Initial commit

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/__main__.py

@@ -0,0 +1,17 @@
+#!/usr/bin/env python3
+
+# Execute with
+# $ python3 -m yt_dlp
+
+import sys
+
+if __package__ is None and not getattr(sys, 'frozen', False):
+    # direct call of __main__.py
+    import os.path
+    path = os.path.realpath(os.path.abspath(__file__))
+    sys.path.insert(0, os.path.dirname(os.path.dirname(path)))
+
+import yt_dlp
+
+if __name__ == '__main__':
+    yt_dlp.main()

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/__pyinstaller/__init__.py

@@ -0,0 +1,5 @@
+import os
+
+
+def get_hook_dirs():
+    return [os.path.dirname(__file__)]

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/__pyinstaller/hook-yt_dlp.py

@@ -0,0 +1,36 @@
+import sys
+
+from PyInstaller.utils.hooks import collect_submodules, collect_data_files
+
+
+def pycryptodome_module():
+    try:
+        import Cryptodome  # noqa: F401
+    except ImportError:
+        try:
+            import Crypto  # noqa: F401
+            print('WARNING: Using Crypto since Cryptodome is not available. '
+                  'Install with: python3 -m pip install pycryptodomex', file=sys.stderr)
+            return 'Crypto'
+        except ImportError:
+            pass
+    return 'Cryptodome'
+
+
+def get_hidden_imports():
+    yield from ('yt_dlp.compat._legacy', 'yt_dlp.compat._deprecated')
+    yield from ('yt_dlp.utils._legacy', 'yt_dlp.utils._deprecated')
+    yield pycryptodome_module()
+    # Only `websockets` is required, others are collected just in case
+    for module in ('websockets', 'requests', 'urllib3'):
+        yield from collect_submodules(module)
+    # These are auto-detected, but explicitly add them just in case
+    yield from ('mutagen', 'brotli', 'certifi', 'secretstorage', 'curl_cffi')
+
+
+hiddenimports = list(get_hidden_imports())
+print(f'Adding imports: {hiddenimports}')
+
+excludedimports = ['youtube_dl', 'youtube_dlc', 'test', 'ytdlp_plugins', 'devscripts', 'bundle']
+
+datas = collect_data_files('curl_cffi', includes=['cacert.pem'])

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/aes.py

@@ -0,0 +1,552 @@
+import base64
+from math import ceil
+
+from .compat import compat_ord
+from .dependencies import Cryptodome
+
+if Cryptodome.AES:
+    def aes_cbc_decrypt_bytes(data, key, iv):
+        """ Decrypt bytes with AES-CBC using pycryptodome """
+        return Cryptodome.AES.new(key, Cryptodome.AES.MODE_CBC, iv).decrypt(data)
+
+    def aes_gcm_decrypt_and_verify_bytes(data, key, tag, nonce):
+        """ Decrypt bytes with AES-GCM using pycryptodome """
+        return Cryptodome.AES.new(key, Cryptodome.AES.MODE_GCM, nonce).decrypt_and_verify(data, tag)
+
+else:
+    def aes_cbc_decrypt_bytes(data, key, iv):
+        """ Decrypt bytes with AES-CBC using native implementation since pycryptodome is unavailable """
+        return bytes(aes_cbc_decrypt(*map(list, (data, key, iv))))
+
+    def aes_gcm_decrypt_and_verify_bytes(data, key, tag, nonce):
+        """ Decrypt bytes with AES-GCM using native implementation since pycryptodome is unavailable """
+        return bytes(aes_gcm_decrypt_and_verify(*map(list, (data, key, tag, nonce))))
+
+
+def aes_cbc_encrypt_bytes(data, key, iv, **kwargs):
+    return bytes(aes_cbc_encrypt(*map(list, (data, key, iv)), **kwargs))
+
+
+BLOCK_SIZE_BYTES = 16
+
+
+def unpad_pkcs7(data):
+    return data[:-compat_ord(data[-1])]
+
+
+def pkcs7_padding(data):
+    """
+    PKCS#7 padding
+
+    @param {int[]} data        cleartext
+    @returns {int[]}           padding data
+    """
+
+    remaining_length = BLOCK_SIZE_BYTES - len(data) % BLOCK_SIZE_BYTES
+    return data + [remaining_length] * remaining_length
+
+
+def pad_block(block, padding_mode):
+    """
+    Pad a block with the given padding mode
+    @param {int[]} block        block to pad
+    @param padding_mode         padding mode
+    """
+    padding_size = BLOCK_SIZE_BYTES - len(block)
+
+    PADDING_BYTE = {
+        'pkcs7': padding_size,
+        'iso7816': 0x0,
+        'whitespace': 0x20,
+        'zero': 0x0,
+    }
+
+    if padding_size < 0:
+        raise ValueError('Block size exceeded')
+    elif padding_mode not in PADDING_BYTE:
+        raise NotImplementedError(f'Padding mode {padding_mode} is not implemented')
+
+    if padding_mode == 'iso7816' and padding_size:
+        block = [*block, 0x80]  # NB: += mutates list
+        padding_size -= 1
+
+    return block + [PADDING_BYTE[padding_mode]] * padding_size
+
+
+def aes_ecb_encrypt(data, key, iv=None):
+    """
+    Encrypt with aes in ECB mode. Using PKCS#7 padding
+
+    @param {int[]} data        cleartext
+    @param {int[]} key         16/24/32-Byte cipher key
+    @param {int[]} iv          Unused for this mode
+    @returns {int[]}           encrypted data
+    """
+    expanded_key = key_expansion(key)
+    block_count = ceil(len(data) / BLOCK_SIZE_BYTES)
+
+    encrypted_data = []
+    for i in range(block_count):
+        block = data[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES]
+        encrypted_data += aes_encrypt(pkcs7_padding(block), expanded_key)
+
+    return encrypted_data
+
+
+def aes_ecb_decrypt(data, key, iv=None):
+    """
+    Decrypt with aes in ECB mode
+
+    @param {int[]} data        cleartext
+    @param {int[]} key         16/24/32-Byte cipher key
+    @param {int[]} iv          Unused for this mode
+    @returns {int[]}           decrypted data
+    """
+    expanded_key = key_expansion(key)
+    block_count = ceil(len(data) / BLOCK_SIZE_BYTES)
+
+    encrypted_data = []
+    for i in range(block_count):
+        block = data[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES]
+        encrypted_data += aes_decrypt(block, expanded_key)
+    return encrypted_data[:len(data)]
+
+
+def aes_ctr_decrypt(data, key, iv):
+    """
+    Decrypt with aes in counter mode
+
+    @param {int[]} data        cipher
+    @param {int[]} key         16/24/32-Byte cipher key
+    @param {int[]} iv          16-Byte initialization vector
+    @returns {int[]}           decrypted data
+    """
+    return aes_ctr_encrypt(data, key, iv)
+
+
+def aes_ctr_encrypt(data, key, iv):
+    """
+    Encrypt with aes in counter mode
+
+    @param {int[]} data        cleartext
+    @param {int[]} key         16/24/32-Byte cipher key
+    @param {int[]} iv          16-Byte initialization vector
+    @returns {int[]}           encrypted data
+    """
+    expanded_key = key_expansion(key)
+    block_count = ceil(len(data) / BLOCK_SIZE_BYTES)
+    counter = iter_vector(iv)
+
+    encrypted_data = []
+    for i in range(block_count):
+        counter_block = next(counter)
+        block = data[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES]
+        block += [0] * (BLOCK_SIZE_BYTES - len(block))
+
+        cipher_counter_block = aes_encrypt(counter_block, expanded_key)
+        encrypted_data += xor(block, cipher_counter_block)
+    return encrypted_data[:len(data)]
+
+
+def aes_cbc_decrypt(data, key, iv):
+    """
+    Decrypt with aes in CBC mode
+
+    @param {int[]} data        cipher
+    @param {int[]} key         16/24/32-Byte cipher key
+    @param {int[]} iv          16-Byte IV
+    @returns {int[]}           decrypted data
+    """
+    expanded_key = key_expansion(key)
+    block_count = ceil(len(data) / BLOCK_SIZE_BYTES)
+
+    decrypted_data = []
+    previous_cipher_block = iv
+    for i in range(block_count):
+        block = data[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES]
+        block += [0] * (BLOCK_SIZE_BYTES - len(block))
+
+        decrypted_block = aes_decrypt(block, expanded_key)
+        decrypted_data += xor(decrypted_block, previous_cipher_block)
+        previous_cipher_block = block
+    return decrypted_data[:len(data)]
+
+
+def aes_cbc_encrypt(data, key, iv, *, padding_mode='pkcs7'):
+    """
+    Encrypt with aes in CBC mode
+
+    @param {int[]} data        cleartext
+    @param {int[]} key         16/24/32-Byte cipher key
+    @param {int[]} iv          16-Byte IV
+    @param padding_mode        Padding mode to use
+    @returns {int[]}           encrypted data
+    """
+    expanded_key = key_expansion(key)
+    block_count = ceil(len(data) / BLOCK_SIZE_BYTES)
+
+    encrypted_data = []
+    previous_cipher_block = iv
+    for i in range(block_count):
+        block = data[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES]
+        block = pad_block(block, padding_mode)
+
+        mixed_block = xor(block, previous_cipher_block)
+
+        encrypted_block = aes_encrypt(mixed_block, expanded_key)
+        encrypted_data += encrypted_block
+
+        previous_cipher_block = encrypted_block
+
+    return encrypted_data
+
+
+def aes_gcm_decrypt_and_verify(data, key, tag, nonce):
+    """
+    Decrypt with aes in GBM mode and checks authenticity using tag
+
+    @param {int[]} data        cipher
+    @param {int[]} key         16-Byte cipher key
+    @param {int[]} tag         authentication tag
+    @param {int[]} nonce       IV (recommended 12-Byte)
+    @returns {int[]}           decrypted data
+    """
+
+    # XXX: check aes, gcm param
+
+    hash_subkey = aes_encrypt([0] * BLOCK_SIZE_BYTES, key_expansion(key))
+
+    if len(nonce) == 12:
+        j0 = [*nonce, 0, 0, 0, 1]
+    else:
+        fill = (BLOCK_SIZE_BYTES - (len(nonce) % BLOCK_SIZE_BYTES)) % BLOCK_SIZE_BYTES + 8
+        ghash_in = nonce + [0] * fill + list((8 * len(nonce)).to_bytes(8, 'big'))
+        j0 = ghash(hash_subkey, ghash_in)
+
+    # TODO: add nonce support to aes_ctr_decrypt
+
+    # nonce_ctr = j0[:12]
+    iv_ctr = inc(j0)
+
+    decrypted_data = aes_ctr_decrypt(data, key, iv_ctr + [0] * (BLOCK_SIZE_BYTES - len(iv_ctr)))
+    pad_len = (BLOCK_SIZE_BYTES - (len(data) % BLOCK_SIZE_BYTES)) % BLOCK_SIZE_BYTES
+    s_tag = ghash(
+        hash_subkey,
+        data
+        + [0] * pad_len                                  # pad
+        + list((0 * 8).to_bytes(8, 'big')                # length of associated data
+               + ((len(data) * 8).to_bytes(8, 'big'))),  # length of data
+    )
+
+    if tag != aes_ctr_encrypt(s_tag, key, j0):
+        raise ValueError('Mismatching authentication tag')
+
+    return decrypted_data
+
+
+def aes_encrypt(data, expanded_key):
+    """
+    Encrypt one block with aes
+
+    @param {int[]} data          16-Byte state
+    @param {int[]} expanded_key  176/208/240-Byte expanded key
+    @returns {int[]}             16-Byte cipher
+    """
+    rounds = len(expanded_key) // BLOCK_SIZE_BYTES - 1
+
+    data = xor(data, expanded_key[:BLOCK_SIZE_BYTES])
+    for i in range(1, rounds + 1):
+        data = sub_bytes(data)
+        data = shift_rows(data)
+        if i != rounds:
+            data = list(iter_mix_columns(data, MIX_COLUMN_MATRIX))
+        data = xor(data, expanded_key[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES])
+
+    return data
+
+
+def aes_decrypt(data, expanded_key):
+    """
+    Decrypt one block with aes
+
+    @param {int[]} data          16-Byte cipher
+    @param {int[]} expanded_key  176/208/240-Byte expanded key
+    @returns {int[]}             16-Byte state
+    """
+    rounds = len(expanded_key) // BLOCK_SIZE_BYTES - 1
+
+    for i in range(rounds, 0, -1):
+        data = xor(data, expanded_key[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES])
+        if i != rounds:
+            data = list(iter_mix_columns(data, MIX_COLUMN_MATRIX_INV))
+        data = shift_rows_inv(data)
+        data = sub_bytes_inv(data)
+    return xor(data, expanded_key[:BLOCK_SIZE_BYTES])
+
+
+def aes_decrypt_text(data, password, key_size_bytes):
+    """
+    Decrypt text
+    - The first 8 Bytes of decoded 'data' are the 8 high Bytes of the counter
+    - The cipher key is retrieved by encrypting the first 16 Byte of 'password'
+      with the first 'key_size_bytes' Bytes from 'password' (if necessary filled with 0's)
+    - Mode of operation is 'counter'
+
+    @param {str} data                    Base64 encoded string
+    @param {str,unicode} password        Password (will be encoded with utf-8)
+    @param {int} key_size_bytes          Possible values: 16 for 128-Bit, 24 for 192-Bit or 32 for 256-Bit
+    @returns {str}                       Decrypted data
+    """
+    NONCE_LENGTH_BYTES = 8
+
+    data = list(base64.b64decode(data))
+    password = list(password.encode())
+
+    key = password[:key_size_bytes] + [0] * (key_size_bytes - len(password))
+    key = aes_encrypt(key[:BLOCK_SIZE_BYTES], key_expansion(key)) * (key_size_bytes // BLOCK_SIZE_BYTES)
+
+    nonce = data[:NONCE_LENGTH_BYTES]
+    cipher = data[NONCE_LENGTH_BYTES:]
+
+    decrypted_data = aes_ctr_decrypt(cipher, key, nonce + [0] * (BLOCK_SIZE_BYTES - NONCE_LENGTH_BYTES))
+    return bytes(decrypted_data)
+
+
+RCON = (0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36)
+SBOX = (0x63, 0x7C, 0x77, 0x7B, 0xF2, 0x6B, 0x6F, 0xC5, 0x30, 0x01, 0x67, 0x2B, 0xFE, 0xD7, 0xAB, 0x76,
+        0xCA, 0x82, 0xC9, 0x7D, 0xFA, 0x59, 0x47, 0xF0, 0xAD, 0xD4, 0xA2, 0xAF, 0x9C, 0xA4, 0x72, 0xC0,
+        0xB7, 0xFD, 0x93, 0x26, 0x36, 0x3F, 0xF7, 0xCC, 0x34, 0xA5, 0xE5, 0xF1, 0x71, 0xD8, 0x31, 0x15,
+        0x04, 0xC7, 0x23, 0xC3, 0x18, 0x96, 0x05, 0x9A, 0x07, 0x12, 0x80, 0xE2, 0xEB, 0x27, 0xB2, 0x75,
+        0x09, 0x83, 0x2C, 0x1A, 0x1B, 0x6E, 0x5A, 0xA0, 0x52, 0x3B, 0xD6, 0xB3, 0x29, 0xE3, 0x2F, 0x84,
+        0x53, 0xD1, 0x00, 0xED, 0x20, 0xFC, 0xB1, 0x5B, 0x6A, 0xCB, 0xBE, 0x39, 0x4A, 0x4C, 0x58, 0xCF,
+        0xD0, 0xEF, 0xAA, 0xFB, 0x43, 0x4D, 0x33, 0x85, 0x45, 0xF9, 0x02, 0x7F, 0x50, 0x3C, 0x9F, 0xA8,
+        0x51, 0xA3, 0x40, 0x8F, 0x92, 0x9D, 0x38, 0xF5, 0xBC, 0xB6, 0xDA, 0x21, 0x10, 0xFF, 0xF3, 0xD2,
+        0xCD, 0x0C, 0x13, 0xEC, 0x5F, 0x97, 0x44, 0x17, 0xC4, 0xA7, 0x7E, 0x3D, 0x64, 0x5D, 0x19, 0x73,
+        0x60, 0x81, 0x4F, 0xDC, 0x22, 0x2A, 0x90, 0x88, 0x46, 0xEE, 0xB8, 0x14, 0xDE, 0x5E, 0x0B, 0xDB,
+        0xE0, 0x32, 0x3A, 0x0A, 0x49, 0x06, 0x24, 0x5C, 0xC2, 0xD3, 0xAC, 0x62, 0x91, 0x95, 0xE4, 0x79,
+        0xE7, 0xC8, 0x37, 0x6D, 0x8D, 0xD5, 0x4E, 0xA9, 0x6C, 0x56, 0xF4, 0xEA, 0x65, 0x7A, 0xAE, 0x08,
+        0xBA, 0x78, 0x25, 0x2E, 0x1C, 0xA6, 0xB4, 0xC6, 0xE8, 0xDD, 0x74, 0x1F, 0x4B, 0xBD, 0x8B, 0x8A,
+        0x70, 0x3E, 0xB5, 0x66, 0x48, 0x03, 0xF6, 0x0E, 0x61, 0x35, 0x57, 0xB9, 0x86, 0xC1, 0x1D, 0x9E,
+        0xE1, 0xF8, 0x98, 0x11, 0x69, 0xD9, 0x8E, 0x94, 0x9B, 0x1E, 0x87, 0xE9, 0xCE, 0x55, 0x28, 0xDF,
+        0x8C, 0xA1, 0x89, 0x0D, 0xBF, 0xE6, 0x42, 0x68, 0x41, 0x99, 0x2D, 0x0F, 0xB0, 0x54, 0xBB, 0x16)
+SBOX_INV = (0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38, 0xbf, 0x40, 0xa3, 0x9e, 0x81, 0xf3, 0xd7, 0xfb,
+            0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87, 0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb,
+            0x54, 0x7b, 0x94, 0x32, 0xa6, 0xc2, 0x23, 0x3d, 0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e,
+            0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2, 0x76, 0x5b, 0xa2, 0x49, 0x6d, 0x8b, 0xd1, 0x25,
+            0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16, 0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92,
+            0x6c, 0x70, 0x48, 0x50, 0xfd, 0xed, 0xb9, 0xda, 0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84,
+            0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a, 0xf7, 0xe4, 0x58, 0x05, 0xb8, 0xb3, 0x45, 0x06,
+            0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02, 0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b,
+            0x3a, 0x91, 0x11, 0x41, 0x4f, 0x67, 0xdc, 0xea, 0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73,
+            0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85, 0xe2, 0xf9, 0x37, 0xe8, 0x1c, 0x75, 0xdf, 0x6e,
+            0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89, 0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b,
+            0xfc, 0x56, 0x3e, 0x4b, 0xc6, 0xd2, 0x79, 0x20, 0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4,
+            0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31, 0xb1, 0x12, 0x10, 0x59, 0x27, 0x80, 0xec, 0x5f,
+            0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d, 0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef,
+            0xa0, 0xe0, 0x3b, 0x4d, 0xae, 0x2a, 0xf5, 0xb0, 0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61,
+            0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26, 0xe1, 0x69, 0x14, 0x63, 0x55, 0x21, 0x0c, 0x7d)
+MIX_COLUMN_MATRIX = ((0x2, 0x3, 0x1, 0x1),
+                     (0x1, 0x2, 0x3, 0x1),
+                     (0x1, 0x1, 0x2, 0x3),
+                     (0x3, 0x1, 0x1, 0x2))
+MIX_COLUMN_MATRIX_INV = ((0xE, 0xB, 0xD, 0x9),
+                         (0x9, 0xE, 0xB, 0xD),
+                         (0xD, 0x9, 0xE, 0xB),
+                         (0xB, 0xD, 0x9, 0xE))
+RIJNDAEL_EXP_TABLE = (0x01, 0x03, 0x05, 0x0F, 0x11, 0x33, 0x55, 0xFF, 0x1A, 0x2E, 0x72, 0x96, 0xA1, 0xF8, 0x13, 0x35,
+                      0x5F, 0xE1, 0x38, 0x48, 0xD8, 0x73, 0x95, 0xA4, 0xF7, 0x02, 0x06, 0x0A, 0x1E, 0x22, 0x66, 0xAA,
+                      0xE5, 0x34, 0x5C, 0xE4, 0x37, 0x59, 0xEB, 0x26, 0x6A, 0xBE, 0xD9, 0x70, 0x90, 0xAB, 0xE6, 0x31,
+                      0x53, 0xF5, 0x04, 0x0C, 0x14, 0x3C, 0x44, 0xCC, 0x4F, 0xD1, 0x68, 0xB8, 0xD3, 0x6E, 0xB2, 0xCD,
+                      0x4C, 0xD4, 0x67, 0xA9, 0xE0, 0x3B, 0x4D, 0xD7, 0x62, 0xA6, 0xF1, 0x08, 0x18, 0x28, 0x78, 0x88,
+                      0x83, 0x9E, 0xB9, 0xD0, 0x6B, 0xBD, 0xDC, 0x7F, 0x81, 0x98, 0xB3, 0xCE, 0x49, 0xDB, 0x76, 0x9A,
+                      0xB5, 0xC4, 0x57, 0xF9, 0x10, 0x30, 0x50, 0xF0, 0x0B, 0x1D, 0x27, 0x69, 0xBB, 0xD6, 0x61, 0xA3,
+                      0xFE, 0x19, 0x2B, 0x7D, 0x87, 0x92, 0xAD, 0xEC, 0x2F, 0x71, 0x93, 0xAE, 0xE9, 0x20, 0x60, 0xA0,
+                      0xFB, 0x16, 0x3A, 0x4E, 0xD2, 0x6D, 0xB7, 0xC2, 0x5D, 0xE7, 0x32, 0x56, 0xFA, 0x15, 0x3F, 0x41,
+                      0xC3, 0x5E, 0xE2, 0x3D, 0x47, 0xC9, 0x40, 0xC0, 0x5B, 0xED, 0x2C, 0x74, 0x9C, 0xBF, 0xDA, 0x75,
+                      0x9F, 0xBA, 0xD5, 0x64, 0xAC, 0xEF, 0x2A, 0x7E, 0x82, 0x9D, 0xBC, 0xDF, 0x7A, 0x8E, 0x89, 0x80,
+                      0x9B, 0xB6, 0xC1, 0x58, 0xE8, 0x23, 0x65, 0xAF, 0xEA, 0x25, 0x6F, 0xB1, 0xC8, 0x43, 0xC5, 0x54,
+                      0xFC, 0x1F, 0x21, 0x63, 0xA5, 0xF4, 0x07, 0x09, 0x1B, 0x2D, 0x77, 0x99, 0xB0, 0xCB, 0x46, 0xCA,
+                      0x45, 0xCF, 0x4A, 0xDE, 0x79, 0x8B, 0x86, 0x91, 0xA8, 0xE3, 0x3E, 0x42, 0xC6, 0x51, 0xF3, 0x0E,
+                      0x12, 0x36, 0x5A, 0xEE, 0x29, 0x7B, 0x8D, 0x8C, 0x8F, 0x8A, 0x85, 0x94, 0xA7, 0xF2, 0x0D, 0x17,
+                      0x39, 0x4B, 0xDD, 0x7C, 0x84, 0x97, 0xA2, 0xFD, 0x1C, 0x24, 0x6C, 0xB4, 0xC7, 0x52, 0xF6, 0x01)
+RIJNDAEL_LOG_TABLE = (0x00, 0x00, 0x19, 0x01, 0x32, 0x02, 0x1a, 0xc6, 0x4b, 0xc7, 0x1b, 0x68, 0x33, 0xee, 0xdf, 0x03,
+                      0x64, 0x04, 0xe0, 0x0e, 0x34, 0x8d, 0x81, 0xef, 0x4c, 0x71, 0x08, 0xc8, 0xf8, 0x69, 0x1c, 0xc1,
+                      0x7d, 0xc2, 0x1d, 0xb5, 0xf9, 0xb9, 0x27, 0x6a, 0x4d, 0xe4, 0xa6, 0x72, 0x9a, 0xc9, 0x09, 0x78,
+                      0x65, 0x2f, 0x8a, 0x05, 0x21, 0x0f, 0xe1, 0x24, 0x12, 0xf0, 0x82, 0x45, 0x35, 0x93, 0xda, 0x8e,
+                      0x96, 0x8f, 0xdb, 0xbd, 0x36, 0xd0, 0xce, 0x94, 0x13, 0x5c, 0xd2, 0xf1, 0x40, 0x46, 0x83, 0x38,
+                      0x66, 0xdd, 0xfd, 0x30, 0xbf, 0x06, 0x8b, 0x62, 0xb3, 0x25, 0xe2, 0x98, 0x22, 0x88, 0x91, 0x10,
+                      0x7e, 0x6e, 0x48, 0xc3, 0xa3, 0xb6, 0x1e, 0x42, 0x3a, 0x6b, 0x28, 0x54, 0xfa, 0x85, 0x3d, 0xba,
+                      0x2b, 0x79, 0x0a, 0x15, 0x9b, 0x9f, 0x5e, 0xca, 0x4e, 0xd4, 0xac, 0xe5, 0xf3, 0x73, 0xa7, 0x57,
+                      0xaf, 0x58, 0xa8, 0x50, 0xf4, 0xea, 0xd6, 0x74, 0x4f, 0xae, 0xe9, 0xd5, 0xe7, 0xe6, 0xad, 0xe8,
+                      0x2c, 0xd7, 0x75, 0x7a, 0xeb, 0x16, 0x0b, 0xf5, 0x59, 0xcb, 0x5f, 0xb0, 0x9c, 0xa9, 0x51, 0xa0,
+                      0x7f, 0x0c, 0xf6, 0x6f, 0x17, 0xc4, 0x49, 0xec, 0xd8, 0x43, 0x1f, 0x2d, 0xa4, 0x76, 0x7b, 0xb7,
+                      0xcc, 0xbb, 0x3e, 0x5a, 0xfb, 0x60, 0xb1, 0x86, 0x3b, 0x52, 0xa1, 0x6c, 0xaa, 0x55, 0x29, 0x9d,
+                      0x97, 0xb2, 0x87, 0x90, 0x61, 0xbe, 0xdc, 0xfc, 0xbc, 0x95, 0xcf, 0xcd, 0x37, 0x3f, 0x5b, 0xd1,
+                      0x53, 0x39, 0x84, 0x3c, 0x41, 0xa2, 0x6d, 0x47, 0x14, 0x2a, 0x9e, 0x5d, 0x56, 0xf2, 0xd3, 0xab,
+                      0x44, 0x11, 0x92, 0xd9, 0x23, 0x20, 0x2e, 0x89, 0xb4, 0x7c, 0xb8, 0x26, 0x77, 0x99, 0xe3, 0xa5,
+                      0x67, 0x4a, 0xed, 0xde, 0xc5, 0x31, 0xfe, 0x18, 0x0d, 0x63, 0x8c, 0x80, 0xc0, 0xf7, 0x70, 0x07)
+
+
+def key_expansion(data):
+    """
+    Generate key schedule
+
+    @param {int[]} data  16/24/32-Byte cipher key
+    @returns {int[]}     176/208/240-Byte expanded key
+    """
+    data = data[:]  # copy
+    rcon_iteration = 1
+    key_size_bytes = len(data)
+    expanded_key_size_bytes = (key_size_bytes // 4 + 7) * BLOCK_SIZE_BYTES
+
+    while len(data) < expanded_key_size_bytes:
+        temp = data[-4:]
+        temp = key_schedule_core(temp, rcon_iteration)
+        rcon_iteration += 1
+        data += xor(temp, data[-key_size_bytes: 4 - key_size_bytes])
+
+        for _ in range(3):
+            temp = data[-4:]
+            data += xor(temp, data[-key_size_bytes: 4 - key_size_bytes])
+
+        if key_size_bytes == 32:
+            temp = data[-4:]
+            temp = sub_bytes(temp)
+            data += xor(temp, data[-key_size_bytes: 4 - key_size_bytes])
+
+        for _ in range(3 if key_size_bytes == 32 else 2 if key_size_bytes == 24 else 0):
+            temp = data[-4:]
+            data += xor(temp, data[-key_size_bytes: 4 - key_size_bytes])
+    return data[:expanded_key_size_bytes]
+
+
+def iter_vector(iv):
+    while True:
+        yield iv
+        iv = inc(iv)
+
+
+def sub_bytes(data):
+    return [SBOX[x] for x in data]
+
+
+def sub_bytes_inv(data):
+    return [SBOX_INV[x] for x in data]
+
+
+def rotate(data):
+    return [*data[1:], data[0]]
+
+
+def key_schedule_core(data, rcon_iteration):
+    data = rotate(data)
+    data = sub_bytes(data)
+    data[0] = data[0] ^ RCON[rcon_iteration]
+
+    return data
+
+
+def xor(data1, data2):
+    return [x ^ y for x, y in zip(data1, data2, strict=False)]
+
+
+def iter_mix_columns(data, matrix):
+    for i in (0, 4, 8, 12):
+        for row in matrix:
+            mixed = 0
+            for j in range(4):
+                # xor is (+) and (-)
+                mixed ^= (0 if data[i:i + 4][j] == 0 or row[j] == 0 else
+                          RIJNDAEL_EXP_TABLE[(RIJNDAEL_LOG_TABLE[data[i + j]] + RIJNDAEL_LOG_TABLE[row[j]]) % 0xFF])
+            yield mixed
+
+
+def shift_rows(data):
+    return [data[((column + row) & 0b11) * 4 + row] for column in range(4) for row in range(4)]
+
+
+def shift_rows_inv(data):
+    return [data[((column - row) & 0b11) * 4 + row] for column in range(4) for row in range(4)]
+
+
+def shift_block(data):
+    data_shifted = []
+
+    bit = 0
+    for n in data:
+        if bit:
+            n |= 0x100
+        bit = n & 1
+        n >>= 1
+        data_shifted.append(n)
+
+    return data_shifted
+
+
+def inc(data):
+    data = data[:]  # copy
+    for i in range(len(data) - 1, -1, -1):
+        if data[i] == 255:
+            data[i] = 0
+        else:
+            data[i] = data[i] + 1
+            break
+    return data
+
+
+def block_product(block_x, block_y):
+    # NIST SP 800-38D, Algorithm 1
+
+    if len(block_x) != BLOCK_SIZE_BYTES or len(block_y) != BLOCK_SIZE_BYTES:
+        raise ValueError(f'Length of blocks need to be {BLOCK_SIZE_BYTES} bytes')
+
+    block_r = [0xE1] + [0] * (BLOCK_SIZE_BYTES - 1)
+    block_v = block_y[:]
+    block_z = [0] * BLOCK_SIZE_BYTES
+
+    for i in block_x:
+        for bit in range(7, -1, -1):
+            if i & (1 << bit):
+                block_z = xor(block_z, block_v)
+
+            do_xor = block_v[-1] & 1
+            block_v = shift_block(block_v)
+            if do_xor:
+                block_v = xor(block_v, block_r)
+
+    return block_z
+
+
+def ghash(subkey, data):
+    # NIST SP 800-38D, Algorithm 2
+
+    if len(data) % BLOCK_SIZE_BYTES:
+        raise ValueError(f'Length of data should be {BLOCK_SIZE_BYTES} bytes')
+
+    last_y = [0] * BLOCK_SIZE_BYTES
+    for i in range(0, len(data), BLOCK_SIZE_BYTES):
+        block = data[i: i + BLOCK_SIZE_BYTES]
+        last_y = block_product(xor(last_y, block), subkey)
+
+    return last_y
+
+
+__all__ = [
+    'aes_cbc_decrypt',
+    'aes_cbc_decrypt_bytes',
+    'aes_cbc_encrypt',
+    'aes_cbc_encrypt_bytes',
+    'aes_ctr_decrypt',
+    'aes_ctr_encrypt',
+    'aes_decrypt',
+    'aes_decrypt_text',
+    'aes_ecb_decrypt',
+    'aes_ecb_encrypt',
+    'aes_encrypt',
+    'aes_gcm_decrypt_and_verify',
+    'aes_gcm_decrypt_and_verify_bytes',
+    'key_expansion',
+    'pad_block',
+    'pkcs7_padding',
+    'unpad_pkcs7',
+]

