[jsinterp] Some optimizations and refactoring

Motivated by: https://github.com/ytdl-org/youtube-dl/issues/30641#issuecomment-1041904912

Authored by: dirkf, pukkandan

.github/workflows/build.yml

@@ -70,14 +70,16 @@ jobs:
 
   build_unix:
     needs: create_release
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-18.04  # Standalone executable should be built on minimum supported OS
     outputs:
       sha256_bin: ${{ steps.get_sha.outputs.sha256_bin }}
       sha512_bin: ${{ steps.get_sha.outputs.sha512_bin }}
       sha256_tar: ${{ steps.get_sha.outputs.sha256_tar }}
       sha512_tar: ${{ steps.get_sha.outputs.sha512_tar }}
-      sha256_unix_zip: ${{ steps.get_sha.outputs.sha256_unix_zip }}
-      sha512_unix_zip: ${{ steps.get_sha.outputs.sha512_unix_zip }}
+      sha256_linux: ${{ steps.get_sha.outputs.sha256_linux }}
+      sha512_linux: ${{ steps.get_sha.outputs.sha512_linux }}
+      sha256_linux_zip: ${{ steps.get_sha.outputs.sha256_linux_zip }}
+      sha512_linux_zip: ${{ steps.get_sha.outputs.sha512_linux_zip }}
 
     steps:
     - uses: actions/checkout@v2
@@ -94,9 +96,12 @@ jobs:
       run: |
           python devscripts/update-version.py ${{ needs.create_release.outputs.version_suffix }}
           python devscripts/make_lazy_extractors.py
-    - name: Build UNIX executables
+    - name: Build Unix executables
       run: |
           make all tar
+          python pyinst.py --onedir
+          (cd ./dist/yt-dlp_linux && zip -r ../yt-dlp_linux.zip .)
+          python pyinst.py
     - name: Get SHA2-SUMS
       id: get_sha
       run: |
@@ -104,6 +109,10 @@ jobs:
           echo "::set-output name=sha512_bin::$(sha512sum yt-dlp | awk '{print $1}')"
           echo "::set-output name=sha256_tar::$(sha256sum yt-dlp.tar.gz | awk '{print $1}')"
           echo "::set-output name=sha512_tar::$(sha512sum yt-dlp.tar.gz | awk '{print $1}')"
+          echo "::set-output name=sha256_linux::$(sha256sum dist/yt-dlp_linux | awk '{print $1}')"
+          echo "::set-output name=sha512_linux::$(sha512sum dist/yt-dlp_linux | awk '{print $1}')"
+          echo "::set-output name=sha256_linux_zip::$(sha256sum dist/yt-dlp_linux.zip | awk '{print $1}')"
+          echo "::set-output name=sha512_linux_zip::$(sha512sum dist/yt-dlp_linux.zip | awk '{print $1}')"
 
     - name: Upload zip binary
       uses: actions/upload-release-asset@v1
@@ -123,6 +132,24 @@ jobs:
         asset_path: ./yt-dlp.tar.gz
         asset_name: yt-dlp.tar.gz
         asset_content_type: application/gzip
+    - name: Upload standalone binary
+      uses: actions/upload-release-asset@v1
+      env:
+        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      with:
+        upload_url: ${{ needs.create_release.outputs.upload_url }}
+        asset_path: ./dist/yt-dlp_linux
+        asset_name: yt-dlp_linux
+        asset_content_type: application/octet-stream
+    - name: Upload onedir binary
+      uses: actions/upload-release-asset@v1
+      env:
+        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      with:
+        upload_url: ${{ needs.create_release.outputs.upload_url }}
+        asset_path: ./dist/yt-dlp_linux.zip
+        asset_name: yt-dlp_linux.zip
+        asset_content_type: application/zip
 
