将字符串日期和时间转换为DateTime

Question

I have to run through thousands of Word document and create XML files out of them. Everything works fine except the Date fields because I'm working in two languages.

Here are a few examples

• DATE: NOVEMBER 24, 2016 TIME: 15:31
• DATE: 28 NOVEMBRE 2016 HEURE: 10H31

I cleanup up the string a bit using the below but I still get the infamous 'String was not recognized as a valid DateTime.'

IFormatProvider culture = null;
if (m.rdoEnglish.IsChecked == true)
{
    culture = new System.Globalization.CultureInfo("en-CA", true);
}
else if (m.rdoFrench.IsChecked == true)
{
    culture = new System.Globalization.CultureInfo("fr-CA", true);
}

string dt = "";
dt = m.txtPublished.Text;

if (dt.IndexOf("HEURE:") != -1)
{
    dt = dt.Replace("HEURE:", "");
}

if (dt.IndexOf("H") != -1)
{
    dt = dt.Replace("H", ":");
}
DateTime dt2;
dt2 = DateTime.ParseExact(dt, "MM/dd/yyyy HH:mm:ss tt", culture);

//Cleaned string looks like this " 28 NOVEMBRE 2016 10:31  "
return dt2;

Answer 1

Looks like your format string is incorrect, in light of the two examples you gave. It should be something like "MMMM dd, yyyy hh:mm" Read more about it at: DateTime.ParseExact Method

Answer 2

The ParseExact, as well TryParseExact, has an overload that accepts an array of formats to use in parsing the string. This will allow you to use something like this

string test = dt.Replace("DATE: ", "")
                .Replace("TIME: ", "")
                .Replace("HEURE: ", "");

string[] formats = new string[]
{
    "MMMM dd, yyyy HH:mm", "dd MMMM yyyy HH\'H\'mm"
};
DateTime dt2 = DateTime.ParseExact(test, formats, culture, DateTimeStyles.None);

Notice that I don't try to replace the single char H inside the french version of your string because I don't know if the same char appears somewhere in your months.

Answer 3

If the strings you need to parse can be in two different cultures in the same document then you need to code more defensively.

Set up both cultures:

IFormatProvider englishCulture = new System.Globalization.CultureInfo("en-CA", true);
IFormatProvider frenchCulture = new System.Globalization.CultureInfo("fr-CA", true);

Then use TryParse with one culture and if that fails try again with the other:

DateTime output;
if (DateTime.TryParseExact(input, frenchFormatString, frenchCulture, out output))
{
    // Read it successfully
}
else if (DateTime.TryParseExact(input, englishFormatString, englishCulture, out output))
{
    // Read it successfully 
}
else
{
    // Unknown format, try something else?
}

If you have information that the text should be in French or English you could arrange the code so that the most likely culture is tried first.

The format strings can contain all the padding words as well as the placeholders for the actual data. There's even an overload of TryParseExact that takes an array of format strings so you can set up all (or at least most) of the formats you'll encounter.