Maybe headers_auth.json.example need update?

[NaSangWon]

Hi, I'm using ytmusicapi in school project. Thank you for this.

When I try "Browser authentication" like

f = open("browser.txt")
cred = ytmusicapi.setup(headers_raw=f.read())
f.close()
yt = YTMusic(str(cred), requests_session=session)

and browser.txt like so, notice there is no Authorization field.
When I add it, works fine.

Accept: */*
Accept-Encoding: gzip, deflate, br
Accept-Language: ko-KR,ko;q=0.9,en-US;q=0.8,en;q=0.7
Content-Length: 2483
Content-Type: application/json
Cookie: <some cookie value>
Origin: https://music.youtube.com
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/114.0.0.0 Safari/537.36
X-Goog-Authuser: 0
X-Goog-Visitor-Id: CgtTaXN6SG5DVXJfbyjGntelBg%3D%3D

If I don't add it, error raised.

  File "C:\Users\***\PycharmProjects\ytmusicapi\tests\my_browser_search.py", line 15, in init
    yt = YTMusic(str(cred), requests_session=session)
  File "C:\Users\***\PycharmProjects\ytmusicapi\ytmusicapi\ytmusic.py", line 97, in __init__
    self.headers = prepare_headers(self._session, proxies, self.input_dict)
  File "C:\Users\***\PycharmProjects\ytmusicapi\ytmusicapi\auth\headers.py", line 42, in prepare_headers
    raise Exception(
Exception: Could not detect credential type. Please ensure your oauth or browser credentials are set up correctly.

I think key Authorization should add in the doc example.

ytmusicapi/docs/source/setup/headers_auth.json.example

Lines 1 to 9 in ef312d7

	{
	"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:72.0) Gecko/20100101 Firefox/72.0",
	"Accept": "*/*",
	"Accept-Language": "en-US,en;q=0.5",
	"Content-Type": "application/json",
	"X-Goog-AuthUser": "0",
	"x-origin": "https://music.youtube.com",
	"Cookie" : "PASTE_COOKIE"
	}

[aont]

I agree.

In addition, on Chrome, "copy everything starting from “accept: */*” to the end of the section" (from the document) results in

Accept:
*/*
Accept-Encoding:
gzip, deflate, br
Accept-Language:
ja,en-US;q=0.9,en;q=0.8
Authorization:
SAPISIDHASH ...
Cache-Control:
no-cache
Content-Length:
3424
Content-Type:
application/json
Cookie:
LOGIN_INFO=...
...

, which will not be parsed correctly.

Instead, one can use "Copy as fetch" of Chrome inspector.
With that one gets

fetch("https://music.youtube.com/youtubei/v1/browse?...", {
  "headers": {
    "accept": "*/*",
    "accept-language": "ja,en-US;q=0.9,en;q=0.8",
    "authorization": "SAPISIDHASH ...",
    "cache-control": "no-cache",
    "content-type": "application/json",
    "pragma": "no-cache",
    "sec-ch-ua": "\"Not/A)Brand\";v=\"99\", \"Google Chrome\";v=\"115\", \"Chromium\";v=\"115\"",
    "sec-ch-ua-arch": "\"x86\"",
    "sec-ch-ua-bitness": "\"64\"",
    "sec-ch-ua-full-version": "\"115.0.5790.171\"",
    "sec-ch-ua-full-version-list": "\"Not/A)Brand\";v=\"99.0.0.0\", \"Google Chrome\";v=\"115.0.5790.171\", \"Chromium\";v=\"115.0.5790.171\"",
    "sec-ch-ua-mobile": "?0",
    "sec-ch-ua-model": "\"\"",
    "sec-ch-ua-platform": "\"Windows\"",
    "sec-ch-ua-platform-version": "\"14.0.0\"",
    "sec-ch-ua-wow64": "?0",
    "sec-fetch-dest": "empty",
    "sec-fetch-mode": "same-origin",
    "sec-fetch-site": "same-origin",
    "x-client-data": "...",
    "x-goog-authuser": "0",
    "x-goog-visitor-id": "...",
    "x-origin": "https://music.youtube.com",
    "x-youtube-bootstrap-logged-in": "true",
    "x-youtube-client-name": "67",
    "x-youtube-client-version": "1.20230731.00.00"
  },
  "referrer": "https://music.youtube.com/",
  "referrerPolicy": "strict-origin-when-cross-origin",
  "body": "...",
  "method": "POST",
  "mode": "cors",
  "credentials": "include"
});

So, one can simply use the headers part as browser.json.
I suggest putting above instructions on the document.

[sigma67]

That is a good suggestion, although a bit more manual effort is required compared to just copy pasting.

Why would the above not be parsed correctly?

[aont]

Thanks for the reply!

The above former copy-paste result has extra line breaks after ':'. So, the "values" of the headers will be empty.
In addition, chrome inspector kindly shows decoded string for "X-Client-Data", which may cause unexpected behavior.

X-Client-Data:
....
Decoded:
message ClientVariations {
  // Active client experiment variation IDs.
  repeated int32 variation_id = [....];
  // Active client experiment variation IDs that trigger server-side behavior.
  repeated int32 trigger_variation_id = [....];
}

[aont]

Let me correct. I found "Copy as fetch" ommits cookie, and "Copy as Node.js fetch" does include cookie. So we should use latter.

fetch("https://music.youtube.com/youtubei/v1/browse?...", {
  "headers": {
    "accept": "*/*",
    "accept-language": "ja,en-US;q=0.9,en;q=0.8",
    "authorization": "SAPISIDHASH ...",
    "cache-control": "no-cache",
    "content-type": "application/json",
    "pragma": "no-cache",
    "sec-ch-ua": "\"Not/A)Brand\";v=\"99\", \"Google Chrome\";v=\"115\", \"Chromium\";v=\"115\"",
    "sec-ch-ua-arch": "\"x86\"",
    "sec-ch-ua-bitness": "\"64\"",
    "sec-ch-ua-full-version": "\"115.0.5790.171\"",
    "sec-ch-ua-full-version-list": "\"Not/A)Brand\";v=\"99.0.0.0\", \"Google Chrome\";v=\"115.0.5790.171\", \"Chromium\";v=\"115.0.5790.171\"",
    "sec-ch-ua-mobile": "?0",
    "sec-ch-ua-model": "\"\"",
    "sec-ch-ua-platform": "\"Windows\"",
    "sec-ch-ua-platform-version": "\"14.0.0\"",
    "sec-ch-ua-wow64": "?0",
    "sec-fetch-dest": "empty",
    "sec-fetch-mode": "same-origin",
    "sec-fetch-site": "same-origin",
    "x-client-data": "...",
    "x-goog-authuser": "0",
    "x-goog-visitor-id": "...",
    "x-origin": "https://music.youtube.com",
    "x-youtube-bootstrap-logged-in": "true",
    "x-youtube-client-name": "67",
    "x-youtube-client-version": "1.20230731.00.00",
    "cookie": "...",
    "Referer": "https://music.youtube.com/",
    "Referrer-Policy": "strict-origin-when-cross-origin"
  },
  "body": "...",
  "method": "POST"
});