修剪所有文件中文本的最快方法

Question

I have a function to trim all the text in all files in a directory. Here is the code

var dbtables = System.IO.Directory.GetFiles(db);

foreach(var table in dbtables)
{
     string text = File.ReadAllText(table);
     File.WriteAllText(table, text.Trim());
}

There are many large files in this directory and it takes about 30 minutes to complete. Do you know of a faster way to do this?

Answer 1

Here's what I would recommend doing for each file:

1. Read the first character.

2. If it's whitespace, you have to rewrite the whole file.

   1. Read the whole file into memory.
   2. Iterate character-by-character from the beginning to find the first non-whitespace character.
   3. Iterate character-by-character from the end to find the first non-whitespace character.
   4. Seek to the beginning of the file.
   5. Write just the middle characters of your string to the file.
   6. Set the file's length to the number of characters you wrote.

3. If the first character is not whitespace, you won't be trimming the beginning, so you can just truncate the end.

   1. Read the file backwards, character-by-character, until you hit a character that isn't whitespace.
   2. If you've found whitespace, set the file's length to its current length minus the number of whitespace characters.

Answer 2

Your code currently trims whitespace from the beginning and end of each entire file, rather than each line. If you want to trim whitespace on a per-line basis, you could use:

var dbtables = System.IO.Directory.GetFiles(db);

foreach (string table in dbtables)
{
    string temp = table + ".tmp";
    using (StreamWriter target = new StreamWriter(temp))
        foreach (string line in File.ReadLines(table))
            target.WriteLine(line.Trim());

    File.Delete(table);
    File.Move(temp, table);
}