可以将文章内容翻译成中文,广告屏蔽插件可能会导致该功能失效(如失效,请关闭广告屏蔽插件后再试):
问题:
Preface: This question is a derivative of this question.
Here is my code:
using System;
using System.Linq;
using System.Text.RegularExpressions;
class MainClass {
public static void Main (string[] args) {
const string rawLine = "\"TeamName\",\"PlayerName\",\"Position\" \"Chargers\",\"Philip Rivers\",\"QB\" \"Colts\",\"Peyton Manning\",\"QB\" \"Patriots\",\"Tom Brady\",\"QB\"";
var parsedLines = Regex.Split(rawLine, "(\".*? \"(?:,\".*? \")*)");
parsedLines.ToList().ForEach(Console.WriteLine);
Console.WriteLine("Press [ENTER] to exit.");
Console.ReadLine();
}
}
Here is my output:
"TeamName","PlayerName","Position" "
Chargers
","Philip Rivers","QB" "
Colts
","Peyton Manning","QB" "
Patriots","Tom Brady","QB"
Press [ENTER] to exit.
And here is my desired output:
"TeamName","PlayerName","Position"
"Chargers","Philip Rivers","QB"
"Colts","Peyton Manning","QB"
"Patriots","Tom Brady","QB"
Press [ENTER] to exit.
How can I fix the regex to generate my desired output?
Relevant:
- line breaks lost in sql server
- How do you view ALL text from an ntext or nvarchar(max) in SSMS?
回答1:
Use negative lookbehind, positive lookbehind, character class with quanitifer, positive lookahead, and negative lookahead.
Working Demo
using System;
using System.Linq;
using System.Text.RegularExpressions;
class MainClass {
public static void Main (string[] args) {
const string rawLine = "\"TeamName\",\"PlayerName\",\"Position\" \"Chargers\",\"Philip Rivers\",\"QB\" \"Colts\",\"Peyton Manning\",\"QB\" \"Patriots\",\"Tom Brady\",\"QB\"";
var parsedLines = Regex.Split(rawLine, "(?<![,])(?<=[\"])[ ]{2}(?=[\"])(?![,])");
parsedLines.ToList().ForEach(Console.WriteLine);
Console.WriteLine("Press [ENTER] to exit.");
Console.ReadLine();
}
}
回答2:
As Amy has already mentioned, your string seems to be something like CSV. If it is really a valid CSV - use special libraries.
If CSVHelper isn't applicable in this case and you really need regex, try something like this one:
(?<=(?:^| ))(.*?)(?=(?: \")|$)
I haven't coded for C#, so regex may need some edits due to c# specific.
Edit. Code example.
using System;
using System.Linq;
using System.Text.RegularExpressions;
class MainClass {
public static void Main (string[] args) {
const string rawLine = "\"TeamName\",\"PlayerName\",\"Position\" \"Chargers\",\"Philip Rivers\",\"QB\" \"Colts\",\"Peyton Manning\",\"QB\" \"Patriots\",\"Tom Brady\",\"QB\"";
//var parsedLines = Regex.Split(rawLine, "(?<=(?:^| ))(.*?)(?=(?: \")|$)");
var parsedLines = Regex.Split(rawLine, "(?<=^)(.*?)(?=(?: \")|$)|(?<= )(.*?)(?=(?: \")|$)");
parsedLines.ToList().ForEach(Console.WriteLine);
Console.WriteLine("Press [ENTER] to exit.");
Console.ReadLine();
}
}
This code with "dirty" fix for assertion error. However, i can't reproduce it with onlinetool :)
Original regex commented in this example.
I hope, this will help you. But i must say again if you working with csv - it is better to use special tools, not regex :)
回答3:
Good comments through-out the thread (I would strongly suggest pursuing one of those options), I wont focus on them. Here's an alternative solution that uses Matches
from the Regex
pattern, skip
how many fields you have (columns) and then take
how many records you want.
I'm using a pattern like (\"(.*?)[^,]")
and explanation can be found here of what it means.
const string rawLine = "\"TeamName\",\"PlayerName\",\"Position\" \"Chargers\",\"Philip Rivers\",\"QB\" \"Colts\",\"Peyton Manning\",\"QB\" \"Patriots\",\"Tom Brady\",\"QB\"";
var matches = new Regex(@"(\""(.*?)[^,]"")").Matches(rawLine).Cast<Match>().ToList();
// loop through our matches
for(int i = 0; i < matches.Count; i++)
{
// join our records we need to output
string str = string.Join(",", matches.Skip(i * 3).Take(3));
if(!string.IsNullOrEmpty(str))
Console.WriteLine(str);
}
Console.WriteLine("Press [ENTER] to exit.");
Console.ReadLine();
Please note, there's no error checking at all, can be improved, but does produces the output you need. *Also make sure you import System.Linq
if not already there.
Output Test