     - name: Build and publish on PyPi
       env:
@@ -330,6 +357,8 @@ jobs:
       run: |
         echo "${{ needs.build_unix.outputs.sha256_bin }}  yt-dlp" >> SHA2-256SUMS
         echo "${{ needs.build_unix.outputs.sha256_tar }}  yt-dlp.tar.gz" >> SHA2-256SUMS
+        echo "${{ needs.build_unix.outputs.sha256_linux }}  yt-dlp_linux" >> SHA2-256SUMS
+        echo "${{ needs.build_unix.outputs.sha256_linux_zip }}  yt-dlp_linux.zip" >> SHA2-256SUMS
         echo "${{ needs.build_windows.outputs.sha256_win }}  yt-dlp.exe" >> SHA2-256SUMS
         echo "${{ needs.build_windows.outputs.sha256_py2exe }}  yt-dlp_min.exe" >> SHA2-256SUMS
         echo "${{ needs.build_windows32.outputs.sha256_win32 }}  yt-dlp_x86.exe" >> SHA2-256SUMS
@@ -338,6 +367,8 @@ jobs:
         echo "${{ needs.build_macos.outputs.sha256_macos_zip }}  yt-dlp_macos.zip" >> SHA2-256SUMS
         echo "${{ needs.build_unix.outputs.sha512_bin }}  yt-dlp" >> SHA2-512SUMS
         echo "${{ needs.build_unix.outputs.sha512_tar }}  yt-dlp.tar.gz" >> SHA2-512SUMS
+        echo "${{ needs.build_unix.outputs.sha512_linux }}  yt-dlp_linux" >> SHA2-512SUMS
+        echo "${{ needs.build_unix.outputs.sha512_linux_zip }}  yt-dlp_linux.zip" >> SHA2-512SUMS
         echo "${{ needs.build_windows.outputs.sha512_win }}  yt-dlp.exe" >> SHA2-512SUMS
         echo "${{ needs.build_windows.outputs.sha512_py2exe }}  yt-dlp_min.exe" >> SHA2-512SUMS
         echo "${{ needs.build_windows32.outputs.sha512_win32 }}  yt-dlp_x86.exe" >> SHA2-512SUMS

README.md

@@ -248,6 +248,8 @@ File|Description
 :---|:---
 [yt-dlp_x86.exe](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_x86.exe)|Windows (Vista SP2+) standalone x86 (32-bit) binary
 [yt-dlp_min.exe](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_min.exe)|Windows (Win7 SP1+) standalone x64 binary built with `py2exe`<br/> ([Not recommended](#standalone-py2exe-builds-windows))
+[yt-dlp_linux](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_linux)|UNIX standalone x64 binary
+[yt-dlp_linux.zip](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_linux.zip)|Unpackaged Unix executable (no auto-update)
 [yt-dlp_win.zip](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_win.zip)|Unpackaged Windows executable (no auto-update)
 [yt-dlp_macos.zip](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_macos.zip)|Unpackaged MacOS (10.15+) executable (no auto-update)
 

pyinst.py

@@ -5,7 +5,6 @@ import sys
 
 from PyInstaller.__main__ import run as run_pyinstaller
 
-
 OS_NAME, ARCH = sys.platform, platform.architecture()[0][:2]
 
 
@@ -64,7 +63,7 @@ def exe(onedir):
     """@returns (name, path)"""
     name = '_'.join(filter(None, (
         'yt-dlp',
-        OS_NAME == 'darwin' and 'macos',
+        {'win32': '', 'darwin': 'macos'}.get(OS_NAME, OS_NAME),
         ARCH == '32' and 'x86'
     )))
     return name, ''.join(filter(None, (

yt_dlp/__init__.py

@@ -26,7 +26,7 @@ from .postprocessor import (
     MetadataFromFieldPP,
     MetadataParserPP,
 )
-from .update import run_update
+from .update import Updater
 from .utils import (
     NO_DEFAULT,
     POSTPROCESS_WHEN,
@@ -879,17 +879,23 @@ def _real_main(argv=None):
         return
 
     with YoutubeDL(ydl_opts) as ydl:
+        pre_process = opts.update_self or opts.rm_cachedir
         actual_use = all_urls or opts.load_info_filename
 
         if opts.rm_cachedir:
             ydl.cache.remove()
 
-        if opts.update_self and run_update(ydl) and actual_use:
-            # If updater returns True, exit. Required for windows
-            return 100, 'ERROR: The program must exit for the update to complete'
+        updater = Updater(ydl)
+        if opts.update_self and updater.update() and actual_use:
+            if updater.cmd:
+                return updater.restart()
+            # This code is reachable only for zip variant in py < 3.10
+            # It makes sense to exit here, but the old behavior is to continue
+            ydl.report_warning('Restart yt-dlp to use the updated version')
+            # return 100, 'ERROR: The program must exit for the update to complete'
 
         if not actual_use:
-            if opts.update_self or opts.rm_cachedir:
+            if pre_process:
                 return ydl._download_retcode
 
             ydl.warn_if_short_id(sys.argv[1:] if argv is None else argv)

yt_dlp/compat/__init__.py

@@ -55,7 +55,7 @@ if compat_os_name == 'nt' and sys.version_info < (3, 8):
     def compat_realpath(path):
         while os.path.islink(path):
             path = os.path.abspath(os.readlink(path))
-        return path
+        return os.path.realpath(path)
 else:
     compat_realpath = os.path.realpath
 