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/cache.py

@@ -0,0 +1,91 @@
+import contextlib
+import json
+import os
+import re
+import shutil
+import traceback
+import urllib.parse
+
+from .utils import expand_path, traverse_obj, version_tuple, write_json_file
+from .version import __version__
+
+
+class Cache:
+    def __init__(self, ydl):
+        self._ydl = ydl
+
+    def _get_root_dir(self):
+        res = self._ydl.params.get('cachedir')
+        if res is None:
+            cache_root = os.getenv('XDG_CACHE_HOME', '~/.cache')
+            res = os.path.join(cache_root, 'yt-dlp')
+        return expand_path(res)
+
+    def _get_cache_fn(self, section, key, dtype):
+        assert re.match(r'^[\w.-]+$', section), f'invalid section {section!r}'
+        key = urllib.parse.quote(key, safe='').replace('%', ',')  # encode non-ascii characters
+        return os.path.join(self._get_root_dir(), section, f'{key}.{dtype}')
+
+    @property
+    def enabled(self):
+        return self._ydl.params.get('cachedir') is not False
+
+    def store(self, section, key, data, dtype='json'):
+        assert dtype in ('json',)
+
+        if not self.enabled:
+            return
+
+        fn = self._get_cache_fn(section, key, dtype)
+        try:
+            os.makedirs(os.path.dirname(fn), exist_ok=True)
+            self._ydl.write_debug(f'Saving {section}.{key} to cache')
+            write_json_file({'yt-dlp_version': __version__, 'data': data}, fn)
+        except Exception:
+            tb = traceback.format_exc()
+            self._ydl.report_warning(f'Writing cache to {fn!r} failed: {tb}')
+
+    def _validate(self, data, min_ver):
+        version = traverse_obj(data, 'yt-dlp_version')
+        if not version:  # Backward compatibility
+            data, version = {'data': data}, '2022.08.19'
+        if not min_ver or version_tuple(version) >= version_tuple(min_ver):
+            return data['data']
+        self._ydl.write_debug(f'Discarding old cache from version {version} (needs {min_ver})')
+
+    def load(self, section, key, dtype='json', default=None, *, min_ver=None):
+        assert dtype in ('json',)
+
+        if not self.enabled:
+            return default
+
+        cache_fn = self._get_cache_fn(section, key, dtype)
+        with contextlib.suppress(OSError):
+            try:
+                with open(cache_fn, encoding='utf-8') as cachef:
+                    self._ydl.write_debug(f'Loading {section}.{key} from cache')
+                    return self._validate(json.load(cachef), min_ver)
+            except (ValueError, KeyError):
+                try:
+                    file_size = os.path.getsize(cache_fn)
+                except OSError as oe:
+                    file_size = str(oe)
+                self._ydl.report_warning(f'Cache retrieval from {cache_fn} failed ({file_size})')
+
+        return default
+
+    def remove(self):
+        if not self.enabled:
+            self._ydl.to_screen('Cache is disabled (Did you combine --no-cache-dir and --rm-cache-dir?)')
+            return
+
+        cachedir = self._get_root_dir()
+        if not any((term in cachedir) for term in ('cache', 'tmp')):
+            raise Exception(f'Not removing directory {cachedir} - this does not look like a cache dir')
+
+        self._ydl.to_screen(
+            f'Removing cache dir {cachedir} .', skip_eol=True)
+        if os.path.exists(cachedir):
+            self._ydl.to_screen('.', skip_eol=True)
+            shutil.rmtree(cachedir)
+        self._ydl.to_screen('.')

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/__init__.py

@@ -0,0 +1,64 @@
+import datetime as dt
+import os
+import xml.etree.ElementTree as etree
+
+from .compat_utils import passthrough_module
+
+passthrough_module(__name__, '._deprecated')
+del passthrough_module
+
+
+# HTMLParseError has been deprecated in Python 3.3 and removed in
+# Python 3.5. Introducing dummy exception for Python >3.5 for compatible
+# and uniform cross-version exception handling
+class compat_HTMLParseError(ValueError):
+    pass
+
+
+class _TreeBuilder(etree.TreeBuilder):
+    def doctype(self, name, pubid, system):
+        pass
+
+
+def compat_etree_fromstring(text):
+    return etree.XML(text, parser=etree.XMLParser(target=_TreeBuilder()))
+
+
+def compat_ord(c):
+    return c if isinstance(c, int) else ord(c)
+
+
+def compat_datetime_from_timestamp(timestamp):
+    # Calling dt.datetime.fromtimestamp with negative timestamps throws error in Windows
+    # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/81708,
+    # https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
+    return (dt.datetime.fromtimestamp(0, dt.timezone.utc) + dt.timedelta(seconds=timestamp))
+
+
+# Python 3.8+ does not honor %HOME% on windows, but this breaks compatibility with youtube-dl
+# See https://github.com/yt-dlp/yt-dlp/issues/792
+# https://docs.python.org/3/library/os.path.html#os.path.expanduser
+if os.name in ('nt', 'ce'):
+    def compat_expanduser(path):
+        HOME = os.environ.get('HOME')
+        if not HOME:
+            return os.path.expanduser(path)
+        elif not path.startswith('~'):
+            return path
+        i = path.replace('\\', '/', 1).find('/')  # ~user
+        if i < 0:
+            i = len(path)
+        userhome = os.path.join(os.path.dirname(HOME), path[1:i]) if i > 1 else HOME
+        return userhome + path[i:]
+else:
+    compat_expanduser = os.path.expanduser
+
+
+def urllib_req_to_req(urllib_request):
+    """Convert urllib Request to a networking Request"""
+    from ..networking import Request
+    from ..utils.networking import HTTPHeaderDict
+    return Request(
+        urllib_request.get_full_url(), data=urllib_request.data, method=urllib_request.get_method(),
+        headers=HTTPHeaderDict(urllib_request.headers, urllib_request.unredirected_hdrs),
+        extensions={'timeout': urllib_request.timeout} if hasattr(urllib_request, 'timeout') else None)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/_deprecated.py

@@ -0,0 +1,21 @@
+"""Deprecated - New code should avoid these"""
+import warnings
+
+from .compat_utils import passthrough_module
+
+# XXX: Implement this the same way as other DeprecationWarnings without circular import
+passthrough_module(__name__, '.._legacy', callback=lambda attr: warnings.warn(
+    DeprecationWarning(f'{__name__}.{attr} is deprecated'), stacklevel=6))
+del passthrough_module
+
+import functools  # noqa: F401
+import os
+
+
+compat_os_name = os.name
+compat_realpath = os.path.realpath
+
+
+def compat_shlex_quote(s):
+    from ..utils import shell_quote
+    return shell_quote(s)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/_legacy.py

@@ -0,0 +1,115 @@
+""" Do not use! """
+
+import base64
+import collections
+import ctypes
+import getpass
+import html.entities
+import html.parser
+import http.client
+import http.cookiejar
+import http.cookies
+import http.server
+import itertools
+import os
+import shlex
+import shutil
+import socket
+import struct
+import subprocess
+import tokenize
+import urllib.error
+import urllib.parse
+import urllib.request
+import xml.etree.ElementTree as etree
+
+# isort: split
+import asyncio  # noqa: F401
+import re  # noqa: F401
+from asyncio import run as compat_asyncio_run  # noqa: F401
+from re import Pattern as compat_Pattern  # noqa: F401
+from re import match as compat_Match  # noqa: F401
+
+from . import compat_expanduser, compat_HTMLParseError
+from .compat_utils import passthrough_module
+from ..dependencies import brotli as compat_brotli  # noqa: F401
+from ..dependencies import websockets as compat_websockets  # noqa: F401
+from ..dependencies.Cryptodome import AES as compat_pycrypto_AES  # noqa: F401
+from ..networking.exceptions import HTTPError as compat_HTTPError
+
+passthrough_module(__name__, '...utils', ('windows_enable_vt_mode',))
+
+
+# compat_ctypes_WINFUNCTYPE = ctypes.WINFUNCTYPE
+# will not work since ctypes.WINFUNCTYPE does not exist in UNIX machines
+def compat_ctypes_WINFUNCTYPE(*args, **kwargs):
+    return ctypes.WINFUNCTYPE(*args, **kwargs)
+
+
+def compat_setenv(key, value, env=os.environ):
+    env[key] = value
+
+
+compat_base64_b64decode = base64.b64decode
+compat_basestring = str
+compat_casefold = str.casefold
+compat_chr = chr
+compat_collections_abc = collections.abc
+compat_cookiejar = compat_http_cookiejar = http.cookiejar
+compat_cookiejar_Cookie = compat_http_cookiejar_Cookie = http.cookiejar.Cookie
+compat_cookies = compat_http_cookies = http.cookies
+compat_cookies_SimpleCookie = compat_http_cookies_SimpleCookie = http.cookies.SimpleCookie
+compat_etree_Element = compat_xml_etree_ElementTree_Element = etree.Element
+compat_etree_register_namespace = compat_xml_etree_register_namespace = etree.register_namespace
+compat_filter = filter
+compat_get_terminal_size = shutil.get_terminal_size
+compat_getenv = os.getenv
+compat_getpass = compat_getpass_getpass = getpass.getpass
+compat_html_entities = html.entities
+compat_html_entities_html5 = html.entities.html5
+compat_html_parser_HTMLParseError = compat_HTMLParseError
+compat_HTMLParser = compat_html_parser_HTMLParser = html.parser.HTMLParser
+compat_http_client = http.client
+compat_http_server = http.server
+compat_input = input
+compat_integer_types = (int, )
+compat_itertools_count = itertools.count
+compat_kwargs = lambda kwargs: kwargs
+compat_map = map
+compat_numeric_types = (int, float, complex)
+compat_os_path_expanduser = compat_expanduser
+compat_os_path_realpath = os.path.realpath
+compat_print = print
+compat_shlex_split = shlex.split
+compat_socket_create_connection = socket.create_connection
+compat_Struct = struct.Struct
+compat_struct_pack = struct.pack
+compat_struct_unpack = struct.unpack
+compat_subprocess_get_DEVNULL = lambda: subprocess.DEVNULL
+compat_tokenize_tokenize = tokenize.tokenize
+compat_urllib_error = urllib.error
+compat_urllib_HTTPError = compat_HTTPError
+compat_urllib_parse = urllib.parse
+compat_urllib_parse_parse_qs = urllib.parse.parse_qs
+compat_urllib_parse_quote = urllib.parse.quote
+compat_urllib_parse_quote_plus = urllib.parse.quote_plus
+compat_urllib_parse_unquote_plus = urllib.parse.unquote_plus
+compat_urllib_parse_unquote_to_bytes = urllib.parse.unquote_to_bytes
+compat_urllib_parse_urlunparse = urllib.parse.urlunparse
+compat_urllib_request = urllib.request
+compat_urllib_request_DataHandler = urllib.request.DataHandler
+compat_urllib_response = urllib.response
+compat_urlretrieve = compat_urllib_request_urlretrieve = urllib.request.urlretrieve
+compat_xml_parse_error = compat_xml_etree_ElementTree_ParseError = etree.ParseError
+compat_xpath = lambda xpath: xpath
+compat_zip = zip
+workaround_optparse_bug9161 = lambda: None
+compat_str = str
+compat_b64decode = base64.b64decode
+compat_urlparse = urllib.parse
+compat_parse_qs = urllib.parse.parse_qs
+compat_urllib_parse_unquote = urllib.parse.unquote
+compat_urllib_parse_urlencode = urllib.parse.urlencode
+compat_urllib_parse_urlparse = urllib.parse.urlparse
+
+legacy = []

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/compat_utils.py

@@ -0,0 +1,83 @@
+import collections
+import contextlib
+import functools
+import importlib
+import sys
+import types
+
+_NO_ATTRIBUTE = object()
+
+_Package = collections.namedtuple('Package', ('name', 'version'))
+
+
+def get_package_info(module):
+    return _Package(
+        name=getattr(module, '_yt_dlp__identifier', module.__name__),
+        version=str(next(filter(None, (
+            getattr(module, attr, None)
+            for attr in ('_yt_dlp__version', '__version__', 'version_string', 'version')
+        )), None)))
+
+
+def _is_package(module):
+    return '__path__' in vars(module)
+
+
+def _is_dunder(name):
+    return name.startswith('__') and name.endswith('__')
+
+
+class EnhancedModule(types.ModuleType):
+    def __bool__(self):
+        return vars(self).get('__bool__', lambda: True)()
+
+    def __getattribute__(self, attr):
+        try:
+            ret = super().__getattribute__(attr)
+        except AttributeError:
+            if _is_dunder(attr):
+                raise
+            getter = getattr(self, '__getattr__', None)
+            if not getter:
+                raise
+            ret = getter(attr)
+        return ret.fget() if isinstance(ret, property) else ret
+
+
+def passthrough_module(parent, child, allowed_attributes=(..., ), *, callback=lambda _: None):
+    """Passthrough parent module into a child module, creating the parent if necessary"""
+    def __getattr__(attr):
+        if _is_package(parent):
+            with contextlib.suppress(ModuleNotFoundError):
+                return importlib.import_module(f'.{attr}', parent.__name__)
+
+        ret = from_child(attr)
+        if ret is _NO_ATTRIBUTE:
+            raise AttributeError(f'module {parent.__name__} has no attribute {attr}')
+        callback(attr)
+        return ret
+
+    @functools.cache
+    def from_child(attr):
+        nonlocal child
+        if attr not in allowed_attributes:
+            if ... not in allowed_attributes or _is_dunder(attr):
+                return _NO_ATTRIBUTE
+
+        if isinstance(child, str):
+            child = importlib.import_module(child, parent.__name__)
+
+        if _is_package(child):
+            with contextlib.suppress(ImportError):
+                return passthrough_module(f'{parent.__name__}.{attr}',
+                                          importlib.import_module(f'.{attr}', child.__name__))
+
+        with contextlib.suppress(AttributeError):
+            return getattr(child, attr)
+
+        return _NO_ATTRIBUTE
+
+    parent = sys.modules.get(parent, types.ModuleType(parent))
+    parent.__class__ = EnhancedModule
+    parent.__getattr__ = __getattr__
+    return parent

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/imghdr.py

@@ -0,0 +1,22 @@
+def what(file=None, h=None):
+    """Detect format of image (Currently supports jpeg, png, webp, gif only)
+    Ref: https://github.com/python/cpython/blob/3.11/Lib/imghdr.py
+    Ref: https://www.w3.org/Graphics/JPEG/itu-t81.pdf
+    """
+    if h is None:
+        with open(file, 'rb') as f:
+            h = f.read(12)
+
+    if h.startswith(b'RIFF') and h.startswith(b'WEBP', 8):
+        return 'webp'
+
+    if h.startswith(b'\x89PNG'):
+        return 'png'
+
+    if h.startswith(b'\xFF\xD8\xFF'):
+        return 'jpeg'
+
+    if h.startswith(b'GIF'):
+        return 'gif'
+
+    return None

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/shutil.py

@@ -0,0 +1,30 @@
+# flake8: noqa: F405
+from shutil import *  # noqa: F403
+
+from .compat_utils import passthrough_module
+
+passthrough_module(__name__, 'shutil')
+del passthrough_module
+
+
+import sys
+
+if sys.platform.startswith('freebsd'):
+    import errno
+    import os
+    import shutil
+
+    # Workaround for PermissionError when using restricted ACL mode on FreeBSD
+    def copy2(src, dst, *args, **kwargs):
+        if os.path.isdir(dst):
+            dst = os.path.join(dst, os.path.basename(src))
+        shutil.copyfile(src, dst, *args, **kwargs)
+        try:
+            shutil.copystat(src, dst, *args, **kwargs)
+        except PermissionError as e:
+            if e.errno != getattr(errno, 'EPERM', None):
+                raise
+        return dst
+
+    def move(*args, copy_function=copy2, **kwargs):
+        return shutil.move(*args, copy_function=copy_function, **kwargs)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/urllib/__init__.py

@@ -0,0 +1,10 @@
+# flake8: noqa: F405
+from urllib import *  # noqa: F403
+
+del request  # noqa: F821
+from . import request  # noqa: F401
+
+from ..compat_utils import passthrough_module
+
+passthrough_module(__name__, 'urllib')
+del passthrough_module

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/compat/urllib/request.py

@@ -0,0 +1,36 @@
+# flake8: noqa: F405
+from urllib.request import *  # noqa: F403
+
+from ..compat_utils import passthrough_module
+
+passthrough_module(__name__, 'urllib.request')
+del passthrough_module
+
+
+import os
+
+if os.name == 'nt':
+    # On older Python versions, proxies are extracted from Windows registry erroneously. [1]
+    # If the https proxy in the registry does not have a scheme, urllib will incorrectly add https:// to it. [2]
+    # It is unlikely that the user has actually set it to be https, so we should be fine to safely downgrade
+    # it to http on these older Python versions to avoid issues
+    # This also applies for ftp proxy type, as ftp:// proxy scheme is not supported.
+    # 1: https://github.com/python/cpython/issues/86793
+    # 2: https://github.com/python/cpython/blob/51f1ae5ceb0673316c4e4b0175384e892e33cc6e/Lib/urllib/request.py#L2683-L2698
+    import sys
+    from urllib.request import getproxies_environment, getproxies_registry
+
+    def getproxies_registry_patched():
+        proxies = getproxies_registry()
+
+        if sys.version_info < (3, 10, 5):  # https://docs.python.org/3.10/whatsnew/changelog.html#python-3-10-5-final
+            for scheme in ('https', 'ftp'):
+                if scheme in proxies and proxies[scheme].startswith(f'{scheme}://'):
+                    proxies[scheme] = 'http' + proxies[scheme][len(scheme):]
+
+        return proxies
+
+    def getproxies():
+        return getproxies_environment() or getproxies_registry_patched()
+
+del os

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/dependencies/Cryptodome.py

@@ -0,0 +1,38 @@
+from ..compat.compat_utils import passthrough_module
+
+try:
+    import Cryptodome as _parent
+except ImportError:
+    try:
+        import Crypto as _parent
+    except (ImportError, SyntaxError):  # Old Crypto gives SyntaxError in newer Python
+        _parent = passthrough_module(__name__, 'no_Cryptodome')
+        __bool__ = lambda: False
+
+del passthrough_module
+
+__version__ = ''
+AES = PKCS1_v1_5 = Blowfish = PKCS1_OAEP = SHA1 = CMAC = RSA = None
+try:
+    if _parent.__name__ == 'Cryptodome':
+        from Cryptodome import __version__
+        from Cryptodome.Cipher import AES, PKCS1_OAEP, Blowfish, PKCS1_v1_5
+        from Cryptodome.Hash import CMAC, SHA1
+        from Cryptodome.PublicKey import RSA
+    elif _parent.__name__ == 'Crypto':
+        from Crypto import __version__
+        from Crypto.Cipher import AES, PKCS1_OAEP, Blowfish, PKCS1_v1_5  # noqa: F401
+        from Crypto.Hash import CMAC, SHA1  # noqa: F401
+        from Crypto.PublicKey import RSA  # noqa: F401
+except (ImportError, OSError):
+    __version__ = f'broken {__version__}'.strip()
+
+
+_yt_dlp__identifier = _parent.__name__
+if AES and _yt_dlp__identifier == 'Crypto':
+    try:
+        # In pycrypto, mode defaults to ECB. See:
+        # https://www.pycryptodome.org/en/latest/src/vs_pycrypto.html#:~:text=not%20have%20ECB%20as%20default%20mode
+        AES.new(b'abcdefghijklmnop')
+    except TypeError:
+        _yt_dlp__identifier = 'pycrypto'

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/dependencies/__init__.py

@@ -0,0 +1,96 @@
+# flake8: noqa: F401
+"""Imports all optional dependencies for the project.
+An attribute "_yt_dlp__identifier" may be inserted into the module if it uses an ambiguous namespace"""
+
+try:
+    import brotlicffi as brotli
+except ImportError:
+    try:
+        import brotli
+    except ImportError:
+        brotli = None
+
+
+try:
+    import certifi
+except ImportError:
+    certifi = None
+else:
+    from os.path import exists as _path_exists
+
+    # The certificate may not be bundled in executable
+    if not _path_exists(certifi.where()):
+        certifi = None
+
+
+try:
+    import mutagen
+except ImportError:
+    mutagen = None
+
+
+secretstorage = None
+try:
+    import secretstorage
+    _SECRETSTORAGE_UNAVAILABLE_REASON = None
+except ImportError:
+    _SECRETSTORAGE_UNAVAILABLE_REASON = (
+        'as the `secretstorage` module is not installed. '
+        'Please install by running `python3 -m pip install secretstorage`')
+except Exception as _err:
+    _SECRETSTORAGE_UNAVAILABLE_REASON = f'as the `secretstorage` module could not be initialized. {_err}'
+
+
+try:
+    import sqlite3
+    # We need to get the underlying `sqlite` version, see https://github.com/yt-dlp/yt-dlp/issues/8152
+    sqlite3._yt_dlp__version = sqlite3.sqlite_version
+except ImportError:
+    # although sqlite3 is part of the standard library, it is possible to compile Python without
+    # sqlite support. See: https://github.com/yt-dlp/yt-dlp/issues/544
+    sqlite3 = None
+
+
+try:
+    import websockets
+except ImportError:
+    websockets = None
+
+try:
+    import urllib3
+except ImportError:
+    urllib3 = None
+
+try:
+    import requests
+except ImportError:
+    requests = None
+
+try:
+    import xattr  # xattr or pyxattr
+except ImportError:
+    xattr = None
+else:
+    if hasattr(xattr, 'set'):  # pyxattr
+        xattr._yt_dlp__identifier = 'pyxattr'
+
+try:
+    import curl_cffi
+except ImportError:
+    curl_cffi = None
+
+from . import Cryptodome
+
+all_dependencies = {k: v for k, v in globals().items() if not k.startswith('_')}
+available_dependencies = {k: v for k, v in all_dependencies.items() if v}
+
+
+# Deprecated
+Cryptodome_AES = Cryptodome.AES
+
+
+__all__ = [
+    'all_dependencies',
+    'available_dependencies',
+    *all_dependencies.keys(),
+]

