Node.js readFile内容转换为字符串会导致“ ??”

Question

var fs = require('fs');
var myNumber = undefined;

function addOne(callback) {
  fs.readFile('./User2.txt', 'utf8', function doneReading(err, fileContents) {
    myNumber = fileContents.toString();
    callback();
  });
}

function logMyNumber() {
  console.log(myNumber);
}

addOne(logMyNumber);

User2.txt only contains one single character, "1".

So when I run it, the output is: "??1". Why does these question marks appear? I originally wanted a number but I just got the message, NaN(not a number, I guess). So I convert the buffer to a string instead, and got this. Any help?

Answer 1

It seems you are not the first one with this issue.

Basically you just need to do something like the following:

fs.readFile(filePath, 'utf8', function (err, fileContents) {
    // Remove BOM character if there is one at the start of the file.
    if(fileContents.charCodeAt(0) == 65279) fileContents = fileContents.substr(1);
}

Here you have many other workarounds taken from that discussion:

• Replace:

  fileContents = fileContents.replace(/^\/, '');

• Use fs.readFileSync instead of fs.readFile

• Use the bomstrip package.

Answer 2

I've fixed it by saving the textfile in wordpad instead of notepad. Saving it as text document - ms-dos format.

Answer 3

Most of IDE support RegExp search, so it's pretty easy to search for that buggy char in the code base:

\uFEFF

Just replace it with empty string