yt_dlp/jsinterp.py

@@ -6,22 +6,19 @@ import re
 
 from .utils import ExtractorError, remove_quotes
 
-_OPERATORS = [
-    ('|', operator.or_),
-    ('^', operator.xor),
-    ('&', operator.and_),
-    ('>>', operator.rshift),
-    ('<<', operator.lshift),
-    ('-', operator.sub),
-    ('+', operator.add),
-    ('%', operator.mod),
-    ('/', operator.truediv),
-    ('*', operator.mul),
-]
-_ASSIGN_OPERATORS = [(op + '=', opfunc) for op, opfunc in _OPERATORS]
-_ASSIGN_OPERATORS.append(('=', (lambda cur, right: right)))
-
-_NAME_RE = r'[a-zA-Z_$][a-zA-Z_$0-9]*'
+_NAME_RE = r'[a-zA-Z_$][\w$]*'
+_OPERATORS = {
+    '|': operator.or_,
+    '^': operator.xor,
+    '&': operator.and_,
+    '>>': operator.rshift,
+    '<<': operator.lshift,
+    '-': operator.sub,
+    '+': operator.add,
+    '%': operator.mod,
+    '/': operator.truediv,
+    '*': operator.mul,
+}
 
 _MATCHING_PARENS = dict(zip('({[', ')}]'))
 _QUOTES = '\'"'
@@ -50,13 +47,11 @@ class LocalNameSpace(collections.ChainMap):
 
 
 class JSInterpreter:
+    __named_object_counter = 0
+
     def __init__(self, code, objects=None):
-        if objects is None:
-            objects = {}
-        self.code = code
-        self._functions = {}
-        self._objects = objects
-        self.__named_object_counter = 0
+        self.code, self._functions = code, {}
+        self._objects = {} if objects is None else objects
 
     def _named_object(self, namespace, obj):
         self.__named_object_counter += 1
@@ -93,9 +88,9 @@ class JSInterpreter:
                 break
         yield expr[start:]
 
-    @staticmethod
-    def _separate_at_paren(expr, delim):
-        separated = list(JSInterpreter._separate(expr, delim, 1))
+    @classmethod
+    def _separate_at_paren(cls, expr, delim):
+        separated = list(cls._separate(expr, delim, 1))
         if len(separated) < 2:
             raise ExtractorError(f'No terminating paren {delim} in {expr}')
         return separated[0][1:].strip(), separated[1].strip()
@@ -104,33 +99,29 @@ class JSInterpreter:
         if allow_recursion < 0:
             raise ExtractorError('Recursion limit reached')
 
-        sub_statements = list(self._separate(stmt, ';'))
-        stmt = (sub_statements or ['']).pop()
+        should_abort = False
+        sub_statements = list(self._separate(stmt, ';')) or ['']
+        stmt = sub_statements.pop().lstrip()
+
         for sub_stmt in sub_statements:
             ret, should_abort = self.interpret_statement(sub_stmt, local_vars, allow_recursion - 1)
             if should_abort:
-                return ret
+                return ret, should_abort
 
-        should_abort = False
-        stmt = stmt.lstrip()
-        stmt_m = re.match(r'var\s', stmt)
-        if stmt_m:
-            expr = stmt[len(stmt_m.group(0)):]
+        m = re.match(r'(?P<var>var\s)|return(?:\s+|$)', stmt)
+        if not m:  # Try interpreting it as an expression
+            expr = stmt
+        elif m.group('var'):
+            expr = stmt[len(m.group(0)):]
         else:
-            return_m = re.match(r'return(?:\s+|$)', stmt)
-            if return_m:
-                expr = stmt[len(return_m.group(0)):]
-                should_abort = True
-            else:
-                # Try interpreting it as an expression
-                expr = stmt
+            expr = stmt[len(m.group(0)):]
+            should_abort = True
 
-        v = self.interpret_expression(expr, local_vars, allow_recursion)
-        return v, should_abort
+        return self.interpret_expression(expr, local_vars, allow_recursion), should_abort
 
     def interpret_expression(self, expr, local_vars, allow_recursion):
         expr = expr.strip()
-        if expr == '':  # Empty expression
+        if not expr:
             return None
 
         if expr.startswith('{'):
@@ -156,8 +147,8 @@ class JSInterpreter:
                 for item in self._separate(inner)])
             expr = name + outer
 
-        m = re.match(r'try\s*', expr)
-        if m:
+        m = re.match(r'(?P<try>try)\s*|(?:(?P<catch>catch)|(?P<for>for)|(?P<switch>switch))\s*\(', expr)
+        if m and m.group('try'):
             if expr[m.end()] == '{':
                 try_expr, expr = self._separate_at_paren(expr[m.end():], '}')
             else:
@@ -167,21 +158,19 @@ class JSInterpreter:
                 return ret
             return self.interpret_statement(expr, local_vars, allow_recursion - 1)[0]
 
-        m = re.match(r'catch\s*\(', expr)
-        if m:
+        elif m and m.group('catch'):
             # We ignore the catch block
             _, expr = self._separate_at_paren(expr, '}')
             return self.interpret_statement(expr, local_vars, allow_recursion - 1)[0]
 