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/__init__.py

@@ -0,0 +1,131 @@
+from ..utils import NO_DEFAULT, determine_protocol
+
+
+def get_suitable_downloader(info_dict, params={}, default=NO_DEFAULT, protocol=None, to_stdout=False):
+    info_dict['protocol'] = determine_protocol(info_dict)
+    info_copy = info_dict.copy()
+    info_copy['to_stdout'] = to_stdout
+
+    protocols = (protocol or info_copy['protocol']).split('+')
+    downloaders = [_get_suitable_downloader(info_copy, proto, params, default) for proto in protocols]
+
+    if set(downloaders) == {FFmpegFD} and FFmpegFD.can_merge_formats(info_copy, params):
+        return FFmpegFD
+    elif (set(downloaders) == {DashSegmentsFD}
+          and not (to_stdout and len(protocols) > 1)
+          and set(protocols) == {'http_dash_segments_generator'}):
+        return DashSegmentsFD
+    elif len(downloaders) == 1:
+        return downloaders[0]
+    return None
+
+
+# Some of these require get_suitable_downloader
+from .common import FileDownloader
+from .dash import DashSegmentsFD
+from .external import FFmpegFD, get_external_downloader
+from .f4m import F4mFD
+from .fc2 import FC2LiveFD
+from .hls import HlsFD
+from .http import HttpFD
+from .ism import IsmFD
+from .mhtml import MhtmlFD
+from .niconico import NiconicoLiveFD
+from .rtmp import RtmpFD
+from .rtsp import RtspFD
+from .websocket import WebSocketFragmentFD
+from .youtube_live_chat import YoutubeLiveChatFD
+from .bunnycdn import BunnyCdnFD
+
+PROTOCOL_MAP = {
+    'rtmp': RtmpFD,
+    'rtmpe': RtmpFD,
+    'rtmp_ffmpeg': FFmpegFD,
+    'm3u8_native': HlsFD,
+    'm3u8': FFmpegFD,
+    'mms': RtspFD,
+    'rtsp': RtspFD,
+    'f4m': F4mFD,
+    'http_dash_segments': DashSegmentsFD,
+    'http_dash_segments_generator': DashSegmentsFD,
+    'ism': IsmFD,
+    'mhtml': MhtmlFD,
+    'niconico_live': NiconicoLiveFD,
+    'fc2_live': FC2LiveFD,
+    'websocket_frag': WebSocketFragmentFD,
+    'youtube_live_chat': YoutubeLiveChatFD,
+    'youtube_live_chat_replay': YoutubeLiveChatFD,
+    'bunnycdn': BunnyCdnFD,
+}
+
+
+def shorten_protocol_name(proto, simplify=False):
+    short_protocol_names = {
+        'm3u8_native': 'm3u8',
+        'm3u8': 'm3u8F',
+        'rtmp_ffmpeg': 'rtmpF',
+        'http_dash_segments': 'dash',
+        'http_dash_segments_generator': 'dashG',
+        'websocket_frag': 'WSfrag',
+    }
+    if simplify:
+        short_protocol_names.update({
+            'https': 'http',
+            'ftps': 'ftp',
+            'm3u8': 'm3u8',  # Reverse above m3u8 mapping
+            'm3u8_native': 'm3u8',
+            'http_dash_segments_generator': 'dash',
+            'rtmp_ffmpeg': 'rtmp',
+            'm3u8_frag_urls': 'm3u8',
+            'dash_frag_urls': 'dash',
+        })
+    return short_protocol_names.get(proto, proto)
+
+
+def _get_suitable_downloader(info_dict, protocol, params, default):
+    """Get the downloader class that can handle the info dict."""
+    if default is NO_DEFAULT:
+        default = HttpFD
+
+    if (info_dict.get('section_start') or info_dict.get('section_end')) and FFmpegFD.can_download(info_dict):
+        return FFmpegFD
+
+    info_dict['protocol'] = protocol
+    downloaders = params.get('external_downloader')
+    external_downloader = (
+        downloaders if isinstance(downloaders, str) or downloaders is None
+        else downloaders.get(shorten_protocol_name(protocol, True), downloaders.get('default')))
+
+    if external_downloader is None:
+        if info_dict['to_stdout'] and FFmpegFD.can_merge_formats(info_dict, params):
+            return FFmpegFD
+    elif external_downloader.lower() != 'native' and info_dict.get('impersonate') is None:
+        ed = get_external_downloader(external_downloader)
+        if ed.can_download(info_dict, external_downloader):
+            return ed
+
+    if protocol == 'http_dash_segments':
+        if info_dict.get('is_live') and (external_downloader or '').lower() != 'native':
+            return FFmpegFD
+
+    if protocol in ('m3u8', 'm3u8_native'):
+        if info_dict.get('is_live'):
+            return FFmpegFD
+        elif (external_downloader or '').lower() == 'native':
+            return HlsFD
+        elif protocol == 'm3u8_native' and get_suitable_downloader(
+                info_dict, params, None, protocol='m3u8_frag_urls', to_stdout=info_dict['to_stdout']):
+            return HlsFD
+        elif params.get('hls_prefer_native') is True:
+            return HlsFD
+        elif params.get('hls_prefer_native') is False:
+            return FFmpegFD
+
+    return PROTOCOL_MAP.get(protocol, default)
+
+
+__all__ = [
+    'FileDownloader',
+    'get_suitable_downloader',
+    'shorten_protocol_name',
+]

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/bunnycdn.py

@@ -0,0 +1,50 @@
+import hashlib
+import random
+import threading
+
+from .common import FileDownloader
+from . import HlsFD
+from ..networking import Request
+from ..networking.exceptions import network_exceptions
+
+
+class BunnyCdnFD(FileDownloader):
+    """
+    Downloads from BunnyCDN with required pings
+    Note, this is not a part of public API, and will be removed without notice.
+    DO NOT USE
+    """
+
+    def real_download(self, filename, info_dict):
+        self.to_screen(f'[{self.FD_NAME}] Downloading from BunnyCDN')
+
+        fd = HlsFD(self.ydl, self.params)
+
+        stop_event = threading.Event()
+        ping_thread = threading.Thread(target=self.ping_thread, args=(stop_event,), kwargs=info_dict['_bunnycdn_ping_data'])
+        ping_thread.start()
+
+        try:
+            return fd.real_download(filename, info_dict)
+        finally:
+            stop_event.set()
+
+    def ping_thread(self, stop_event, url, headers, secret, context_id):
+        # Site sends ping every 4 seconds, but this throttles the download. Pinging every 2 seconds seems to work.
+        ping_interval = 2
+        # Hard coded resolution as it doesn't seem to matter
+        res = 1080
+        paused = 'false'
+        current_time = 0
+
+        while not stop_event.wait(ping_interval):
+            current_time += ping_interval
+
+            time = current_time + round(random.random(), 6)
+            md5_hash = hashlib.md5(f'{secret}_{context_id}_{time}_{paused}_{res}'.encode()).hexdigest()
+            ping_url = f'{url}?hash={md5_hash}&time={time}&paused={paused}&resolution={res}'
+
+            try:
+                self.ydl.urlopen(Request(ping_url, headers=headers)).read()
+            except network_exceptions as e:
+                self.to_screen(f'[{self.FD_NAME}] Ping failed: {e}')

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/common.py

@@ -0,0 +1,519 @@
+import contextlib
+import errno
+import functools
+import os
+import random
+import re
+import threading
+import time
+
+from ..minicurses import (
+    BreaklineStatusPrinter,
+    MultilineLogger,
+    MultilinePrinter,
+    QuietMultilinePrinter,
+)
+from ..utils import (
+    IDENTITY,
+    NO_DEFAULT,
+    LockingUnsupportedError,
+    Namespace,
+    RetryManager,
+    classproperty,
+    deprecation_warning,
+    format_bytes,
+    join_nonempty,
+    parse_bytes,
+    remove_start,
+    sanitize_open,
+    shell_quote,
+    timeconvert,
+    timetuple_from_msec,
+    try_call,
+)
+from ..utils._utils import _ProgressState
+
+
+class FileDownloader:
+    """File Downloader class.
+
+    File downloader objects are the ones responsible of downloading the
+    actual video file and writing it to disk.
+
+    File downloaders accept a lot of parameters. In order not to saturate
+    the object constructor with arguments, it receives a dictionary of
+    options instead.
+
+    Available options:
+
+    verbose:            Print additional info to stdout.
+    quiet:              Do not print messages to stdout.
+    ratelimit:          Download speed limit, in bytes/sec.
+    throttledratelimit: Assume the download is being throttled below this speed (bytes/sec)
+    retries:            Number of times to retry for expected network errors.
+                        Default is 0 for API, but 10 for CLI
+    file_access_retries:   Number of times to retry on file access error (default: 3)
+    buffersize:         Size of download buffer in bytes.
+    noresizebuffer:     Do not automatically resize the download buffer.
+    continuedl:         Try to continue downloads if possible.
+    noprogress:         Do not print the progress bar.
+    nopart:             Do not use temporary .part files.
+    updatetime:         Use the Last-modified header to set output file timestamps.
+    test:               Download only first bytes to test the downloader.
+    min_filesize:       Skip files smaller than this size
+    max_filesize:       Skip files larger than this size
+    progress_delta:     The minimum time between progress output, in seconds
+    external_downloader_args:  A dictionary of downloader keys (in lower case)
+                        and a list of additional command-line arguments for the
+                        executable. Use 'default' as the name for arguments to be
+                        passed to all downloaders. For compatibility with youtube-dl,
+                        a single list of args can also be used
+    hls_use_mpegts:     Use the mpegts container for HLS videos.
+    http_chunk_size:    Size of a chunk for chunk-based HTTP downloading. May be
+                        useful for bypassing bandwidth throttling imposed by
+                        a webserver (experimental)
+    progress_template:  See YoutubeDL.py
+    retry_sleep_functions: See YoutubeDL.py
+
+    Subclasses of this one must re-define the real_download method.
+    """
+
+    _TEST_FILE_SIZE = 10241
+    params = None
+
+    def __init__(self, ydl, params):
+        """Create a FileDownloader object with the given options."""
+        self._set_ydl(ydl)
+        self._progress_hooks = []
+        self.params = params
+        self._prepare_multiline_status()
+        self.add_progress_hook(self.report_progress)
+        if self.params.get('progress_delta'):
+            self._progress_delta_lock = threading.Lock()
+            self._progress_delta_time = time.monotonic()
+
+    def _set_ydl(self, ydl):
+        self.ydl = ydl
+
+        for func in (
+            'deprecation_warning',
+            'deprecated_feature',
+            'report_error',
+            'report_file_already_downloaded',
+            'report_warning',
+            'to_console_title',
+            'to_stderr',
+            'trouble',
+            'write_debug',
+        ):
+            if not hasattr(self, func):
+                setattr(self, func, getattr(ydl, func))
+
+    def to_screen(self, *args, **kargs):
+        self.ydl.to_screen(*args, quiet=self.params.get('quiet'), **kargs)
+
+    __to_screen = to_screen
+
+    @classproperty
+    def FD_NAME(cls):
+        return re.sub(r'(?<=[a-z])(?=[A-Z])', '_', cls.__name__[:-2]).lower()
+
+    @staticmethod
+    def format_seconds(seconds):
+        if seconds is None:
+            return ' Unknown'
+        time = timetuple_from_msec(seconds * 1000)
+        if time.hours > 99:
+            return '--:--:--'
+        return '%02d:%02d:%02d' % time[:-1]
+
+    @classmethod
+    def format_eta(cls, seconds):
+        return f'{remove_start(cls.format_seconds(seconds), "00:"):>8s}'
+
+    @staticmethod
+    def calc_percent(byte_counter, data_len):
+        if data_len is None:
+            return None
+        return float(byte_counter) / float(data_len) * 100.0
+
+    @staticmethod
+    def format_percent(percent):
+        return '  N/A%' if percent is None else f'{percent:>5.1f}%'
+
+    @classmethod
+    def calc_eta(cls, start_or_rate, now_or_remaining, total=NO_DEFAULT, current=NO_DEFAULT):
+        if total is NO_DEFAULT:
+            rate, remaining = start_or_rate, now_or_remaining
+            if None in (rate, remaining):
+                return None
+            return int(float(remaining) / rate)
+
+        start, now = start_or_rate, now_or_remaining
+        if total is None:
+            return None
+        if now is None:
+            now = time.time()
+        rate = cls.calc_speed(start, now, current)
+        return rate and int((float(total) - float(current)) / rate)
+
+    @staticmethod
+    def calc_speed(start, now, bytes):
+        dif = now - start
+        if bytes == 0 or dif < 0.001:  # One millisecond
+            return None
+        return float(bytes) / dif
+
+    @staticmethod
+    def format_speed(speed):
+        return ' Unknown B/s' if speed is None else f'{format_bytes(speed):>10s}/s'
+
+    @staticmethod
+    def format_retries(retries):
+        return 'inf' if retries == float('inf') else int(retries)
+
+    @staticmethod
+    def filesize_or_none(unencoded_filename):
+        if os.path.isfile(unencoded_filename):
+            return os.path.getsize(unencoded_filename)
+        return 0
+
+    @staticmethod
+    def best_block_size(elapsed_time, bytes):
+        new_min = max(bytes / 2.0, 1.0)
+        new_max = min(max(bytes * 2.0, 1.0), 4194304)  # Do not surpass 4 MB
+        if elapsed_time < 0.001:
+            return int(new_max)
+        rate = bytes / elapsed_time
+        if rate > new_max:
+            return int(new_max)
+        if rate < new_min:
+            return int(new_min)
+        return int(rate)
+
+    @staticmethod
+    def parse_bytes(bytestr):
+        """Parse a string indicating a byte quantity into an integer."""
+        deprecation_warning('yt_dlp.FileDownloader.parse_bytes is deprecated and '
+                            'may be removed in the future. Use yt_dlp.utils.parse_bytes instead')
+        return parse_bytes(bytestr)
+
+    def slow_down(self, start_time, now, byte_counter):
+        """Sleep if the download speed is over the rate limit."""
+        rate_limit = self.params.get('ratelimit')
+        if rate_limit is None or byte_counter == 0:
+            return
+        if now is None:
+            now = time.time()
+        elapsed = now - start_time
+        if elapsed <= 0.0:
+            return
+        speed = float(byte_counter) / elapsed
+        if speed > rate_limit:
+            sleep_time = float(byte_counter) / rate_limit - elapsed
+            if sleep_time > 0:
+                time.sleep(sleep_time)
+
+    def temp_name(self, filename):
+        """Returns a temporary filename for the given filename."""
+        if self.params.get('nopart', False) or filename == '-' or \
+                (os.path.exists(filename) and not os.path.isfile(filename)):
+            return filename
+        return filename + '.part'
+
+    def undo_temp_name(self, filename):
+        if filename.endswith('.part'):
+            return filename[:-len('.part')]
+        return filename
+
+    def ytdl_filename(self, filename):
+        return filename + '.ytdl'
+
+    def wrap_file_access(action, *, fatal=False):
+        def error_callback(err, count, retries, *, fd):
+            return RetryManager.report_retry(
+                err, count, retries, info=fd.__to_screen,
+                warn=lambda e: (time.sleep(0.01), fd.to_screen(f'[download] Unable to {action} file: {e}')),
+                error=None if fatal else lambda e: fd.report_error(f'Unable to {action} file: {e}'),
+                sleep_func=fd.params.get('retry_sleep_functions', {}).get('file_access'))
+
+        def wrapper(self, func, *args, **kwargs):
+            for retry in RetryManager(self.params.get('file_access_retries', 3), error_callback, fd=self):
+                try:
+                    return func(self, *args, **kwargs)
+                except OSError as err:
+                    if err.errno in (errno.EACCES, errno.EINVAL):
+                        retry.error = err
+                        continue
+                    retry.error_callback(err, 1, 0)
+
+        return functools.partial(functools.partialmethod, wrapper)
+
+    @wrap_file_access('open', fatal=True)
+    def sanitize_open(self, filename, open_mode):
+        f, filename = sanitize_open(filename, open_mode)
+        if not getattr(f, 'locked', None):
+            self.write_debug(f'{LockingUnsupportedError.msg}. Proceeding without locking', only_once=True)
+        return f, filename
+
+    @wrap_file_access('remove')
+    def try_remove(self, filename):
+        if os.path.isfile(filename):
+            os.remove(filename)
+
+    @wrap_file_access('rename')
+    def try_rename(self, old_filename, new_filename):
+        if old_filename == new_filename:
+            return
+        os.replace(old_filename, new_filename)
+
+    def try_utime(self, filename, last_modified_hdr):
+        """Try to set the last-modified time of the given file."""
+        if last_modified_hdr is None:
+            return
+        if not os.path.isfile(filename):
+            return
+        timestr = last_modified_hdr
+        if timestr is None:
+            return
+        filetime = timeconvert(timestr)
+        if filetime is None:
+            return filetime
+        # Ignore obviously invalid dates
+        if filetime == 0:
+            return
+        with contextlib.suppress(Exception):
+            os.utime(filename, (time.time(), filetime))
+        return filetime
+
+    def report_destination(self, filename):
+        """Report destination filename."""
+        self.to_screen('[download] Destination: ' + filename)
+
+    def _prepare_multiline_status(self, lines=1):
+        if self.params.get('noprogress'):
+            self._multiline = QuietMultilinePrinter()
+        elif self.ydl.params.get('logger'):
+            self._multiline = MultilineLogger(self.ydl.params['logger'], lines)
+        elif self.params.get('progress_with_newline'):
+            self._multiline = BreaklineStatusPrinter(self.ydl._out_files.out, lines)
+        else:
+            self._multiline = MultilinePrinter(self.ydl._out_files.out, lines, not self.params.get('quiet'))
+        self._multiline.allow_colors = self.ydl._allow_colors.out and self.ydl._allow_colors.out != 'no_color'
+        self._multiline._HAVE_FULLCAP = self.ydl._allow_colors.out
+
+    def _finish_multiline_status(self):
+        self._multiline.end()
+
+    ProgressStyles = Namespace(
+        downloaded_bytes='light blue',
+        percent='light blue',
+        eta='yellow',
+        speed='green',
+        elapsed='bold white',
+        total_bytes='',
+        total_bytes_estimate='',
+    )
+
+    def _report_progress_status(self, s, default_template):
+        for name, style in self.ProgressStyles.items_:
+            name = f'_{name}_str'
+            if name not in s:
+                continue
+            s[name] = self._format_progress(s[name], style)
+        s['_default_template'] = default_template % s
+
+        progress_dict = s.copy()
+        progress_dict.pop('info_dict')
+        progress_dict = {'info': s['info_dict'], 'progress': progress_dict}
+
+        progress_template = self.params.get('progress_template', {})
+        self._multiline.print_at_line(self.ydl.evaluate_outtmpl(
+            progress_template.get('download') or '[download] %(progress._default_template)s',
+            progress_dict), s.get('progress_idx') or 0)
+        self.to_console_title(self.ydl.evaluate_outtmpl(
+            progress_template.get('download-title') or 'yt-dlp %(progress._default_template)s',
+            progress_dict), _ProgressState.from_dict(s), s.get('_percent'))
+
+    def _format_progress(self, *args, **kwargs):
+        return self.ydl._format_text(
+            self._multiline.stream, self._multiline.allow_colors, *args, **kwargs)
+
+    def report_progress(self, s):
+        def with_fields(*tups, default=''):
+            for *fields, tmpl in tups:
+                if all(s.get(f) is not None for f in fields):
+                    return tmpl
+            return default
+
+        _format_bytes = lambda k: f'{format_bytes(s.get(k)):>10s}'
+
+        if s['status'] == 'finished':
+            if self.params.get('noprogress'):
+                self.to_screen('[download] Download completed')
+            speed = try_call(lambda: s['total_bytes'] / s['elapsed'])
+            s.update({
+                'speed': speed,
+                '_speed_str': self.format_speed(speed).strip(),
+                '_total_bytes_str': _format_bytes('total_bytes'),
+                '_elapsed_str': self.format_seconds(s.get('elapsed')),
+                '_percent': 100.0,
+                '_percent_str': self.format_percent(100),
+            })
+            self._report_progress_status(s, join_nonempty(
+                '100%%',
+                with_fields(('total_bytes', 'of %(_total_bytes_str)s')),
+                with_fields(('elapsed', 'in %(_elapsed_str)s')),
+                with_fields(('speed', 'at %(_speed_str)s')),
+                delim=' '))
+
+        if s['status'] != 'downloading':
+            return
+
+        if update_delta := self.params.get('progress_delta'):
+            with self._progress_delta_lock:
+                if time.monotonic() < self._progress_delta_time:
+                    return
+                self._progress_delta_time += update_delta
+
+        progress = try_call(
+            lambda: 100 * s['downloaded_bytes'] / s['total_bytes'],
+            lambda: 100 * s['downloaded_bytes'] / s['total_bytes_estimate'],
+            lambda: s['downloaded_bytes'] == 0 and 0)
+        s.update({
+            '_eta_str': self.format_eta(s.get('eta')).strip(),
+            '_speed_str': self.format_speed(s.get('speed')),
+            '_percent': progress,
+            '_percent_str': self.format_percent(progress),
+            '_total_bytes_str': _format_bytes('total_bytes'),
+            '_total_bytes_estimate_str': _format_bytes('total_bytes_estimate'),
+            '_downloaded_bytes_str': _format_bytes('downloaded_bytes'),
+            '_elapsed_str': self.format_seconds(s.get('elapsed')),
+        })
+
+        msg_template = with_fields(
+            ('total_bytes', '%(_percent_str)s of %(_total_bytes_str)s at %(_speed_str)s ETA %(_eta_str)s'),
+            ('total_bytes_estimate', '%(_percent_str)s of ~%(_total_bytes_estimate_str)s at %(_speed_str)s ETA %(_eta_str)s'),
+            ('downloaded_bytes', 'elapsed', '%(_downloaded_bytes_str)s at %(_speed_str)s (%(_elapsed_str)s)'),
+            ('downloaded_bytes', '%(_downloaded_bytes_str)s at %(_speed_str)s'),
+            default='%(_percent_str)s at %(_speed_str)s ETA %(_eta_str)s')
+
+        msg_template += with_fields(
+            ('fragment_index', 'fragment_count', ' (frag %(fragment_index)s/%(fragment_count)s)'),
+            ('fragment_index', ' (frag %(fragment_index)s)'))
+        self._report_progress_status(s, msg_template)
+
+    def report_resuming_byte(self, resume_len):
+        """Report attempt to resume at given byte."""
+        self.to_screen(f'[download] Resuming download at byte {resume_len}')
+
+    def report_retry(self, err, count, retries, frag_index=NO_DEFAULT, fatal=True):
+        """Report retry"""
+        is_frag = False if frag_index is NO_DEFAULT else 'fragment'
+        RetryManager.report_retry(
+            err, count, retries, info=self.__to_screen,
+            warn=lambda msg: self.__to_screen(f'[download] Got error: {msg}'),
+            error=IDENTITY if not fatal else lambda e: self.report_error(f'\r[download] Got error: {e}'),
+            sleep_func=self.params.get('retry_sleep_functions', {}).get(is_frag or 'http'),
+            suffix=f'fragment{"s" if frag_index is None else f" {frag_index}"}' if is_frag else None)
+
+    def report_unable_to_resume(self):
+        """Report it was impossible to resume download."""
+        self.to_screen('[download] Unable to resume')
+
+    @staticmethod
+    def supports_manifest(manifest):
+        """ Whether the downloader can download the fragments from the manifest.
+        Redefine in subclasses if needed. """
+        pass
+
+    def download(self, filename, info_dict, subtitle=False):
+        """Download to a filename using the info from info_dict
+        Return True on success and False otherwise
+        """
+        nooverwrites_and_exists = (
+            not self.params.get('overwrites', True)
+            and os.path.exists(filename)
+        )
+
+        if not hasattr(filename, 'write'):
+            continuedl_and_exists = (
+                self.params.get('continuedl', True)
+                and os.path.isfile(filename)
+                and not self.params.get('nopart', False)
+            )
+
+            # Check file already present
+            if filename != '-' and (nooverwrites_and_exists or continuedl_and_exists):
+                self.report_file_already_downloaded(filename)
+                self._hook_progress({
+                    'filename': filename,
+                    'status': 'finished',
+                    'total_bytes': os.path.getsize(filename),
+                }, info_dict)
+                self._finish_multiline_status()
+                return True, False
+
+        sleep_note = ''
+        if subtitle:
+            sleep_interval = self.params.get('sleep_interval_subtitles') or 0
+        else:
+            min_sleep_interval = self.params.get('sleep_interval') or 0
+            max_sleep_interval = self.params.get('max_sleep_interval') or 0
+
+            if available_at := info_dict.get('available_at'):
+                forced_sleep_interval = available_at - int(time.time())
+                if forced_sleep_interval > min_sleep_interval:
+                    sleep_note = 'as required by the site'
+                    min_sleep_interval = forced_sleep_interval
+                if forced_sleep_interval > max_sleep_interval:
+                    max_sleep_interval = forced_sleep_interval
+
+            sleep_interval = random.uniform(
+                min_sleep_interval, max_sleep_interval or min_sleep_interval)
+
+        if sleep_interval > 0:
+            self.to_screen(f'[download] Sleeping {sleep_interval:.2f} seconds {sleep_note}...')
+            time.sleep(sleep_interval)
+
+        ret = self.real_download(filename, info_dict)
+        self._finish_multiline_status()
+        return ret, True
+
+    def real_download(self, filename, info_dict):
+        """Real download process. Redefine in subclasses."""
+        raise NotImplementedError('This method must be implemented by subclasses')
+
+    def _hook_progress(self, status, info_dict):
+        # Ideally we want to make a copy of the dict, but that is too slow
+        status['info_dict'] = info_dict
+        # youtube-dl passes the same status object to all the hooks.
+        # Some third party scripts seems to be relying on this.
+        # So keep this behavior if possible
+        for ph in self._progress_hooks:
+            ph(status)
+
+    def add_progress_hook(self, ph):
+        # See YoutubeDl.py (search for progress_hooks) for a description of
+        # this interface
+        self._progress_hooks.append(ph)
+
+    def _debug_cmd(self, args, exe=None):
+        if not self.params.get('verbose', False):
+            return
+
+        if exe is None:
+            exe = os.path.basename(args[0])
+
+        self.write_debug(f'{exe} command line: {shell_quote(args)}')
+
+    def _get_impersonate_target(self, info_dict):
+        impersonate = info_dict.get('impersonate')
+        if impersonate is None:
+            return None
+        available_target, requested_targets = self.ydl._parse_impersonate_targets(impersonate)
+        if available_target:
+            return available_target
+        elif requested_targets:
+            self.report_warning(self.ydl._unavailable_targets_message(requested_targets))
+        return None

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/dash.py

@@ -0,0 +1,95 @@
+import time
+import urllib.parse
+
+from . import get_suitable_downloader
+from .fragment import FragmentFD
+from ..utils import ReExtractInfo, update_url_query, urljoin
+
+
+class DashSegmentsFD(FragmentFD):
+    """
+    Download segments in a DASH manifest. External downloaders can take over
+    the fragment downloads by supporting the 'dash_frag_urls' protocol
+    """
+
+    FD_NAME = 'dashsegments'
+
+    def real_download(self, filename, info_dict):
+        if 'http_dash_segments_generator' in info_dict['protocol'].split('+'):
+            real_downloader = None  # No external FD can support --live-from-start
+        else:
+            if info_dict.get('is_live'):
+                self.report_error('Live DASH videos are not supported')
+            real_downloader = get_suitable_downloader(
+                info_dict, self.params, None, protocol='dash_frag_urls', to_stdout=(filename == '-'))
+
+        real_start = time.time()
+
+        requested_formats = [{**info_dict, **fmt} for fmt in info_dict.get('requested_formats', [])]
+        args = []
+        for fmt in requested_formats or [info_dict]:
+            # Re-extract if --load-info-json is used and 'fragments' was originally a generator
+            # See https://github.com/yt-dlp/yt-dlp/issues/13906
+            if isinstance(fmt['fragments'], str):
+                raise ReExtractInfo('the stream needs to be re-extracted', expected=True)
+
+            try:
+                fragment_count = 1 if self.params.get('test') else len(fmt['fragments'])
+            except TypeError:
+                fragment_count = None
+            ctx = {
+                'filename': fmt.get('filepath') or filename,
+                'live': 'is_from_start' if fmt.get('is_from_start') else fmt.get('is_live'),
+                'total_frags': fragment_count,
+            }
+
+            if real_downloader:
+                self._prepare_external_frag_download(ctx)
+            else:
+                self._prepare_and_start_frag_download(ctx, fmt)
+            ctx['start'] = real_start
+
+            extra_query = None
+            extra_param_to_segment_url = info_dict.get('extra_param_to_segment_url')
+            if extra_param_to_segment_url:
+                extra_query = urllib.parse.parse_qs(extra_param_to_segment_url)
+
+            fragments_to_download = self._get_fragments(fmt, ctx, extra_query)
+
+            if real_downloader:
+                self.to_screen(
+                    f'[{self.FD_NAME}] Fragment downloads will be delegated to {real_downloader.get_basename()}')
+                info_dict['fragments'] = list(fragments_to_download)
+                fd = real_downloader(self.ydl, self.params)
+                return fd.real_download(filename, info_dict)
+
+            args.append([ctx, fragments_to_download, fmt])
+
+        return self.download_and_append_fragments_multiple(*args, is_fatal=lambda idx: idx == 0)
+
+    def _resolve_fragments(self, fragments, ctx):
+        fragments = fragments(ctx) if callable(fragments) else fragments
+        return [next(iter(fragments))] if self.params.get('test') else fragments
+
+    def _get_fragments(self, fmt, ctx, extra_query):
+        fragment_base_url = fmt.get('fragment_base_url')
+        fragments = self._resolve_fragments(fmt['fragments'], ctx)
+
+        frag_index = 0
+        for i, fragment in enumerate(fragments):
+            frag_index += 1
+            if frag_index <= ctx['fragment_index']:
+                continue
+            fragment_url = fragment.get('url')
+            if not fragment_url:
+                assert fragment_base_url
+                fragment_url = urljoin(fragment_base_url, fragment['path'])
+            if extra_query:
+                fragment_url = update_url_query(fragment_url, extra_query)
+
+            yield {
+                'frag_index': frag_index,
+                'fragment_count': fragment.get('fragment_count'),
+                'index': i,
+                'url': fragment_url,
+            }

