devart-dl 4.4.2__tar.gz → 4.4.3__tar.gz

{devart_dl-4.4.2/devart_dl.egg-info → devart_dl-4.4.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: devart-dl
-Version: 4.4.2
+Version: 4.4.3
 Summary: 专业的 DeviantArt 作品批量下载工具 - Professional DeviantArt batch downloader with anti-ban features
 Author: DeviantArt Downloader Team
 License-Expression: MIT

{devart_dl-4.4.2 → devart_dl-4.4.3}/da_downloader/api.py

@@ -12,6 +12,7 @@ import requests
 from requests.exceptions import RequestException
 
 from .models import Deviation, ActionType
+from .wix_media import append_media_token, bounded_fullview_url
 
 logger = logging.getLogger(__name__)
 
@@ -324,9 +325,13 @@ class DeviantArtAPI:
             None,
         )
         
-        # 现代 DA 的 baseUri 本身就是原始文件（以图片/视频扩展名结尾）：
-        # 直接 baseUri + token 取原图即可。别拿它拼 fullview 的 /v1/fit|fill
-        # 变体模板——那些 URL 即使带 token 也可能返回 400/404（2026-09 实测）。
+        bounded = bounded_fullview_url(media)
+        if bounded:
+            logger.warning("Using DA fullview display image, not the uploaded original")
+            return bounded
+
+        # Preserve bare files for unrestricted/unknown signatures and GIFs.
+        # An extension alone does not establish that a bounded token permits it.
         def _is_file_url(u: str) -> bool:
             path = str(u).split('?', 1)[0].rstrip('/')
             return bool(re.search(r'\.(?:png|jpe?g|gif|webp|mp4|m4v)$', path, re.I))
@@ -344,10 +349,7 @@ class DeviantArtAPI:
             logger.warning("Full view not found, using base URI")
             url = base_uri
         
-        if token:
-            url += f"{'&' if '?' in url else '?'}token={token}"
-        
-        return url
+        return append_media_token(url, token)
     
     def _get_preview_url(self, media: Dict, base_uri: str, 
                         token: str, pretty_name: str) -> Optional[str]:
@@ -368,7 +370,4 @@ class DeviantArtAPI:
         
         url = base_uri + str(preview['c']).replace('<prettyName>', pretty_name)
         
-        if token:
-            url += f"{'&' if '?' in url else '?'}token={token}"
-        
-        return url
+        return append_media_token(url, token)

{devart_dl-4.4.2 → devart_dl-4.4.3}/da_downloader/download.py

@@ -85,6 +85,7 @@ class Downloader:
                     item.media_url,
                     extension=item.extension,
                     mature=item.mature,
+                    representation=item.metadata.get("representation"),
                 )
             )
         return [

{devart_dl-4.4.2 → devart_dl-4.4.3}/da_downloader/models.py

@@ -220,6 +220,7 @@ class MediaAsset:
     width: int | None = None
     height: int | None = None
     mature: bool = False
+    representation: str | None = None
 
     def to_dict(self) -> dict:
         data: dict = {
@@ -233,6 +234,7 @@ class MediaAsset:
             ("width", self.width),
             ("height", self.height),
             ("mature", self.mature),
+            ("representation", self.representation),
         ):
             if value is not None:
                 data[key] = value
@@ -270,6 +272,7 @@ def make_media_asset(
     *,
     extension: str | None = None,
     mature: bool = False,
+    representation: str | None = None,
 ) -> MediaAsset:
     ext = (extension or "").lstrip(".").lower() or None
     if ext == "txt":
@@ -285,4 +288,5 @@ def make_media_asset(
         extension=ext,
         mime_type=_MIME_BY_EXT.get(ext) or _DEFAULT_MIME[kind],
         mature=mature,
+        representation=representation,
     )

{devart_dl-4.4.2 → devart_dl-4.4.3}/da_downloader/official_api.py

@@ -24,6 +24,7 @@ from .errors import (
     ParseError,
 )
 from .oauth import OAuthSession
+from .wix_media import append_media_token, bounded_fullview_url
 
 logger = logging.getLogger(__name__)
 
@@ -337,12 +338,13 @@ def is_blurred(url: str) -> bool:
     return "blur_" in url
 
 def additional_media_urls(init_data: dict) -> list:
-    """Original-file URLs of a multimedia deviation's extra pages
+    """Media URLs of a multimedia deviation's extra pages
     (``deviation.extended.additionalMedia``, each entry nests its Wix
     descriptor under ``media``). Mirrors DeviantDrop's pick order: a playable
     video ``types[].b`` (best quality) wins over ``baseUri`` — a video page's
-    baseUri is its poster — then the ``baseUri`` file itself, then a
-    fullview template. A page whose only representation is a small preview is
+    baseUri is its poster — then a verified bounded fullview or the bare file.
+    A fullview is a display representation, not an uploaded original.
+    A page whose only representation is a small preview is
     skipped rather than saved as a fake original.
 