-        m = re.match(r'for\s*\(', expr)
-        if m:
+        elif m and m.group('for'):
             constructor, remaining = self._separate_at_paren(expr[m.end() - 1:], ')')
             if remaining.startswith('{'):
                 body, expr = self._separate_at_paren(remaining, '}')
             else:
-                m = re.match(r'switch\s*\(', remaining)  # FIXME
-                if m:
-                    switch_val, remaining = self._separate_at_paren(remaining[m.end() - 1:], ')')
+                switch_m = re.match(r'switch\s*\(', remaining)  # FIXME
+                if switch_m:
+                    switch_val, remaining = self._separate_at_paren(remaining[switch_m.end() - 1:], ')')
                     body, expr = self._separate_at_paren(remaining, '}')
                     body = 'switch(%s){%s}' % (switch_val, body)
                 else:
@@ -206,8 +195,7 @@ class JSInterpreter:
                         f'Premature return in the initialization of a for loop in {constructor!r}')
             return self.interpret_statement(expr, local_vars, allow_recursion - 1)[0]
 
-        m = re.match(r'switch\s*\(', expr)
-        if m:
+        elif m and m.group('switch'):
             switch_val, remaining = self._separate_at_paren(expr[m.end() - 1:], ')')
             switch_val = self.interpret_expression(switch_val, local_vars, allow_recursion)
             body, expr = self._separate_at_paren(remaining, '}')
@@ -250,55 +238,63 @@ class JSInterpreter:
                 ret = local_vars[var]
             expr = expr[:start] + json.dumps(ret) + expr[end:]
 
-        for op, opfunc in _ASSIGN_OPERATORS:
-            m = re.match(rf'''(?x)
-                (?P<out>{_NAME_RE})(?:\[(?P<index>[^\]]+?)\])?
-                \s*{re.escape(op)}
-                (?P<expr>.*)$''', expr)
-            if not m:
-                continue
-            right_val = self.interpret_expression(m.group('expr'), local_vars, allow_recursion)
+        if not expr:
+            return None
 
-            if m.groupdict().get('index'):
-                lvar = local_vars[m.group('out')]
-                idx = self.interpret_expression(m.group('index'), local_vars, allow_recursion)
-                if not isinstance(idx, int):
-                    raise ExtractorError(f'List indices must be integers: {idx}')
-                cur = lvar[idx]
-                val = opfunc(cur, right_val)
-                lvar[idx] = val
-                return val
+        m = re.match(fr'''(?x)
+            (?P<assign>
+                (?P<out>{_NAME_RE})(?:\[(?P<index>[^\]]+?)\])?\s*
+                (?P<op>{"|".join(map(re.escape, _OPERATORS))})?
+                =(?P<expr>.*)$
+            )|(?P<return>
+                (?!if|return|true|false|null)(?P<name>{_NAME_RE})$
+            )|(?P<indexing>
+                (?P<in>{_NAME_RE})\[(?P<idx>.+)\]$
+            )|(?P<attribute>
+                (?P<var>{_NAME_RE})(?:\.(?P<member>[^(]+)|\[(?P<member2>[^\]]+)\])\s*
+            )|(?P<function>
+                (?P<fname>{_NAME_RE})\((?P<args>[\w$,]*)\)$
+            )''', expr)
+        if m and m.group('assign'):
+            if not m.group('op'):
+                opfunc = lambda curr, right: right
             else:
-                cur = local_vars.get(m.group('out'))
-                val = opfunc(cur, right_val)
-                local_vars[m.group('out')] = val
-                return val
+                opfunc = _OPERATORS[m.group('op')]
+            right_val = self.interpret_expression(m.group('expr'), local_vars, allow_recursion)
+            left_val = local_vars.get(m.group('out'))
 
-        if expr.isdigit():
+            if not m.group('index'):
+                local_vars[m.group('out')] = opfunc(left_val, right_val)
+                return local_vars[m.group('out')]
+            elif left_val is None:
+                raise ExtractorError(f'Cannot index undefined variable: {m.group("out")}')
+
+            idx = self.interpret_expression(m.group('index'), local_vars, allow_recursion)
+            if not isinstance(idx, int):
+                raise ExtractorError(f'List indices must be integers: {idx}')
+            left_val[idx] = opfunc(left_val[idx], right_val)
+            return left_val[idx]
+
+        elif expr.isdigit():
             return int(expr)
 
-        if expr == 'break':
+        elif expr == 'break':
             raise JS_Break()
         elif expr == 'continue':
             raise JS_Continue()
 
-        var_m = re.match(
-            r'(?!if|return|true|false|null)(?P<name>%s)$' % _NAME_RE,
-            expr)
-        if var_m:
-            return local_vars[var_m.group('name')]
+        elif m and m.group('return'):
+            return local_vars[m.group('name')]
 
         with contextlib.suppress(ValueError):
             return json.loads(expr)
 
