UTF8 解码汉字

Question

I am using node.js and express to build an API which transforms chinese characters into their phonetic spelling (Pinyin) but I am having some utf8 decoding/encoding issues. My PHP Curl request to this API encodes the characters like this:

你好 (nǐ​hǎo) => ä½ å¥½

...so I have to decode them in my node application. I am using the following function:

function decode_utf8(s) {
    return decodeURIComponent(escape(s));
}

and it works perfectly fine in most cases. However, I noticed some weird behavior. Here are two inputs, the value after escape() and the value after decodeURIComponent() :

你好 (nǐ​hǎo): ä½ å¥½ => %E4%BD%A0%E5%A5%BD => 你好
你 (nǐ​): ä½ => %E4%BD => URIError: URI malformed

The first one (nǐ​hǎo) works, but when only using the first of the two characters (nǐ​) it gives me an URIError? How is this possible? The input to the decodeURIComponent() function is the exact same for the nǐ​-part but it only works when combining it with another character. What's wrong here?

Answer 1

The string you're trying to decode doesn't resolve to valid utf8. Something is wrong with your PHP Curl request because it's not encoding你 (nǐ​) correctly. The percent encoded value should be你 => %E4%BD%A0