     Mature pages the server still returns with a ``blur_`` censored image are
@@ -378,6 +380,12 @@ def _descriptor_media_url(media: dict) -> Optional[str]:
     if videos:
         best = max(videos, key=lambda item: _VIDEO_QUALITY_RANK.get(item.get("q"), 0))
         return _append_token(str(best["b"]), media.get("token"))
+    bounded = bounded_fullview_url(media)
+    if bounded:
+        logger.warning(
+            "Using DA fullview display image for an additional page, not the uploaded original"
+        )
+        return bounded
     base = media.get("baseUri")
     if isinstance(base, str) and base and _FILE_EXT_RE.search(base.split("?", 1)[0]):
         return _append_token(base, media.get("token"))
@@ -399,7 +407,7 @@ def _append_token(url: str, raw_token: Any) -> str:
     token = raw_token[0] if isinstance(raw_token, list) and raw_token else raw_token
     if not token:
         return url
-    return f"{url}{'&' if '?' in url else '?'}token={token}"
+    return append_media_token(url, token)
 
 
 __all__ = ["OfficialApiClient", "OriginalDownload", "resolve_uuid", "deviation_init", "deviation_uuid", "additional_media_urls"]

{devart_dl-4.4.2 → devart_dl-4.4.3}/da_downloader/provider.py

@@ -15,6 +15,7 @@ from .api import ActionType, DeviantArtAPI
 from .errors import MediaUnavailableError, ParseError
 from .errors import AuthenticationError, NetworkError
 from .models import Deviation, DownloadItem, literature_text
+from .wix_media import bounded_fullview_url
 from .official_api import (
     OfficialApiClient,
     additional_media_urls,
@@ -152,7 +153,10 @@ class WebProvider:
             media_url=media_url,
             extension=_extension(deviation, media_url),
             mature=deviation.is_mature,
-            metadata={"quality": self.quality},
+            metadata={"quality": self.quality, **(
+                {"representation": "fullview", "original_available": False}
+                if media_url == bounded_fullview_url(deviation.media or {}) else {}
+            )},
         )
 
 
@@ -251,6 +255,11 @@ class OfficialProvider:
                 return
         main = self._item(self.client.deviation(uuid))
         yield main
+        bounded_urls = {
+            bounded_fullview_url(entry.get("media") or {})
+            for entry in ((init_data.get("deviation", {}).get("extended") or {}).get("additionalMedia") or [])
+            if isinstance(entry, dict) and isinstance(entry.get("media"), dict)
+        }
         for index, url in enumerate(additional_media_urls(init_data), start=1):
             item = DownloadItem(
                 artwork_id=f"{main.artwork_id}-{index}",
@@ -261,7 +270,10 @@ class OfficialProvider:
                 extension=_ext_from_url(url),
                 published_at=main.published_at,
                 mature=main.mature,
-                metadata={**main.metadata, "index": index},
+                metadata={**main.metadata, "index": index, **(
+                    {"representation": "fullview", "original_available": False}
+                    if url in bounded_urls else {}
+                )},
             )
             yield item
 