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/external.py

@@ -0,0 +1,673 @@
+import enum
+import functools
+import json
+import os
+import re
+import subprocess
+import sys
+import tempfile
+import time
+import uuid
+
+from .fragment import FragmentFD
+from ..networking import Request
+from ..postprocessor.ffmpeg import EXT_TO_OUT_FORMATS, FFmpegPostProcessor
+from ..utils import (
+    Popen,
+    RetryManager,
+    _configuration_args,
+    check_executable,
+    classproperty,
+    cli_bool_option,
+    cli_option,
+    cli_valueless_option,
+    determine_ext,
+    encodeArgument,
+    find_available_port,
+    remove_end,
+    traverse_obj,
+)
+
+
+class Features(enum.Enum):
+    TO_STDOUT = enum.auto()
+    MULTIPLE_FORMATS = enum.auto()
+
+
+class ExternalFD(FragmentFD):
+    SUPPORTED_PROTOCOLS = ('http', 'https', 'ftp', 'ftps')
+    SUPPORTED_FEATURES = ()
+    _CAPTURE_STDERR = True
+
+    def real_download(self, filename, info_dict):
+        self.report_destination(filename)
+        tmpfilename = self.temp_name(filename)
+        self._cookies_tempfile = None
+
+        try:
+            started = time.time()
+            retval = self._call_downloader(tmpfilename, info_dict)
+        except KeyboardInterrupt:
+            if not info_dict.get('is_live'):
+                raise
+            # Live stream downloading cancellation should be considered as
+            # correct and expected termination thus all postprocessing
+            # should take place
+            retval = 0
+            self.to_screen(f'[{self.get_basename()}] Interrupted by user')
+        finally:
+            if self._cookies_tempfile:
+                self.try_remove(self._cookies_tempfile)
+
+        if retval == 0:
+            status = {
+                'filename': filename,
+                'status': 'finished',
+                'elapsed': time.time() - started,
+            }
+            if filename != '-':
+                fsize = os.path.getsize(tmpfilename)
+                self.try_rename(tmpfilename, filename)
+                status.update({
+                    'downloaded_bytes': fsize,
+                    'total_bytes': fsize,
+                })
+            self._hook_progress(status, info_dict)
+            return True
+        else:
+            self.to_stderr('\n')
+            self.report_error('%s exited with code %d' % (
+                self.get_basename(), retval))
+            return False
+
+    @classmethod
+    def get_basename(cls):
+        return cls.__name__[:-2].lower()
+
+    @classproperty
+    def EXE_NAME(cls):
+        return cls.get_basename()
+
+    @functools.cached_property
+    def exe(self):
+        return self.EXE_NAME
+
+    @classmethod
+    def available(cls, path=None):
+        path = check_executable(
+            cls.EXE_NAME if path in (None, cls.get_basename()) else path,
+            [cls.AVAILABLE_OPT])
+        if not path:
+            return False
+        cls.exe = path
+        return path
+
+    @classmethod
+    def supports(cls, info_dict):
+        return all((
+            not info_dict.get('to_stdout') or Features.TO_STDOUT in cls.SUPPORTED_FEATURES,
+            '+' not in info_dict['protocol'] or Features.MULTIPLE_FORMATS in cls.SUPPORTED_FEATURES,
+            not traverse_obj(info_dict, ('hls_aes', ...), 'extra_param_to_segment_url', 'extra_param_to_key_url'),
+            all(proto in cls.SUPPORTED_PROTOCOLS for proto in info_dict['protocol'].split('+')),
+        ))
+
+    @classmethod
+    def can_download(cls, info_dict, path=None):
+        return cls.available(path) and cls.supports(info_dict)
+
+    def _option(self, command_option, param):
+        return cli_option(self.params, command_option, param)
+
+    def _bool_option(self, command_option, param, true_value='true', false_value='false', separator=None):
+        return cli_bool_option(self.params, command_option, param, true_value, false_value, separator)
+
+    def _valueless_option(self, command_option, param, expected_value=True):
+        return cli_valueless_option(self.params, command_option, param, expected_value)
+
+    def _configuration_args(self, keys=None, *args, **kwargs):
+        return _configuration_args(
+            self.get_basename(), self.params.get('external_downloader_args'), self.EXE_NAME,
+            keys, *args, **kwargs)
+
+    def _write_cookies(self):
+        if not self.ydl.cookiejar.filename:
+            tmp_cookies = tempfile.NamedTemporaryFile(suffix='.cookies', delete=False)
+            tmp_cookies.close()
+            self._cookies_tempfile = tmp_cookies.name
+            self.to_screen(f'[download] Writing temporary cookies file to "{self._cookies_tempfile}"')
+        # real_download resets _cookies_tempfile; if it's None then save() will write to cookiejar.filename
+        self.ydl.cookiejar.save(self._cookies_tempfile)
+        return self.ydl.cookiejar.filename or self._cookies_tempfile
+
+    def _call_downloader(self, tmpfilename, info_dict):
+        """ Either overwrite this or implement _make_cmd """
+        cmd = [encodeArgument(a) for a in self._make_cmd(tmpfilename, info_dict)]
+
+        self._debug_cmd(cmd)
+
+        if 'fragments' not in info_dict:
+            _, stderr, returncode = self._call_process(cmd, info_dict)
+            if returncode and stderr:
+                self.to_stderr(stderr)
+            return returncode
+
+        skip_unavailable_fragments = self.params.get('skip_unavailable_fragments', True)
+
+        retry_manager = RetryManager(self.params.get('fragment_retries'), self.report_retry,
+                                     frag_index=None, fatal=not skip_unavailable_fragments)
+        for retry in retry_manager:
+            _, stderr, returncode = self._call_process(cmd, info_dict)
+            if not returncode:
+                break
+            # TODO: Decide whether to retry based on error code
+            # https://aria2.github.io/manual/en/html/aria2c.html#exit-status
+            if stderr:
+                self.to_stderr(stderr)
+            retry.error = Exception()
+            continue
+        if not skip_unavailable_fragments and retry_manager.error:
+            return -1
+
+        decrypt_fragment = self.decrypter(info_dict)
+        dest, _ = self.sanitize_open(tmpfilename, 'wb')
+        for frag_index, fragment in enumerate(info_dict['fragments']):
+            fragment_filename = f'{tmpfilename}-Frag{frag_index}'
+            try:
+                src, _ = self.sanitize_open(fragment_filename, 'rb')
+            except OSError as err:
+                if skip_unavailable_fragments and frag_index > 1:
+                    self.report_skip_fragment(frag_index, err)
+                    continue
+                self.report_error(f'Unable to open fragment {frag_index}; {err}')
+                return -1
+            dest.write(decrypt_fragment(fragment, src.read()))
+            src.close()
+            if not self.params.get('keep_fragments', False):
+                self.try_remove(fragment_filename)
+        dest.close()
+        self.try_remove(f'{tmpfilename}.frag.urls')
+        return 0
+
+    def _call_process(self, cmd, info_dict):
+        return Popen.run(cmd, text=True, stderr=subprocess.PIPE if self._CAPTURE_STDERR else None)
+
+
+class CurlFD(ExternalFD):
+    AVAILABLE_OPT = '-V'
+    _CAPTURE_STDERR = False  # curl writes the progress to stderr
+
+    def _make_cmd(self, tmpfilename, info_dict):
+        cmd = [self.exe, '--location', '-o', tmpfilename, '--compressed']
+        cookie_header = self.ydl.cookiejar.get_cookie_header(info_dict['url'])
+        if cookie_header:
+            cmd += ['--cookie', cookie_header]
+        if info_dict.get('http_headers') is not None:
+            for key, val in info_dict['http_headers'].items():
+                cmd += ['--header', f'{key}: {val}']
+
+        cmd += self._bool_option('--continue-at', 'continuedl', '-', '0')
+        cmd += self._valueless_option('--silent', 'noprogress')
+        cmd += self._valueless_option('--verbose', 'verbose')
+        cmd += self._option('--limit-rate', 'ratelimit')
+        retry = self._option('--retry', 'retries')
+        if len(retry) == 2:
+            if retry[1] in ('inf', 'infinite'):
+                retry[1] = '2147483647'
+            cmd += retry
+        cmd += self._option('--max-filesize', 'max_filesize')
+        cmd += self._option('--interface', 'source_address')
+        cmd += self._option('--proxy', 'proxy')
+        cmd += self._valueless_option('--insecure', 'nocheckcertificate')
+        cmd += self._configuration_args()
+        cmd += ['--', info_dict['url']]
+        return cmd
+
+
+class AxelFD(ExternalFD):
+    AVAILABLE_OPT = '-V'
+
+    def _make_cmd(self, tmpfilename, info_dict):
+        cmd = [self.exe, '-o', tmpfilename]
+        if info_dict.get('http_headers') is not None:
+            for key, val in info_dict['http_headers'].items():
+                cmd += ['-H', f'{key}: {val}']
+        cookie_header = self.ydl.cookiejar.get_cookie_header(info_dict['url'])
+        if cookie_header:
+            cmd += ['-H', f'Cookie: {cookie_header}', '--max-redirect=0']
+        cmd += self._configuration_args()
+        cmd += ['--', info_dict['url']]
+        return cmd
+
+
+class WgetFD(ExternalFD):
+    AVAILABLE_OPT = '--version'
+
+    def _make_cmd(self, tmpfilename, info_dict):
+        cmd = [self.exe, '-O', tmpfilename, '-nv', '--compression=auto']
+        if self.ydl.cookiejar.get_cookie_header(info_dict['url']):
+            cmd += ['--load-cookies', self._write_cookies()]
+        if info_dict.get('http_headers') is not None:
+            for key, val in info_dict['http_headers'].items():
+                cmd += ['--header', f'{key}: {val}']
+        cmd += self._option('--limit-rate', 'ratelimit')
+        retry = self._option('--tries', 'retries')
+        if len(retry) == 2:
+            if retry[1] in ('inf', 'infinite'):
+                retry[1] = '0'
+            cmd += retry
+        cmd += self._option('--bind-address', 'source_address')
+        proxy = self.params.get('proxy')
+        if proxy:
+            for var in ('http_proxy', 'https_proxy'):
+                cmd += ['--execute', f'{var}={proxy}']
+        cmd += self._valueless_option('--no-check-certificate', 'nocheckcertificate')
+        cmd += self._configuration_args()
+        cmd += ['--', info_dict['url']]
+        return cmd
+
+
+class Aria2cFD(ExternalFD):
+    AVAILABLE_OPT = '-v'
+    SUPPORTED_PROTOCOLS = ('http', 'https', 'ftp', 'ftps', 'dash_frag_urls', 'm3u8_frag_urls')
+
+    @staticmethod
+    def supports_manifest(manifest):
+        UNSUPPORTED_FEATURES = [
+            r'#EXT-X-BYTERANGE',  # playlists composed of byte ranges of media files [1]
+            # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.2.2
+        ]
+        check_results = (not re.search(feature, manifest) for feature in UNSUPPORTED_FEATURES)
+        return all(check_results)
+
+    @staticmethod
+    def _aria2c_filename(fn):
+        return fn if os.path.isabs(fn) else f'.{os.path.sep}{fn}'
+
+    def _call_downloader(self, tmpfilename, info_dict):
+        # FIXME: Disabled due to https://github.com/yt-dlp/yt-dlp/issues/5931
+        if False and 'no-external-downloader-progress' not in self.params.get('compat_opts', []):
+            info_dict['__rpc'] = {
+                'port': find_available_port() or 19190,
+                'secret': str(uuid.uuid4()),
+            }
+        return super()._call_downloader(tmpfilename, info_dict)
+
+    def _make_cmd(self, tmpfilename, info_dict):
+        cmd = [self.exe, '-c', '--no-conf',
+               '--console-log-level=warn', '--summary-interval=0', '--download-result=hide',
+               '--http-accept-gzip=true', '--file-allocation=none', '-x16', '-j16', '-s16']
+        if 'fragments' in info_dict:
+            cmd += ['--allow-overwrite=true', '--allow-piece-length-change=true']
+        else:
+            cmd += ['--min-split-size', '1M']
+
+        if self.ydl.cookiejar.get_cookie_header(info_dict['url']):
+            cmd += [f'--load-cookies={self._write_cookies()}']
+        if info_dict.get('http_headers') is not None:
+            for key, val in info_dict['http_headers'].items():
+                cmd += ['--header', f'{key}: {val}']
+        cmd += self._option('--max-overall-download-limit', 'ratelimit')
+        cmd += self._option('--interface', 'source_address')
+        cmd += self._option('--all-proxy', 'proxy')
+        cmd += self._bool_option('--check-certificate', 'nocheckcertificate', 'false', 'true', '=')
+        cmd += self._bool_option('--remote-time', 'updatetime', 'true', 'false', '=')
+        cmd += self._bool_option('--show-console-readout', 'noprogress', 'false', 'true', '=')
+        cmd += self._configuration_args()
+
+        if '__rpc' in info_dict:
+            cmd += [
+                '--enable-rpc',
+                f'--rpc-listen-port={info_dict["__rpc"]["port"]}',
+                f'--rpc-secret={info_dict["__rpc"]["secret"]}']
+
+        # aria2c strips out spaces from the beginning/end of filenames and paths.
+        # We work around this issue by adding a "./" to the beginning of the
+        # filename and relative path, and adding a "/" at the end of the path.
+        # See: https://github.com/yt-dlp/yt-dlp/issues/276
+        # https://github.com/ytdl-org/youtube-dl/issues/20312
+        # https://github.com/aria2/aria2/issues/1373
+        dn = os.path.dirname(tmpfilename)
+        if dn:
+            cmd += ['--dir', self._aria2c_filename(dn) + os.path.sep]
+        if 'fragments' not in info_dict:
+            cmd += ['--out', self._aria2c_filename(os.path.basename(tmpfilename))]
+        cmd += ['--auto-file-renaming=false']
+
+        if 'fragments' in info_dict:
+            cmd += ['--uri-selector=inorder']
+            url_list_file = f'{tmpfilename}.frag.urls'
+            url_list = []
+            for frag_index, fragment in enumerate(info_dict['fragments']):
+                fragment_filename = f'{os.path.basename(tmpfilename)}-Frag{frag_index}'
+                url_list.append('{}\n\tout={}'.format(fragment['url'], self._aria2c_filename(fragment_filename)))
+            stream, _ = self.sanitize_open(url_list_file, 'wb')
+            stream.write('\n'.join(url_list).encode())
+            stream.close()
+            cmd += ['-i', self._aria2c_filename(url_list_file)]
+        else:
+            cmd += ['--', info_dict['url']]
+        return cmd
+
+    def aria2c_rpc(self, rpc_port, rpc_secret, method, params=()):
+        # Does not actually need to be UUID, just unique
+        sanitycheck = str(uuid.uuid4())
+        d = json.dumps({
+            'jsonrpc': '2.0',
+            'id': sanitycheck,
+            'method': method,
+            'params': [f'token:{rpc_secret}', *params],
+        }).encode()
+        request = Request(
+            f'http://localhost:{rpc_port}/jsonrpc',
+            data=d, headers={
+                'Content-Type': 'application/json',
+                'Content-Length': f'{len(d)}',
+            }, proxies={'all': None})
+        with self.ydl.urlopen(request) as r:
+            resp = json.load(r)
+        assert resp.get('id') == sanitycheck, 'Something went wrong with RPC server'
+        return resp['result']
+
+    def _call_process(self, cmd, info_dict):
+        if '__rpc' not in info_dict:
+            return super()._call_process(cmd, info_dict)
+
+        send_rpc = functools.partial(self.aria2c_rpc, info_dict['__rpc']['port'], info_dict['__rpc']['secret'])
+        started = time.time()
+
+        fragmented = 'fragments' in info_dict
+        frag_count = len(info_dict['fragments']) if fragmented else 1
+        status = {
+            'filename': info_dict.get('_filename'),
+            'status': 'downloading',
+            'elapsed': 0,
+            'downloaded_bytes': 0,
+            'fragment_count': frag_count if fragmented else None,
+            'fragment_index': 0 if fragmented else None,
+        }
+        self._hook_progress(status, info_dict)
+
+        def get_stat(key, *obj, average=False):
+            val = tuple(filter(None, map(float, traverse_obj(obj, (..., ..., key))))) or [0]
+            return sum(val) / (len(val) if average else 1)
+
+        with Popen(cmd, text=True, stdout=subprocess.DEVNULL, stderr=subprocess.PIPE) as p:
+            # Add a small sleep so that RPC client can receive response,
+            # or the connection stalls infinitely
+            time.sleep(0.2)
+            retval = p.poll()
+            while retval is None:
+                # We don't use tellStatus as we won't know the GID without reading stdout
+                # Ref: https://aria2.github.io/manual/en/html/aria2c.html#aria2.tellActive
+                active = send_rpc('aria2.tellActive')
+                completed = send_rpc('aria2.tellStopped', [0, frag_count])
+
+                downloaded = get_stat('totalLength', completed) + get_stat('completedLength', active)
+                speed = get_stat('downloadSpeed', active)
+                total = frag_count * get_stat('totalLength', active, completed, average=True)
+                if total < downloaded:
+                    total = None
+
+                status.update({
+                    'downloaded_bytes': int(downloaded),
+                    'speed': speed,
+                    'total_bytes': None if fragmented else total,
+                    'total_bytes_estimate': total,
+                    'eta': (total - downloaded) / (speed or 1),
+                    'fragment_index': min(frag_count, len(completed) + 1) if fragmented else None,
+                    'elapsed': time.time() - started,
+                })
+                self._hook_progress(status, info_dict)
+
+                if not active and len(completed) >= frag_count:
+                    send_rpc('aria2.shutdown')
+                    retval = p.wait()
+                    break
+
+                time.sleep(0.1)
+                retval = p.poll()
+
+            return '', p.stderr.read(), retval
+
+
+class HttpieFD(ExternalFD):
+    AVAILABLE_OPT = '--version'
+    EXE_NAME = 'http'
+
+    def _make_cmd(self, tmpfilename, info_dict):
+        cmd = ['http', '--download', '--output', tmpfilename, info_dict['url']]
+
+        if info_dict.get('http_headers') is not None:
+            for key, val in info_dict['http_headers'].items():
+                cmd += [f'{key}:{val}']
+
+        # httpie 3.1.0+ removes the Cookie header on redirect, so this should be safe for now. [1]
+        # If we ever need cookie handling for redirects, we can export the cookiejar into a session. [2]
+        # 1: https://github.com/httpie/httpie/security/advisories/GHSA-9w4w-cpc8-h2fq
+        # 2: https://httpie.io/docs/cli/sessions
+        cookie_header = self.ydl.cookiejar.get_cookie_header(info_dict['url'])
+        if cookie_header:
+            cmd += [f'Cookie:{cookie_header}']
+        return cmd
+
+
+class FFmpegFD(ExternalFD):
+    SUPPORTED_PROTOCOLS = ('http', 'https', 'ftp', 'ftps', 'm3u8', 'm3u8_native', 'rtsp', 'rtmp', 'rtmp_ffmpeg', 'mms', 'http_dash_segments')
+    SUPPORTED_FEATURES = (Features.TO_STDOUT, Features.MULTIPLE_FORMATS)
+
+    @classmethod
+    def available(cls, path=None):
+        return FFmpegPostProcessor().available
+
+    def on_process_started(self, proc, stdin):
+        """ Override this in subclasses  """
+        pass
+
+    @classmethod
+    def can_merge_formats(cls, info_dict, params):
+        return (
+            info_dict.get('requested_formats')
+            and info_dict.get('protocol')
+            and not params.get('allow_unplayable_formats')
+            and 'no-direct-merge' not in params.get('compat_opts', [])
+            and cls.can_download(info_dict))
+
+    def _call_downloader(self, tmpfilename, info_dict):
+        ffpp = FFmpegPostProcessor(downloader=self)
+        if not ffpp.available:
+            self.report_error('m3u8 download detected but ffmpeg could not be found. Please install')
+            return False
+        ffpp.check_version()
+
+        args = [ffpp.executable, '-y']
+
+        for log_level in ('quiet', 'verbose'):
+            if self.params.get(log_level, False):
+                args += ['-loglevel', log_level]
+                break
+        if not self.params.get('verbose'):
+            args += ['-hide_banner']
+
+        args += traverse_obj(info_dict, ('downloader_options', 'ffmpeg_args', ...))
+
+        # These exists only for compatibility. Extractors should use
+        # info_dict['downloader_options']['ffmpeg_args'] instead
+        args += info_dict.get('_ffmpeg_args') or []
+        seekable = info_dict.get('_seekable')
+        if seekable is not None:
+            # setting -seekable prevents ffmpeg from guessing if the server
+            # supports seeking(by adding the header `Range: bytes=0-`), which
+            # can cause problems in some cases
+            # https://github.com/ytdl-org/youtube-dl/issues/11800#issuecomment-275037127
+            # http://trac.ffmpeg.org/ticket/6125#comment:10
+            args += ['-seekable', '1' if seekable else '0']
+
+        env = None
+        proxy = self.params.get('proxy')
+        if proxy:
+            if not re.match(r'[\da-zA-Z]+://', proxy):
+                proxy = f'http://{proxy}'
+
+            if proxy.startswith('socks'):
+                self.report_warning(
+                    f'{self.get_basename()} does not support SOCKS proxies. Downloading is likely to fail. '
+                    'Consider adding --hls-prefer-native to your command.')
+
+            # Since December 2015 ffmpeg supports -http_proxy option (see
+            # http://git.videolan.org/?p=ffmpeg.git;a=commit;h=b4eb1f29ebddd60c41a2eb39f5af701e38e0d3fd)
+            # We could switch to the following code if we are able to detect version properly
+            # args += ['-http_proxy', proxy]
+            env = os.environ.copy()
+            env['HTTP_PROXY'] = proxy
+            env['http_proxy'] = proxy
+
+        protocol = info_dict.get('protocol')
+
+        if protocol == 'rtmp':
+            player_url = info_dict.get('player_url')
+            page_url = info_dict.get('page_url')
+            app = info_dict.get('app')
+            play_path = info_dict.get('play_path')
+            tc_url = info_dict.get('tc_url')
+            flash_version = info_dict.get('flash_version')
+            live = info_dict.get('rtmp_live', False)
+            conn = info_dict.get('rtmp_conn')
+            if player_url is not None:
+                args += ['-rtmp_swfverify', player_url]
+            if page_url is not None:
+                args += ['-rtmp_pageurl', page_url]
+            if app is not None:
+                args += ['-rtmp_app', app]
+            if play_path is not None:
+                args += ['-rtmp_playpath', play_path]
+            if tc_url is not None:
+                args += ['-rtmp_tcurl', tc_url]
+            if flash_version is not None:
+                args += ['-rtmp_flashver', flash_version]
+            if live:
+                args += ['-rtmp_live', 'live']
+            if isinstance(conn, list):
+                for entry in conn:
+                    args += ['-rtmp_conn', entry]
+            elif isinstance(conn, str):
+                args += ['-rtmp_conn', conn]
+
+        start_time, end_time = info_dict.get('section_start') or 0, info_dict.get('section_end')
+
+        selected_formats = info_dict.get('requested_formats') or [info_dict]
+        for i, fmt in enumerate(selected_formats):
+            is_http = re.match(r'https?://', fmt['url'])
+            cookies = self.ydl.cookiejar.get_cookies_for_url(fmt['url']) if is_http else []
+            if cookies:
+                args.extend(['-cookies', ''.join(
+                    f'{cookie.name}={cookie.value}; path={cookie.path}; domain={cookie.domain};\r\n'
+                    for cookie in cookies)])
+            if fmt.get('http_headers') and is_http:
+                # Trailing \r\n after each HTTP header is important to prevent warning from ffmpeg:
+                # [http @ 00000000003d2fa0] No trailing CRLF found in HTTP header.
+                args.extend(['-headers', ''.join(f'{key}: {val}\r\n' for key, val in fmt['http_headers'].items())])
+
+            if start_time:
+                args += ['-ss', str(start_time)]
+            if end_time:
+                args += ['-t', str(end_time - start_time)]
+
+            url = fmt['url']
+            if self.params.get('enable_file_urls') and url.startswith('file:'):
+                # The default protocol_whitelist is 'file,crypto,data' when reading local m3u8 URLs,
+                # so only local segments can be read unless we also include 'http,https,tcp,tls'
+                args += ['-protocol_whitelist', 'file,crypto,data,http,https,tcp,tls']
+                # ffmpeg incorrectly handles 'file:' URLs by only removing the
+                # 'file:' prefix and treating the rest as if it's a normal filepath.
+                # FFmpegPostProcessor also depends on this behavior, so we need to fixup the URLs:
+                # - On Windows/Cygwin, replace 'file:///' and 'file://localhost/' with 'file:'
+                # - On *nix, replace 'file://localhost/' with 'file:/'
+                # Ref: https://github.com/yt-dlp/yt-dlp/issues/13781
+                #      https://trac.ffmpeg.org/ticket/2702
+                url = re.sub(r'^file://(?:localhost)?/', 'file:' if os.name == 'nt' else 'file:/', url)
+
+            args += [*self._configuration_args((f'_i{i + 1}', '_i')), '-i', url]
+
+        if not (start_time or end_time) or not self.params.get('force_keyframes_at_cuts'):
+            args += ['-c', 'copy']
+
+        if info_dict.get('requested_formats') or protocol == 'http_dash_segments':
+            for i, fmt in enumerate(selected_formats):
+                stream_number = fmt.get('manifest_stream_number', 0)
+                args.extend(['-map', f'{i}:{stream_number}'])
+
+        if self.params.get('test', False):
+            args += ['-fs', str(self._TEST_FILE_SIZE)]
+
+        ext = info_dict['ext']
+        if protocol in ('m3u8', 'm3u8_native'):
+            use_mpegts = (tmpfilename == '-') or self.params.get('hls_use_mpegts')
+            if use_mpegts is None:
+                use_mpegts = info_dict.get('is_live')
+            if use_mpegts:
+                args += ['-f', 'mpegts']
+            else:
+                args += ['-f', 'mp4']
+                if (ffpp.basename == 'ffmpeg' and ffpp._features.get('needs_adtstoasc')) and (not info_dict.get('acodec') or info_dict['acodec'].split('.')[0] in ('aac', 'mp4a')):
+                    args += ['-bsf:a', 'aac_adtstoasc']
+        elif protocol == 'rtmp':
+            args += ['-f', 'flv']
+        elif ext == 'mp4' and tmpfilename == '-':
+            args += ['-f', 'mpegts']
+        elif ext == 'unknown_video':
+            ext = determine_ext(remove_end(tmpfilename, '.part'))
+            if ext == 'unknown_video':
+                self.report_warning(
+                    'The video format is unknown and cannot be downloaded by ffmpeg. '
+                    'Explicitly set the extension in the filename to attempt download in that format')
+            else:
+                self.report_warning(f'The video format is unknown. Trying to download as {ext} according to the filename')
+                args += ['-f', EXT_TO_OUT_FORMATS.get(ext, ext)]
+        else:
+            args += ['-f', EXT_TO_OUT_FORMATS.get(ext, ext)]
+
+        args += traverse_obj(info_dict, ('downloader_options', 'ffmpeg_args_out', ...))
+
+        args += self._configuration_args(('_o1', '_o', ''))
+
+        args = [encodeArgument(opt) for opt in args]
+        args.append(ffpp._ffmpeg_filename_argument(tmpfilename))
+        self._debug_cmd(args)
+
+        piped = any(fmt['url'] in ('-', 'pipe:') for fmt in selected_formats)
+        with Popen(args, stdin=subprocess.PIPE, env=env) as proc:
+            if piped:
+                self.on_process_started(proc, proc.stdin)
+            try:
+                retval = proc.wait()
+            except BaseException as e:
+                # subprocces.run would send the SIGKILL signal to ffmpeg and the
+                # mp4 file couldn't be played, but if we ask ffmpeg to quit it
+                # produces a file that is playable (this is mostly useful for live
+                # streams). Note that Windows is not affected and produces playable
+                # files (see https://github.com/ytdl-org/youtube-dl/issues/8300).
+                if isinstance(e, KeyboardInterrupt) and sys.platform != 'win32' and not piped:
+                    proc.communicate_or_kill(b'q')
+                else:
+                    proc.kill(timeout=None)
+                raise
+            return retval
+
+
+_BY_NAME = {
+    klass.get_basename(): klass
+    for name, klass in globals().items()
+    if name.endswith('FD') and name not in ('ExternalFD', 'FragmentFD')
+}
+
+
+def list_external_downloaders():
+    return sorted(_BY_NAME.keys())
+
+
+def get_external_downloader(external_downloader):
+    """ Given the name of the executable, see whether we support the given downloader """
+    bn = os.path.splitext(os.path.basename(external_downloader))[0]
+    return _BY_NAME.get(bn) or next((
+        klass for klass in _BY_NAME.values() if klass.EXE_NAME in bn
+    ), None)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/f4m.py