-        m = re.match(
-            r'(?P<in>%s)\[(?P<idx>.+)\]$' % _NAME_RE, expr)
-        if m:
+        if m and m.group('indexing'):
             val = local_vars[m.group('in')]
             idx = self.interpret_expression(m.group('idx'), local_vars, allow_recursion)
             return val[idx]
 
-        for op, opfunc in _OPERATORS:
+        for op, opfunc in _OPERATORS.items():
             separated = list(self._separate(expr, op))
             if len(separated) < 2:
                 continue
@@ -314,10 +310,7 @@ class JSInterpreter:
                 raise ExtractorError(f'Premature right-side return of {op} in {expr!r}')
             return opfunc(left_val or 0, right_val)
 
-        m = re.match(
-            r'(?P<var>%s)(?:\.(?P<member>[^(]+)|\[(?P<member2>[^]]+)\])\s*' % _NAME_RE,
-            expr)
-        if m:
+        if m and m.group('attribute'):
             variable = m.group('var')
             member = remove_quotes(m.group('member') or m.group('member2'))
             arg_str = expr[m.end():]
@@ -332,7 +325,6 @@ class JSInterpreter:
                     raise ExtractorError(f'{member} {msg}: {expr}')
 
             def eval_method():
-                nonlocal member
                 if variable == 'String':
                     obj = str
                 elif variable in local_vars:
@@ -342,8 +334,8 @@ class JSInterpreter:
                         self._objects[variable] = self.extract_object(variable)
                     obj = self._objects[variable]
 
+                # Member access
                 if arg_str is None:
-                    # Member access
                     if member == 'length':
                         return len(obj)
                     return obj[member]
@@ -418,9 +410,7 @@ class JSInterpreter:
                     except ValueError:
                         return -1
 
-                if isinstance(obj, list):
-                    member = int(member)
-                return obj[member](argvals)
+                return obj[int(member) if isinstance(obj, list) else member](argvals)
 
             if remaining:
                 return self.interpret_expression(
@@ -429,9 +419,8 @@ class JSInterpreter:
             else:
                 return eval_method()
 
-        m = re.match(r'^(?P<func>%s)\((?P<args>[a-zA-Z0-9_$,]*)\)$' % _NAME_RE, expr)
-        if m:
-            fname = m.group('func')
+        elif m and m.group('function'):
+            fname = m.group('fname')
             argvals = tuple(
                 int(v) if v.isdigit() else local_vars[v]
                 for v in self._separate(m.group('args')))
@@ -441,8 +430,7 @@ class JSInterpreter:
                 self._functions[fname] = self.extract_function(fname)
             return self._functions[fname](argvals)
 
-        if expr:
-            raise ExtractorError('Unsupported JS expression %r' % expr)
+        raise ExtractorError(f'Unsupported JS expression {expr!r}')
 
     def extract_object(self, objname):
         _FUNC_NAME_RE = r'''(?:[a-zA-Z$0-9]+|"[a-zA-Z$0-9]+"|'[a-zA-Z$0-9]+')'''
@@ -471,14 +459,17 @@ class JSInterpreter:
         """ @returns argnames, code """
         func_m = re.search(
             r'''(?x)
-                (?:function\s+%s|[{;,]\s*%s\s*=\s*function|var\s+%s\s*=\s*function)\s*
+                (?:
+                    function\s+%(name)s|
+                    [{;,]\s*%(name)s\s*=\s*function|
+                    var\s+%(name)s\s*=\s*function
+                )\s*
                 \((?P<args>[^)]*)\)\s*
-                (?P<code>\{(?:(?!};)[^"]|"([^"]|\\")*")+\})''' % (
-                re.escape(funcname), re.escape(funcname), re.escape(funcname)),
+                (?P<code>{(?:(?!};)[^"]|"([^"]|\\")*")+})''' % {'name': re.escape(funcname)},
             self.code)
         code, _ = self._separate_at_paren(func_m.group('code'), '}')  # refine the match
         if func_m is None:
-            raise ExtractorError('Could not find JS function %r' % funcname)
+            raise ExtractorError(f'Could not find JS function "{funcname}"')
         return func_m.group('args').split(','), code
 
     def extract_function(self, funcname):
@@ -492,11 +483,9 @@ class JSInterpreter:
                 break
             start, body_start = mobj.span()
             body, remaining = self._separate_at_paren(code[body_start - 1:], '}')
-            name = self._named_object(
-                local_vars,
-                self.extract_function_from_code(
-                    [str.strip(x) for x in mobj.group('args').split(',')],
-                    body, local_vars, *global_stack))
+            name = self._named_object(local_vars, self.extract_function_from_code(
+                [x.strip() for x in mobj.group('args').split(',')],
+                body, local_vars, *global_stack))
             code = code[:start] + name + remaining
         return self.build_function(argnames, code, local_vars, *global_stack)
 

yt_dlp/update.py

@@ -1,3 +1,4 @@
+import atexit
 import hashlib
 import json
 import os
@@ -8,11 +9,11 @@ from zipimport import zipimporter
 
 from .compat import functools  # isort: split
 from .compat import compat_realpath
-from .utils import Popen, traverse_obj, version_tuple
+from .utils import Popen, shell_quote, traverse_obj, version_tuple
 from .version import __version__
 
