fix(dl): delete old file after repackage (#114)

* fix(dl): delete old file after repackage

* fix(dl): using original_path instead of self.path in repackage method

.gitignore

@@ -1,4 +1,6 @@
 # devine
+devine.yaml
+devine.yml
 *.mkv
 *.mp4
 *.exe

CHANGELOG.md

@@ -7,6 +7,25 @@ This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm
 Versions [3.0.0] and older use a format based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 but versions thereafter use a custom changelog format using [git-cliff](https://git-cliff.org).
 
+## [3.3.3] - 2024-05-07
+
+### Bug Fixes
+
+- *dl*: Automatically convert TTML Subs to WebVTT for MKV support
+- *Subtitle*: Correct timestamps when merging fragmented WebVTT
+
+### Changes
+
+- *env*: List all directories as table in info
+- *env*: List possible config path locations when not found
+- *binaries*: Move all binary definitions to core/binaries file
+- *curl-impersonate*: Remove manual fix for curl proxy SSL
+- *curl-impersonate*: Update the default browser to chrome124
+- *Config*: Move possible config paths out of func to constant
+- *utilities*: Remove get_binary_path, use binaries.find instead
+- *dl*: Improve readability of download worker errors
+- *env*: Shorten paths on Windows with env vars
+
 ## [3.3.2] - 2024-04-16
 
 ### Bug Fixes
@@ -795,6 +814,7 @@ This release brings a huge change to the fundamentals of Devine's logging, UI, a
 
 Initial public release under the name Devine.
 
+[3.3.3]: https://github.com/devine-dl/devine/releases/tag/v3.3.3
 [3.3.2]: https://github.com/devine-dl/devine/releases/tag/v3.3.2
 [3.3.1]: https://github.com/devine-dl/devine/releases/tag/v3.3.1
 [3.3.0]: https://github.com/devine-dl/devine/releases/tag/v3.3.0

README.md

@@ -343,6 +343,8 @@ Please refrain from spam or asking for questions that infringe upon a Service's
 <a href="https://github.com/Shivelight"><img src="https://images.weserv.nl/?url=avatars.githubusercontent.com/u/20620780?v=4&h=25&w=25&fit=cover&mask=circle&maxage=7d" alt="Shivelight"/></a>
 <a href="https://github.com/knowhere01"><img src="https://images.weserv.nl/?url=avatars.githubusercontent.com/u/113712042?v=4&h=25&w=25&fit=cover&mask=circle&maxage=7d" alt="knowhere01"/></a>
 <a href="https://github.com/retouching"><img src="https://images.weserv.nl/?url=avatars.githubusercontent.com/u/33735357?v=4&h=25&w=25&fit=cover&mask=circle&maxage=7d" alt="retouching"/></a>
+<a href="https://github.com/pandamoon21"><img src="https://images.weserv.nl/?url=avatars.githubusercontent.com/u/33972938?v=4&h=25&w=25&fit=cover&mask=circle&maxage=7d" alt="pandamoon21"/></a>
+<a href="https://github.com/adbbbb"><img src="https://images.weserv.nl/?url=avatars.githubusercontent.com/u/56319336?v=4&h=25&w=25&fit=cover&mask=circle&maxage=7d" alt="adbbbb"/></a>
 
 ## Licensing
 

devine/commands/dl.py

@@ -38,6 +38,7 @@ from rich.table import Table
 from rich.text import Text
 from rich.tree import Tree
 
+from devine.core import binaries
 from devine.core.config import config
 from devine.core.console import console
 from devine.core.constants import DOWNLOAD_LICENCE_ONLY, AnyTrack, context_settings
@@ -51,7 +52,7 @@ from devine.core.titles import Movie, Song, Title_T
 from devine.core.titles.episode import Episode
 from devine.core.tracks import Audio, Subtitle, Tracks, Video
 from devine.core.tracks.attachment import Attachment
-from devine.core.utilities import get_binary_path, get_system_fonts, is_close_match, time_elapsed_since
+from devine.core.utilities import get_system_fonts, is_close_match, time_elapsed_since
 from devine.core.utils.click_types import LANGUAGE_RANGE, QUALITY_LIST, SEASON_RANGE, ContextData, MultipleChoice
 from devine.core.utils.collections import merge_dict
 from devine.core.utils.subprocess import ffprobe
@@ -177,6 +178,7 @@ class dl:
             except ValueError as e:
                 self.log.error(f"Failed to load Widevine CDM, {e}")
                 sys.exit(1)
+            if self.cdm:
                 self.log.info(
                     f"Loaded {self.cdm.__class__.__name__} Widevine CDM: {self.cdm.system_id} (L{self.cdm.security_level})"
                 )
@@ -198,7 +200,7 @@ class dl:
                     self.proxy_providers.append(Basic(**config.proxy_providers["basic"]))
                 if config.proxy_providers.get("nordvpn"):
                     self.proxy_providers.append(NordVPN(**config.proxy_providers["nordvpn"]))
-                if get_binary_path("hola-proxy"):
+                if binaries.HolaProxy:
                     self.proxy_providers.append(Hola())
                 for proxy_provider in self.proxy_providers:
                     self.log.info(f"Loaded {proxy_provider.__class__.__name__}: {proxy_provider}")
@@ -546,13 +548,16 @@ class dl:
             except Exception as e:  # noqa
                 error_messages = [
                     ":x: Download Failed...",
-                    "   One of the track downloads had an error!",
-                    "   See the error trace above for more information."
                 ]
-                if isinstance(e, subprocess.CalledProcessError):
-                    # ignore process exceptions as proper error logs are already shown
-                    error_messages.append(f"   Process exit code: {e.returncode}")
+                if isinstance(e, EnvironmentError):
+                    error_messages.append(f"   {e}")
                 else:
+                    error_messages.append("   An unexpected error occurred in one of the download workers.",)
+                    if hasattr(e, "returncode"):
+                        error_messages.append(f"   Binary call failed, Process exit code: {e.returncode}")
+                    error_messages.append("   See the error trace above for more information.")
+                    if isinstance(e, subprocess.CalledProcessError):
+                        # CalledProcessError already lists the exception trace
                         console.print_exception()
                 console.print(Padding(
                     Group(*error_messages),
@@ -610,11 +615,14 @@ class dl:
                             break
                     video_track_n += 1
 
-                if sub_format:
-                    with console.status(f"Converting Subtitles to {sub_format.name}..."):
+                with console.status("Converting Subtitles..."):
                     for subtitle in title.tracks.subtitles:
+                        if sub_format:
                             if subtitle.codec != sub_format:
                                 subtitle.convert(sub_format)
+                        elif subtitle.codec == Subtitle.Codec.TimedTextMarkupLang:
+                            # MKV does not support TTML, VTT is the next best option
+                            subtitle.convert(Subtitle.Codec.WebVTT)
 
                 with console.status("Checking Subtitles for Fonts..."):
                     font_names = []
@@ -694,16 +702,22 @@ class dl:
                     ):
                         for task_id, task_tracks in multiplex_tasks:
                             progress.start_task(task_id)  # TODO: Needed?
-                            muxed_path, return_code = task_tracks.mux(
+                            muxed_path, return_code, errors = task_tracks.mux(
                                 str(title),
                                 progress=partial(progress.update, task_id=task_id),
                                 delete=False
                             )
                             muxed_paths.append(muxed_path)
-                            if return_code == 1:
-                                self.log.warning("mkvmerge had at least one warning, will continue anyway...")
-                            elif return_code >= 2:
-                                self.log.error(f"Failed to Mux video to Matroska file ({return_code})")
+                            if return_code >= 2:
+                                self.log.error(f"Failed to Mux video to Matroska file ({return_code}):")
+                            elif return_code == 1 or errors:
+                                self.log.warning("mkvmerge had at least one warning or error, continuing anyway...")
+                            for line in errors:
+                                if line.startswith("#GUI#error"):
+                                    self.log.error(line)
+                                else:
+                                    self.log.warning(line)
+                            if return_code >= 2:
                                 sys.exit(1)
                             for video_track in task_tracks.videos:
                                 video_track.delete()
@@ -923,21 +937,21 @@ class dl:
                 return Credential.loads(credentials)  # type: ignore
 