@@ -0,0 +1,427 @@
+import base64
+import io
+import itertools
+import struct
+import time
+import urllib.parse
+
+from .fragment import FragmentFD
+from ..compat import compat_etree_fromstring
+from ..networking.exceptions import HTTPError
+from ..utils import fix_xml_ampersands, xpath_text
+
+
+class DataTruncatedError(Exception):
+    pass
+
+
+class FlvReader(io.BytesIO):
+    """
+    Reader for Flv files
+    The file format is documented in https://www.adobe.com/devnet/f4v.html
+    """
+
+    def read_bytes(self, n):
+        data = self.read(n)
+        if len(data) < n:
+            raise DataTruncatedError(
+                'FlvReader error: need %d bytes while only %d bytes got' % (
+                    n, len(data)))
+        return data
+
+    # Utility functions for reading numbers and strings
+    def read_unsigned_long_long(self):
+        return struct.unpack('!Q', self.read_bytes(8))[0]
+
+    def read_unsigned_int(self):
+        return struct.unpack('!I', self.read_bytes(4))[0]
+
+    def read_unsigned_char(self):
+        return struct.unpack('!B', self.read_bytes(1))[0]
+
+    def read_string(self):
+        res = b''
+        while True:
+            char = self.read_bytes(1)
+            if char == b'\x00':
+                break
+            res += char
+        return res
+
+    def read_box_info(self):
+        """
+        Read a box and return the info as a tuple: (box_size, box_type, box_data)
+        """
+        real_size = size = self.read_unsigned_int()
+        box_type = self.read_bytes(4)
+        header_end = 8
+        if size == 1:
+            real_size = self.read_unsigned_long_long()
+            header_end = 16
+        return real_size, box_type, self.read_bytes(real_size - header_end)
+
+    def read_asrt(self):
+        # version
+        self.read_unsigned_char()
+        # flags
+        self.read_bytes(3)
+        quality_entry_count = self.read_unsigned_char()
+        # QualityEntryCount
+        for _ in range(quality_entry_count):
+            self.read_string()
+
+        segment_run_count = self.read_unsigned_int()
+        segments = []
+        for _ in range(segment_run_count):
+            first_segment = self.read_unsigned_int()
+            fragments_per_segment = self.read_unsigned_int()
+            segments.append((first_segment, fragments_per_segment))
+
+        return {
+            'segment_run': segments,
+        }
+
+    def read_afrt(self):
+        # version
+        self.read_unsigned_char()
+        # flags
+        self.read_bytes(3)
+        # time scale
+        self.read_unsigned_int()
+
+        quality_entry_count = self.read_unsigned_char()
+        # QualitySegmentUrlModifiers
+        for _ in range(quality_entry_count):
+            self.read_string()
+
+        fragments_count = self.read_unsigned_int()
+        fragments = []
+        for _ in range(fragments_count):
+            first = self.read_unsigned_int()
+            first_ts = self.read_unsigned_long_long()
+            duration = self.read_unsigned_int()
+            if duration == 0:
+                discontinuity_indicator = self.read_unsigned_char()
+            else:
+                discontinuity_indicator = None
+            fragments.append({
+                'first': first,
+                'ts': first_ts,
+                'duration': duration,
+                'discontinuity_indicator': discontinuity_indicator,
+            })
+
+        return {
+            'fragments': fragments,
+        }
+
+    def read_abst(self):
+        # version
+        self.read_unsigned_char()
+        # flags
+        self.read_bytes(3)
+
+        self.read_unsigned_int()  # BootstrapinfoVersion
+        # Profile,Live,Update,Reserved
+        flags = self.read_unsigned_char()
+        live = flags & 0x20 != 0
+        # time scale
+        self.read_unsigned_int()
+        # CurrentMediaTime
+        self.read_unsigned_long_long()
+        # SmpteTimeCodeOffset
+        self.read_unsigned_long_long()
+
+        self.read_string()  # MovieIdentifier
+        server_count = self.read_unsigned_char()
+        # ServerEntryTable
+        for _ in range(server_count):
+            self.read_string()
+        quality_count = self.read_unsigned_char()
+        # QualityEntryTable
+        for _ in range(quality_count):
+            self.read_string()
+        # DrmData
+        self.read_string()
+        # MetaData
+        self.read_string()
+
+        segments_count = self.read_unsigned_char()
+        segments = []
+        for _ in range(segments_count):
+            _box_size, box_type, box_data = self.read_box_info()
+            assert box_type == b'asrt'
+            segment = FlvReader(box_data).read_asrt()
+            segments.append(segment)
+        fragments_run_count = self.read_unsigned_char()
+        fragments = []
+        for _ in range(fragments_run_count):
+            _box_size, box_type, box_data = self.read_box_info()
+            assert box_type == b'afrt'
+            fragments.append(FlvReader(box_data).read_afrt())
+
+        return {
+            'segments': segments,
+            'fragments': fragments,
+            'live': live,
+        }
+
+    def read_bootstrap_info(self):
+        _, box_type, box_data = self.read_box_info()
+        assert box_type == b'abst'
+        return FlvReader(box_data).read_abst()
+
+
+def read_bootstrap_info(bootstrap_bytes):
+    return FlvReader(bootstrap_bytes).read_bootstrap_info()
+
+
+def build_fragments_list(boot_info):
+    """ Return a list of (segment, fragment) for each fragment in the video """
+    res = []
+    segment_run_table = boot_info['segments'][0]
+    fragment_run_entry_table = boot_info['fragments'][0]['fragments']
+    first_frag_number = fragment_run_entry_table[0]['first']
+    fragments_counter = itertools.count(first_frag_number)
+    for segment, fragments_count in segment_run_table['segment_run']:
+        # In some live HDS streams (e.g. Rai), `fragments_count` is
+        # abnormal and causing out-of-memory errors. It's OK to change the
+        # number of fragments for live streams as they are updated periodically
+        if fragments_count == 4294967295 and boot_info['live']:
+            fragments_count = 2
+        for _ in range(fragments_count):
+            res.append((segment, next(fragments_counter)))
+
+    if boot_info['live']:
+        res = res[-2:]
+
+    return res
+
+
+def write_unsigned_int(stream, val):
+    stream.write(struct.pack('!I', val))
+
+
+def write_unsigned_int_24(stream, val):
+    stream.write(struct.pack('!I', val)[1:])
+
+
+def write_flv_header(stream):
+    """Writes the FLV header to stream"""
+    # FLV header
+    stream.write(b'FLV\x01')
+    stream.write(b'\x05')
+    stream.write(b'\x00\x00\x00\x09')
+    stream.write(b'\x00\x00\x00\x00')
+
+
+def write_metadata_tag(stream, metadata):
+    """Writes optional metadata tag to stream"""
+    SCRIPT_TAG = b'\x12'
+    FLV_TAG_HEADER_LEN = 11
+
+    if metadata:
+        stream.write(SCRIPT_TAG)
+        write_unsigned_int_24(stream, len(metadata))
+        stream.write(b'\x00\x00\x00\x00\x00\x00\x00')
+        stream.write(metadata)
+        write_unsigned_int(stream, FLV_TAG_HEADER_LEN + len(metadata))
+
+
+def remove_encrypted_media(media):
+    return list(filter(lambda e: 'drmAdditionalHeaderId' not in e.attrib
+                                 and 'drmAdditionalHeaderSetId' not in e.attrib,
+                       media))
+
+
+def _add_ns(prop, ver=1):
+    return '{http://ns.adobe.com/f4m/%d.0}%s' % (ver, prop)
+
+
+def get_base_url(manifest):
+    base_url = xpath_text(
+        manifest, [_add_ns('baseURL'), _add_ns('baseURL', 2)],
+        'base URL', default=None)
+    if base_url:
+        base_url = base_url.strip()
+    return base_url
+
+
+class F4mFD(FragmentFD):
+    """
+    A downloader for f4m manifests or AdobeHDS.
+    """
+
+    def _get_unencrypted_media(self, doc):
+        media = doc.findall(_add_ns('media'))
+        if not media:
+            self.report_error('No media found')
+        if not self.params.get('allow_unplayable_formats'):
+            for e in (doc.findall(_add_ns('drmAdditionalHeader'))
+                      + doc.findall(_add_ns('drmAdditionalHeaderSet'))):
+                # If id attribute is missing it's valid for all media nodes
+                # without drmAdditionalHeaderId or drmAdditionalHeaderSetId attribute
+                if 'id' not in e.attrib:
+                    self.report_error('Missing ID in f4m DRM')
+            media = remove_encrypted_media(media)
+        if not media:
+            self.report_error('Unsupported DRM')
+        return media
+
+    def _get_bootstrap_from_url(self, bootstrap_url):
+        bootstrap = self.ydl.urlopen(bootstrap_url).read()
+        return read_bootstrap_info(bootstrap)
+
+    def _update_live_fragments(self, bootstrap_url, latest_fragment):
+        fragments_list = []
+        retries = 30
+        while (not fragments_list) and (retries > 0):
+            boot_info = self._get_bootstrap_from_url(bootstrap_url)
+            fragments_list = build_fragments_list(boot_info)
+            fragments_list = [f for f in fragments_list if f[1] > latest_fragment]
+            if not fragments_list:
+                # Retry after a while
+                time.sleep(5.0)
+                retries -= 1
+
+        if not fragments_list:
+            self.report_error('Failed to update fragments')
+
+        return fragments_list
+
+    def _parse_bootstrap_node(self, node, base_url):
+        # Sometimes non empty inline bootstrap info can be specified along
+        # with bootstrap url attribute (e.g. dummy inline bootstrap info
+        # contains whitespace characters in [1]). We will prefer bootstrap
+        # url over inline bootstrap info when present.
+        # 1. http://live-1-1.rutube.ru/stream/1024/HDS/SD/C2NKsS85HQNckgn5HdEmOQ/1454167650/S-s604419906/move/four/dirs/upper/1024-576p.f4m
+        bootstrap_url = node.get('url')
+        if bootstrap_url:
+            bootstrap_url = urllib.parse.urljoin(
+                base_url, bootstrap_url)
+            boot_info = self._get_bootstrap_from_url(bootstrap_url)
+        else:
+            bootstrap_url = None
+            bootstrap = base64.b64decode(node.text)
+            boot_info = read_bootstrap_info(bootstrap)
+        return boot_info, bootstrap_url
+
+    def real_download(self, filename, info_dict):
+        man_url = info_dict['url']
+        requested_bitrate = info_dict.get('tbr')
+        self.to_screen(f'[{self.FD_NAME}] Downloading f4m manifest')
+
+        urlh = self.ydl.urlopen(self._prepare_url(info_dict, man_url))
+        man_url = urlh.url
+        # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
+        # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244
+        # and https://github.com/ytdl-org/youtube-dl/issues/7823)
+        manifest = fix_xml_ampersands(urlh.read().decode('utf-8', 'ignore')).strip()
+
+        doc = compat_etree_fromstring(manifest)
+        formats = [(int(f.attrib.get('bitrate', -1)), f)
+                   for f in self._get_unencrypted_media(doc)]
+        if requested_bitrate is None or len(formats) == 1:
+            # get the best format
+            formats = sorted(formats, key=lambda f: f[0])
+            _, media = formats[-1]
+        else:
+            _, media = next(filter(
+                lambda f: int(f[0]) == requested_bitrate, formats))
+
+        # Prefer baseURL for relative URLs as per 11.2 of F4M 3.0 spec.
+        man_base_url = get_base_url(doc) or man_url
+
+        base_url = urllib.parse.urljoin(man_base_url, media.attrib['url'])
+        bootstrap_node = doc.find(_add_ns('bootstrapInfo'))
+        boot_info, bootstrap_url = self._parse_bootstrap_node(
+            bootstrap_node, man_base_url)
+        live = boot_info['live']
+        metadata_node = media.find(_add_ns('metadata'))
+        if metadata_node is not None:
+            metadata = base64.b64decode(metadata_node.text)
+        else:
+            metadata = None
+
+        fragments_list = build_fragments_list(boot_info)
+        test = self.params.get('test', False)
+        if test:
+            # We only download the first fragment
+            fragments_list = fragments_list[:1]
+        total_frags = len(fragments_list)
+        # For some akamai manifests we'll need to add a query to the fragment url
+        akamai_pv = xpath_text(doc, _add_ns('pv-2.0'))
+
+        ctx = {
+            'filename': filename,
+            'total_frags': total_frags,
+            'live': bool(live),
+        }
+
+        self._prepare_frag_download(ctx)
+
+        dest_stream = ctx['dest_stream']
+
+        if ctx['complete_frags_downloaded_bytes'] == 0:
+            write_flv_header(dest_stream)
+            if not live:
+                write_metadata_tag(dest_stream, metadata)
+
+        base_url_parsed = urllib.parse.urlparse(base_url)
+
+        self._start_frag_download(ctx, info_dict)
+
+        frag_index = 0
+        while fragments_list:
+            seg_i, frag_i = fragments_list.pop(0)
+            frag_index += 1
+            if frag_index <= ctx['fragment_index']:
+                continue
+            name = 'Seg%d-Frag%d' % (seg_i, frag_i)
+            query = []
+            if base_url_parsed.query:
+                query.append(base_url_parsed.query)
+            if akamai_pv:
+                query.append(akamai_pv.strip(';'))
+            if info_dict.get('extra_param_to_segment_url'):
+                query.append(info_dict['extra_param_to_segment_url'])
+            url_parsed = base_url_parsed._replace(path=base_url_parsed.path + name, query='&'.join(query))
+            try:
+                success = self._download_fragment(ctx, url_parsed.geturl(), info_dict)
+                if not success:
+                    return False
+                down_data = self._read_fragment(ctx)
+                reader = FlvReader(down_data)
+                while True:
+                    try:
+                        _, box_type, box_data = reader.read_box_info()
+                    except DataTruncatedError:
+                        if test:
+                            # In tests, segments may be truncated, and thus
+                            # FlvReader may not be able to parse the whole
+                            # chunk. If so, write the segment as is
+                            # See https://github.com/ytdl-org/youtube-dl/issues/9214
+                            dest_stream.write(down_data)
+                            break
+                        raise
+                    if box_type == b'mdat':
+                        self._append_fragment(ctx, box_data)
+                        break
+            except HTTPError as err:
+                if live and (err.status == 404 or err.status == 410):
+                    # We didn't keep up with the live window. Continue
+                    # with the next available fragment.
+                    msg = 'Fragment %d unavailable' % frag_i
+                    self.report_warning(msg)
+                    fragments_list = []
+                else:
+                    raise
+
+            if not fragments_list and not test and live and bootstrap_url:
+                fragments_list = self._update_live_fragments(bootstrap_url, frag_i)
+                total_frags += len(fragments_list)
+                if fragments_list and (fragments_list[0][1] > frag_i + 1):
+                    msg = 'Missed %d fragments' % (fragments_list[0][1] - (frag_i + 1))
+                    self.report_warning(msg)
+
+        return self._finish_frag_download(ctx, info_dict)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/fc2.py

@@ -0,0 +1,46 @@
+import threading
+
+from .common import FileDownloader
+from .external import FFmpegFD
+
+
+class FC2LiveFD(FileDownloader):
+    """
+    Downloads FC2 live without being stopped. <br>
+    Note, this is not a part of public API, and will be removed without notice.
+    DO NOT USE
+    """
+
+    def real_download(self, filename, info_dict):
+        ws = info_dict['ws']
+
+        heartbeat_lock = threading.Lock()
+        heartbeat_state = [None, 1]
+
+        def heartbeat():
+            if heartbeat_state[1] < 0:
+                return
+
+            try:
+                heartbeat_state[1] += 1
+                ws.send('{"name":"heartbeat","arguments":{},"id":%d}' % heartbeat_state[1])
+            except Exception:
+                self.to_screen('[fc2:live] Heartbeat failed')
+
+            with heartbeat_lock:
+                heartbeat_state[0] = threading.Timer(30, heartbeat)
+                heartbeat_state[0]._daemonic = True
+                heartbeat_state[0].start()
+
+        heartbeat()
+
+        new_info_dict = info_dict.copy()
+        new_info_dict.update({
+            'ws': None,
+            'protocol': 'live_ffmpeg',
+        })
+        try:
+            return FFmpegFD(self.ydl, self.params or {}).download(filename, new_info_dict)
+        finally:
+            # stop heartbeating
+            heartbeat_state[1] = -1

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/fragment.py

@@ -0,0 +1,526 @@
+import concurrent.futures
+import contextlib
+import json
+import math
+import os
+import struct
+import time
+
+from .common import FileDownloader
+from .http import HttpFD
+from ..aes import aes_cbc_decrypt_bytes, unpad_pkcs7
+from ..networking import Request
+from ..networking.exceptions import HTTPError, IncompleteRead
+from ..utils import DownloadError, RetryManager, traverse_obj
+from ..utils.networking import HTTPHeaderDict
+from ..utils.progress import ProgressCalculator
+
+
+class HttpQuietDownloader(HttpFD):
+    def to_screen(self, *args, **kargs):
+        pass
+
+    to_console_title = to_screen
+
+
+class FragmentFD(FileDownloader):
+    """
+    A base file downloader class for fragmented media (e.g. f4m/m3u8 manifests).
+
+    Available options:
+
+    fragment_retries:   Number of times to retry a fragment for HTTP error
+                        (DASH and hlsnative only). Default is 0 for API, but 10 for CLI
+    skip_unavailable_fragments:
+                        Skip unavailable fragments (DASH and hlsnative only)
+    keep_fragments:     Keep downloaded fragments on disk after downloading is
+                        finished
+    concurrent_fragment_downloads:  The number of threads to use for native hls and dash downloads
+    _no_ytdl_file:      Don't use .ytdl file
+
+    For each incomplete fragment download yt-dlp keeps on disk a special
+    bookkeeping file with download state and metadata (in future such files will
+    be used for any incomplete download handled by yt-dlp). This file is
+    used to properly handle resuming, check download file consistency and detect
+    potential errors. The file has a .ytdl extension and represents a standard
+    JSON file of the following format:
+
+    extractor:
+        Dictionary of extractor related data. TBD.
+
+    downloader:
+        Dictionary of downloader related data. May contain following data:
+            current_fragment:
+                Dictionary with current (being downloaded) fragment data:
+                index:  0-based index of current fragment among all fragments
+            fragment_count:
+                Total count of fragments
+
+    This feature is experimental and file format may change in future.
+    """
+
+    def report_retry_fragment(self, err, frag_index, count, retries):
+        self.deprecation_warning('yt_dlp.downloader.FragmentFD.report_retry_fragment is deprecated. '
+                                 'Use yt_dlp.downloader.FileDownloader.report_retry instead')
+        return self.report_retry(err, count, retries, frag_index)
+
+    def report_skip_fragment(self, frag_index, err=None):
+        err = f' {err};' if err else ''
+        self.to_screen(f'[download]{err} Skipping fragment {frag_index:d} ...')
+
+    def _prepare_url(self, info_dict, url):
+        headers = info_dict.get('http_headers')
+        return Request(url, None, headers) if headers else url
+
+    def _prepare_and_start_frag_download(self, ctx, info_dict):
+        self._prepare_frag_download(ctx)
+        self._start_frag_download(ctx, info_dict)
+
+    def __do_ytdl_file(self, ctx):
+        return ctx['live'] is not True and ctx['tmpfilename'] != '-' and not self.params.get('_no_ytdl_file')
+
+    def _read_ytdl_file(self, ctx):
+        assert 'ytdl_corrupt' not in ctx
+        stream, _ = self.sanitize_open(self.ytdl_filename(ctx['filename']), 'r')
+        try:
+            ytdl_data = json.loads(stream.read())
+            ctx['fragment_index'] = ytdl_data['downloader']['current_fragment']['index']
+            if 'extra_state' in ytdl_data['downloader']:
+                ctx['extra_state'] = ytdl_data['downloader']['extra_state']
+        except Exception:
+            ctx['ytdl_corrupt'] = True
+        finally:
+            stream.close()
+
+    def _write_ytdl_file(self, ctx):
+        frag_index_stream, _ = self.sanitize_open(self.ytdl_filename(ctx['filename']), 'w')
+        try:
+            downloader = {
+                'current_fragment': {
+                    'index': ctx['fragment_index'],
+                },
+            }
+            if 'extra_state' in ctx:
+                downloader['extra_state'] = ctx['extra_state']
+            if ctx.get('fragment_count') is not None:
+                downloader['fragment_count'] = ctx['fragment_count']
+            frag_index_stream.write(json.dumps({'downloader': downloader}))
+        finally:
+            frag_index_stream.close()
+
+    def _download_fragment(self, ctx, frag_url, info_dict, headers=None, request_data=None):
+        fragment_filename = '%s-Frag%d' % (ctx['tmpfilename'], ctx['fragment_index'])
+        fragment_info_dict = {
+            'url': frag_url,
+            'http_headers': headers or info_dict.get('http_headers'),
+            'request_data': request_data,
+            'ctx_id': ctx.get('ctx_id'),
+        }
+        frag_resume_len = 0
+        if ctx['dl'].params.get('continuedl', True):
+            frag_resume_len = self.filesize_or_none(self.temp_name(fragment_filename))
+        fragment_info_dict['frag_resume_len'] = ctx['frag_resume_len'] = frag_resume_len
+
+        success, _ = ctx['dl'].download(fragment_filename, fragment_info_dict)
+        if not success:
+            return False
+        if fragment_info_dict.get('filetime'):
+            ctx['fragment_filetime'] = fragment_info_dict.get('filetime')
+        ctx['fragment_filename_sanitized'] = fragment_filename
+        return True
+
+    def _read_fragment(self, ctx):
+        if not ctx.get('fragment_filename_sanitized'):
+            return None
+        try:
+            down, frag_sanitized = self.sanitize_open(ctx['fragment_filename_sanitized'], 'rb')
+        except FileNotFoundError:
+            if ctx.get('live'):
+                return None
+            raise
+        ctx['fragment_filename_sanitized'] = frag_sanitized
+        frag_content = down.read()
+        down.close()
+        return frag_content
+
+    def _append_fragment(self, ctx, frag_content):
+        try:
+            ctx['dest_stream'].write(frag_content)
+            ctx['dest_stream'].flush()
+        finally:
+            if self.__do_ytdl_file(ctx):
+                self._write_ytdl_file(ctx)
+            if not self.params.get('keep_fragments', False):
+                self.try_remove(ctx['fragment_filename_sanitized'])
+            del ctx['fragment_filename_sanitized']
+
+    def _prepare_frag_download(self, ctx):
+        if not ctx.setdefault('live', False):
+            total_frags_str = '%d' % ctx['total_frags']
+            ad_frags = ctx.get('ad_frags', 0)
+            if ad_frags:
+                total_frags_str += ' (not including %d ad)' % ad_frags
+        else:
+            total_frags_str = 'unknown (live)'
+        self.to_screen(f'[{self.FD_NAME}] Total fragments: {total_frags_str}')
+        self.report_destination(ctx['filename'])
+        dl = HttpQuietDownloader(self.ydl, {
+            **self.params,
+            'noprogress': True,
+            'test': False,
+            'sleep_interval': 0,
+            'max_sleep_interval': 0,
+            'sleep_interval_subtitles': 0,
+        })
+        tmpfilename = self.temp_name(ctx['filename'])
+        open_mode = 'wb'
+
+        # Establish possible resume length
+        resume_len = self.filesize_or_none(tmpfilename)
+        if resume_len > 0:
+            open_mode = 'ab'
+
+        # Should be initialized before ytdl file check
+        ctx.update({
+            'tmpfilename': tmpfilename,
+            'fragment_index': 0,
+        })
+
+        if self.__do_ytdl_file(ctx):
+            ytdl_file_exists = os.path.isfile(self.ytdl_filename(ctx['filename']))
+            continuedl = self.params.get('continuedl', True)
+            if continuedl and ytdl_file_exists:
+                self._read_ytdl_file(ctx)
+                is_corrupt = ctx.get('ytdl_corrupt') is True
+                is_inconsistent = ctx['fragment_index'] > 0 and resume_len == 0
+                if is_corrupt or is_inconsistent:
+                    message = (
+                        '.ytdl file is corrupt' if is_corrupt else
+                        'Inconsistent state of incomplete fragment download')
+                    self.report_warning(
+                        f'{message}. Restarting from the beginning ...')
+                    ctx['fragment_index'] = resume_len = 0
+                    if 'ytdl_corrupt' in ctx:
+                        del ctx['ytdl_corrupt']
+                    self._write_ytdl_file(ctx)
+
+            else:
+                if not continuedl:
+                    if ytdl_file_exists:
+                        self._read_ytdl_file(ctx)
+                    ctx['fragment_index'] = resume_len = 0
+                self._write_ytdl_file(ctx)
+                assert ctx['fragment_index'] == 0
+
+        dest_stream, tmpfilename = self.sanitize_open(tmpfilename, open_mode)
+
+        ctx.update({
+            'dl': dl,
+            'dest_stream': dest_stream,
+            'tmpfilename': tmpfilename,
+            # Total complete fragments downloaded so far in bytes
+            'complete_frags_downloaded_bytes': resume_len,
+        })
+
+    def _start_frag_download(self, ctx, info_dict):
+        resume_len = ctx['complete_frags_downloaded_bytes']
+        total_frags = ctx['total_frags']
+        ctx_id = ctx.get('ctx_id')
+        # Stores the download progress, updated by the progress hook
+        state = {
+            'status': 'downloading',
+            'downloaded_bytes': resume_len,
+            'fragment_index': ctx['fragment_index'],
+            'fragment_count': total_frags,
+            'filename': ctx['filename'],
+            'tmpfilename': ctx['tmpfilename'],
+        }
+
+        ctx['started'] = time.time()
+        progress = ProgressCalculator(resume_len)
+
+        def frag_progress_hook(s):
+            if s['status'] not in ('downloading', 'finished'):
+                return
+
+            if not total_frags and ctx.get('fragment_count'):
+                state['fragment_count'] = ctx['fragment_count']
+
+            if ctx_id is not None and s.get('ctx_id') != ctx_id:
+                return
+
+            state['max_progress'] = ctx.get('max_progress')
+            state['progress_idx'] = ctx.get('progress_idx')
+
+            state['elapsed'] = progress.elapsed
+            frag_total_bytes = s.get('total_bytes') or 0
+            s['fragment_info_dict'] = s.pop('info_dict', {})
+
+            # XXX: Fragment resume is not accounted for here
+            if not ctx['live']:
+                estimated_size = (
+                    (ctx['complete_frags_downloaded_bytes'] + frag_total_bytes)
+                    / (state['fragment_index'] + 1) * total_frags)
+                progress.total = estimated_size
+                progress.update(s.get('downloaded_bytes'))
+                state['total_bytes_estimate'] = progress.total
+            else:
+                progress.update(s.get('downloaded_bytes'))
+
+            if s['status'] == 'finished':
+                state['fragment_index'] += 1
+                ctx['fragment_index'] = state['fragment_index']
+                progress.thread_reset()
+
+            state['downloaded_bytes'] = ctx['complete_frags_downloaded_bytes'] = progress.downloaded
+            state['speed'] = ctx['speed'] = progress.speed.smooth
+            state['eta'] = progress.eta.smooth
+
+            self._hook_progress(state, info_dict)
+
+        ctx['dl'].add_progress_hook(frag_progress_hook)
+
+        return ctx['started']
+
+    def _finish_frag_download(self, ctx, info_dict):
+        ctx['dest_stream'].close()
+        if self.__do_ytdl_file(ctx):
+            self.try_remove(self.ytdl_filename(ctx['filename']))
+        elapsed = time.time() - ctx['started']
+
+        to_file = ctx['tmpfilename'] != '-'
+        if to_file:
+            downloaded_bytes = self.filesize_or_none(ctx['tmpfilename'])
+        else:
+            downloaded_bytes = ctx['complete_frags_downloaded_bytes']
+
+        if not downloaded_bytes:
+            if to_file:
+                self.try_remove(ctx['tmpfilename'])
+            self.report_error('The downloaded file is empty')
+            return False
+        elif to_file:
+            self.try_rename(ctx['tmpfilename'], ctx['filename'])
+            filetime = ctx.get('fragment_filetime')
+            if self.params.get('updatetime') and filetime:
+                with contextlib.suppress(Exception):
+                    os.utime(ctx['filename'], (time.time(), filetime))
+
+        self._hook_progress({
+            'downloaded_bytes': downloaded_bytes,
+            'total_bytes': downloaded_bytes,
+            'filename': ctx['filename'],
+            'status': 'finished',
+            'elapsed': elapsed,
+            'ctx_id': ctx.get('ctx_id'),
+            'max_progress': ctx.get('max_progress'),
+            'progress_idx': ctx.get('progress_idx'),
+        }, info_dict)
+        return True
+
+    def _prepare_external_frag_download(self, ctx):
+        if 'live' not in ctx:
+            ctx['live'] = False
+        if not ctx['live']:
+            total_frags_str = '%d' % ctx['total_frags']
+            ad_frags = ctx.get('ad_frags', 0)
+            if ad_frags:
+                total_frags_str += ' (not including %d ad)' % ad_frags
+        else:
+            total_frags_str = 'unknown (live)'
+        self.to_screen(f'[{self.FD_NAME}] Total fragments: {total_frags_str}')
+
+        tmpfilename = self.temp_name(ctx['filename'])
+
+        # Should be initialized before ytdl file check
+        ctx.update({
+            'tmpfilename': tmpfilename,
+            'fragment_index': 0,
+        })
+
+    def decrypter(self, info_dict):
+        _key_cache = {}
+
+        def _get_key(url):
+            if url not in _key_cache:
+                _key_cache[url] = self.ydl.urlopen(self._prepare_url(info_dict, url)).read()
+            return _key_cache[url]
+
+        def decrypt_fragment(fragment, frag_content):
+            if frag_content is None:
+                return
+            decrypt_info = fragment.get('decrypt_info')
+            if not decrypt_info or decrypt_info['METHOD'] != 'AES-128':
+                return frag_content
+            iv = decrypt_info.get('IV') or struct.pack('>8xq', fragment['media_sequence'])
+            decrypt_info['KEY'] = (decrypt_info.get('KEY')
+                                   or _get_key(traverse_obj(info_dict, ('hls_aes', 'uri')) or decrypt_info['URI']))
+            # Don't decrypt the content in tests since the data is explicitly truncated and it's not to a valid block
+            # size (see https://github.com/ytdl-org/youtube-dl/pull/27660). Tests only care that the correct data downloaded,
+            # not what it decrypts to.
+            if self.params.get('test', False):
+                return frag_content
+            return unpad_pkcs7(aes_cbc_decrypt_bytes(frag_content, decrypt_info['KEY'], iv))
+
+        return decrypt_fragment
+
+    def download_and_append_fragments_multiple(self, *args, **kwargs):
+        """
+        @params (ctx1, fragments1, info_dict1), (ctx2, fragments2, info_dict2), ...
+                all args must be either tuple or list
+        """
+        interrupt_trigger = [True]
+        max_progress = len(args)
+        if max_progress == 1:
+            return self.download_and_append_fragments(*args[0], **kwargs)
+        max_workers = self.params.get('concurrent_fragment_downloads', 1)
+        if max_progress > 1:
+            self._prepare_multiline_status(max_progress)
+        is_live = any(traverse_obj(args, (..., 2, 'is_live')))
+
+        def thread_func(idx, ctx, fragments, info_dict, tpe):
+            ctx['max_progress'] = max_progress
+            ctx['progress_idx'] = idx
+            return self.download_and_append_fragments(
+                ctx, fragments, info_dict, **kwargs, tpe=tpe, interrupt_trigger=interrupt_trigger)
+
+        class FTPE(concurrent.futures.ThreadPoolExecutor):
+            # has to stop this or it's going to wait on the worker thread itself
+            def __exit__(self, exc_type, exc_val, exc_tb):
+                pass
+
+        if os.name == 'nt':
+            def future_result(future):
+                while True:
+                    try:
+                        return future.result(0.1)
+                    except KeyboardInterrupt:
+                        raise
+                    except concurrent.futures.TimeoutError:
+                        continue
+        else:
+            def future_result(future):
+                return future.result()
+
+        def interrupt_trigger_iter(fg):
+            for f in fg:
+                if not interrupt_trigger[0]:
+                    break
+                yield f
+
+        spins = []
+        for idx, (ctx, fragments, info_dict) in enumerate(args):
+            tpe = FTPE(math.ceil(max_workers / max_progress))
+            job = tpe.submit(thread_func, idx, ctx, interrupt_trigger_iter(fragments), info_dict, tpe)
+            spins.append((tpe, job))
+
+        result = True
+        for tpe, job in spins:
+            try:
+                result = result and future_result(job)
+            except KeyboardInterrupt:
+                interrupt_trigger[0] = False
+            finally:
+                tpe.shutdown(wait=True)
+        if not interrupt_trigger[0] and not is_live:
+            raise KeyboardInterrupt
+        # we expect the user wants to stop and DO WANT the preceding postprocessors to run;
+        # so returning a intermediate result here instead of KeyboardInterrupt on live
+        return result
+
+    def download_and_append_fragments(
+            self, ctx, fragments, info_dict, *, is_fatal=(lambda idx: False),
+            pack_func=(lambda content, idx: content), finish_func=None,
+            tpe=None, interrupt_trigger=(True, )):
+
+        if not self.params.get('skip_unavailable_fragments', True):
+            is_fatal = lambda _: True
+
+        def download_fragment(fragment, ctx):
+            if not interrupt_trigger[0]:
+                return
+
+            frag_index = ctx['fragment_index'] = fragment['frag_index']
+            ctx['last_error'] = None
+            headers = HTTPHeaderDict(info_dict.get('http_headers'))
+            byte_range = fragment.get('byte_range')
+            if byte_range:
+                headers['Range'] = 'bytes=%d-%d' % (byte_range['start'], byte_range['end'] - 1)
+
+            # Never skip the first fragment
+            fatal = is_fatal(fragment.get('index') or (frag_index - 1))
+
+            def error_callback(err, count, retries):
+                if fatal and count > retries:
+                    ctx['dest_stream'].close()
+                self.report_retry(err, count, retries, frag_index, fatal)
+                ctx['last_error'] = err
+
+            for retry in RetryManager(self.params.get('fragment_retries'), error_callback):
+                try:
+                    ctx['fragment_count'] = fragment.get('fragment_count')
+                    if not self._download_fragment(
+                            ctx, fragment['url'], info_dict, headers, info_dict.get('request_data')):
+                        return
+                except (HTTPError, IncompleteRead) as err:
+                    retry.error = err
+                    continue
+                except DownloadError:  # has own retry settings
+                    if fatal:
+                        raise
+
+        def append_fragment(frag_content, frag_index, ctx):
+            if frag_content:
+                self._append_fragment(ctx, pack_func(frag_content, frag_index))
+            elif not is_fatal(frag_index - 1):
+                self.report_skip_fragment(frag_index, 'fragment not found')
+            else:
+                ctx['dest_stream'].close()
+                self.report_error(f'fragment {frag_index} not found, unable to continue')
+                return False
+            return True
+
+        decrypt_fragment = self.decrypter(info_dict)
+
+        max_workers = math.ceil(
+            self.params.get('concurrent_fragment_downloads', 1) / ctx.get('max_progress', 1))
+        if max_workers > 1:
+            def _download_fragment(fragment):
+                ctx_copy = ctx.copy()
+                download_fragment(fragment, ctx_copy)
+                return fragment, fragment['frag_index'], ctx_copy.get('fragment_filename_sanitized')
+
+            with tpe or concurrent.futures.ThreadPoolExecutor(max_workers) as pool:
+                try:
+                    for fragment, frag_index, frag_filename in pool.map(_download_fragment, fragments):
+                        ctx.update({
+                            'fragment_filename_sanitized': frag_filename,
+                            'fragment_index': frag_index,
+                        })
+                        if not append_fragment(decrypt_fragment(fragment, self._read_fragment(ctx)), frag_index, ctx):
+                            return False
+                except KeyboardInterrupt:
+                    self._finish_multiline_status()
+                    self.report_error(
+                        'Interrupted by user. Waiting for all threads to shutdown...', is_error=False, tb=False)
+                    pool.shutdown(wait=False)
+                    raise
+        else:
+            for fragment in fragments:
+                if not interrupt_trigger[0]:
+                    break
+                try:
+                    download_fragment(fragment, ctx)
+                    result = append_fragment(
+                        decrypt_fragment(fragment, self._read_fragment(ctx)), fragment['frag_index'], ctx)
+                except KeyboardInterrupt:
+                    if info_dict.get('is_live'):
+                        break
+                    raise
+                if not result:
+                    return False
+
+        if finish_func is not None:
+            ctx['dest_stream'].write(finish_func())
+            ctx['dest_stream'].flush()
+        return self._finish_frag_download(ctx, info_dict)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/hls.py