-
-RELEASE_JSON_URL = 'https://api.github.com/repos/yt-dlp/yt-dlp/releases/latest'
+REPOSITORY = 'yt-dlp/yt-dlp'
+API_URL = f'https://api.github.com/repos/{REPOSITORY}/releases/latest'
 
 
 @functools.cache
@@ -44,12 +45,13 @@ _FILE_SUFFIXES = {
     'py2exe': '_min.exe',
     'win32_exe': '.exe',
     'darwin_exe': '_macos',
+    'linux_exe': '_linux',
 }
 
 _NON_UPDATEABLE_REASONS = {
     **{variant: None for variant in _FILE_SUFFIXES},  # Updatable
     **{variant: f'Auto-update is not supported for unpackaged {name} executable; Re-download the latest release'
-       for variant, name in {'win32_dir': 'Windows', 'darwin_dir': 'MacOS'}.items()},
+       for variant, name in {'win32_dir': 'Windows', 'darwin_dir': 'MacOS', 'linux_dir': 'Linux'}.items()},
     'source': 'You cannot update when running from source code; Use git to pull the latest changes',
     'unknown': 'It looks like you installed yt-dlp with a package manager, pip or setup.py; Use that to update',
     'other': 'It looks like you are using an unofficial build of yt-dlp; Build the executable again',
@@ -60,137 +62,187 @@ def is_non_updateable():
     return _NON_UPDATEABLE_REASONS.get(detect_variant(), _NON_UPDATEABLE_REASONS['other'])
 
 
-def run_update(ydl):
-    """
-    Update the program file with the latest version from the repository
-    Returns whether the program should terminate
-    """
+def _sha256_file(path):
+    h = hashlib.sha256()
+    mv = memoryview(bytearray(128 * 1024))
+    with open(os.path.realpath(path), 'rb', buffering=0) as f:
+        for n in iter(lambda: f.readinto(mv), 0):
+            h.update(mv[:n])
+    return h.hexdigest()
 
-    def report_error(msg, expected=False):
-        ydl.report_error(msg, tb=False if expected else None)
 
-    def report_unable(action, expected=False):
-        report_error(f'Unable to {action}', expected)
+class Updater:
+    def __init__(self, ydl):
+        self.ydl = ydl
 
-    def report_permission_error(file):
-        report_unable(f'write to {file}; Try running as administrator', True)
+    @functools.cached_property
+    def _new_version_info(self):
+        self.ydl.write_debug(f'Fetching release info: {API_URL}')
+        return json.loads(self.ydl.urlopen(API_URL).read().decode())
 
-    def report_network_error(action, delim=';'):
-        report_unable(f'{action}{delim} Visit  https://github.com/yt-dlp/yt-dlp/releases/latest', True)
+    @property
+    def current_version(self):
+        """Current version"""
+        return __version__
 
-    def calc_sha256sum(path):
-        h = hashlib.sha256()
-        mv = memoryview(bytearray(128 * 1024))
-        with open(os.path.realpath(path), 'rb', buffering=0) as f:
-            for n in iter(lambda: f.readinto(mv), 0):
-                h.update(mv[:n])
-        return h.hexdigest()
+    @property
+    def new_version(self):
+        """Version of the latest release"""
+        return self._new_version_info['tag_name']
 
-    try:
-        version_info = json.loads(ydl.urlopen(RELEASE_JSON_URL).read().decode())
-    except Exception:
-        return report_network_error('obtain version info', delim='; Please try again later or')
+    @property
+    def has_update(self):
+        """Whether there is an update available"""
+        return version_tuple(__version__) < version_tuple(self.new_version)
 
-    version_id = version_info['tag_name']
-    ydl.to_screen(f'Latest version: {version_id}, Current version: {__version__}')
-    if version_tuple(__version__) >= version_tuple(version_id):
-        ydl.to_screen(f'yt-dlp is up to date ({__version__})')
-        return
+    @functools.cached_property
+    def filename(self):
+        """Filename of the executable"""
+        return compat_realpath(_get_variant_and_executable_path()[1])
 