     @staticmethod
-    def get_cdm(service: str, profile: Optional[str] = None) -> WidevineCdm:
+    def get_cdm(service: str, profile: Optional[str] = None) -> Optional[WidevineCdm]:
         """
         Get CDM for a specified service (either Local or Remote CDM).
         Raises a ValueError if there's a problem getting a CDM.
         """
         cdm_name = config.cdm.get(service) or config.cdm.get("default")
         if not cdm_name:
-            raise ValueError("A CDM to use wasn't listed in the config")
+            return None
 
         if isinstance(cdm_name, dict):
             if not profile:
-                raise ValueError("CDM config is mapped for profiles, but no profile was chosen")
+                return None
             cdm_name = cdm_name.get(profile) or config.cdm.get("default")
             if not cdm_name:
-                raise ValueError(f"A CDM to use was not mapped for the profile {profile}")
+                return None
 
         cdm_api = next(iter(x for x in config.remote_cdm if x["name"] == cdm_name), None)
         if cdm_api:

devine/commands/env.py

@@ -1,10 +1,17 @@
 import logging
+import os
 import shutil
+import sys
+from pathlib import Path
 from typing import Optional
 
 import click
+from rich.padding import Padding
+from rich.table import Table
+from rich.tree import Tree
 
-from devine.core.config import config, config_path
+from devine.core.config import POSSIBLE_CONFIG_PATHS, config, config_path
+from devine.core.console import console
 from devine.core.constants import context_settings
 from devine.core.services import Services
 
@@ -18,13 +25,42 @@ def env() -> None:
 def info() -> None:
     """Displays information about the current environment."""
     log = logging.getLogger("env")
-    log.info(f"[Config]     : {config_path or '--'}")
-    log.info(f"[Cookies]    : {config.directories.cookies}")
-    log.info(f"[WVDs]       : {config.directories.wvds}")
-    log.info(f"[Cache]      : {config.directories.cache}")
-    log.info(f"[Logs]       : {config.directories.logs}")
-    log.info(f"[Temp Files] : {config.directories.temp}")
-    log.info(f"[Downloads]  : {config.directories.downloads}")
+
+    if config_path:
+        log.info(f"Config loaded from {config_path}")
+    else:
+        tree = Tree("No config file found, you can use any of the following locations:")
+        for i, path in enumerate(POSSIBLE_CONFIG_PATHS, start=1):
+            tree.add(f"[repr.number]{i}.[/] [text2]{path.resolve()}[/]")
+        console.print(Padding(
+            tree,
+            (0, 5)
+        ))
+
+    table = Table(title="Directories", expand=True)
+    table.add_column("Name", no_wrap=True)
+    table.add_column("Path")
+
+    path_vars = {
+        x: Path(os.getenv(x))
+        for x in ("TEMP", "APPDATA", "LOCALAPPDATA", "USERPROFILE")
+        if sys.platform == "win32" and os.getenv(x)
+    }
+
+    for name in sorted(dir(config.directories)):
+        if name.startswith("__") or name == "app_dirs":
+            continue
+        path = getattr(config.directories, name).resolve()
+        for var, var_path in path_vars.items():
+            if path.is_relative_to(var_path):
+                path = rf"%{var}%\{path.relative_to(var_path)}"
+                break
+        table.add_row(name.title(), str(path))
+
+    console.print(Padding(
+        table,
+        (1, 5)
+    ))
 
 
 @env.group(name="clear", short_help="Clear an environment directory.", context_settings=context_settings)

devine/commands/search.py

@@ -12,13 +12,13 @@ from rich.rule import Rule
 from rich.tree import Tree
 
 from devine.commands.dl import dl
+from devine.core import binaries
 from devine.core.config import config
 from devine.core.console import console
 from devine.core.constants import context_settings
 from devine.core.proxies import Basic, Hola, NordVPN
 from devine.core.service import Service
 from devine.core.services import Services
-from devine.core.utilities import get_binary_path
 from devine.core.utils.click_types import ContextData
 from devine.core.utils.collections import merge_dict
 
@@ -72,7 +72,7 @@ def search(
                 proxy_providers.append(Basic(**config.proxy_providers["basic"]))
             if config.proxy_providers.get("nordvpn"):
                 proxy_providers.append(NordVPN(**config.proxy_providers["nordvpn"]))
-            if get_binary_path("hola-proxy"):
+            if binaries.HolaProxy:
                 proxy_providers.append(Hola())
             for proxy_provider in proxy_providers:
                 log.info(f"Loaded {proxy_provider.__class__.__name__}: {proxy_provider}")

devine/commands/serve.py

@@ -2,9 +2,9 @@ import subprocess
 
 import click
 
+from devine.core import binaries
 from devine.core.config import config
 from devine.core.constants import context_settings
-from devine.core.utilities import get_binary_path
 
 
 @click.command(
@@ -29,11 +29,10 @@ def serve(host: str, port: int, caddy: bool) -> None:
     from pywidevine import serve
 
     if caddy:
-        executable = get_binary_path("caddy")
-        if not executable:
+        if not binaries.Caddy:
             raise click.ClickException("Caddy executable \"caddy\" not found but is required for --caddy.")
         caddy_p = subprocess.Popen([
-            executable,
+            binaries.Caddy,
             "run",
             "--config", str(config.directories.user_configs / "Caddyfile")
         ])

devine/commands/util.py

@@ -4,8 +4,8 @@ from pathlib import Path
 import click
 from pymediainfo import MediaInfo
 
+from devine.core import binaries
 from devine.core.constants import context_settings
-from devine.core.utilities import get_binary_path
 
 
 @click.group(short_help="Various helper scripts and programs.", context_settings=context_settings)
@@ -38,8 +38,7 @@ def crop(path: Path, aspect: str, letter: bool, offset: int, preview: bool) -> N
     as it may go from being 2px away from a perfect crop, to 20px over-cropping
     again due to sub-sampled chroma.
     """
-    executable = get_binary_path("ffmpeg")
-    if not executable:
+    if not binaries.FFMPEG:
         raise click.ClickException("FFmpeg executable \"ffmpeg\" not found but is required.")
 
     if path.is_dir():
@@ -87,7 +86,7 @@ def crop(path: Path, aspect: str, letter: bool, offset: int, preview: bool) -> N
             ]))))]
 
         ffmpeg_call = subprocess.Popen([
-            executable, "-y",
+            binaries.FFMPEG, "-y",
             "-i", str(video_path),
             "-map", "0:v:0",
             "-c", "copy",
@@ -95,7 +94,7 @@ def crop(path: Path, aspect: str, letter: bool, offset: int, preview: bool) -> N
         ] + out_path, stdout=subprocess.PIPE)
         try:
             if preview:
-                previewer = get_binary_path("mpv", "ffplay")
+                previewer = binaries.MPV or binaries.FFPlay
                 if not previewer:
                     raise click.ClickException("MPV/FFplay executables weren't found but are required for previewing.")
                 subprocess.Popen((previewer, "-"), stdin=ffmpeg_call.stdout)
@@ -120,8 +119,7 @@ def range_(path: Path, full: bool, preview: bool) -> None:
     then you're video may have the range set to the wrong value. Flip its range to the
     opposite value and see if that fixes it.
     """
-    executable = get_binary_path("ffmpeg")
-    if not executable:
+    if not binaries.FFMPEG:
         raise click.ClickException("FFmpeg executable \"ffmpeg\" not found but is required.")
 
     if path.is_dir():
@@ -157,7 +155,7 @@ def range_(path: Path, full: bool, preview: bool) -> None:
             ]))))]
 
         ffmpeg_call = subprocess.Popen([
-            executable, "-y",
+            binaries.FFMPEG, "-y",
             "-i", str(video_path),
             "-map", "0:v:0",
             "-c", "copy",
@@ -165,7 +163,7 @@ def range_(path: Path, full: bool, preview: bool) -> None:
         ] + out_path, stdout=subprocess.PIPE)
         try:
             if preview:
-                previewer = get_binary_path("mpv", "ffplay")
+                previewer = binaries.MPV or binaries.FFPlay
                 if not previewer:
                     raise click.ClickException("MPV/FFplay executables weren't found but are required for previewing.")
                 subprocess.Popen((previewer, "-"), stdin=ffmpeg_call.stdout)
@@ -188,8 +186,7 @@ def test(path: Path, map_: str) -> None:
     You may choose specific streams using the -m/--map parameter. E.g.,
     '0:v:0' to test the first video stream, or '0:a' to test all audio streams.
     """
-    executable = get_binary_path("ffmpeg")
-    if not executable:
+    if not binaries.FFMPEG:
         raise click.ClickException("FFmpeg executable \"ffmpeg\" not found but is required.")
 
