将Pandas系列的先前值与字符串和整数混合使用

我正在尝试用当天的日期覆盖时间。该列表长约100行,下面是一个示例:

    Date
0   May-21-20 #Gets passed
1   02:51PM #(should read May-21-20)
2   01:59PM #(should read May-21-20)
3   01:29PM #etc
4   12:45PM #etc
5   12:42PM
6   11:55AM
7   10:02AM
8   09:37AM #(should read May-21-20)
9   May-20-20 #gets passed
10  02:47PM #(should read May-20-20)
11  02:30PM #(should read May-20-20)
12  02:29PM #(should read May-20-20)
13  02:01PM #(should read May-20-20)

这是我目前使用我的代码的位置:

for i in headline_table['Date']:
date_list = headline_table['Date'].tolist() #Make the pd Sereies a List
index_value = date_list.index(i) #Now a list so I can reference index value
previous = index_value - 1 #index of current minus one = previous value

if re.search(r'^[A-Z]', i):
    pass
else:
    headline_table['Date'][i] = headline_table.loc[previous, 'Date']

我尝试了多种方法来解决此问题,但似乎无法弄清楚。我的代码没有任何错误,但是日期没有覆盖时间,相反,似乎没有任何反应。