-    err = is_non_updateable()
-    if err:
-        return report_error(err, True)
-
-    variant, filename = _get_variant_and_executable_path()
-    filename = compat_realpath(filename)  # Absolute path, following symlinks
-
-    label = _FILE_SUFFIXES[variant]
-    if label and platform.architecture()[0][:2] == '32':
-        label = f'_x86{label}'
-    release_name = f'yt-dlp{label}'
-
-    ydl.to_screen(f'Current Build Hash {calc_sha256sum(filename)}')
-    ydl.to_screen(f'Updating to version {version_id} ...')
-
-    def get_file(name, fatal=True):
-        error = report_network_error if fatal else lambda _: None
-        url = traverse_obj(
-            version_info, ('assets', lambda _, v: v['name'] == name, 'browser_download_url'), get_all=False)
+    def _download(self, name=None):
+        name = name or self.release_name
+        url = traverse_obj(self._new_version_info, (
+            'assets', lambda _, v: v['name'] == name, 'browser_download_url'), get_all=False)
         if not url:
-            return error('fetch updates')
-        try:
-            return ydl.urlopen(url).read()
-        except OSError:
-            return error('download latest version')
+            raise Exception('Unable to find download URL')
+        self.ydl.write_debug(f'Downloading {name} from {url}')
+        return self.ydl.urlopen(url).read()
 
-    def verify(content):
-        if not content:
-            return False
-        hash_data = get_file('SHA2-256SUMS', fatal=False) or b''
-        expected = dict(ln.split()[::-1] for ln in hash_data.decode().splitlines()).get(release_name)
-        if not expected:
-            ydl.report_warning('no hash information found for the release')
-        elif hashlib.sha256(content).hexdigest() != expected:
-            return report_network_error('verify the new executable')
+    @functools.cached_property
+    def release_name(self):
+        """The release filename"""
+        label = _FILE_SUFFIXES[detect_variant()]
+        if label and platform.architecture()[0][:2] == '32':
+            label = f'_x86{label}'
+        return f'yt-dlp{label}'
+
+    @functools.cached_property
+    def release_hash(self):
+        """Hash of the latest release"""
+        hash_data = dict(ln.split()[::-1] for ln in self._download('SHA2-256SUMS').decode().splitlines())
+        return hash_data[self.release_name]
+
+    def _report_error(self, msg, expected=False):
+        self.ydl.report_error(msg, tb=False if expected else None)
+
+    def _report_permission_error(self, file):
+        self._report_error(f'Unable to write to {file}; Try running as administrator', True)
+
+    def _report_network_error(self, action, delim=';'):
+        self._report_error(f'Unable to {action}{delim} Visit  https://github.com/{REPOSITORY}/releases/latest', True)
+
+    def check_update(self):
+        """Report whether there is an update available"""
+        try:
+            self.ydl.to_screen(
+                f'Latest version: {self.new_version}, Current version: {self.current_version}')
+        except Exception:
+            return self._report_network_error('obtain version info', delim='; Please try again later or')
+
+        if not self.has_update:
+            return self.ydl.to_screen(f'yt-dlp is up to date ({__version__})')
+
+        if not is_non_updateable():
+            self.ydl.to_screen(f'Current Build Hash {_sha256_file(self.filename)}')
         return True
 
-    directory = os.path.dirname(filename)
-    if not os.access(filename, os.W_OK):
-        return report_permission_error(filename)
-    elif not os.access(directory, os.W_OK):
-        return report_permission_error(directory)
+    def update(self):
+        """Update yt-dlp executable to the latest version"""
+        if not self.check_update():
+            return
+        err = is_non_updateable()
+        if err:
+            return self._report_error(err, True)
+        self.ydl.to_screen(f'Updating to version {self.new_version} ...')
 
-    new_filename, old_filename = f'{filename}.new', f'{filename}.old'
-    if variant == 'zip':  # Can be replaced in-place
-        new_filename, old_filename = filename, None
+        directory = os.path.dirname(self.filename)
+        if not os.access(self.filename, os.W_OK):
+            return self._report_permission_error(self.filename)
+        elif not os.access(directory, os.W_OK):
+            return self._report_permission_error(directory)
 
-    try:
-        if os.path.exists(old_filename or ''):
-            os.remove(old_filename)
-    except OSError:
-        return report_unable('remove the old version')
+        new_filename, old_filename = f'{self.filename}.new', f'{self.filename}.old'
+        if detect_variant() == 'zip':  # Can be replaced in-place
+            new_filename, old_filename = self.filename, None
 
-    newcontent = get_file(release_name)
-    if not verify(newcontent):
-        return
-    try:
-        with open(new_filename, 'wb') as outf:
-            outf.write(newcontent)
-    except OSError:
-        return report_permission_error(new_filename)
+        try:
+            if os.path.exists(old_filename or ''):
+                os.remove(old_filename)
+        except OSError:
+            return self._report_error('Unable to remove the old version')
 