     if path.is_dir():
@@ -199,7 +196,7 @@ def test(path: Path, map_: str) -> None:
     for video_path in paths:
         print("Starting...")
         p = subprocess.Popen([
-            executable, "-hide_banner",
+            binaries.FFMPEG, "-hide_banner",
             "-benchmark",
             "-i", str(video_path),
             "-map", map_,

devine/core/__init__.py

@@ -1 +1 @@
-__version__ = "3.3.2"
+__version__ = "3.3.3"

devine/core/binaries.py

@@ -0,0 +1,46 @@
+import shutil
+import sys
+from pathlib import Path
+from typing import Optional
+
+__shaka_platform = {
+    "win32": "win",
+    "darwin": "osx"
+}.get(sys.platform, sys.platform)
+
+
+def find(*names: str) -> Optional[Path]:
+    """Find the path of the first found binary name."""
+    for name in names:
+        path = shutil.which(name)
+        if path:
+            return Path(path)
+    return None
+
+
+FFMPEG = find("ffmpeg")
+FFProbe = find("ffprobe")
+FFPlay = find("ffplay")
+SubtitleEdit = find("SubtitleEdit")
+ShakaPackager = find(
+    "shaka-packager",
+    "packager",
+    f"packager-{__shaka_platform}",
+    f"packager-{__shaka_platform}-arm64",
+    f"packager-{__shaka_platform}-x64"
+)
+Aria2 = find("aria2c", "aria2")
+CCExtractor = find(
+    "ccextractor",
+    "ccextractorwin",
+    "ccextractorwinfull"
+)
+HolaProxy = find("hola-proxy")
+MPV = find("mpv")
+Caddy = find("caddy")
+
+
+__all__ = (
+    "FFMPEG", "FFProbe", "FFPlay", "SubtitleEdit", "ShakaPackager",
+    "Aria2", "CCExtractor", "HolaProxy", "MPV", "Caddy", "find"
+)

devine/core/config.py

@@ -77,29 +77,27 @@ class Config:
         return cls(**yaml.safe_load(path.read_text(encoding="utf8")) or {})
 
 
+# noinspection PyProtectedMember
+POSSIBLE_CONFIG_PATHS = (
+    # The Devine Namespace Folder (e.g., %appdata%/Python/Python311/site-packages/devine)
+    Config._Directories.namespace_dir / Config._Filenames.root_config,
+    # The Parent Folder to the Devine Namespace Folder (e.g., %appdata%/Python/Python311/site-packages)
+    Config._Directories.namespace_dir.parent / Config._Filenames.root_config,
+    # The AppDirs User Config Folder (e.g., %localappdata%/devine)
+    Config._Directories.user_configs / Config._Filenames.root_config
+)
+
+
 def get_config_path() -> Optional[Path]:
     """
-    Get Path to Config from various locations.
-
-    Looks for a config file in the following folders in order:
-
-    1. The Devine Namespace Folder (e.g., %appdata%/Python/Python311/site-packages/devine)
-    2. The Parent Folder to the Devine Namespace Folder (e.g., %appdata%/Python/Python311/site-packages)
-    3. The AppDirs User Config Folder (e.g., %localappdata%/devine)
+    Get Path to Config from any one of the possible locations.
 
     Returns None if no config file could be found.
     """
-    # noinspection PyProtectedMember
-    path = Config._Directories.namespace_dir / Config._Filenames.root_config
-    if not path.exists():
-        # noinspection PyProtectedMember
-        path = Config._Directories.namespace_dir.parent / Config._Filenames.root_config
-    if not path.exists():
-        # noinspection PyProtectedMember
-        path = Config._Directories.user_configs / Config._Filenames.root_config
-    if not path.exists():
-        path = None
+    for path in POSSIBLE_CONFIG_PATHS:
+        if path.exists():
             return path
+    return None
 
 
 config_path = get_config_path()

devine/core/downloaders/aria2c.py

@@ -15,10 +15,11 @@ from requests.cookies import cookiejar_from_dict, get_cookie_header
 from rich import filesize
 from rich.text import Text
 
+from devine.core import binaries
 from devine.core.config import config
 from devine.core.console import console
 from devine.core.constants import DOWNLOAD_CANCELLED
-from devine.core.utilities import get_binary_path, get_extension, get_free_port
+from devine.core.utilities import get_extension, get_free_port
 
 
 def rpc(caller: Callable, secret: str, method: str, params: Optional[list[Any]] = None) -> Any:
@@ -87,8 +88,7 @@ def download(
     if not isinstance(urls, list):
         urls = [urls]
 
-    executable = get_binary_path("aria2c", "aria2")
-    if not executable:
+    if not binaries.Aria2:
         raise EnvironmentError("Aria2c executable not found...")
 
     if proxy and not proxy.lower().startswith("http://"):
@@ -186,7 +186,7 @@ def download(
     try:
         p = subprocess.Popen(
             [
-                executable,
+                binaries.Aria2,
                 *arguments
             ],
             stdin=subprocess.PIPE,

devine/core/downloaders/curl_impersonate.py

@@ -6,7 +6,6 @@ from http.cookiejar import CookieJar
 from pathlib import Path
 from typing import Any, Generator, MutableMapping, Optional, Union
 
-from curl_cffi import CurlOpt
 from curl_cffi.requests import Session
 from rich import filesize
 
@@ -18,7 +17,7 @@ MAX_ATTEMPTS = 5
 RETRY_WAIT = 2
 CHUNK_SIZE = 1024
 PROGRESS_WINDOW = 5
-BROWSER = config.curl_impersonate.get("browser", "chrome120")
+BROWSER = config.curl_impersonate.get("browser", "chrome124")
 
 
 def download(
@@ -53,11 +52,6 @@ def download(
             for one-time request changes like a header, cookie, or proxy. For example,
             to request Byte-ranges use e.g., `headers={"Range": "bytes=0-128"}`.
     """
-    # https://github.com/yifeikong/curl_cffi/issues/6#issuecomment-2028518677
-    # must be applied here since the `session.curl` is thread-localized
-    # noinspection PyProtectedMember
-    session.curl.setopt(CurlOpt.PROXY_CAINFO, session.curl._cacert)
-
     save_dir = save_path.parent
     control_file = save_path.with_name(f"{save_path.name}.!dev")
 

devine/core/drm/clearkey.py

@@ -7,7 +7,7 @@ from typing import Optional, Union
 from urllib.parse import urljoin
 
 from Cryptodome.Cipher import AES
-from Cryptodome.Util.Padding import pad, unpad
+from Cryptodome.Util.Padding import unpad
 from m3u8.model import Key
 from requests import Session
 
@@ -43,7 +43,7 @@ class ClearKey:
 
         decrypted = AES. \
             new(self.key, AES.MODE_CBC, self.iv). \
-            decrypt(pad(path.read_bytes(), AES.block_size))
+            decrypt(path.read_bytes())
 
         try:
             decrypted = unpad(decrypted, AES.block_size)

devine/core/drm/widevine.py

@@ -3,7 +3,6 @@ from __future__ import annotations
 import base64
 import shutil
 import subprocess
-import sys
 import textwrap
 from pathlib import Path
 from typing import Any, Callable, Optional, Union
@@ -17,10 +16,11 @@ from pywidevine.pssh import PSSH
 from requests import Session
 from rich.text import Text
 
+from devine.core import binaries
 from devine.core.config import config
 from devine.core.console import console
 from devine.core.constants import AnyTrack
-from devine.core.utilities import get_binary_path, get_boxes
+from devine.core.utilities import get_boxes
 from devine.core.utils.subprocess import ffprobe
 
 
@@ -223,9 +223,7 @@ class Widevine:
         if not self.content_keys:
             raise ValueError("Cannot decrypt a Track without any Content Keys...")
 