@@ -0,0 +1,409 @@
+import binascii
+import io
+import re
+import urllib.parse
+
+from . import get_suitable_downloader
+from .external import FFmpegFD
+from .fragment import FragmentFD
+from .. import webvtt
+from ..dependencies import Cryptodome
+from ..utils import (
+    bug_reports_message,
+    parse_m3u8_attributes,
+    remove_start,
+    traverse_obj,
+    update_url_query,
+    urljoin,
+)
+from ..utils._utils import _request_dump_filename
+
+
+class HlsFD(FragmentFD):
+    """
+    Download segments in a m3u8 manifest. External downloaders can take over
+    the fragment downloads by supporting the 'm3u8_frag_urls' protocol and
+    re-defining 'supports_manifest' function
+    """
+
+    FD_NAME = 'hlsnative'
+
+    @staticmethod
+    def _has_drm(manifest):  # TODO: https://github.com/yt-dlp/yt-dlp/pull/5039
+        return bool(re.search('|'.join((
+            r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd://',  # Apple FairPlay
+            r'#EXT-X-(?:SESSION-)?KEY:.*?KEYFORMAT="com\.apple\.streamingkeydelivery"',  # Apple FairPlay
+            r'#EXT-X-(?:SESSION-)?KEY:.*?KEYFORMAT="com\.microsoft\.playready"',  # Microsoft PlayReady
+            r'#EXT-X-FAXS-CM:',  # Adobe Flash Access
+        )), manifest))
+
+    @classmethod
+    def can_download(cls, manifest, info_dict, allow_unplayable_formats=False):
+        UNSUPPORTED_FEATURES = [
+            # r'#EXT-X-BYTERANGE',  # playlists composed of byte ranges of media files [2]
+
+            # Live streams heuristic does not always work (e.g. geo restricted to Germany
+            # http://hls-geo.daserste.de/i/videoportal/Film/c_620000/622873/format,716451,716457,716450,716458,716459,.mp4.csmil/index_4_av.m3u8?null=0)
+            # r'#EXT-X-MEDIA-SEQUENCE:(?!0$)',  # live streams [3]
+
+            # This heuristic also is not correct since segments may not be appended as well.
+            # Twitch vods of finished streams have EXT-X-PLAYLIST-TYPE:EVENT despite
+            # no segments will definitely be appended to the end of the playlist.
+            # r'#EXT-X-PLAYLIST-TYPE:EVENT',  # media segments may be appended to the end of
+            #                                 # event media playlists [4]
+            # r'#EXT-X-MAP:',  # media initialization [5]
+            # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.2.4
+            # 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.2.2
+            # 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.2
+            # 4. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.5
+            # 5. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.2.5
+        ]
+        if not allow_unplayable_formats:
+            UNSUPPORTED_FEATURES += [
+                r'#EXT-X-KEY:METHOD=(?!NONE|AES-128)',  # encrypted streams [1], but not necessarily DRM
+            ]
+
+        def check_results():
+            yield not info_dict.get('is_live')
+            for feature in UNSUPPORTED_FEATURES:
+                yield not re.search(feature, manifest)
+            if not allow_unplayable_formats:
+                yield not cls._has_drm(manifest)
+        return all(check_results())
+
+    def real_download(self, filename, info_dict):
+        man_url = info_dict['url']
+
+        s = info_dict.get('hls_media_playlist_data')
+        if s:
+            self.to_screen(f'[{self.FD_NAME}] Using m3u8 manifest from extracted info')
+        else:
+            self.to_screen(f'[{self.FD_NAME}] Downloading m3u8 manifest')
+            urlh = self.ydl.urlopen(self._prepare_url(info_dict, man_url))
+            man_url = urlh.url
+            s_bytes = urlh.read()
+            if self.params.get('write_pages'):
+                dump_filename = _request_dump_filename(
+                    man_url, info_dict['id'], None,
+                    trim_length=self.params.get('trim_file_name'))
+                self.to_screen(f'[{self.FD_NAME}] Saving request to {dump_filename}')
+                with open(dump_filename, 'wb') as outf:
+                    outf.write(s_bytes)
+            s = s_bytes.decode('utf-8', 'ignore')
+
+        can_download, message = self.can_download(s, info_dict, self.params.get('allow_unplayable_formats')), None
+        if can_download:
+            has_ffmpeg = FFmpegFD.available()
+            if not Cryptodome.AES and '#EXT-X-KEY:METHOD=AES-128' in s:
+                # Even if pycryptodomex isn't available, force HlsFD for m3u8s that won't work with ffmpeg
+                ffmpeg_can_dl = not traverse_obj(info_dict, ((
+                    'extra_param_to_segment_url', 'extra_param_to_key_url',
+                    'hls_media_playlist_data', ('hls_aes', ('uri', 'key', 'iv')),
+                ), any))
+                message = 'The stream has AES-128 encryption and {} available'.format(
+                    'neither ffmpeg nor pycryptodomex are' if ffmpeg_can_dl and not has_ffmpeg else
+                    'pycryptodomex is not')
+                if has_ffmpeg and ffmpeg_can_dl:
+                    can_download = False
+                else:
+                    message += '; decryption will be performed natively, but will be extremely slow'
+            elif info_dict.get('extractor_key') == 'Generic' and re.search(r'(?m)#EXT-X-MEDIA-SEQUENCE:(?!0$)', s):
+                install_ffmpeg = '' if has_ffmpeg else 'install ffmpeg and '
+                message = ('Live HLS streams are not supported by the native downloader. If this is a livestream, '
+                           f'please {install_ffmpeg}add "--downloader ffmpeg --hls-use-mpegts" to your command')
+        if not can_download:
+            if self._has_drm(s) and not self.params.get('allow_unplayable_formats'):
+                if info_dict.get('has_drm') and self.params.get('test'):
+                    self.to_screen(f'[{self.FD_NAME}] This format is DRM protected', skip_eol=True)
+                else:
+                    self.report_error(
+                        'This format is DRM protected; Try selecting another format with --format or '
+                        'add --check-formats to automatically fallback to the next best format', tb=False)
+                return False
+            message = message or 'Unsupported features have been detected'
+            fd = FFmpegFD(self.ydl, self.params)
+            self.report_warning(f'{message}; extraction will be delegated to {fd.get_basename()}')
+            return fd.real_download(filename, info_dict)
+        elif message:
+            self.report_warning(message)
+
+        is_webvtt = info_dict['ext'] == 'vtt'
+        if is_webvtt:
+            real_downloader = None  # Packing the fragments is not currently supported for external downloader
+        else:
+            real_downloader = get_suitable_downloader(
+                info_dict, self.params, None, protocol='m3u8_frag_urls', to_stdout=(filename == '-'))
+        if real_downloader and not real_downloader.supports_manifest(s):
+            real_downloader = None
+        if real_downloader:
+            self.to_screen(f'[{self.FD_NAME}] Fragment downloads will be delegated to {real_downloader.get_basename()}')
+
+        def is_ad_fragment_start(s):
+            return ((s.startswith('#ANVATO-SEGMENT-INFO') and 'type=ad' in s)
+                    or (s.startswith('#UPLYNK-SEGMENT') and s.endswith(',ad')))
+
+        def is_ad_fragment_end(s):
+            return ((s.startswith('#ANVATO-SEGMENT-INFO') and 'type=master' in s)
+                    or (s.startswith('#UPLYNK-SEGMENT') and s.endswith(',segment')))
+
+        fragments = []
+
+        media_frags = 0
+        ad_frags = 0
+        ad_frag_next = False
+        for line in s.splitlines():
+            line = line.strip()
+            if not line:
+                continue
+            if line.startswith('#'):
+                if is_ad_fragment_start(line):
+                    ad_frag_next = True
+                elif is_ad_fragment_end(line):
+                    ad_frag_next = False
+                continue
+            if ad_frag_next:
+                ad_frags += 1
+                continue
+            media_frags += 1
+
+        ctx = {
+            'filename': filename,
+            'total_frags': media_frags,
+            'ad_frags': ad_frags,
+        }
+
+        if real_downloader:
+            self._prepare_external_frag_download(ctx)
+        else:
+            self._prepare_and_start_frag_download(ctx, info_dict)
+
+        extra_state = ctx.setdefault('extra_state', {})
+
+        format_index = info_dict.get('format_index')
+        extra_segment_query = None
+        if extra_param_to_segment_url := info_dict.get('extra_param_to_segment_url'):
+            extra_segment_query = urllib.parse.parse_qs(extra_param_to_segment_url)
+        extra_key_query = None
+        if extra_param_to_key_url := info_dict.get('extra_param_to_key_url'):
+            extra_key_query = urllib.parse.parse_qs(extra_param_to_key_url)
+        i = 0
+        media_sequence = 0
+        decrypt_info = {'METHOD': 'NONE'}
+        external_aes_key = traverse_obj(info_dict, ('hls_aes', 'key'))
+        if external_aes_key:
+            external_aes_key = binascii.unhexlify(remove_start(external_aes_key, '0x'))
+            assert len(external_aes_key) in (16, 24, 32), 'Invalid length for HLS AES-128 key'
+        external_aes_iv = traverse_obj(info_dict, ('hls_aes', 'iv'))
+        if external_aes_iv:
+            external_aes_iv = binascii.unhexlify(remove_start(external_aes_iv, '0x').zfill(32))
+        byte_range = {}
+        byte_range_offset = 0
+        discontinuity_count = 0
+        frag_index = 0
+        ad_frag_next = False
+        for line in s.splitlines():
+            line = line.strip()
+            if line:
+                if not line.startswith('#'):
+                    if format_index is not None and discontinuity_count != format_index:
+                        continue
+                    if ad_frag_next:
+                        continue
+                    frag_index += 1
+                    if frag_index <= ctx['fragment_index']:
+                        continue
+                    frag_url = urljoin(man_url, line)
+                    if extra_segment_query:
+                        frag_url = update_url_query(frag_url, extra_segment_query)
+
+                    fragments.append({
+                        'frag_index': frag_index,
+                        'url': frag_url,
+                        'decrypt_info': decrypt_info,
+                        'byte_range': byte_range,
+                        'media_sequence': media_sequence,
+                    })
+                    media_sequence += 1
+
+                    # If the byte_range is truthy, reset it after appending a fragment that uses it
+                    if byte_range:
+                        byte_range_offset = byte_range['end']
+                        byte_range = {}
+
+                elif line.startswith('#EXT-X-MAP'):
+                    if format_index is not None and discontinuity_count != format_index:
+                        continue
+                    if frag_index > 0:
+                        self.report_error(
+                            'Initialization fragment found after media fragments, unable to download')
+                        return False
+                    frag_index += 1
+                    map_info = parse_m3u8_attributes(line[11:])
+                    frag_url = urljoin(man_url, map_info.get('URI'))
+                    if extra_segment_query:
+                        frag_url = update_url_query(frag_url, extra_segment_query)
+
+                    map_byte_range = {}
+
+                    if map_info.get('BYTERANGE'):
+                        splitted_byte_range = map_info.get('BYTERANGE').split('@')
+                        sub_range_start = int(splitted_byte_range[1]) if len(splitted_byte_range) == 2 else 0
+                        map_byte_range = {
+                            'start': sub_range_start,
+                            'end': sub_range_start + int(splitted_byte_range[0]),
+                        }
+
+                    fragments.append({
+                        'frag_index': frag_index,
+                        'url': frag_url,
+                        'decrypt_info': decrypt_info,
+                        'byte_range': map_byte_range,
+                        'media_sequence': media_sequence,
+                    })
+                    media_sequence += 1
+
+                elif line.startswith('#EXT-X-KEY'):
+                    decrypt_url = decrypt_info.get('URI')
+                    decrypt_info = parse_m3u8_attributes(line[11:])
+                    if decrypt_info['METHOD'] == 'AES-128':
+                        if external_aes_iv:
+                            decrypt_info['IV'] = external_aes_iv
+                        elif 'IV' in decrypt_info:
+                            decrypt_info['IV'] = binascii.unhexlify(decrypt_info['IV'][2:].zfill(32))
+                        if external_aes_key:
+                            decrypt_info['KEY'] = external_aes_key
+                        else:
+                            decrypt_info['URI'] = urljoin(man_url, decrypt_info['URI'])
+                            if extra_key_query or extra_segment_query:
+                                # Fall back to extra_segment_query to key for backwards compat
+                                decrypt_info['URI'] = update_url_query(
+                                    decrypt_info['URI'], extra_key_query or extra_segment_query)
+                            if decrypt_url != decrypt_info['URI']:
+                                decrypt_info['KEY'] = None
+
+                elif line.startswith('#EXT-X-MEDIA-SEQUENCE'):
+                    media_sequence = int(line[22:])
+                elif line.startswith('#EXT-X-BYTERANGE'):
+                    splitted_byte_range = line[17:].split('@')
+                    sub_range_start = int(splitted_byte_range[1]) if len(splitted_byte_range) == 2 else byte_range_offset
+                    byte_range = {
+                        'start': sub_range_start,
+                        'end': sub_range_start + int(splitted_byte_range[0]),
+                    }
+                elif is_ad_fragment_start(line):
+                    ad_frag_next = True
+                elif is_ad_fragment_end(line):
+                    ad_frag_next = False
+                elif line.startswith('#EXT-X-DISCONTINUITY'):
+                    discontinuity_count += 1
+                i += 1
+
+        # We only download the first fragment during the test
+        if self.params.get('test', False):
+            fragments = [fragments[0] if fragments else None]
+
+        if real_downloader:
+            info_dict['fragments'] = fragments
+            fd = real_downloader(self.ydl, self.params)
+            # TODO: Make progress updates work without hooking twice
+            # for ph in self._progress_hooks:
+            #     fd.add_progress_hook(ph)
+            return fd.real_download(filename, info_dict)
+
+        if is_webvtt:
+            def pack_fragment(frag_content, frag_index):
+                output = io.StringIO()
+                adjust = 0
+                overflow = False
+                mpegts_last = None
+                for block in webvtt.parse_fragment(frag_content):
+                    if isinstance(block, webvtt.CueBlock):
+                        extra_state['webvtt_mpegts_last'] = mpegts_last
+                        if overflow:
+                            extra_state['webvtt_mpegts_adjust'] += 1
+                            overflow = False
+                        block.start += adjust
+                        block.end += adjust
+
+                        dedup_window = extra_state.setdefault('webvtt_dedup_window', [])
+
+                        ready = []
+
+                        i = 0
+                        is_new = True
+                        while i < len(dedup_window):
+                            wcue = dedup_window[i]
+                            wblock = webvtt.CueBlock.from_json(wcue)
+                            i += 1
+                            if wblock.hinges(block):
+                                wcue['end'] = block.end
+                                is_new = False
+                                continue
+                            if wblock == block:
+                                is_new = False
+                                continue
+                            if wblock.end > block.start:
+                                continue
+                            ready.append(wblock)
+                            i -= 1
+                            del dedup_window[i]
+
+                        if is_new:
+                            dedup_window.append(block.as_json)
+                        for block in ready:
+                            block.write_into(output)
+
+                        # we only emit cues once they fall out of the duplicate window
+                        continue
+                    elif isinstance(block, webvtt.Magic):
+                        # take care of MPEG PES timestamp overflow
+                        if block.mpegts is None:
+                            block.mpegts = 0
+                        extra_state.setdefault('webvtt_mpegts_adjust', 0)
+                        block.mpegts += extra_state['webvtt_mpegts_adjust'] << 33
+                        if block.mpegts < extra_state.get('webvtt_mpegts_last', 0):
+                            overflow = True
+                            block.mpegts += 1 << 33
+                        mpegts_last = block.mpegts
+
+                        if frag_index == 1:
+                            extra_state['webvtt_mpegts'] = block.mpegts or 0
+                            extra_state['webvtt_local'] = block.local or 0
+                            # XXX: block.local = block.mpegts = None ?
+                        else:
+                            if block.mpegts is not None and block.local is not None:
+                                adjust = (
+                                    (block.mpegts - extra_state.get('webvtt_mpegts', 0))
+                                    - (block.local - extra_state.get('webvtt_local', 0))
+                                )
+                            continue
+                    elif isinstance(block, webvtt.HeaderBlock):
+                        if frag_index != 1:
+                            # XXX: this should probably be silent as well
+                            # or verify that all segments contain the same data
+                            self.report_warning(bug_reports_message(
+                                f'Discarding a {type(block).__name__} block found in the middle of the stream; '
+                                'if the subtitles display incorrectly,'))
+                            continue
+                    block.write_into(output)
+
+                return output.getvalue().encode()
+
+            def fin_fragments():
+                dedup_window = extra_state.get('webvtt_dedup_window')
+                if not dedup_window:
+                    return b''
+
+                output = io.StringIO()
+                for cue in dedup_window:
+                    webvtt.CueBlock.from_json(cue).write_into(output)
+
+                return output.getvalue().encode()
+
+            if len(fragments) == 1:
+                self.download_and_append_fragments(ctx, fragments, info_dict)
+            else:
+                self.download_and_append_fragments(
+                    ctx, fragments, info_dict, pack_func=pack_fragment, finish_func=fin_fragments)
+        else:
+            return self.download_and_append_fragments(ctx, fragments, info_dict)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/http.py