-    try:
-        if old_filename:
-            os.rename(filename, old_filename)
-    except OSError:
-        return report_unable('move current version')
-    try:
-        if old_filename:
-            os.rename(new_filename, filename)
-    except OSError:
-        report_unable('overwrite current version')
-        os.rename(old_filename, filename)
-        return
+        try:
+            newcontent = self._download()
+        except OSError:
+            return self._report_network_error('download latest version')
+        except Exception:
+            return self._report_network_error('fetch updates')
 
-    if variant not in ('win32_exe', 'py2exe'):
-        if old_filename:
-            os.remove(old_filename)
-        ydl.to_screen(f'Updated yt-dlp to version {version_id}; Restart yt-dlp to use the new version')
-        return
+        try:
+            expected_hash = self.release_hash
+        except Exception:
+            self.ydl.report_warning('no hash information found for the release')
+        else:
+            if hashlib.sha256(newcontent).hexdigest() != expected_hash:
+                return self._report_network_error('verify the new executable')
 
-    try:
-        # Continues to run in the background
-        Popen(f'ping 127.0.0.1 -n 5 -w 1000 & del /F "{old_filename}"',
-              shell=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
-        ydl.to_screen(f'Updated yt-dlp to version {version_id}')
-        return True  # Exit app
-    except OSError:
-        report_unable('delete the old version')
+        try:
+            with open(new_filename, 'wb') as outf:
+                outf.write(newcontent)
+        except OSError:
+            return self._report_permission_error(new_filename)
+
+        try:
+            if old_filename:
+                os.rename(self.filename, old_filename)
+        except OSError:
+            return self._report_error('Unable to move current version')
+        try:
+            if old_filename:
+                os.rename(new_filename, self.filename)
+        except OSError:
+            self._report_error('Unable to overwrite current version')
+            return os.rename(old_filename, self.filename)
+
+        if detect_variant() not in ('win32_exe', 'py2exe'):
+            if old_filename:
+                os.remove(old_filename)
+        else:
+            atexit.register(Popen, f'ping 127.0.0.1 -n 5 -w 1000 & del /F "{old_filename}"',
+                            shell=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+
+        self.ydl.to_screen(f'Updated yt-dlp to version {self.new_version}')
+        return True
+
+    @functools.cached_property
+    def cmd(self):
+        """The command-line to run the executable, if known"""
+        # There is no sys.orig_argv in py < 3.10. Also, it can be [] when frozen
+        if getattr(sys, 'orig_argv', None):
+            return sys.orig_argv
+        elif hasattr(sys, 'frozen'):
+            return sys.argv
+
+    def restart(self):
+        """Restart the executable"""
+        assert self.cmd, 'Must be frozen or Py >= 3.10'
+        self.ydl.write_debug(f'Restarting: {shell_quote(self.cmd)}')
+        _, _, returncode = Popen.run(self.cmd)
+        return returncode
+
+
+def run_update(ydl):
+    """Update the program file with the latest version from the repository
+    @returns    Whether there was a successfull update (No update = False)
+    """
+    return Updater(ydl).update()
 
 
 # Deprecated
 def update_self(to_screen, verbose, opener):
     import traceback
+
     from .utils import write_string
 
     write_string(
@@ -202,12 +254,10 @@ def update_self(to_screen, verbose, opener):
     class FakeYDL():
         to_screen = printfn
 
-        @staticmethod
-        def report_warning(msg, *args, **kwargs):
+        def report_warning(self, msg, *args, **kwargs):
             return printfn(f'WARNING: {msg}', *args, **kwargs)
 
-        @staticmethod
-        def report_error(msg, tb=None):
+        def report_error(self, msg, tb=None):
             printfn(f'ERROR: {msg}')
             if not verbose:
                 return
@@ -224,6 +274,9 @@ def update_self(to_screen, verbose, opener):
             if tb:
                 printfn(tb)
 
+        def write_debug(self, msg, *args, **kwargs):
+            printfn(f'[debug] {msg}', *args, **kwargs)
+
         def urlopen(self, url):
             return opener.open(url)
 

yt_dlp/utils.py

@@ -991,9 +991,10 @@ def make_HTTPS_handler(params, **kwargs):
 
 
 def bug_reports_message(before=';'):
-    msg = ('please report this issue on  https://github.com/yt-dlp/yt-dlp/issues?q= , '
-           'filling out the appropriate issue template. '
-           'Confirm you are on the latest version using  yt-dlp -U')
+    from .update import REPOSITORY
+
+    msg = (f'please report this issue on  https://github.com/{REPOSITORY}/issues?q= , '
+           'filling out the appropriate issue template. Confirm you are on the latest version using  yt-dlp -U')
 
     before = before.rstrip()
     if not before or before.endswith(('.', '!', '?')):