-        platform = {"win32": "win", "darwin": "osx"}.get(sys.platform, sys.platform)
-        executable = get_binary_path("shaka-packager", "packager", f"packager-{platform}", f"packager-{platform}-x64")
-        if not executable:
+        if not binaries.ShakaPackager:
             raise EnvironmentError("Shaka Packager executable not found but is required.")
         if not path or not path.exists():
             raise ValueError("Tried to decrypt a file that does not exist.")
@@ -252,7 +250,7 @@ class Widevine:
             ]
 
             p = subprocess.Popen(
-                [executable, *arguments],
+                [binaries.ShakaPackager, *arguments],
                 stdout=subprocess.DEVNULL,
                 stderr=subprocess.PIPE,
                 universal_newlines=True

devine/core/manifests/dash.py

@@ -285,12 +285,16 @@ class DASH:
             segment_base = adaptation_set.find("SegmentBase")
 
         segments: list[tuple[str, Optional[str]]] = []
+        segment_timescale: float = 0
+        segment_durations: list[int] = []
         track_kid: Optional[UUID] = None
 
         if segment_template is not None:
             segment_template = copy(segment_template)
             start_number = int(segment_template.get("startNumber") or 1)
+            end_number = int(segment_template.get("endNumber") or 0) or None
             segment_timeline = segment_template.find("SegmentTimeline")
+            segment_timescale = float(segment_template.get("timescale") or 1)
 
             for item in ("initialization", "media"):
                 value = segment_template.get(item)
@@ -318,17 +322,18 @@ class DASH:
                 track_kid = track.get_key_id(init_data)
 
             if segment_timeline is not None:
-                seg_time_list = []
                 current_time = 0
                 for s in segment_timeline.findall("S"):
                     if s.get("t"):
                         current_time = int(s.get("t"))
                     for _ in range(1 + (int(s.get("r") or 0))):
-                        seg_time_list.append(current_time)
+                        segment_durations.append(current_time)
                         current_time += int(s.get("d"))
-                seg_num_list = list(range(start_number, len(seg_time_list) + start_number))
 
-                for t, n in zip(seg_time_list, seg_num_list):
+                if not end_number:
+                    end_number = len(segment_durations)
+
+                for t, n in zip(segment_durations, range(start_number, end_number + 1)):
                     segments.append((
                         DASH.replace_fields(
                             segment_template.get("media"),
@@ -342,11 +347,12 @@ class DASH:
                 if not period_duration:
                     raise ValueError("Duration of the Period was unable to be determined.")
                 period_duration = DASH.pt_to_sec(period_duration)
-                segment_duration = float(segment_template.get("duration"))
-                segment_timescale = float(segment_template.get("timescale") or 1)
-                total_segments = math.ceil(period_duration / (segment_duration / segment_timescale))
+                segment_duration = float(segment_template.get("duration")) or 1
 
-                for s in range(start_number, start_number + total_segments):
+                if not end_number:
+                    end_number = math.ceil(period_duration / (segment_duration / segment_timescale))
+
+                for s in range(start_number, end_number + 1):
                     segments.append((
                         DASH.replace_fields(
                             segment_template.get("media"),
@@ -356,7 +362,11 @@ class DASH:
                             Time=s
                         ), None
                     ))
+                    # TODO: Should we floor/ceil/round, or is int() ok?
+                    segment_durations.append(int(segment_duration))
         elif segment_list is not None:
+            segment_timescale = float(segment_list.get("timescale") or 1)
+
             init_data = None
             initialization = segment_list.find("Initialization")
             if initialization is not None:
@@ -388,6 +398,7 @@ class DASH:
                     media_url,
                     segment_url.get("mediaRange")
                 ))
+                segment_durations.append(int(segment_url.get("duration") or 1))
         elif segment_base is not None:
             media_range = None
             init_data = None
@@ -420,6 +431,10 @@ class DASH:
             log.debug(track.url)
             sys.exit(1)
 
+        # TODO: Should we floor/ceil/round, or is int() ok?
+        track.data["dash"]["timescale"] = int(segment_timescale)
+        track.data["dash"]["segment_durations"] = segment_durations
+
         if not track.drm and isinstance(track, (Video, Audio)):
             try:
                 track.drm = [Widevine.from_init_data(init_data)]
@@ -457,6 +472,7 @@ class DASH:
         if downloader.__name__ == "aria2c" and any(bytes_range is not None for url, bytes_range in segments):
             # aria2(c) is shit and doesn't support the Range header, fallback to the requests downloader
             downloader = requests_downloader
+            log.warning("Falling back to the requests downloader as aria2(c) doesn't support the Range header")
 
         for status_update in downloader(
             urls=[

devine/core/manifests/hls.py

@@ -19,12 +19,13 @@ from pywidevine.cdm import Cdm as WidevineCdm
 from pywidevine.pssh import PSSH
 from requests import Session
 
+from devine.core import binaries
 from devine.core.constants import DOWNLOAD_CANCELLED, DOWNLOAD_LICENCE_ONLY, AnyTrack
 from devine.core.downloaders import requests as requests_downloader
 from devine.core.drm import DRM_T, ClearKey, Widevine
 from devine.core.events import events
 from devine.core.tracks import Audio, Subtitle, Tracks, Video
-from devine.core.utilities import get_binary_path, get_extension, is_close_match, try_ensure_utf8
+from devine.core.utilities import get_extension, is_close_match, try_ensure_utf8
 
 
 class HLS:
@@ -253,17 +254,24 @@ class HLS:
         progress(total=total_segments)
 
         downloader = track.downloader
+        if (
+            downloader.__name__ == "aria2c" and
+            any(x.byterange for x in master.segments if x not in unwanted_segments)
+        ):
+            downloader = requests_downloader
+            log.warning("Falling back to the requests downloader as aria2(c) doesn't support the Range header")
 
         urls: list[dict[str, Any]] = []
+        segment_durations: list[int] = []
+
         range_offset = 0
         for segment in master.segments:
             if segment in unwanted_segments:
                 continue
 
+            segment_durations.append(int(segment.duration))
+
             if segment.byterange:
-                if downloader.__name__ == "aria2c":
-                    # aria2(c) is shit and doesn't support the Range header, fallback to the requests downloader
-                    downloader = requests_downloader
                 byte_range = HLS.calculate_byte_range(segment.byterange, range_offset)
                 range_offset = byte_range.split("-")[0]
             else:
@@ -276,6 +284,8 @@ class HLS:
                 } if byte_range else {}
             })
 
+        track.data["hls"]["segment_durations"] = segment_durations
+
         segment_save_dir = save_dir / "segments"
 
         for status_update in downloader(
@@ -377,15 +387,27 @@ class HLS:
                 elif len(files) != range_len:
                     raise ValueError(f"Missing {range_len - len(files)} segment files for {segment_range}...")
 
+                if isinstance(drm, Widevine):
+                    # with widevine we can merge all segments and decrypt once
                     merge(
                         to=merged_path,
                         via=files,
                         delete=True,
                         include_map_data=True
                     )
-
                     drm.decrypt(merged_path)
                     merged_path.rename(decrypted_path)
+                else:
+                    # with other drm we must decrypt separately and then merge them
+                    # for aes this is because each segment likely has 16-byte padding
+                    for file in files:
+                        drm.decrypt(file)
+                    merge(
+                        to=merged_path,
+                        via=files,
+                        delete=True,
+                        include_map_data=True
+                    )
 
                 events.emit(
                     events.Types.TRACK_DECRYPTED,
@@ -556,8 +578,7 @@ class HLS:
 
         Returns the file size of the merged file.
         """
-        ffmpeg = get_binary_path("ffmpeg")
-        if not ffmpeg:
+        if not binaries.FFMPEG:
             raise EnvironmentError("FFmpeg executable was not found but is required to merge HLS segments.")
 
         demuxer_file = segments[0].parent / "ffmpeg_concat_demuxer.txt"
@@ -567,7 +588,7 @@ class HLS:
         ]))
 
         subprocess.check_call([
-            ffmpeg, "-hide_banner",
+            binaries.FFMPEG, "-hide_banner",
             "-loglevel", "panic",
             "-f", "concat",
             "-safe", "0",

devine/core/proxies/hola.py

@@ -3,8 +3,8 @@ import re
 import subprocess
 from typing import Optional
 
+from devine.core import binaries
 from devine.core.proxies.proxy import Proxy
-from devine.core.utilities import get_binary_path
 
 
 class Hola(Proxy):
@@ -13,7 +13,7 @@ class Hola(Proxy):
         Proxy Service using Hola's direct connections via the hola-proxy project.
         https://github.com/Snawoot/hola-proxy
         """
-        self.binary = get_binary_path("hola-proxy")
+        self.binary = binaries.HolaProxy
         if not self.binary:
             raise EnvironmentError("hola-proxy executable not found but is required for the Hola proxy provider.")
 

devine/core/tracks/attachment.py

@@ -37,7 +37,7 @@ class Attachment:
             mime_type = {
                 ".ttf": "application/x-truetype-font",
                 ".otf": "application/vnd.ms-opentype"
-            }.get(path.suffix, mimetypes.guess_type(path)[0])
+            }.get(path.suffix.lower(), mimetypes.guess_type(path)[0])
             if not mime_type:
                 raise ValueError("The attachment mime-type could not be automatically detected.")
 