@@ -0,0 +1,376 @@
+import os
+import random
+import time
+
+from .common import FileDownloader
+from ..networking import Request
+from ..networking.exceptions import (
+    CertificateVerifyError,
+    HTTPError,
+    TransportError,
+)
+from ..utils import (
+    ContentTooShortError,
+    RetryManager,
+    ThrottledDownload,
+    int_or_none,
+    parse_http_range,
+    try_call,
+)
+from ..utils.networking import HTTPHeaderDict
+
+
+class HttpFD(FileDownloader):
+    def real_download(self, filename, info_dict):
+        url = info_dict['url']
+        request_data = info_dict.get('request_data', None)
+        request_extensions = {}
+        impersonate_target = self._get_impersonate_target(info_dict)
+        if impersonate_target is not None:
+            request_extensions['impersonate'] = impersonate_target
+
+        class DownloadContext(dict):
+            __getattr__ = dict.get
+            __setattr__ = dict.__setitem__
+            __delattr__ = dict.__delitem__
+
+        ctx = DownloadContext()
+        ctx.filename = filename
+        ctx.tmpfilename = self.temp_name(filename)
+        ctx.stream = None
+
+        # Disable compression
+        headers = HTTPHeaderDict({'Accept-Encoding': 'identity'}, info_dict.get('http_headers'))
+
+        is_test = self.params.get('test', False)
+        chunk_size = self._TEST_FILE_SIZE if is_test else (
+            self.params.get('http_chunk_size')
+            or info_dict.get('downloader_options', {}).get('http_chunk_size')
+            or 0)
+
+        ctx.open_mode = 'wb'
+        ctx.resume_len = 0
+        ctx.block_size = self.params.get('buffersize', 1024)
+        ctx.start_time = time.time()
+
+        # parse given Range
+        req_start, req_end, _ = parse_http_range(headers.get('Range'))
+
+        if self.params.get('continuedl', True):
+            # Establish possible resume length
+            if os.path.isfile(ctx.tmpfilename):
+                ctx.resume_len = os.path.getsize(ctx.tmpfilename)
+
+        ctx.is_resume = ctx.resume_len > 0
+
+        class SucceedDownload(Exception):
+            pass
+
+        class RetryDownload(Exception):
+            def __init__(self, source_error):
+                self.source_error = source_error
+
+        class NextFragment(Exception):
+            pass
+
+        def establish_connection():
+            ctx.chunk_size = (random.randint(int(chunk_size * 0.95), chunk_size)
+                              if not is_test and chunk_size else chunk_size)
+            if ctx.resume_len > 0:
+                range_start = ctx.resume_len
+                if req_start is not None:
+                    # offset the beginning of Range to be within request
+                    range_start += req_start
+                if ctx.is_resume:
+                    self.report_resuming_byte(ctx.resume_len)
+                ctx.open_mode = 'ab'
+            elif req_start is not None:
+                range_start = req_start
+            elif ctx.chunk_size > 0:
+                range_start = 0
+            else:
+                range_start = None
+            ctx.is_resume = False
+
+            if ctx.chunk_size:
+                chunk_aware_end = range_start + ctx.chunk_size - 1
+                # we're not allowed to download outside Range
+                range_end = chunk_aware_end if req_end is None else min(chunk_aware_end, req_end)
+            elif req_end is not None:
+                # there's no need for chunked downloads, so download until the end of Range
+                range_end = req_end
+            else:
+                range_end = None
+
+            if try_call(lambda: range_start > range_end):
+                ctx.resume_len = 0
+                ctx.open_mode = 'wb'
+                raise RetryDownload(Exception(f'Conflicting range. (start={range_start} > end={range_end})'))
+
+            if try_call(lambda: range_end >= ctx.content_len):
+                range_end = ctx.content_len - 1
+
+            request = Request(url, request_data, headers, extensions=request_extensions)
+            has_range = range_start is not None
+            if has_range:
+                request.headers['Range'] = f'bytes={int(range_start)}-{int_or_none(range_end) or ""}'
+            # Establish connection
+            try:
+                ctx.data = self.ydl.urlopen(request)
+                # When trying to resume, Content-Range HTTP header of response has to be checked
+                # to match the value of requested Range HTTP header. This is due to a webservers
+                # that don't support resuming and serve a whole file with no Content-Range
+                # set in response despite of requested Range (see
+                # https://github.com/ytdl-org/youtube-dl/issues/6057#issuecomment-126129799)
+                if has_range:
+                    content_range = ctx.data.headers.get('Content-Range')
+                    content_range_start, content_range_end, content_len = parse_http_range(content_range)
+                    # Content-Range is present and matches requested Range, resume is possible
+                    if range_start == content_range_start and (
+                            # Non-chunked download
+                            not ctx.chunk_size
+                            # Chunked download and requested piece or
+                            # its part is promised to be served
+                            or content_range_end == range_end
+                            or content_len < range_end):
+                        ctx.content_len = content_len
+                        if content_len or req_end:
+                            ctx.data_len = min(content_len or req_end, req_end or content_len) - (req_start or 0)
+                        return
+                    # Content-Range is either not present or invalid. Assuming remote webserver is
+                    # trying to send the whole file, resume is not possible, so wiping the local file
+                    # and performing entire redownload
+                    elif range_start > 0:
+                        self.report_unable_to_resume()
+                    ctx.resume_len = 0
+                    ctx.open_mode = 'wb'
+                ctx.data_len = ctx.content_len = int_or_none(ctx.data.headers.get('Content-length', None))
+            except HTTPError as err:
+                if err.status == 416:
+                    # Unable to resume (requested range not satisfiable)
+                    try:
+                        # Open the connection again without the range header
+                        ctx.data = self.ydl.urlopen(
+                            Request(url, request_data, headers))
+                        content_length = ctx.data.headers['Content-Length']
+                    except HTTPError as err:
+                        if err.status < 500 or err.status >= 600:
+                            raise
+                    else:
+                        # Examine the reported length
+                        if (content_length is not None
+                                and (ctx.resume_len - 100 < int(content_length) < ctx.resume_len + 100)):
+                            # The file had already been fully downloaded.
+                            # Explanation to the above condition: in issue #175 it was revealed that
+                            # YouTube sometimes adds or removes a few bytes from the end of the file,
+                            # changing the file size slightly and causing problems for some users. So
+                            # I decided to implement a suggested change and consider the file
+                            # completely downloaded if the file size differs less than 100 bytes from
+                            # the one in the hard drive.
+                            self.report_file_already_downloaded(ctx.filename)
+                            self.try_rename(ctx.tmpfilename, ctx.filename)
+                            self._hook_progress({
+                                'filename': ctx.filename,
+                                'status': 'finished',
+                                'downloaded_bytes': ctx.resume_len,
+                                'total_bytes': ctx.resume_len,
+                            }, info_dict)
+                            raise SucceedDownload
+                        else:
+                            # The length does not match, we start the download over
+                            self.report_unable_to_resume()
+                            ctx.resume_len = 0
+                            ctx.open_mode = 'wb'
+                            return
+                elif err.status < 500 or err.status >= 600:
+                    # Unexpected HTTP error
+                    raise
+                raise RetryDownload(err)
+            except CertificateVerifyError:
+                raise
+            except TransportError as err:
+                raise RetryDownload(err)
+
+        def close_stream():
+            if ctx.stream is not None:
+                if ctx.tmpfilename != '-':
+                    ctx.stream.close()
+                ctx.stream = None
+
+        def download():
+            data_len = ctx.data.headers.get('Content-length')
+
+            if ctx.data.headers.get('Content-encoding'):
+                # Content-encoding is present, Content-length is not reliable anymore as we are
+                # doing auto decompression. (See: https://github.com/yt-dlp/yt-dlp/pull/6176)
+                data_len = None
+
+            # Range HTTP header may be ignored/unsupported by a webserver
+            # (e.g. extractor/scivee.py, extractor/bambuser.py).
+            # However, for a test we still would like to download just a piece of a file.
+            # To achieve this we limit data_len to _TEST_FILE_SIZE and manually control
+            # block size when downloading a file.
+            if is_test and (data_len is None or int(data_len) > self._TEST_FILE_SIZE):
+                data_len = self._TEST_FILE_SIZE
+
+            if data_len is not None:
+                data_len = int(data_len) + ctx.resume_len
+                min_data_len = self.params.get('min_filesize')
+                max_data_len = self.params.get('max_filesize')
+                if min_data_len is not None and data_len < min_data_len:
+                    self.to_screen(
+                        f'\r[download] File is smaller than min-filesize ({data_len} bytes < {min_data_len} bytes). Aborting.')
+                    return False
+                if max_data_len is not None and data_len > max_data_len:
+                    self.to_screen(
+                        f'\r[download] File is larger than max-filesize ({data_len} bytes > {max_data_len} bytes). Aborting.')
+                    return False
+
+            byte_counter = 0 + ctx.resume_len
+            block_size = ctx.block_size
+            start = time.time()
+
+            # measure time over whole while-loop, so slow_down() and best_block_size() work together properly
+            now = None  # needed for slow_down() in the first loop run
+            before = start  # start measuring
+
+            def retry(e):
+                close_stream()
+                if ctx.tmpfilename == '-':
+                    ctx.resume_len = byte_counter
+                else:
+                    try:
+                        ctx.resume_len = os.path.getsize(ctx.tmpfilename)
+                    except FileNotFoundError:
+                        ctx.resume_len = 0
+                raise RetryDownload(e)
+
+            while True:
+                try:
+                    # Download and write
+                    data_block = ctx.data.read(block_size if not is_test else min(block_size, data_len - byte_counter))
+                except TransportError as err:
+                    retry(err)
+
+                byte_counter += len(data_block)
+
+                # exit loop when download is finished
+                if len(data_block) == 0:
+                    break
+
+                # Open destination file just in time
+                if ctx.stream is None:
+                    try:
+                        ctx.stream, ctx.tmpfilename = self.sanitize_open(
+                            ctx.tmpfilename, ctx.open_mode)
+                        assert ctx.stream is not None
+                        ctx.filename = self.undo_temp_name(ctx.tmpfilename)
+                        self.report_destination(ctx.filename)
+                    except OSError as err:
+                        self.report_error(f'unable to open for writing: {err}')
+                        return False
+
+                try:
+                    ctx.stream.write(data_block)
+                except OSError as err:
+                    self.to_stderr('\n')
+                    self.report_error(f'unable to write data: {err}')
+                    return False
+
+                # Apply rate limit
+                self.slow_down(start, now, byte_counter - ctx.resume_len)
+
+                # end measuring of one loop run
+                now = time.time()
+                after = now
+
+                # Adjust block size
+                if not self.params.get('noresizebuffer', False):
+                    block_size = self.best_block_size(after - before, len(data_block))
+
+                before = after
+
+                # Progress message
+                speed = self.calc_speed(start, now, byte_counter - ctx.resume_len)
+                if ctx.data_len is None:
+                    eta = None
+                else:
+                    eta = self.calc_eta(start, time.time(), ctx.data_len - ctx.resume_len, byte_counter - ctx.resume_len)
+
+                self._hook_progress({
+                    'status': 'downloading',
+                    'downloaded_bytes': byte_counter,
+                    'total_bytes': ctx.data_len,
+                    'tmpfilename': ctx.tmpfilename,
+                    'filename': ctx.filename,
+                    'eta': eta,
+                    'speed': speed,
+                    'elapsed': now - ctx.start_time,
+                    'ctx_id': info_dict.get('ctx_id'),
+                }, info_dict)
+
+                if data_len is not None and byte_counter == data_len:
+                    break
+
+                if speed and speed < (self.params.get('throttledratelimit') or 0):
+                    # The speed must stay below the limit for 3 seconds
+                    # This prevents raising error when the speed temporarily goes down
+                    if ctx.throttle_start is None:
+                        ctx.throttle_start = now
+                    elif now - ctx.throttle_start > 3:
+                        if ctx.stream is not None and ctx.tmpfilename != '-':
+                            ctx.stream.close()
+                        raise ThrottledDownload
+                elif speed:
+                    ctx.throttle_start = None
+
+            if ctx.stream is None:
+                self.to_stderr('\n')
+                self.report_error('Did not get any data blocks')
+                return False
+
+            if not is_test and ctx.chunk_size and ctx.content_len is not None and byte_counter < ctx.content_len:
+                ctx.resume_len = byte_counter
+                raise NextFragment
+
+            if ctx.tmpfilename != '-':
+                ctx.stream.close()
+
+            if data_len is not None and byte_counter != data_len:
+                err = ContentTooShortError(byte_counter, int(data_len))
+                retry(err)
+
+            self.try_rename(ctx.tmpfilename, ctx.filename)
+
+            # Update file modification time
+            if self.params.get('updatetime'):
+                info_dict['filetime'] = self.try_utime(ctx.filename, ctx.data.headers.get('last-modified', None))
+
+            self._hook_progress({
+                'downloaded_bytes': byte_counter,
+                'total_bytes': byte_counter,
+                'filename': ctx.filename,
+                'status': 'finished',
+                'elapsed': time.time() - ctx.start_time,
+                'ctx_id': info_dict.get('ctx_id'),
+            }, info_dict)
+
+            return True
+
+        for retry in RetryManager(self.params.get('retries'), self.report_retry):
+            try:
+                establish_connection()
+                return download()
+            except RetryDownload as err:
+                retry.error = err.source_error
+                continue
+            except NextFragment:
+                retry.error = None
+                retry.attempt -= 1
+                continue
+            except SucceedDownload:
+                return True
+            except:  # noqa: E722
+                close_stream()
+                raise
+        return False

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/ism.py

@@ -0,0 +1,283 @@
+import binascii
+import io
+import struct
+import time
+
+from .fragment import FragmentFD
+from ..networking.exceptions import HTTPError
+from ..utils import RetryManager
+
+u8 = struct.Struct('>B')
+u88 = struct.Struct('>Bx')
+u16 = struct.Struct('>H')
+u1616 = struct.Struct('>Hxx')
+u32 = struct.Struct('>I')
+u64 = struct.Struct('>Q')
+
+s88 = struct.Struct('>bx')
+s16 = struct.Struct('>h')
+s1616 = struct.Struct('>hxx')
+s32 = struct.Struct('>i')
+
+unity_matrix = (s32.pack(0x10000) + s32.pack(0) * 3) * 2 + s32.pack(0x40000000)
+
+TRACK_ENABLED = 0x1
+TRACK_IN_MOVIE = 0x2
+TRACK_IN_PREVIEW = 0x4
+
+SELF_CONTAINED = 0x1
+
+
+def box(box_type, payload):
+    return u32.pack(8 + len(payload)) + box_type + payload
+
+
+def full_box(box_type, version, flags, payload):
+    return box(box_type, u8.pack(version) + u32.pack(flags)[1:] + payload)
+
+
+def write_piff_header(stream, params):
+    track_id = params['track_id']
+    fourcc = params['fourcc']
+    duration = params['duration']
+    timescale = params.get('timescale', 10000000)
+    language = params.get('language', 'und')
+    height = params.get('height', 0)
+    width = params.get('width', 0)
+    stream_type = params['stream_type']
+    creation_time = modification_time = int(time.time())
+
+    ftyp_payload = b'isml'  # major brand
+    ftyp_payload += u32.pack(1)  # minor version
+    ftyp_payload += b'piff' + b'iso2'  # compatible brands
+    stream.write(box(b'ftyp', ftyp_payload))  # File Type Box
+
+    mvhd_payload = u64.pack(creation_time)
+    mvhd_payload += u64.pack(modification_time)
+    mvhd_payload += u32.pack(timescale)
+    mvhd_payload += u64.pack(duration)
+    mvhd_payload += s1616.pack(1)  # rate
+    mvhd_payload += s88.pack(1)  # volume
+    mvhd_payload += u16.pack(0)  # reserved
+    mvhd_payload += u32.pack(0) * 2  # reserved
+    mvhd_payload += unity_matrix
+    mvhd_payload += u32.pack(0) * 6  # pre defined
+    mvhd_payload += u32.pack(0xffffffff)  # next track id
+    moov_payload = full_box(b'mvhd', 1, 0, mvhd_payload)  # Movie Header Box
+
+    tkhd_payload = u64.pack(creation_time)
+    tkhd_payload += u64.pack(modification_time)
+    tkhd_payload += u32.pack(track_id)  # track id
+    tkhd_payload += u32.pack(0)  # reserved
+    tkhd_payload += u64.pack(duration)
+    tkhd_payload += u32.pack(0) * 2  # reserved
+    tkhd_payload += s16.pack(0)  # layer
+    tkhd_payload += s16.pack(0)  # alternate group
+    tkhd_payload += s88.pack(1 if stream_type == 'audio' else 0)  # volume
+    tkhd_payload += u16.pack(0)  # reserved
+    tkhd_payload += unity_matrix
+    tkhd_payload += u1616.pack(width)
+    tkhd_payload += u1616.pack(height)
+    trak_payload = full_box(b'tkhd', 1, TRACK_ENABLED | TRACK_IN_MOVIE | TRACK_IN_PREVIEW, tkhd_payload)  # Track Header Box
+
+    mdhd_payload = u64.pack(creation_time)
+    mdhd_payload += u64.pack(modification_time)
+    mdhd_payload += u32.pack(timescale)
+    mdhd_payload += u64.pack(duration)
+    mdhd_payload += u16.pack(((ord(language[0]) - 0x60) << 10) | ((ord(language[1]) - 0x60) << 5) | (ord(language[2]) - 0x60))
+    mdhd_payload += u16.pack(0)  # pre defined
+    mdia_payload = full_box(b'mdhd', 1, 0, mdhd_payload)  # Media Header Box
+
+    hdlr_payload = u32.pack(0)  # pre defined
+    if stream_type == 'audio':  # handler type
+        hdlr_payload += b'soun'
+        hdlr_payload += u32.pack(0) * 3  # reserved
+        hdlr_payload += b'SoundHandler\0'  # name
+    elif stream_type == 'video':
+        hdlr_payload += b'vide'
+        hdlr_payload += u32.pack(0) * 3  # reserved
+        hdlr_payload += b'VideoHandler\0'  # name
+    elif stream_type == 'text':
+        hdlr_payload += b'subt'
+        hdlr_payload += u32.pack(0) * 3  # reserved
+        hdlr_payload += b'SubtitleHandler\0'  # name
+    else:
+        assert False
+    mdia_payload += full_box(b'hdlr', 0, 0, hdlr_payload)  # Handler Reference Box
+
+    if stream_type == 'audio':
+        smhd_payload = s88.pack(0)  # balance
+        smhd_payload += u16.pack(0)  # reserved
+        media_header_box = full_box(b'smhd', 0, 0, smhd_payload)  # Sound Media Header
+    elif stream_type == 'video':
+        vmhd_payload = u16.pack(0)  # graphics mode
+        vmhd_payload += u16.pack(0) * 3  # opcolor
+        media_header_box = full_box(b'vmhd', 0, 1, vmhd_payload)  # Video Media Header
+    elif stream_type == 'text':
+        media_header_box = full_box(b'sthd', 0, 0, b'')  # Subtitle Media Header
+    else:
+        assert False
+    minf_payload = media_header_box
+
+    dref_payload = u32.pack(1)  # entry count
+    dref_payload += full_box(b'url ', 0, SELF_CONTAINED, b'')  # Data Entry URL Box
+    dinf_payload = full_box(b'dref', 0, 0, dref_payload)  # Data Reference Box
+    minf_payload += box(b'dinf', dinf_payload)  # Data Information Box
+
+    stsd_payload = u32.pack(1)  # entry count
+
+    sample_entry_payload = u8.pack(0) * 6  # reserved
+    sample_entry_payload += u16.pack(1)  # data reference index
+    if stream_type == 'audio':
+        sample_entry_payload += u32.pack(0) * 2  # reserved
+        sample_entry_payload += u16.pack(params.get('channels', 2))
+        sample_entry_payload += u16.pack(params.get('bits_per_sample', 16))
+        sample_entry_payload += u16.pack(0)  # pre defined
+        sample_entry_payload += u16.pack(0)  # reserved
+        sample_entry_payload += u1616.pack(params['sampling_rate'])
+
+        if fourcc == 'AACL':
+            sample_entry_box = box(b'mp4a', sample_entry_payload)
+        if fourcc == 'EC-3':
+            sample_entry_box = box(b'ec-3', sample_entry_payload)
+    elif stream_type == 'video':
+        sample_entry_payload += u16.pack(0)  # pre defined
+        sample_entry_payload += u16.pack(0)  # reserved
+        sample_entry_payload += u32.pack(0) * 3  # pre defined
+        sample_entry_payload += u16.pack(width)
+        sample_entry_payload += u16.pack(height)
+        sample_entry_payload += u1616.pack(0x48)  # horiz resolution 72 dpi
+        sample_entry_payload += u1616.pack(0x48)  # vert resolution 72 dpi
+        sample_entry_payload += u32.pack(0)  # reserved
+        sample_entry_payload += u16.pack(1)  # frame count
+        sample_entry_payload += u8.pack(0) * 32  # compressor name
+        sample_entry_payload += u16.pack(0x18)  # depth
+        sample_entry_payload += s16.pack(-1)  # pre defined
+
+        codec_private_data = binascii.unhexlify(params['codec_private_data'].encode())
+        if fourcc in ('H264', 'AVC1'):
+            sps, pps = codec_private_data.split(u32.pack(1))[1:]
+            avcc_payload = u8.pack(1)  # configuration version
+            avcc_payload += sps[1:4]  # avc profile indication + profile compatibility + avc level indication
+            avcc_payload += u8.pack(0xfc | (params.get('nal_unit_length_field', 4) - 1))  # complete representation (1) + reserved (11111) + length size minus one
+            avcc_payload += u8.pack(1)  # reserved (0) + number of sps (0000001)
+            avcc_payload += u16.pack(len(sps))
+            avcc_payload += sps
+            avcc_payload += u8.pack(1)  # number of pps
+            avcc_payload += u16.pack(len(pps))
+            avcc_payload += pps
+            sample_entry_payload += box(b'avcC', avcc_payload)  # AVC Decoder Configuration Record
+            sample_entry_box = box(b'avc1', sample_entry_payload)  # AVC Simple Entry
+        else:
+            assert False
+    elif stream_type == 'text':
+        if fourcc == 'TTML':
+            sample_entry_payload += b'http://www.w3.org/ns/ttml\0'  # namespace
+            sample_entry_payload += b'\0'  # schema location
+            sample_entry_payload += b'\0'  # auxilary mime types(??)
+            sample_entry_box = box(b'stpp', sample_entry_payload)
+        else:
+            assert False
+    else:
+        assert False
+    stsd_payload += sample_entry_box
+
+    stbl_payload = full_box(b'stsd', 0, 0, stsd_payload)  # Sample Description Box
+
+    stts_payload = u32.pack(0)  # entry count
+    stbl_payload += full_box(b'stts', 0, 0, stts_payload)  # Decoding Time to Sample Box
+
+    stsc_payload = u32.pack(0)  # entry count
+    stbl_payload += full_box(b'stsc', 0, 0, stsc_payload)  # Sample To Chunk Box
+
+    stco_payload = u32.pack(0)  # entry count
+    stbl_payload += full_box(b'stco', 0, 0, stco_payload)  # Chunk Offset Box
+
+    minf_payload += box(b'stbl', stbl_payload)  # Sample Table Box
+
+    mdia_payload += box(b'minf', minf_payload)  # Media Information Box
+
+    trak_payload += box(b'mdia', mdia_payload)  # Media Box
+
+    moov_payload += box(b'trak', trak_payload)  # Track Box
+
+    mehd_payload = u64.pack(duration)
+    mvex_payload = full_box(b'mehd', 1, 0, mehd_payload)  # Movie Extends Header Box
+
+    trex_payload = u32.pack(track_id)  # track id
+    trex_payload += u32.pack(1)  # default sample description index
+    trex_payload += u32.pack(0)  # default sample duration
+    trex_payload += u32.pack(0)  # default sample size
+    trex_payload += u32.pack(0)  # default sample flags
+    mvex_payload += full_box(b'trex', 0, 0, trex_payload)  # Track Extends Box
+
+    moov_payload += box(b'mvex', mvex_payload)  # Movie Extends Box
+    stream.write(box(b'moov', moov_payload))  # Movie Box
+
+
+def extract_box_data(data, box_sequence):
+    data_reader = io.BytesIO(data)
+    while True:
+        box_size = u32.unpack(data_reader.read(4))[0]
+        box_type = data_reader.read(4)
+        if box_type == box_sequence[0]:
+            box_data = data_reader.read(box_size - 8)
+            if len(box_sequence) == 1:
+                return box_data
+            return extract_box_data(box_data, box_sequence[1:])
+        data_reader.seek(box_size - 8, 1)
+
+
+class IsmFD(FragmentFD):
+    """
+    Download segments in a ISM manifest
+    """
+
+    def real_download(self, filename, info_dict):
+        segments = info_dict['fragments'][:1] if self.params.get(
+            'test', False) else info_dict['fragments']
+
+        ctx = {
+            'filename': filename,
+            'total_frags': len(segments),
+        }
+
+        self._prepare_and_start_frag_download(ctx, info_dict)
+
+        extra_state = ctx.setdefault('extra_state', {
+            'ism_track_written': False,
+        })
+
+        skip_unavailable_fragments = self.params.get('skip_unavailable_fragments', True)
+
+        frag_index = 0
+        for segment in segments:
+            frag_index += 1
+            if frag_index <= ctx['fragment_index']:
+                continue
+
+            retry_manager = RetryManager(self.params.get('fragment_retries'), self.report_retry,
+                                         frag_index=frag_index, fatal=not skip_unavailable_fragments)
+            for retry in retry_manager:
+                try:
+                    success = self._download_fragment(ctx, segment['url'], info_dict)
+                    if not success:
+                        return False
+                    frag_content = self._read_fragment(ctx)
+
+                    if not extra_state['ism_track_written']:
+                        tfhd_data = extract_box_data(frag_content, [b'moof', b'traf', b'tfhd'])
+                        info_dict['_download_params']['track_id'] = u32.unpack(tfhd_data[4:8])[0]
+                        write_piff_header(ctx['dest_stream'], info_dict['_download_params'])
+                        extra_state['ism_track_written'] = True
+                    self._append_fragment(ctx, frag_content)
+                except HTTPError as err:
+                    retry.error = err
+                    continue
+
+            if retry_manager.error:
+                if not skip_unavailable_fragments:
+                    return False
+                self.report_skip_fragment(frag_index)
+
+        return self._finish_frag_download(ctx, info_dict)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/mhtml.py

@@ -0,0 +1,174 @@
+import io
+import quopri
+import re
+import uuid
+
+from .fragment import FragmentFD
+from ..compat import imghdr
+from ..utils import escapeHTML, formatSeconds, srt_subtitles_timecode, urljoin
+from ..version import __version__ as YT_DLP_VERSION
+
+
+class MhtmlFD(FragmentFD):
+    _STYLESHEET = '''\
+html, body {
+    margin: 0;
+    padding: 0;
+    height: 100vh;
+}
+
+html {
+    overflow-y: scroll;
+    scroll-snap-type: y mandatory;
+}
+
+body {
+    scroll-snap-type: y mandatory;
+    display: flex;
+    flex-flow: column;
+}
+
+body > figure {
+    max-width: 100vw;
+    max-height: 100vh;
+    scroll-snap-align: center;
+}
+
+body > figure > figcaption {
+    text-align: center;
+    height: 2.5em;
+}
+
+body > figure > img {
+    display: block;
+    margin: auto;
+    max-width: 100%;
+    max-height: calc(100vh - 5em);
+}
+'''
+    _STYLESHEET = re.sub(r'\s+', ' ', _STYLESHEET)
+    _STYLESHEET = re.sub(r'\B \B|(?<=[\w\-]) (?=[^\w\-])|(?<=[^\w\-]) (?=[\w\-])', '', _STYLESHEET)
+
+    @staticmethod
+    def _escape_mime(s):
+        return '=?utf-8?Q?' + (b''.join(
+            bytes((b,)) if b >= 0x20 else b'=%02X' % b
+            for b in quopri.encodestring(s.encode(), header=True)
+        )).decode('us-ascii') + '?='
+
+    def _gen_cid(self, i, fragment, frag_boundary):
+        return f'{i}.{frag_boundary}@yt-dlp.github.io.invalid'
+
+    def _gen_stub(self, *, fragments, frag_boundary, title):
+        output = io.StringIO()
+
+        output.write(
+            '<!DOCTYPE html>'
+            '<html>'
+            '<head>'
+            f'<meta name="generator" content="yt-dlp {escapeHTML(YT_DLP_VERSION)}">'
+            f'<title>{escapeHTML(title)}</title>'
+            f'<style>{self._STYLESHEET}</style>'
+            '<body>')
+
+        t0 = 0
+        for i, frag in enumerate(fragments):
+            output.write('<figure>')
+            try:
+                t1 = t0 + frag['duration']
+                output.write((
+                    '<figcaption>Slide #{num}: {t0} – {t1} (duration: {duration})</figcaption>'
+                ).format(
+                    num=i + 1,
+                    t0=srt_subtitles_timecode(t0),
+                    t1=srt_subtitles_timecode(t1),
+                    duration=formatSeconds(frag['duration'], msec=True),
+                ))
+            except (KeyError, ValueError, TypeError):
+                t1 = None
+                output.write(f'<figcaption>Slide #{i + 1}</figcaption>')
+            output.write(f'<img src="cid:{self._gen_cid(i, frag, frag_boundary)}">')
+            output.write('</figure>')
+            t0 = t1
+
+        return output.getvalue()
+
+    def real_download(self, filename, info_dict):
+        fragment_base_url = info_dict.get('fragment_base_url')
+        fragments = info_dict['fragments'][:1] if self.params.get(
+            'test', False) else info_dict['fragments']
+        title = info_dict.get('title', info_dict['format_id'])
+        origin = info_dict.get('webpage_url', info_dict['url'])
+
+        ctx = {
+            'filename': filename,
+            'total_frags': len(fragments),
+        }
+
+        self._prepare_and_start_frag_download(ctx, info_dict)
+
+        extra_state = ctx.setdefault('extra_state', {
+            'header_written': False,
+            'mime_boundary': str(uuid.uuid4()).replace('-', ''),
+        })
+
+        frag_boundary = extra_state['mime_boundary']
+
+        if not extra_state['header_written']:
+            stub = self._gen_stub(
+                fragments=fragments,
+                frag_boundary=frag_boundary,
+                title=title,
+            )
+
+            ctx['dest_stream'].write((
+                'MIME-Version: 1.0\r\n'
+                'From: <nowhere@yt-dlp.github.io.invalid>\r\n'
+                'To: <nowhere@yt-dlp.github.io.invalid>\r\n'
+                f'Subject: {self._escape_mime(title)}\r\n'
+                'Content-type: multipart/related; '
+                f'boundary="{frag_boundary}"; '
+                'type="text/html"\r\n'
+                f'X.yt-dlp.Origin: {origin}\r\n'
+                '\r\n'
+                f'--{frag_boundary}\r\n'
+                'Content-Type: text/html; charset=utf-8\r\n'
+                f'Content-Length: {len(stub)}\r\n'
+                '\r\n'
+                f'{stub}\r\n').encode())
+            extra_state['header_written'] = True
+
+        for i, fragment in enumerate(fragments):
+            if (i + 1) <= ctx['fragment_index']:
+                continue
+
+            fragment_url = fragment.get('url')
+            if not fragment_url:
+                assert fragment_base_url
+                fragment_url = urljoin(fragment_base_url, fragment['path'])
+
+            success = self._download_fragment(ctx, fragment_url, info_dict)
+            if not success:
+                continue
+            frag_content = self._read_fragment(ctx)
+
+            frag_header = io.BytesIO()
+            frag_header.write(
+                b'--%b\r\n' % frag_boundary.encode('us-ascii'))
+            frag_header.write(
+                b'Content-ID: <%b>\r\n' % self._gen_cid(i, fragment, frag_boundary).encode('us-ascii'))
+            frag_header.write(
+                b'Content-type: %b\r\n' % f'image/{imghdr.what(h=frag_content) or "jpeg"}'.encode())
+            frag_header.write(
+                b'Content-length: %u\r\n' % len(frag_content))
+            frag_header.write(
+                b'Content-location: %b\r\n' % fragment_url.encode('us-ascii'))
+            frag_header.write(
+                b'X.yt-dlp.Duration: %f\r\n' % fragment['duration'])
+            frag_header.write(b'\r\n')
+            self._append_fragment(
+                ctx, frag_header.getvalue() + frag_content + b'\r\n')
+
+        ctx['dest_stream'].write(
+            b'--%b--\r\n\r\n' % frag_boundary.encode('us-ascii'))
+        return self._finish_frag_download(ctx, info_dict)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/niconico.py