devart_dl-4.4.3/da_downloader/wix_media.py

@@ -0,0 +1,55 @@
+"""Bounded Wix URL selection, shared with DeviantDrop's verified 2026-10-29 fix.
+
+See tests/fixtures/deviantart-bounded-media-evidence.md. Like DAViewer/DAKit,
+keep display representations distinct from uploaded originals. Token payloads
+are selection hints only, never proof of authentication or permission.
+"""
+
+import base64
+import json
+import re
+from urllib.parse import parse_qsl, urlsplit
+
+
+def append_media_token(url: str, token: str) -> str:
+    if not token or any(key == "token" for key, _ in parse_qsl(urlsplit(url).query, keep_blank_values=True)):
+        return url
+    return f"{url}{'&' if '?' in url else '?'}token={token}"
+
+
+def bounded_fullview_url(media: dict) -> str | None:
+    """Return the supplied fullview only for a bounded static-image signature."""
+    try:
+        base = media.get("baseUri", "")
+        path = urlsplit(base).path
+        if not re.search(r"\.(?:png|jpe?g|webp|avif)$", path, re.I):
+            return None
+        tokens = media.get("token")
+        token = tokens[0] if isinstance(tokens, list) else tokens
+        payload = token.split(".")[1]
+        claims = json.loads(base64.urlsafe_b64decode(payload + "=" * (-len(payload) % 4)))
+        groups = claims.get("obj")
+        if not isinstance(groups, list) or not any(
+            isinstance(group, list) and any(
+                isinstance(rule, dict) and rule.get("path") == path and any(
+                    isinstance(rule.get(key), str) and re.fullmatch(r"<=\d+", rule[key])
+                    for key in ("width", "height")
+                ) for rule in group
+            ) for group in groups
+        ):
+            return None
+        types = media.get("types") or []
+        # Video sources take precedence over static posters.
+        if any(isinstance(t, dict) and t.get("t") == "video" and t.get("b") for t in types):
+            return None
+        full = next((t for t in types if isinstance(t, dict) and t.get("t") == "fullview"), {})
+        if full.get("r") not in (None, 0):
+            return None  # Only index 0 was independently verified.
+        url = full.get("b")
+        if not url and full.get("c"):
+            url = base + full["c"].replace("<prettyName>", media.get("prettyName") or "image")
+        if not url:
+            return None
+        return append_media_token(url, token)
+    except (ValueError, TypeError, AttributeError, IndexError, KeyError):
+        return None

{devart_dl-4.4.2 → devart_dl-4.4.3/devart_dl.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: devart-dl
-Version: 4.4.2
+Version: 4.4.3
 Summary: 专业的 DeviantArt 作品批量下载工具 - Professional DeviantArt batch downloader with anti-ban features
 Author: DeviantArt Downloader Team
 License-Expression: MIT

{devart_dl-4.4.2 → devart_dl-4.4.3}/devart_dl.egg-info/SOURCES.txt

@@ -24,6 +24,7 @@ da_downloader/official_api.py
 da_downloader/path.py
 da_downloader/provider.py
 da_downloader/targets.py
+da_downloader/wix_media.py
 devart_dl.egg-info/PKG-INFO
 devart_dl.egg-info/SOURCES.txt
 devart_dl.egg-info/dependency_links.txt
@@ -32,6 +33,7 @@ devart_dl.egg-info/requires.txt
 devart_dl.egg-info/top_level.txt
 tests/test_api.py
 tests/test_archive.py
+tests/test_bounded_media.py
 tests/test_cli.py
 tests/test_conformance.py
 tests/test_download.py

{devart_dl-4.4.2 → devart_dl-4.4.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "devart-dl"
-version = "4.4.2"
+version = "4.4.3"
 description = "专业的 DeviantArt 作品批量下载工具 - Professional DeviantArt batch downloader with anti-ban features"
 readme = "README.md"
 requires-python = ">=3.10"