devine/core/tracks/subtitle.py

@@ -7,7 +7,7 @@ from enum import Enum
 from functools import partial
 from io import BytesIO
 from pathlib import Path
-from typing import Any, Callable, Iterable, Optional
+from typing import Any, Callable, Iterable, Optional, Union
 
 import pycaption
 import requests
@@ -17,8 +17,10 @@ from pycaption.geometry import Layout
 from pymp4.parser import MP4
 from subtitle_filter import Subtitles
 
+from devine.core import binaries
 from devine.core.tracks.track import Track
-from devine.core.utilities import get_binary_path, try_ensure_utf8
+from devine.core.utilities import try_ensure_utf8
+from devine.core.utils.webvtt import merge_segmented_webvtt
 
 
 class Subtitle(Track):
@@ -201,6 +203,26 @@ class Subtitle(Track):
             self.convert(Subtitle.Codec.TimedTextMarkupLang)
         elif self.codec == Subtitle.Codec.fVTT:
             self.convert(Subtitle.Codec.WebVTT)
+        elif self.codec == Subtitle.Codec.WebVTT:
+            text = self.path.read_text("utf8")
+            if self.descriptor == Track.Descriptor.DASH:
+                if len(self.data["dash"]["segment_durations"]) > 1:
+                    text = merge_segmented_webvtt(
+                        text,
+                        segment_durations=self.data["dash"]["segment_durations"],
+                        timescale=self.data["dash"]["timescale"]
+                    )
+            elif self.descriptor == Track.Descriptor.HLS:
+                if len(self.data["hls"]["segment_durations"]) > 1:
+                    text = merge_segmented_webvtt(
+                        text,
+                        segment_durations=self.data["hls"]["segment_durations"],
+                        timescale=1  # ?
+                    )
+            caption_set = pycaption.WebVTTReader().read(text)
+            Subtitle.merge_same_cues(caption_set)
+            subtitle_text = pycaption.WebVTTWriter().write(caption_set)
+            self.path.write_text(subtitle_text, encoding="utf8")
 
     def convert(self, codec: Subtitle.Codec) -> Path:
         """
@@ -233,14 +255,13 @@ class Subtitle(Track):
 
         output_path = self.path.with_suffix(f".{codec.value.lower()}")
 
-        sub_edit_executable = get_binary_path("SubtitleEdit")
-        if sub_edit_executable and self.codec not in (Subtitle.Codec.fTTML, Subtitle.Codec.fVTT):
+        if binaries.SubtitleEdit and self.codec not in (Subtitle.Codec.fTTML, Subtitle.Codec.fVTT):
             sub_edit_format = {
                 Subtitle.Codec.SubStationAlphav4: "AdvancedSubStationAlpha",
                 Subtitle.Codec.TimedTextMarkupLang: "TimedText1.0"
             }.get(codec, codec.name)
             sub_edit_args = [
-                sub_edit_executable,
+                binaries.SubtitleEdit,
                 "/Convert", self.path, sub_edit_format,
                 f"/outputfilename:{output_path.name}",
                 "/encoding:utf8"
@@ -308,14 +329,7 @@ class Subtitle(Track):
                 caption_lists[language] = caption_list
                 caption_set: pycaption.CaptionSet = pycaption.CaptionSet(caption_lists)
             elif codec == Subtitle.Codec.WebVTT:
-                text = try_ensure_utf8(data).decode("utf8")
-                # Segmented VTT when merged may have the WEBVTT headers part of the next caption
-                # if they are not separated far enough from the previous caption, hence the \n\n
-                text = text. \
-                    replace("WEBVTT", "\n\nWEBVTT"). \
-                    replace("\r", ""). \
-                    replace("\n\n\n", "\n \n\n"). \
-                    replace("\n\n<", "\n<")
+                text = Subtitle.space_webvtt_headers(data)
                 caption_set = pycaption.WebVTTReader().read(text)
             else:
                 raise ValueError(f"Unknown Subtitle format \"{codec}\"...")
@@ -332,6 +346,27 @@ class Subtitle(Track):
 
         return caption_set
 
+    @staticmethod
+    def space_webvtt_headers(data: Union[str, bytes]):
+        """
+        Space out the WEBVTT Headers from Captions.
+
+        Segmented VTT when merged may have the WEBVTT headers part of the next caption
+        as they were not separated far enough from the previous caption and ended up
+        being considered as caption text rather than the header for the next segment.
+        """
+        if isinstance(data, bytes):
+            data = try_ensure_utf8(data).decode("utf8")
+        elif not isinstance(data, str):
+            raise ValueError(f"Expecting data to be a str, not {data!r}")
+
+        text = data.replace("WEBVTT", "\n\nWEBVTT").\
+            replace("\r", "").\
+            replace("\n\n\n", "\n \n\n").\
+            replace("\n\n<", "\n<")
+
+        return text
+
     @staticmethod
     def merge_same_cues(caption_set: pycaption.CaptionSet):
         """Merge captions with the same timecodes and text as one in-place."""
@@ -500,8 +535,7 @@ class Subtitle(Track):
         if not self.path or not self.path.exists():
             raise ValueError("You must download the subtitle track first.")
 
-        executable = get_binary_path("SubtitleEdit")
-        if executable:
+        if binaries.SubtitleEdit:
             if self.codec == Subtitle.Codec.SubStationAlphav4:
                 output_format = "AdvancedSubStationAlpha"
             elif self.codec == Subtitle.Codec.TimedTextMarkupLang:
@@ -510,7 +544,7 @@ class Subtitle(Track):
                 output_format = self.codec.name
             subprocess.run(
                 [
-                    executable,
+                    binaries.SubtitleEdit,
                     "/Convert", self.path, output_format,
                     "/encoding:utf8",
                     "/overwrite",
@@ -539,8 +573,7 @@ class Subtitle(Track):
         if not self.path or not self.path.exists():
             raise ValueError("You must download the subtitle track first.")
 
-        executable = get_binary_path("SubtitleEdit")
-        if not executable:
+        if not binaries.SubtitleEdit:
             raise EnvironmentError("SubtitleEdit executable not found...")
 
         if self.codec == Subtitle.Codec.SubStationAlphav4:
@@ -552,7 +585,7 @@ class Subtitle(Track):
 
         subprocess.run(
             [
-                executable,
+                binaries.SubtitleEdit,
                 "/Convert", self.path, output_format,
                 "/ReverseRtlStartEnd",
                 "/encoding:utf8",

devine/core/tracks/track.py

@@ -4,6 +4,7 @@ import logging
 import re
 import shutil
 import subprocess
+from collections import defaultdict
 from copy import copy
 from enum import Enum
 from functools import partial
@@ -15,12 +16,13 @@ from zlib import crc32
 from langcodes import Language
 from requests import Session
 
+from devine.core import binaries
 from devine.core.config import config
 from devine.core.constants import DOWNLOAD_CANCELLED, DOWNLOAD_LICENCE_ONLY
 from devine.core.downloaders import aria2c, curl_impersonate, requests
 from devine.core.drm import DRM_T, Widevine
 from devine.core.events import events
-from devine.core.utilities import get_binary_path, get_boxes, try_ensure_utf8
+from devine.core.utilities import get_boxes, try_ensure_utf8
 from devine.core.utils.subprocess import ffprobe
 
 
@@ -41,7 +43,7 @@ class Track:
         drm: Optional[Iterable[DRM_T]] = None,
         edition: Optional[str] = None,
         downloader: Optional[Callable] = None,
-        data: Optional[dict] = None,
+        data: Optional[Union[dict, defaultdict]] = None,
         id_: Optional[str] = None,
     ) -> None:
         if not isinstance(url, (str, list)):
@@ -62,8 +64,8 @@ class Track:
             raise TypeError(f"Expected edition to be a {str}, not {type(edition)}")
         if not isinstance(downloader, (Callable, type(None))):
             raise TypeError(f"Expected downloader to be a {Callable}, not {type(downloader)}")
-        if not isinstance(data, (dict, type(None))):
-            raise TypeError(f"Expected data to be a {dict}, not {type(data)}")
+        if not isinstance(data, (dict, defaultdict, type(None))):
+            raise TypeError(f"Expected data to be a {dict} or {defaultdict}, not {type(data)}")
 
         invalid_urls = ", ".join(set(type(x) for x in url if not isinstance(x, str)))
         if invalid_urls:
@@ -92,6 +94,7 @@ class Track:
         self.drm = drm
         self.edition: str = edition
         self.downloader = downloader
+        self._data: defaultdict[Any, Any] = defaultdict(dict)
         self.data = data or {}
 
         if self.name is None:
@@ -131,6 +134,42 @@ class Track:
     def __eq__(self, other: Any) -> bool:
         return isinstance(other, Track) and self.id == other.id
 