@@ -0,0 +1,90 @@
+import json
+import threading
+import time
+
+from .common import FileDownloader
+from .external import FFmpegFD
+from ..networking import Request
+from ..networking.websocket import WebSocketResponse
+from ..utils import DownloadError, str_or_none, truncate_string
+from ..utils.traversal import traverse_obj
+
+
+class NiconicoLiveFD(FileDownloader):
+    """ Downloads niconico live without being stopped """
+
+    def real_download(self, filename, info_dict):
+        video_id = info_dict['id']
+        opts = info_dict['downloader_options']
+        quality, ws_extractor, ws_url = opts['max_quality'], opts['ws'], opts['ws_url']
+        dl = FFmpegFD(self.ydl, self.params or {})
+
+        new_info_dict = info_dict.copy()
+        new_info_dict['protocol'] = 'm3u8'
+
+        def communicate_ws(reconnect):
+            # Support --load-info-json as if it is a reconnect attempt
+            if reconnect or not isinstance(ws_extractor, WebSocketResponse):
+                ws = self.ydl.urlopen(Request(
+                    ws_url, headers={'Origin': 'https://live.nicovideo.jp'}))
+                if self.ydl.params.get('verbose', False):
+                    self.write_debug('Sending startWatching request')
+                ws.send(json.dumps({
+                    'data': {
+                        'reconnect': True,
+                        'room': {
+                            'commentable': True,
+                            'protocol': 'webSocket',
+                        },
+                        'stream': {
+                            'accessRightMethod': 'single_cookie',
+                            'chasePlay': False,
+                            'latency': 'high',
+                            'protocol': 'hls',
+                            'quality': quality,
+                        },
+                    },
+                    'type': 'startWatching',
+                }))
+            else:
+                ws = ws_extractor
+            with ws:
+                while True:
+                    recv = ws.recv()
+                    if not recv:
+                        continue
+                    data = json.loads(recv)
+                    if not data or not isinstance(data, dict):
+                        continue
+                    if data.get('type') == 'ping':
+                        ws.send(r'{"type":"pong"}')
+                        ws.send(r'{"type":"keepSeat"}')
+                    elif data.get('type') == 'disconnect':
+                        self.write_debug(data)
+                        return True
+                    elif data.get('type') == 'error':
+                        self.write_debug(data)
+                        message = traverse_obj(data, ('body', 'code', {str_or_none}), default=recv)
+                        return DownloadError(message)
+                    elif self.ydl.params.get('verbose', False):
+                        self.write_debug(f'Server response: {truncate_string(recv, 100)}')
+
+        def ws_main():
+            reconnect = False
+            while True:
+                try:
+                    ret = communicate_ws(reconnect)
+                    if ret is True:
+                        return
+                except BaseException as e:
+                    self.to_screen(
+                        f'[niconico:live] {video_id}: Connection error occured, reconnecting after 10 seconds: {e}')
+                    time.sleep(10)
+                    continue
+                finally:
+                    reconnect = True
+
+        thread = threading.Thread(target=ws_main, daemon=True)
+        thread.start()
+
+        return dl.download(filename, new_info_dict)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/rtmp.py

@@ -0,0 +1,212 @@
+import os
+import re
+import subprocess
+import time
+
+from .common import FileDownloader
+from ..utils import (
+    Popen,
+    check_executable,
+    encodeArgument,
+    get_exe_version,
+)
+
+
+def rtmpdump_version():
+    return get_exe_version(
+        'rtmpdump', ['--help'], r'(?i)RTMPDump\s*v?([0-9a-zA-Z._-]+)')
+
+
+class RtmpFD(FileDownloader):
+    def real_download(self, filename, info_dict):
+        def run_rtmpdump(args):
+            start = time.time()
+            resume_percent = None
+            resume_downloaded_data_len = None
+            proc = Popen(args, stderr=subprocess.PIPE)
+            cursor_in_new_line = True
+            proc_stderr_closed = False
+            try:
+                while not proc_stderr_closed:
+                    # read line from stderr
+                    line = ''
+                    while True:
+                        char = proc.stderr.read(1)
+                        if not char:
+                            proc_stderr_closed = True
+                            break
+                        if char in [b'\r', b'\n']:
+                            break
+                        line += char.decode('ascii', 'replace')
+                    if not line:
+                        # proc_stderr_closed is True
+                        continue
+                    mobj = re.search(r'([0-9]+\.[0-9]{3}) kB / [0-9]+\.[0-9]{2} sec \(([0-9]{1,2}\.[0-9])%\)', line)
+                    if mobj:
+                        downloaded_data_len = int(float(mobj.group(1)) * 1024)
+                        percent = float(mobj.group(2))
+                        if not resume_percent:
+                            resume_percent = percent
+                            resume_downloaded_data_len = downloaded_data_len
+                        time_now = time.time()
+                        eta = self.calc_eta(start, time_now, 100 - resume_percent, percent - resume_percent)
+                        speed = self.calc_speed(start, time_now, downloaded_data_len - resume_downloaded_data_len)
+                        data_len = None
+                        if percent > 0:
+                            data_len = int(downloaded_data_len * 100 / percent)
+                        self._hook_progress({
+                            'status': 'downloading',
+                            'downloaded_bytes': downloaded_data_len,
+                            'total_bytes_estimate': data_len,
+                            'tmpfilename': tmpfilename,
+                            'filename': filename,
+                            'eta': eta,
+                            'elapsed': time_now - start,
+                            'speed': speed,
+                        }, info_dict)
+                        cursor_in_new_line = False
+                    else:
+                        # no percent for live streams
+                        mobj = re.search(r'([0-9]+\.[0-9]{3}) kB / [0-9]+\.[0-9]{2} sec', line)
+                        if mobj:
+                            downloaded_data_len = int(float(mobj.group(1)) * 1024)
+                            time_now = time.time()
+                            speed = self.calc_speed(start, time_now, downloaded_data_len)
+                            self._hook_progress({
+                                'downloaded_bytes': downloaded_data_len,
+                                'tmpfilename': tmpfilename,
+                                'filename': filename,
+                                'status': 'downloading',
+                                'elapsed': time_now - start,
+                                'speed': speed,
+                            }, info_dict)
+                            cursor_in_new_line = False
+                        elif self.params.get('verbose', False):
+                            if not cursor_in_new_line:
+                                self.to_screen('')
+                            cursor_in_new_line = True
+                            self.to_screen('[rtmpdump] ' + line)
+                if not cursor_in_new_line:
+                    self.to_screen('')
+                return proc.wait()
+            except BaseException:  # Including KeyboardInterrupt
+                proc.kill(timeout=None)
+                raise
+
+        url = info_dict['url']
+        player_url = info_dict.get('player_url')
+        page_url = info_dict.get('page_url')
+        app = info_dict.get('app')
+        play_path = info_dict.get('play_path')
+        tc_url = info_dict.get('tc_url')
+        flash_version = info_dict.get('flash_version')
+        live = info_dict.get('rtmp_live', False)
+        conn = info_dict.get('rtmp_conn')
+        protocol = info_dict.get('rtmp_protocol')
+        real_time = info_dict.get('rtmp_real_time', False)
+        no_resume = info_dict.get('no_resume', False)
+        continue_dl = self.params.get('continuedl', True)
+
+        self.report_destination(filename)
+        tmpfilename = self.temp_name(filename)
+        test = self.params.get('test', False)
+
+        # Check for rtmpdump first
+        if not check_executable('rtmpdump', ['-h']):
+            self.report_error('RTMP download detected but "rtmpdump" could not be run. Please install')
+            return False
+
+        # Download using rtmpdump. rtmpdump returns exit code 2 when
+        # the connection was interrupted and resuming appears to be
+        # possible. This is part of rtmpdump's normal usage, AFAIK.
+        basic_args = [
+            'rtmpdump', '--verbose', '-r', url,
+            '-o', tmpfilename]
+        if player_url is not None:
+            basic_args += ['--swfVfy', player_url]
+        if page_url is not None:
+            basic_args += ['--pageUrl', page_url]
+        if app is not None:
+            basic_args += ['--app', app]
+        if play_path is not None:
+            basic_args += ['--playpath', play_path]
+        if tc_url is not None:
+            basic_args += ['--tcUrl', tc_url]
+        if test:
+            basic_args += ['--stop', '1']
+        if flash_version is not None:
+            basic_args += ['--flashVer', flash_version]
+        if live:
+            basic_args += ['--live']
+        if isinstance(conn, list):
+            for entry in conn:
+                basic_args += ['--conn', entry]
+        elif isinstance(conn, str):
+            basic_args += ['--conn', conn]
+        if protocol is not None:
+            basic_args += ['--protocol', protocol]
+        if real_time:
+            basic_args += ['--realtime']
+
+        args = basic_args
+        if not no_resume and continue_dl and not live:
+            args += ['--resume']
+        if not live and continue_dl:
+            args += ['--skip', '1']
+
+        args = [encodeArgument(a) for a in args]
+
+        self._debug_cmd(args, exe='rtmpdump')
+
+        RD_SUCCESS = 0
+        RD_FAILED = 1
+        RD_INCOMPLETE = 2
+        RD_NO_CONNECT = 3
+
+        started = time.time()
+
+        try:
+            retval = run_rtmpdump(args)
+        except KeyboardInterrupt:
+            if not info_dict.get('is_live'):
+                raise
+            retval = RD_SUCCESS
+            self.to_screen('\n[rtmpdump] Interrupted by user')
+
+        if retval == RD_NO_CONNECT:
+            self.report_error('[rtmpdump] Could not connect to RTMP server.')
+            return False
+
+        while retval in (RD_INCOMPLETE, RD_FAILED) and not test and not live:
+            prevsize = os.path.getsize(tmpfilename)
+            self.to_screen(f'[rtmpdump] Downloaded {prevsize} bytes')
+            time.sleep(5.0)  # This seems to be needed
+            args = [*basic_args, '--resume']
+            if retval == RD_FAILED:
+                args += ['--skip', '1']
+            args = [encodeArgument(a) for a in args]
+            retval = run_rtmpdump(args)
+            cursize = os.path.getsize(tmpfilename)
+            if prevsize == cursize and retval == RD_FAILED:
+                break
+            # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
+            if prevsize == cursize and retval == RD_INCOMPLETE and cursize > 1024:
+                self.to_screen('[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
+                retval = RD_SUCCESS
+                break
+        if retval == RD_SUCCESS or (test and retval == RD_INCOMPLETE):
+            fsize = os.path.getsize(tmpfilename)
+            self.to_screen(f'[rtmpdump] Downloaded {fsize} bytes')
+            self.try_rename(tmpfilename, filename)
+            self._hook_progress({
+                'downloaded_bytes': fsize,
+                'total_bytes': fsize,
+                'filename': filename,
+                'status': 'finished',
+                'elapsed': time.time() - started,
+            }, info_dict)
+            return True
+        else:
+            self.to_stderr('\n')
+            self.report_error('rtmpdump exited with code %d' % retval)
+            return False

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/rtsp.py

@@ -0,0 +1,42 @@
+import os
+import subprocess
+
+from .common import FileDownloader
+from ..utils import check_executable
+
+
+class RtspFD(FileDownloader):
+    def real_download(self, filename, info_dict):
+        url = info_dict['url']
+        self.report_destination(filename)
+        tmpfilename = self.temp_name(filename)
+
+        if check_executable('mplayer', ['-h']):
+            args = [
+                'mplayer', '-really-quiet', '-vo', 'null', '-vc', 'dummy',
+                '-dumpstream', '-dumpfile', tmpfilename, url]
+        elif check_executable('mpv', ['-h']):
+            args = [
+                'mpv', '-really-quiet', '--vo=null', '--stream-dump=' + tmpfilename, url]
+        else:
+            self.report_error('MMS or RTSP download detected but neither "mplayer" nor "mpv" could be run. Please install one')
+            return False
+
+        self._debug_cmd(args)
+
+        retval = subprocess.call(args)
+        if retval == 0:
+            fsize = os.path.getsize(tmpfilename)
+            self.to_screen(f'\r[{args[0]}] {fsize} bytes')
+            self.try_rename(tmpfilename, filename)
+            self._hook_progress({
+                'downloaded_bytes': fsize,
+                'total_bytes': fsize,
+                'filename': filename,
+                'status': 'finished',
+            }, info_dict)
+            return True
+        else:
+            self.to_stderr('\n')
+            self.report_error('%s exited with code %d' % (args[0], retval))
+            return False

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/websocket.py

@@ -0,0 +1,53 @@
+import asyncio
+import contextlib
+import os
+import signal
+import threading
+
+from .common import FileDownloader
+from .external import FFmpegFD
+from ..dependencies import websockets
+
+
+class FFmpegSinkFD(FileDownloader):
+    """ A sink to ffmpeg for downloading fragments in any form """
+
+    def real_download(self, filename, info_dict):
+        info_copy = info_dict.copy()
+        info_copy['url'] = '-'
+
+        async def call_conn(proc, stdin):
+            try:
+                await self.real_connection(stdin, info_dict)
+            except OSError:
+                pass
+            finally:
+                with contextlib.suppress(OSError):
+                    stdin.flush()
+                    stdin.close()
+                os.kill(os.getpid(), signal.SIGINT)
+
+        class FFmpegStdinFD(FFmpegFD):
+            @classmethod
+            def get_basename(cls):
+                return FFmpegFD.get_basename()
+
+            def on_process_started(self, proc, stdin):
+                thread = threading.Thread(target=asyncio.run, daemon=True, args=(call_conn(proc, stdin), ))
+                thread.start()
+
+        return FFmpegStdinFD(self.ydl, self.params or {}).download(filename, info_copy)
+
+    async def real_connection(self, sink, info_dict):
+        """ Override this in subclasses """
+        raise NotImplementedError('This method must be implemented by subclasses')
+
+
+class WebSocketFragmentFD(FFmpegSinkFD):
+    async def real_connection(self, sink, info_dict):
+        async with websockets.connect(info_dict['url'], extra_headers=info_dict.get('http_headers', {})) as ws:
+            while True:
+                recv = await ws.recv()
+                if isinstance(recv, str):
+                    recv = recv.encode('utf8')
+                sink.write(recv)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/downloader/youtube_live_chat.py

@@ -0,0 +1,228 @@
+import json
+import time
+
+from .fragment import FragmentFD
+from ..networking.exceptions import HTTPError
+from ..utils import (
+    RegexNotFoundError,
+    RetryManager,
+    dict_get,
+    int_or_none,
+    try_get,
+)
+from ..utils.networking import HTTPHeaderDict
+
+
+class YoutubeLiveChatFD(FragmentFD):
+    """ Downloads YouTube live chats fragment by fragment """
+
+    def real_download(self, filename, info_dict):
+        video_id = info_dict['video_id']
+        self.to_screen(f'[{self.FD_NAME}] Downloading live chat')
+        if not self.params.get('skip_download') and info_dict['protocol'] == 'youtube_live_chat':
+            self.report_warning('Live chat download runs until the livestream ends. '
+                                'If you wish to download the video simultaneously, run a separate yt-dlp instance')
+
+        test = self.params.get('test', False)
+
+        ctx = {
+            'filename': filename,
+            'live': True,
+            'total_frags': None,
+        }
+
+        from ..extractor.youtube import YoutubeBaseInfoExtractor
+
+        ie = YoutubeBaseInfoExtractor(self.ydl)
+
+        start_time = int(time.time() * 1000)
+
+        def dl_fragment(url, data=None, headers=None):
+            http_headers = HTTPHeaderDict(info_dict.get('http_headers'), headers)
+            return self._download_fragment(ctx, url, info_dict, http_headers, data)
+
+        def parse_actions_replay(live_chat_continuation):
+            offset = continuation_id = click_tracking_params = None
+            processed_fragment = bytearray()
+            for action in live_chat_continuation.get('actions', []):
+                if 'replayChatItemAction' in action:
+                    replay_chat_item_action = action['replayChatItemAction']
+                    offset = int(replay_chat_item_action['videoOffsetTimeMsec'])
+                processed_fragment.extend(
+                    json.dumps(action, ensure_ascii=False).encode() + b'\n')
+            if offset is not None:
+                continuation = try_get(
+                    live_chat_continuation,
+                    lambda x: x['continuations'][0]['liveChatReplayContinuationData'], dict)
+                if continuation:
+                    continuation_id = continuation.get('continuation')
+                    click_tracking_params = continuation.get('clickTrackingParams')
+            self._append_fragment(ctx, processed_fragment)
+            return continuation_id, offset, click_tracking_params
+
+        def try_refresh_replay_beginning(live_chat_continuation):
+            # choose the second option that contains the unfiltered live chat replay
+            refresh_continuation = try_get(
+                live_chat_continuation,
+                lambda x: x['header']['liveChatHeaderRenderer']['viewSelector']['sortFilterSubMenuRenderer']['subMenuItems'][1]['continuation']['reloadContinuationData'], dict)
+            if refresh_continuation:
+                # no data yet but required to call _append_fragment
+                self._append_fragment(ctx, b'')
+                refresh_continuation_id = refresh_continuation.get('continuation')
+                offset = 0
+                click_tracking_params = refresh_continuation.get('trackingParams')
+                return refresh_continuation_id, offset, click_tracking_params
+            return parse_actions_replay(live_chat_continuation)
+
+        live_offset = 0
+
+        def parse_actions_live(live_chat_continuation):
+            nonlocal live_offset
+            continuation_id = click_tracking_params = None
+            processed_fragment = bytearray()
+            for action in live_chat_continuation.get('actions', []):
+                timestamp = self.parse_live_timestamp(action)
+                if timestamp is not None:
+                    live_offset = timestamp - start_time
+                # compatibility with replay format
+                pseudo_action = {
+                    'replayChatItemAction': {'actions': [action]},
+                    'videoOffsetTimeMsec': str(live_offset),
+                    'isLive': True,
+                }
+                processed_fragment.extend(
+                    json.dumps(pseudo_action, ensure_ascii=False).encode() + b'\n')
+            continuation_data_getters = [
+                lambda x: x['continuations'][0]['invalidationContinuationData'],
+                lambda x: x['continuations'][0]['timedContinuationData'],
+            ]
+            continuation_data = try_get(live_chat_continuation, continuation_data_getters, dict)
+            if continuation_data:
+                continuation_id = continuation_data.get('continuation')
+                click_tracking_params = continuation_data.get('clickTrackingParams')
+                timeout_ms = int_or_none(continuation_data.get('timeoutMs'))
+                if timeout_ms is not None:
+                    time.sleep(timeout_ms / 1000)
+            self._append_fragment(ctx, processed_fragment)
+            return continuation_id, live_offset, click_tracking_params
+
+        def download_and_parse_fragment(url, frag_index, request_data=None, headers=None):
+            for retry in RetryManager(self.params.get('fragment_retries'), self.report_retry, frag_index=frag_index):
+                try:
+                    success = dl_fragment(url, request_data, headers)
+                    if not success:
+                        return False, None, None, None
+                    raw_fragment = self._read_fragment(ctx)
+                    try:
+                        data = ie.extract_yt_initial_data(video_id, raw_fragment.decode('utf-8', 'replace'))
+                    except RegexNotFoundError:
+                        data = None
+                    if not data:
+                        data = json.loads(raw_fragment)
+                    live_chat_continuation = try_get(
+                        data,
+                        lambda x: x['continuationContents']['liveChatContinuation'], dict) or {}
+
+                    func = ((info_dict['protocol'] == 'youtube_live_chat' and parse_actions_live)
+                            or (frag_index == 1 and try_refresh_replay_beginning)
+                            or parse_actions_replay)
+                    return (True, *func(live_chat_continuation))
+                except HTTPError as err:
+                    retry.error = err
+                    continue
+            return False, None, None, None
+
+        self._prepare_and_start_frag_download(ctx, info_dict)
+
+        success = dl_fragment(info_dict['url'])
+        if not success:
+            return False
+        raw_fragment = self._read_fragment(ctx)
+        try:
+            data = ie.extract_yt_initial_data(video_id, raw_fragment.decode('utf-8', 'replace'))
+        except RegexNotFoundError:
+            return False
+        continuation_id = try_get(
+            data,
+            lambda x: x['contents']['twoColumnWatchNextResults']['conversationBar']['liveChatRenderer']['continuations'][0]['reloadContinuationData']['continuation'])
+        # no data yet but required to call _append_fragment
+        self._append_fragment(ctx, b'')
+
+        ytcfg = ie.extract_ytcfg(video_id, raw_fragment.decode('utf-8', 'replace'))
+
+        if not ytcfg:
+            return False
+        api_key = try_get(ytcfg, lambda x: x['INNERTUBE_API_KEY'])
+        innertube_context = try_get(ytcfg, lambda x: x['INNERTUBE_CONTEXT'])
+        if not api_key or not innertube_context:
+            return False
+        visitor_data = try_get(innertube_context, lambda x: x['client']['visitorData'], str)
+        if info_dict['protocol'] == 'youtube_live_chat_replay':
+            url = 'https://www.youtube.com/youtubei/v1/live_chat/get_live_chat_replay?key=' + api_key
+            chat_page_url = 'https://www.youtube.com/live_chat_replay?continuation=' + continuation_id
+        elif info_dict['protocol'] == 'youtube_live_chat':
+            url = 'https://www.youtube.com/youtubei/v1/live_chat/get_live_chat?key=' + api_key
+            chat_page_url = 'https://www.youtube.com/live_chat?continuation=' + continuation_id
+
+        frag_index = offset = 0
+        click_tracking_params = None
+        while continuation_id is not None:
+            frag_index += 1
+            request_data = {
+                'context': innertube_context,
+                'continuation': continuation_id,
+            }
+            if frag_index > 1:
+                request_data['currentPlayerState'] = {'playerOffsetMs': str(max(offset - 5000, 0))}
+                if click_tracking_params:
+                    request_data['context']['clickTracking'] = {'clickTrackingParams': click_tracking_params}
+                headers = ie.generate_api_headers(ytcfg=ytcfg, visitor_data=visitor_data)
+                headers.update({'content-type': 'application/json'})
+                fragment_request_data = json.dumps(request_data, ensure_ascii=False).encode() + b'\n'
+                success, continuation_id, offset, click_tracking_params = download_and_parse_fragment(
+                    url, frag_index, fragment_request_data, headers)
+            else:
+                success, continuation_id, offset, click_tracking_params = download_and_parse_fragment(
+                    chat_page_url, frag_index)
+            if not success:
+                return False
+            if test:
+                break
+
+        return self._finish_frag_download(ctx, info_dict)
+
+    @staticmethod
+    def parse_live_timestamp(action):
+        action_content = dict_get(
+            action,
+            ['addChatItemAction', 'addLiveChatTickerItemAction', 'addBannerToLiveChatCommand'])
+        if not isinstance(action_content, dict):
+            return None
+        item = dict_get(action_content, ['item', 'bannerRenderer'])
+        if not isinstance(item, dict):
+            return None
+        renderer = dict_get(item, [
+            # text
+            'liveChatTextMessageRenderer', 'liveChatPaidMessageRenderer',
+            'liveChatMembershipItemRenderer', 'liveChatPaidStickerRenderer',
+            # ticker
+            'liveChatTickerPaidMessageItemRenderer',
+            'liveChatTickerSponsorItemRenderer',
+            # banner
+            'liveChatBannerRenderer',
+        ])
+        if not isinstance(renderer, dict):
+            return None
+        parent_item_getters = [
+            lambda x: x['showItemEndpoint']['showLiveChatItemEndpoint']['renderer'],
+            lambda x: x['contents'],
+        ]
+        parent_item = try_get(renderer, parent_item_getters, dict)
+        if parent_item:
+            renderer = dict_get(parent_item, [
+                'liveChatTextMessageRenderer', 'liveChatPaidMessageRenderer',
+                'liveChatMembershipItemRenderer', 'liveChatPaidStickerRenderer',
+            ])
+            if not isinstance(renderer, dict):
+                return None
+        return int_or_none(renderer.get('timestampUsec'), 1000)

download-backend/.venv/lib/python3.12/site-packages/yt_dlp/extractor/__init__.py

@@ -0,0 +1,54 @@
+from ..compat.compat_utils import passthrough_module
+from ..globals import extractors as _extractors_context
+from ..globals import plugin_ies as _plugin_ies_context
+from ..plugins import PluginSpec, register_plugin_spec
+
+passthrough_module(__name__, '.extractors')
+del passthrough_module
+
+register_plugin_spec(PluginSpec(
+    module_name='extractor',
+    suffix='IE',
+    destination=_extractors_context,
+    plugin_destination=_plugin_ies_context,
+))
+
+
+def gen_extractor_classes():
+    """ Return a list of supported extractors.
+    The order does matter; the first extractor matched is the one handling the URL.
+    """
+    import_extractors()
+    return list(_extractors_context.value.values())
+
+
+def gen_extractors():
+    """ Return a list of an instance of every supported extractor.
+    The order does matter; the first extractor matched is the one handling the URL.
+    """
+    return [klass() for klass in gen_extractor_classes()]
+
+
+def list_extractor_classes(age_limit=None):
+    """Return a list of extractors that are suitable for the given age, sorted by extractor name"""
+    from .generic import GenericIE
+
+    yield from sorted(filter(
+        lambda ie: ie.is_suitable(age_limit) and ie != GenericIE,
+        gen_extractor_classes()), key=lambda ie: ie.IE_NAME.lower())
+    yield GenericIE
+
+
+def list_extractors(age_limit=None):
+    """Return a list of extractor instances that are suitable for the given age, sorted by extractor name"""
+    return [ie() for ie in list_extractor_classes(age_limit)]
+
+
+def get_info_extractor(ie_name):
+    """Returns the info extractor class with the given ie_name"""
+    import_extractors()
+    return _extractors_context.value[f'{ie_name}IE']
+
+
+def import_extractors():
+    from . import extractors  # noqa: F401