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

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

@@ -178,6 +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})"
                 )
@@ -701,16 +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()
@@ -930,21 +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/core/binaries.py

@@ -26,6 +26,7 @@ ShakaPackager = find(
     "shaka-packager",
     "packager",
     f"packager-{__shaka_platform}",
+    f"packager-{__shaka_platform}-arm64",
     f"packager-{__shaka_platform}-x64"
 )
 Aria2 = find("aria2c", "aria2")

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/manifests/dash.py

@@ -292,6 +292,7 @@ class DASH:
         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)
 
@@ -328,9 +329,11 @@ class DASH:
                     for _ in range(1 + (int(s.get("r") or 0))):
                         segment_durations.append(current_time)
                         current_time += int(s.get("d"))
-                seg_num_list = list(range(start_number, len(segment_durations) + start_number))
 
-                for t, n in zip(segment_durations, 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"),
@@ -345,9 +348,11 @@ class DASH:
                     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")) or 1
-                total_segments = math.ceil(period_duration / (segment_duration / segment_timescale))
 
-                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"),
@@ -467,6 +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

@@ -254,6 +254,12 @@ 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] = []
@@ -266,9 +272,6 @@ class HLS:
             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:
@@ -384,15 +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,

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

@@ -206,12 +206,14 @@ class Subtitle(Track):
         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"],

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
@@ -42,7 +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)):
@@ -63,8 +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:
@@ -93,6 +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:
@@ -132,6 +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,
@@ -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/utilities.py

@@ -123,18 +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
@@ -147,6 +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