+    @property
+    def data(self) -> defaultdict[Any, Any]:
+        """
+        Arbitrary track data dictionary.
+
+        A defaultdict is used with a dict as the factory for easier
+        nested saving and safer exists-checks.
+
+        Reserved keys:
+
+        - "hls" used by the HLS class.
+          - playlist: m3u8.model.Playlist - The primary track information.
+          - media: m3u8.model.Media - The audio/subtitle track information.
+          - segment_durations: list[int] - A list of each segment's duration.
+        - "dash" used by the DASH class.
+          - manifest: lxml.ElementTree - DASH MPD manifest.
+          - period: lxml.Element - The period of this track.
+          - adaptation_set: lxml.Element - The adaptation set of this track.
+          - representation: lxml.Element - The representation of this track.
+          - timescale: int - The timescale of the track's segments.
+          - segment_durations: list[int] - A list of each segment's duration.
+
+        You should not add, change, or remove any data within reserved keys.
+        You may use their data but do note that the values of them may change
+        or be removed at any point.
+        """
+        return self._data
+
+    @data.setter
+    def data(self, value: Union[dict, defaultdict]) -> None:
+        if not isinstance(value, (dict, defaultdict)):
+            raise TypeError(f"Expected data to be a {dict} or {defaultdict}, not {type(value)}")
+        if isinstance(value, dict):
+            value = defaultdict(dict, **value)
+        self._data = value
+
     def download(
         self,
         session: Session,
@@ -470,8 +509,7 @@ class Track:
         if not self.path or not self.path.exists():
             raise ValueError("Cannot repackage a Track that has not been downloaded.")
 
-        executable = get_binary_path("ffmpeg")
-        if not executable:
+        if not binaries.FFMPEG:
             raise EnvironmentError("FFmpeg executable \"ffmpeg\" was not found but is required for this call.")
 
         original_path = self.path
@@ -480,7 +518,7 @@ class Track:
         def _ffmpeg(extra_args: list[str] = None):
             subprocess.run(
                 [
-                    executable, "-hide_banner",
+                    binaries.FFMPEG, "-hide_banner",
                     "-loglevel", "error",
                     "-i", original_path,
                     *(extra_args or []),
@@ -504,6 +542,7 @@ class Track:
             else:
                 raise
 
+        original_path.unlink()
         self.path = output_path
 
 

devine/core/tracks/tracks.py

@@ -316,7 +316,7 @@ class Tracks:
             ][:per_language or None])
         return selected
 
-    def mux(self, title: str, delete: bool = True, progress: Optional[partial] = None) -> tuple[Path, int]:
+    def mux(self, title: str, delete: bool = True, progress: Optional[partial] = None) -> tuple[Path, int, list[str]]:
         """
         Multiplex all the Tracks into a Matroska Container file.
 
@@ -410,15 +410,18 @@ class Tracks:
 
         # let potential failures go to caller, caller should handle
         try:
+            errors = []
             p = subprocess.Popen([
                 *cl,
                 "--output", str(output_path),
                 "--gui-mode"
             ], text=True, stdout=subprocess.PIPE)
             for line in iter(p.stdout.readline, ""):
+                if line.startswith("#GUI#error") or line.startswith("#GUI#warning"):
+                    errors.append(line)
                 if "progress" in line:
                     progress(total=100, completed=int(line.strip()[14:-1]))
-            return output_path, p.wait()
+            return output_path, p.wait(), errors
         finally:
             if chapters_path:
                 # regardless of delete param, we delete as it's a file we made during muxing

devine/core/tracks/video.py

@@ -10,10 +10,11 @@ from typing import Any, Optional, Union
 
 from langcodes import Language
 
+from devine.core import binaries
 from devine.core.config import config
 from devine.core.tracks.subtitle import Subtitle
 from devine.core.tracks.track import Track
-from devine.core.utilities import FPS, get_binary_path, get_boxes
+from devine.core.utilities import FPS, get_boxes
 
 
 class Video(Track):
@@ -257,8 +258,7 @@ class Video(Track):
                 f"it's codec, {self.codec.value}, is not yet supported."
             )
 
-        executable = get_binary_path("ffmpeg")
-        if not executable:
+        if not binaries.FFMPEG:
             raise EnvironmentError("FFmpeg executable \"ffmpeg\" was not found but is required for this call.")
 
         filter_key = {
@@ -270,7 +270,7 @@ class Video(Track):
         output_path = original_path.with_stem(f"{original_path.stem}_{['limited', 'full'][range_]}_range")
 
         subprocess.run([
-            executable, "-hide_banner",
+            binaries.FFMPEG, "-hide_banner",
             "-loglevel", "panic",
             "-i", original_path,
             "-codec", "copy",
@@ -288,8 +288,7 @@ class Video(Track):
         if not self.path:
             raise ValueError("You must download the track first.")
 
-        executable = get_binary_path("ccextractor", "ccextractorwin", "ccextractorwinfull")
-        if not executable:
+        if not binaries.CCExtractor:
             raise EnvironmentError("ccextractor executable was not found.")
 
         # ccextractor often fails in weird ways unless we repack
@@ -299,7 +298,7 @@ class Video(Track):
 
         try:
             subprocess.run([
-                executable,
+                binaries.CCExtractor,
                 "-trim",
                 "-nobom",
                 "-noru", "-ru1",
@@ -380,8 +379,7 @@ class Video(Track):
         if not self.path or not self.path.exists():
             raise ValueError("Cannot clean a Track that has not been downloaded.")
 
-        executable = get_binary_path("ffmpeg")
-        if not executable:
+        if not binaries.FFMPEG:
             raise EnvironmentError("FFmpeg executable \"ffmpeg\" was not found but is required for this call.")
 
         log = logging.getLogger("x264-clean")
@@ -402,7 +400,7 @@ class Video(Track):
         original_path = self.path
         cleaned_path = original_path.with_suffix(f".cleaned{original_path.suffix}")
         subprocess.run([
-            executable, "-hide_banner",
+            binaries.FFMPEG, "-hide_banner",
             "-loglevel", "panic",
             "-i", original_path,
             "-map_metadata", "-1",

devine/core/utilities.py

@@ -3,7 +3,6 @@ import contextlib
 import importlib.util
 import os
 import re
-import shutil
 import socket
 import sys
 import time
@@ -87,15 +86,6 @@ def import_module_by_path(path: Path) -> ModuleType:
     return module
 
 
-def get_binary_path(*names: str) -> Optional[Path]:
-    """Find the path of the first found binary name."""
-    for name in names:
-        path = shutil.which(name)
-        if path:
-            return Path(path)
-    return None
-
-
 def sanitize_filename(filename: str, spacer: str = ".") -> str:
     """
     Sanitize a string to be filename safe.
@@ -133,18 +123,18 @@ def get_boxes(data: bytes, box_type: bytes, as_bytes: bool = False) -> Box:
     # since it doesn't care what child box the wanted box is from, this works fine.
     if not isinstance(data, (bytes, bytearray)):
         raise ValueError("data must be bytes")
+
+    offset = 0
     while True:
         try:
-            index = data.index(box_type)
+            index = data[offset:].index(box_type)
         except ValueError:
             break
         if index < 0:
             break
-        if index > 4:
         index -= 4  # size is before box type and is 4 bytes long
-        data = data[index:]
         try:
-            box = Box.parse(data)
+            box = Box.parse(data[offset:][index:])
         except IOError:
             # since get_init_segment might cut off unexpectedly, pymp4 may be unable to read
             # the expected amounts of data and complain, so let's just end the function here
@@ -157,6 +147,7 @@ def get_boxes(data: bytes, box_type: bytes, as_bytes: bool = False) -> Box:
             raise e
         if as_bytes:
             box = Box.build(box)
+        offset += index + len(Box.build(box))
         yield box
 
 

devine/core/utils/subprocess.py

@@ -3,11 +3,16 @@ import subprocess
 from pathlib import Path
 from typing import Union
 
