)
     if isinstance(header, list):
         df0.columns = header
     assert df0.equals(df1)
 
 
+@pytest.mark.parametrize("mode", ["append", "overwrite"])
+def test_csv_dataset_header_modes(path, mode, glue_database, glue_table):
+    path0 = f"{path}test_csv_dataset0.csv"
+    dfs = [
+        pd.DataFrame({"id": [1, 2, 3]}),
+        pd.DataFrame({"id": [4, 5, 6]}),
+    ]
+    for df in dfs:
+        wr.s3.to_csv(
+            df=df,
+            path=path0,
+            dataset=True,
+            database=glue_database,
+            table=glue_table,
+            mode=mode,
+            index=False,
+            header=True,
+        )
+    dfs_conc = pd.concat(dfs)
+    df_res = wr.s3.read_csv(path=path0)
+
+    if mode == "append":
+        assert len(df_res) == len(dfs_conc)
+    else:
+        assert df_res.equals(dfs[-1])
+
+
 def test_json(path):
     df0 = pd.DataFrame({"id": [1, 2, 3]})
     path0 = f"{path}test_json0.json"