From c6f49c6c01fb8f9f46d48ef2f9dcb7e342ce262b Mon Sep 17 00:00:00 2001
From: Jacob Mims <122570226+jtmims@users.noreply.github.com>
Date: Tue, 23 Apr 2024 11:06:56 -0400
Subject: [PATCH 1/4] add start_time and end_time attrs to catalog_builder.py

---
 tools/catalog_builder/catalog_builder.py | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/tools/catalog_builder/catalog_builder.py b/tools/catalog_builder/catalog_builder.py
index a83ebe1ab..de5c8c6a1 100644
--- a/tools/catalog_builder/catalog_builder.py
+++ b/tools/catalog_builder/catalog_builder.py
@@ -73,6 +73,8 @@ def parse_gfdl_pp_ts(file_name: str):
         cell_methods = ""
         cell_measures = ""
         time_range = split[1]
+        start_time = time_range.split('-')[0]
+        end_time = time_range.split('-')[1]
         variable_id = split[2]
         source_type = ""
         member_id = ""
@@ -141,6 +143,8 @@ def parse_gfdl_pp_ts(file_name: str):
                 'grid_label': grid_label,
                 'units': units,
                 'time_range': time_range,
+                'start_time': start_time,
+                'end_time': end_time,
                 'chunk_freq': chunk_freq,
                 'standard_name': standard_name,
                 'long_name': long_name,

From ddd7b472dc85364afa5a45b82d2db0ef8845100d Mon Sep 17 00:00:00 2001
From: Jacob Mims <122570226+jtmims@users.noreply.github.com>
Date: Fri, 26 Apr 2024 12:04:11 -0400
Subject: [PATCH 2/4] add int to date conversion in check_group_daterange in
 preprocessor.py

---
 src/preprocessor.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/src/preprocessor.py b/src/preprocessor.py
index 01b2664ac..6ff72873e 100644
--- a/src/preprocessor.py
+++ b/src/preprocessor.py
@@ -797,6 +797,16 @@ def check_group_daterange(self, group_df: pd.DataFrame, log=_log) -> pd.DataFram
         """
         date_col = "date_range"
         try:
+            # convert int to date type
+            date_format = ''
+            date_digits = math.floor(math.log10(group_df['start_time'].values[0]))+1
+            match date_digits:
+                case 8:
+                    date_format = '%Y%m%d'
+                case 14:
+                    date_format = '%Y%m%d%H%M%S'
+            group_df['start_time'] = pd.to_datetime(group_df['start_time'].values[0], format=date_format)
+            group_df['end_time'] = pd.to_datetime(group_df['end_time'].values[0], format=date_format)
             # method throws ValueError if ranges aren't contiguous
             dates_df = group_df.loc[:, ['start_time', 'end_time']]
             date_range_vals = []

From 58aafd8d8cd04de9b4c701c97d9ba9c0627e8cb5 Mon Sep 17 00:00:00 2001
From: Jacob Mims <122570226+jtmims@users.noreply.github.com>
Date: Fri, 26 Apr 2024 15:18:18 -0400
Subject: [PATCH 3/4] add conditional to check for date type

---
 src/preprocessor.py | 21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)

diff --git a/src/preprocessor.py b/src/preprocessor.py
index 6ff72873e..48284aa60 100644
--- a/src/preprocessor.py
+++ b/src/preprocessor.py
@@ -797,16 +797,17 @@ def check_group_daterange(self, group_df: pd.DataFrame, log=_log) -> pd.DataFram
         """
         date_col = "date_range"
         try:
-            # convert int to date type
-            date_format = ''
-            date_digits = math.floor(math.log10(group_df['start_time'].values[0]))+1
-            match date_digits:
-                case 8:
-                    date_format = '%Y%m%d'
-                case 14:
-                    date_format = '%Y%m%d%H%M%S'
-            group_df['start_time'] = pd.to_datetime(group_df['start_time'].values[0], format=date_format)
-            group_df['end_time'] = pd.to_datetime(group_df['end_time'].values[0], format=date_format)
+            if not isinstance(group_df['start_time'], datetime.date):
+                # convert int to date type
+                date_format = ''
+                date_digits = math.floor(math.log10(group_df['start_time'].values[0]))+1
+                match date_digits:
+                    case 8:
+                        date_format = '%Y%m%d'
+                    case 14:
+                        date_format = '%Y%m%d%H%M%S'
+                group_df['start_time'] = pd.to_datetime(group_df['start_time'].values[0], format=date_format)
+                group_df['end_time'] = pd.to_datetime(group_df['end_time'].values[0], format=date_format)
             # method throws ValueError if ranges aren't contiguous
             dates_df = group_df.loc[:, ['start_time', 'end_time']]
             date_range_vals = []

From 92d09b5f8ca8f87aae95bbba40932e08217d2cfb Mon Sep 17 00:00:00 2001
From: Jacob Mims <122570226+jtmims@users.noreply.github.com>
Date: Fri, 26 Apr 2024 15:19:53 -0400
Subject: [PATCH 4/4] point to pd value

---
 src/preprocessor.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/preprocessor.py b/src/preprocessor.py
index 48284aa60..960f8e25f 100644
--- a/src/preprocessor.py
+++ b/src/preprocessor.py
@@ -797,7 +797,7 @@ def check_group_daterange(self, group_df: pd.DataFrame, log=_log) -> pd.DataFram
         """
         date_col = "date_range"
         try:
-            if not isinstance(group_df['start_time'], datetime.date):
+            if not isinstance(group_df['start_time'].values[0], datetime.date):
                 # convert int to date type
                 date_format = ''
                 date_digits = math.floor(math.log10(group_df['start_time'].values[0]))+1