+from devine.core import binaries
+
 
 def ffprobe(uri: Union[bytes, Path]) -> dict:
     """Use ffprobe on the provided data to get stream information."""
+    if not binaries.FFProbe:
+        raise EnvironmentError("FFProbe executable \"ffprobe\" not found but is required.")
+
     args = [
-        "ffprobe",
+        binaries.FFProbe,
         "-v", "quiet",
         "-of", "json",
         "-show_streams"

devine/core/utils/webvtt.py

@@ -0,0 +1,191 @@
+import re
+import sys
+import typing
+from typing import Optional
+
+from pycaption import Caption, CaptionList, CaptionNode, CaptionReadError, WebVTTReader, WebVTTWriter
+
+
+class CaptionListExt(CaptionList):
+    @typing.no_type_check
+    def __init__(self, iterable=None, layout_info=None):
+        self.first_segment_mpegts = 0
+        super().__init__(iterable, layout_info)
+
+
+class CaptionExt(Caption):
+    @typing.no_type_check
+    def __init__(self, start, end, nodes, style=None, layout_info=None, segment_index=0, mpegts=0, cue_time=0.0):
+        style = style or {}
+        self.segment_index: int = segment_index
+        self.mpegts: float = mpegts
+        self.cue_time: float = cue_time
+        super().__init__(start, end, nodes, style, layout_info)
+
+
+class WebVTTReaderExt(WebVTTReader):
+    # HLS extension support <https://datatracker.ietf.org/doc/html/rfc8216#section-3.5>
+    RE_TIMESTAMP_MAP = re.compile(r"X-TIMESTAMP-MAP.*")
+    RE_MPEGTS = re.compile(r"MPEGTS:(\d+)")
+    RE_LOCAL = re.compile(r"LOCAL:((?:(\d{1,}):)?(\d{2}):(\d{2})\.(\d{3}))")
+
+    def _parse(self, lines: list[str]) -> CaptionList:
+        captions = CaptionListExt()
+        start = None
+        end = None
+        nodes: list[CaptionNode] = []
+        layout_info = None
+        found_timing = False
+        segment_index = -1
+        mpegts = 0
+        cue_time = 0.0
+
+        # The first segment MPEGTS is needed to calculate the rest. It is possible that
+        # the first segment contains no cue and is ignored by pycaption, this acts as a fallback.
+        captions.first_segment_mpegts = 0
+
+        for i, line in enumerate(lines):
+            if "-->" in line:
+                found_timing = True
+                timing_line = i
+                last_start_time = captions[-1].start if captions else 0
+                try:
+                    start, end, layout_info = self._parse_timing_line(line, last_start_time)
+                except CaptionReadError as e:
+                    new_msg = f"{e.args[0]} (line {timing_line})"
+                    tb = sys.exc_info()[2]
+                    raise type(e)(new_msg).with_traceback(tb) from None
+
+            elif "" == line:
+                if found_timing and nodes:
+                    found_timing = False
+                    caption = CaptionExt(
+                        start,
+                        end,
+                        nodes,
+                        layout_info=layout_info,
+                        segment_index=segment_index,
+                        mpegts=mpegts,
+                        cue_time=cue_time,
+                    )
+                    captions.append(caption)
+                    nodes = []
+
+            elif "WEBVTT" in line:
+                # Merged segmented VTT doesn't have index information, track manually.
+                segment_index += 1
+                mpegts = 0
+                cue_time = 0.0
+            elif m := self.RE_TIMESTAMP_MAP.match(line):
+                if r := self.RE_MPEGTS.search(m.group()):
+                    mpegts = int(r.group(1))
+
+                cue_time = self._parse_local(m.group())
+
+                # Early assignment in case the first segment contains no cue.
+                if segment_index == 0:
+                    captions.first_segment_mpegts = mpegts
+
+            else:
+                if found_timing:
+                    if nodes:
+                        nodes.append(CaptionNode.create_break())
+                    nodes.append(CaptionNode.create_text(self._decode(line)))
+                else:
+                    # it's a comment or some metadata; ignore it
+                    pass
+
+        # Add a last caption if there are remaining nodes
+        if nodes:
+            caption = CaptionExt(start, end, nodes, layout_info=layout_info, segment_index=segment_index, mpegts=mpegts)
+            captions.append(caption)
+
+        return captions
+
+    @staticmethod
+    def _parse_local(string: str) -> float:
+        """
+        Parse WebVTT LOCAL time and convert it to seconds.
+        """
+        m = WebVTTReaderExt.RE_LOCAL.search(string)
+        if not m:
+            return 0
+
+        parsed = m.groups()
+        if not parsed:
+            return 0
+        hours = int(parsed[1])
+        minutes = int(parsed[2])
+        seconds = int(parsed[3])
+        milliseconds = int(parsed[4])
+        return (milliseconds / 1000) + seconds + (minutes * 60) + (hours * 3600)
+
+
+def merge_segmented_webvtt(vtt_raw: str, segment_durations: Optional[list[int]] = None, timescale: int = 1) -> str:
+    """
+    Merge Segmented WebVTT data.
+
+    Parameters:
+        vtt_raw: The concatenated WebVTT files to merge. All WebVTT headers must be
+            appropriately spaced apart, or it may produce unwanted effects like
+            considering headers as captions, timestamp lines, etc.
+        segment_durations: A list of each segment's duration. If not provided it will try
+            to get it from the X-TIMESTAMP-MAP headers, specifically the MPEGTS number.
+        timescale: The number of time units per second.
+
+    This parses the X-TIMESTAMP-MAP data to compute new absolute timestamps, replacing
+    the old start and end timestamp values. All X-TIMESTAMP-MAP header information will
+    be removed from the output as they are no longer of concern. Consider this function
+    the opposite of a WebVTT Segmenter, a WebVTT Joiner of sorts.
+
+    Algorithm borrowed from N_m3u8DL-RE and shaka-player.
+    """
+    MPEG_TIMESCALE = 90_000
+
+    vtt = WebVTTReaderExt().read(vtt_raw)
+    for lang in vtt.get_languages():
+        prev_caption = None
+        duplicate_index: list[int] = []
+        captions = vtt.get_captions(lang)
+
+        if captions[0].segment_index == 0:
+            first_segment_mpegts = captions[0].mpegts
+        else:
+            first_segment_mpegts = segment_durations[0] if segment_durations else captions.first_segment_mpegts
+
+        caption: CaptionExt
+        for i, caption in enumerate(captions):
+            # DASH WebVTT doesn't have MPEGTS timestamp like HLS. Instead,
+            # calculate the timestamp from SegmentTemplate/SegmentList duration.
+            likely_dash = first_segment_mpegts == 0 and caption.mpegts == 0
+            if likely_dash and segment_durations:
+                duration = segment_durations[caption.segment_index]
+                caption.mpegts = MPEG_TIMESCALE * (duration / timescale)
+
+            if caption.mpegts == 0:
+                continue
+
+            seconds = (caption.mpegts - first_segment_mpegts) / MPEG_TIMESCALE - caption.cue_time
+            offset = seconds * 1_000_000  # pycaption use microseconds
+
+            if caption.start < offset:
+                caption.start += offset
+                caption.end += offset
+
+            # If the difference between current and previous captions is <=1ms
+            # and the payload is equal then splice.
+            if (
+                prev_caption
+                and not caption.is_empty()
+                and (caption.start - prev_caption.end) <= 1000  # 1ms in microseconds
+                and caption.get_text() == prev_caption.get_text()
+            ):
+                prev_caption.end = caption.end
+                duplicate_index.append(i)
+
+            prev_caption = caption
+
+        # Remove duplicate
+        captions[:] = [c for c_index, c in enumerate(captions) if c_index not in set(duplicate_index)]
+
+    return WebVTTWriter().write(vtt)

poetry.lock

@@ -523,28 +523,31 @@ testing = ["cssselect", "importlib-resources", "jaraco.test (>=5.1)", "lxml", "p
 
