From 063618e5ce037f4426df7425f4a55da0d738fa60 Mon Sep 17 00:00:00 2001
From: Davis Muro <dmuro@ona.io>
Date: Tue, 17 Jan 2023 13:30:01 +0300
Subject: [PATCH] Update `sheet_to_csv` function

- Ensure headers are not duplicated
- Ensure that integers & datetimes are correctly handler.
- Ensure empty values are not processed
---
 onadata/apps/viewer/models/data_dictionary.py | 49 ++++++++-----------
 1 file changed, 21 insertions(+), 28 deletions(-)

diff --git a/onadata/apps/viewer/models/data_dictionary.py b/onadata/apps/viewer/models/data_dictionary.py
index 4de9a2eeaa..73122124df 100644
--- a/onadata/apps/viewer/models/data_dictionary.py
+++ b/onadata/apps/viewer/models/data_dictionary.py
@@ -5,12 +5,13 @@
 import os
 from io import BytesIO, StringIO
 
-import unicodecsv as csv
-import openpyxl
 from django.core.files.uploadedfile import InMemoryUploadedFile
 from django.db.models.signals import post_save, pre_save
 from django.utils import timezone
 from django.utils.translation import gettext as _
+
+import openpyxl
+import unicodecsv as csv
 from floip import FloipSurvey
 from kombu.exceptions import OperationalError
 from pyxform.builder import create_survey_element_from_dict
@@ -89,38 +90,30 @@ def sheet_to_csv(xls_content, sheet_name):
     header = [v for v, m in zip(list(sheet.values)[0], mask) if m]
     writer.writerow(header)
 
-    name_column = None
-    try:
-        name_column = header.index("name")
-    except ValueError:
-        pass
-
-    integer_fields = False
-    date_fields = False
-    if name_column:
-        for index in range(1, sheet.max_column):
-            if sheet.cell(index, name_column).data_type == "n":
-                integer_fields = True
-            elif sheet.cell(index, name_column).is_date:
-                date_fields = True
-
-    for row, value in enumerate(sheet.iter_rows()):
-        if integer_fields or date_fields:
-            # convert integers to string/datetime if name has numbers/dates
-            row_values = []
-            for index, val in enumerate(value):
-                if sheet.cell(row, index).data_type == "n":
+    # Start at 1 since the columns & rows in a sheet are 1-based arrays
+    for row, value in enumerate(sheet.iter_rows(values_only=True), start=1):
+        # Skip first row as it's just the headers
+        if row > 1:
+            continue
+
+        row_values = []
+
+        # Start at 1 since the columns & rows in a sheet a 1-based arrays
+        for col, val in enumerate(value, start=1):
+            cell = sheet.cell(row, col)
+            if val is not None:
+                # Try and convert cell to string if it's numeric or a date
+                if cell.data_type == "n":
                     try:
                         val = str(float(val) if (float(val) > int(val)) else int(val))
                     except ValueError:
                         pass
-                elif sheet.cell(row, index).is_date:
+                elif cell.is_date:
                     val = val.strftime("%Y-%m-%d").isoformat()
+
                 row_values.append(val)
-            writer.writerow([v for v, m in zip(row_values, mask) if m])
-        else:
-            single_row = [cell.value for cell in value]
-            writer.writerow([v for v, m in zip(single_row, mask) if m])
+
+        writer.writerow([v for v, m in zip(row_values, mask) if m])
     return csv_file