devart_dl-4.4.3/tests/test_bounded_media.py

@@ -0,0 +1,104 @@
+"""Real captured CDN 403 regression; provenance is beside the fixture."""
+
+import base64
+import copy
+import json
+from pathlib import Path
+from types import SimpleNamespace
+
+import pytest
+
+from da_downloader.api import DeviantArtAPI
+from da_downloader.models import Deviation, DownloadItem, make_media_asset
+from da_downloader.official_api import _descriptor_media_url, additional_media_urls
+from da_downloader.provider import OfficialProvider, WebProvider
+from da_downloader.targets import TargetParser
+from da_downloader.wix_media import append_media_token, bounded_fullview_url
+
+
+def test_existing_query_and_signature_are_preserved():
+    assert append_media_token("https://example.test/a.jpg?download=1", "fresh") == "https://example.test/a.jpg?download=1&token=fresh"
+    signed = "https://example.test/a.jpg?download=1&token=existing"
+    assert append_media_token(signed, "fresh") == signed
+
+SAMPLES = json.loads((Path(__file__).parent / "fixtures/deviantart-bounded-media.json").read_text(encoding="utf-8-sig"))["samples"]
+DESCRIPTORS = [m for d in SAMPLES for m in [d["media"], *[e["media"] for e in d["extended"]["additionalMedia"]]]]
+
+
+@pytest.mark.parametrize("media", DESCRIPTORS)
+def test_captured_bounded_media_selects_fullview_in_both_paths(media, caplog):
+    full = next(t for t in media["types"] if t["t"] == "fullview")
+    expected = media["baseUri"] + full["c"].replace("<prettyName>", media["prettyName"]) + "?token=" + media["token"][0]
+    api = DeviantArtAPI({}, {})
+    assert api._get_full_view_url(media, media["baseUri"], media["token"][0], media["prettyName"]) == expected
+    assert _descriptor_media_url(media) == expected
+    assert "not the uploaded original" in caplog.text
+
+
+def test_extra_page_is_not_lost_or_replaced_by_small_preview():
+    result = additional_media_urls({"deviation": SAMPLES[0]})
+    assert len(result) == 1
+    assert "-fullview.jpg?token=" in result[0]
+    assert "-pre.jpg" not in result[0]
+
+
+@pytest.mark.parametrize("token", [None, [], "opaque", "a.bad.c"])
+def test_unknown_token_preserves_original_selection(token):
+    media = {**SAMPLES[1]["media"], "token": token}
+    assert bounded_fullview_url(media) is None
+
+
+def test_unbounded_original_gif_and_video_stay_unchanged():
+    media = copy.deepcopy(SAMPLES[1]["media"])
+    payload = media["token"][0].split(".")[1]
+    claims = json.loads(base64.urlsafe_b64decode(payload + "=" * (-len(payload) % 4)))
+    claims["obj"][0][0].pop("width")
+    claims["obj"][0][0].pop("height")
+    payload = base64.urlsafe_b64encode(json.dumps(claims).encode()).decode().rstrip("=")
+    media["token"] = [f"test.{payload}.REDACTED"]
+    assert bounded_fullview_url(media) is None
+    assert _descriptor_media_url(media).startswith(media["baseUri"] + "?token=")
+    gif = {**media, "baseUri": media["baseUri"].replace(".jpg", ".gif")}
+    assert _descriptor_media_url(gif).startswith(gif["baseUri"] + "?token=")
+    video = {**SAMPLES[1]["media"], "types": [{"t": "video", "b": "https://example.test/movie.mp4", "q": "1080p"}]}
+    assert _descriptor_media_url(video).startswith("https://example.test/movie.mp4")
+
+
+def test_no_fullview_does_not_substitute_preview():
+    media = copy.deepcopy(SAMPLES[1]["media"])
+    media["types"] = [t for t in media["types"] if t["t"] != "fullview"]
+    assert bounded_fullview_url(media) is None
+    assert _descriptor_media_url(media).startswith(media["baseUri"] + "?token=")
+
+
+def test_original_endpoint_still_wins_and_fullview_metadata_is_explicit(monkeypatch):
+    media = SAMPLES[1]["media"]
+    deviation = Deviation(deviation_id="1403090181", title="Untitled", author="lol1729", url="https://www.deviantart.com/lol1729/art/Untitled-1403090181", media=media, is_downloadable=True, is_mature=False, deviation_type="image")
+    api = DeviantArtAPI({}, {})
+    monkeypatch.setattr(api, "_get_original_download_url", lambda _: "https://example.test/original.jpg")
+    assert api.get_download_url(deviation, "o") == "https://example.test/original.jpg"
+    item = WebProvider(api, quality="f")._to_item(deviation)
+    assert item.metadata["representation"] == "fullview"
+    assert item.metadata["original_available"] is False
+    assert item.extension == "jpg"
+    asset = make_media_asset(item.artwork_id, 0, item.media_url, extension=item.extension, representation=item.metadata["representation"])
+    assert asset.to_dict()["representation"] == "fullview"
+
+
+def test_additional_page_metadata_does_not_claim_the_main_original_quality(monkeypatch):
+    import da_downloader.provider as provider_mod
+
+    sample = SAMPLES[0]
+    uuid = sample["extended"]["deviationUuid"]
+    monkeypatch.setattr(provider_mod, "deviation_init", lambda *a, **kw: {"deviation": sample})
+    provider = OfficialProvider(SimpleNamespace(deviation=lambda _: {}), quality="o")
+    main = DownloadItem(artwork_id=uuid, url="https://example.test/art", title="Work", author="Artist", media_url="https://example.test/original.png", extension="png", metadata={"quality": "o"})
+    monkeypatch.setattr(provider, "_item", lambda _: main)
+    items = list(provider.resolve(TargetParser.parse("1402511705")))
+    assert len(items) == 2
+    assert items[0] is main
+    assert "representation" not in main.metadata
+    assert items[1].artwork_id == uuid + "-1"
+    assert items[1].metadata["representation"] == "fullview"
+    assert items[1].metadata["original_available"] is False
+    assert items[1].extension == "jpg"