 [[package]]
 name = "curl-cffi"
-version = "0.6.2"
-description = "libcurl ffi bindings for Python, with impersonation support"
+version = "0.7.0b4"
+description = "libcurl ffi bindings for Python, with impersonation support."
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "curl_cffi-0.6.2-cp38-abi3-macosx_10_9_x86_64.whl", hash = "sha256:23b8a2872b160718c04b06b1f8aa4fb1a2f4f94bce7040493515e081a27cad19"},
-    {file = "curl_cffi-0.6.2-cp38-abi3-macosx_11_0_arm64.whl", hash = "sha256:ad3c1cf5360810825ec4bc3da425f26ee4098878a615dab9d309a99afd883ba9"},
-    {file = "curl_cffi-0.6.2-cp38-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:3d01de6ed737ad1924aaa0198195b9020c38e77ce90ea3d72b9eacf4938c7adf"},
-    {file = "curl_cffi-0.6.2-cp38-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:37e513cc149d024a2d625e202f2cc9d4423d2937343ea2e06f797d99779e62dc"},
-    {file = "curl_cffi-0.6.2-cp38-abi3-win32.whl", hash = "sha256:12e829af97cbf7c1d5afef177e786f6f404ddf163b08897a1ed087cadbeb4837"},
-    {file = "curl_cffi-0.6.2-cp38-abi3-win_amd64.whl", hash = "sha256:3791b7a9ae4cb1298165300f2dc2d60a86779f055570ae83163fc2d8a74bf714"},
-    {file = "curl_cffi-0.6.2.tar.gz", hash = "sha256:9ee519e960b5fc6e0bbf13d0ecba9ce5f6306cb929354504bf03cc30f59a8f63"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-macosx_10_9_x86_64.whl", hash = "sha256:694d88f7065c59c651970f14bc415431f65ac601a9ba537463d70f432a48ccfc"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-macosx_11_0_arm64.whl", hash = "sha256:6faf01aa8d98d322b877d3d801544692c73729ea6eb4a45af83514a4ecd1c8fe"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5d39849371bbf3eab048113693715a8da5c729c494cccfa1128d768d96fdc31e"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:e3a5099b98c4bf12cc1afecb3409a9c57e7ebce9447a03c96dfb661ad8fa5e79"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:7e3616141a2a0be7896e7dc5da1ed3965e1a78aa2e563d8aba7a641135aeaf1b"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-musllinux_1_1_aarch64.whl", hash = "sha256:bd16cccc0d3e93c2fbc4f4cb7cce0e10cb2ef7f8957352f3f0d770f0d6e05702"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-musllinux_1_1_x86_64.whl", hash = "sha256:d65aa649abb24020c2ad7b3ce45e2816d1ffe25df06f1a6b0f52fbf353af82e0"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-win32.whl", hash = "sha256:b55c53bb6dff713cb63f76e2f147e2d54c984b1b09df66b08f52f3acae1aeca0"},
+    {file = "curl_cffi-0.7.0b4-cp38-abi3-win_amd64.whl", hash = "sha256:449ab07e07335558997cd62296b5c4f16ce27630de7830e4ad22441049a0ef1e"},
+    {file = "curl_cffi-0.7.0b4.tar.gz", hash = "sha256:c09a062b8aac93d4890d2c33b7053c0e1a5cf275328b80c1fb1a950310df75f2"},
 ]
 
 [package.dependencies]
-certifi = "*"
+certifi = ">=2024.2.2"
 cffi = ">=1.12.0"
 
 [package.extras]
 build = ["cibuildwheel", "wheel"]
-dev = ["autoflake (==1.4)", "coverage (==6.4.1)", "cryptography (==38.0.3)", "flake8 (==6.0.0)", "flake8-bugbear (==22.7.1)", "flake8-pie (==0.15.0)", "httpx (==0.23.1)", "mypy (==0.971)", "nest-asyncio (==1.6.0)", "pytest (==7.1.2)", "pytest-asyncio (==0.19.0)", "pytest-trio (==0.7.0)", "ruff (==0.1.14)", "trio (==0.21.0)", "trio-typing (==0.7.0)", "trustme (==0.9.0)", "types-certifi (==2021.10.8.2)", "uvicorn (==0.18.3)", "websockets (==11.0.3)"]
-test = ["cryptography (==38.0.3)", "fastapi (==0.100.0)", "httpx (==0.23.1)", "nest-asyncio (==1.6.0)", "proxy.py (==2.4.3)", "pytest (==7.1.2)", "pytest-asyncio (==0.19.0)", "pytest-trio (==0.7.0)", "python-multipart (==0.0.6)", "trio (==0.21.0)", "trio-typing (==0.7.0)", "trustme (==0.9.0)", "types-certifi (==2021.10.8.2)", "uvicorn (==0.18.3)", "websockets (==11.0.3)"]
+dev = ["charset-normalizer (>=3.3.2,<4.0)", "coverage (>=6.4.1,<7.0)", "cryptography (>=42.0.5,<43.0)", "httpx (==0.23.1)", "mypy (>=1.9.0,<2.0)", "pytest (>=8.1.1,<9.0)", "pytest-asyncio (>=0.23.6,<1.0)", "pytest-trio (>=0.8.0,<1.0)", "ruff (>=0.3.5,<1.0)", "trio (>=0.25.0,<1.0)", "trustme (>=1.1.0,<2.0)", "uvicorn (>=0.29.0,<1.0)", "websockets (>=12.0,<13.0)"]
+test = ["charset-normalizer (>=3.3.2,<4.0)", "cryptography (>=42.0.5,<43.0)", "fastapi (==0.110.0)", "httpx (==0.23.1)", "proxy.py (>=2.4.3,<3.0)", "pytest (>=8.1.1,<9.0)", "pytest-asyncio (>=0.23.6,<1.0)", "pytest-trio (>=0.8.0,<1.0)", "python-multipart (>=0.0.9,<1.0)", "trio (>=0.25.0,<1.0)", "trustme (>=1.1.0,<2.0)", "uvicorn (>=0.29.0,<1.0)", "websockets (>=12.0,<13.0)"]
 
 [[package]]
 name = "distlib"
@@ -727,20 +730,21 @@ testing = ["bson", "ecdsa", "feedparser", "gmpy2", "numpy", "pandas", "pymongo",
 
 [[package]]
 name = "langcodes"
-version = "3.3.0"
+version = "3.4.0"
 description = "Tools for labeling human languages with IETF language tags"
 optional = false
-python-versions = ">=3.6"
+python-versions = ">=3.8"
 files = [
-    {file = "langcodes-3.3.0-py3-none-any.whl", hash = "sha256:4d89fc9acb6e9c8fdef70bcdf376113a3db09b67285d9e1d534de6d8818e7e69"},
-    {file = "langcodes-3.3.0.tar.gz", hash = "sha256:794d07d5a28781231ac335a1561b8442f8648ca07cd518310aeb45d6f0807ef6"},
+    {file = "langcodes-3.4.0-py3-none-any.whl", hash = "sha256:10a4cc078b8e8937d8485d3352312a0a89a3125190db9f2bb2074250eef654e9"},
+    {file = "langcodes-3.4.0.tar.gz", hash = "sha256:ae5a77d1a01d0d1e91854a671890892b7ce9abb601ab7327fc5c874f899e1979"},
 ]
 
 [package.dependencies]
-language-data = {version = ">=1.1,<2.0", optional = true, markers = "extra == \"data\""}
+language-data = ">=1.2"
 
 [package.extras]
-data = ["language-data (>=1.1,<2.0)"]
+build = ["build", "twine"]
+test = ["pytest", "pytest-cov"]
 
 [[package]]
 name = "language-data"
@@ -2004,4 +2008,4 @@ multidict = ">=4.0"
 [metadata]
 lock-version = "2.0"
 python-versions = ">=3.9,<4.0"
-content-hash = "db110c0b1b9e30309fcd4e0f80e4369f20b055651f1bef81d0f5e6153a250dec"
+content-hash = "8bbbd788ab179a0669e8d7c6f45c0746e79f11c24ac39f6d4856563e76ec2f94"

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "devine"
-version = "3.3.2"
+version = "3.3.3"
 description = "Modular Movie, TV, and Music Archival Software."
 license = "GPL-3.0-only"
 authors = ["rlaphoenix <rlaphoenix@pm.me>"]
@@ -40,7 +40,7 @@ click = "^8.1.7"
 construct = "^2.8.8"
 crccheck = "^1.3.0"
 jsonpickle = "^3.0.4"
-langcodes = { extras = ["data"], version = "^3.3.0" }
+langcodes = { extras = ["data"], version = "^3.4.0" }
 lxml = "^5.2.1"
 pproxy = "^2.7.9"
 protobuf = "^4.25.3"
@@ -61,9 +61,7 @@ subtitle-filter = "^1.4.9"
 Unidecode = "^1.3.8"
 urllib3 = "^2.2.1"
 chardet = "^5.2.0"
-curl-cffi = "^0.6.2"
-language-data = "^1.2.0"
-marisa-trie = "^1.1.0"
+curl-cffi = "^0.7.0b4"
 
 [tool.poetry.dev-dependencies]
 pre-commit = "^3.7.0"