Skip to content

v0.11.12 - Read-only cloud drives (.strm), live scan progress & scraper upgrades

Choose a tag to compare

View all tags
@github-actions github-actions released this 12 Jul 12:47
· 807 commits to main since this release
v0.11.12
f169dc0

v0.11.12 - Read-only cloud drives (.strm), live scan progress & scraper upgrades

English version below ↓

一行安裝（推薦）

macOS:

curl -fsSL https://raw.githubusercontent.com/slive777/OpenAver/main/install.sh | bash

Windows (PowerShell):

irm https://raw.githubusercontent.com/slive777/OpenAver/main/install.ps1 | iex

💡 不想開 PowerShell 的 Windows 用戶：也可以到下方 Assets 下載 OpenAver-Windows-Setup.bat，雙擊即可跑完同一套安裝。

自上次發布（v0.10.11）以來累積了整個 v0.11 系列的更新，主軸有五個：新增唯讀雲端硬碟支援（並可用 .strm 掛進媒體伺服器）、掃描補資料看得見每一片、命名設定圖像化更難打錯、刮削資訊更完整，以及搜尋更準更順。

☁️ 唯讀雲端硬碟支援

影片放在掛載的網盤（115、Google Drive 等）上？現在可以把來源標成「唯讀」——OpenAver 只讀檔名、線上刮元數據，把 NFO、封面、劇照全寫到你指定的本地資料夾，雲端完全不寫入、也不下載原圖。

  • 唯讀來源刮好的資料一片一夾放在本地，雲端資料夾一個位元組都不動。
  • OpenAver 自己就能直接瀏覽、播放雲端上的原檔（從資料庫記的原始路徑串流）。

如果你也用 Emby / Jellyfin / Kodi，去設定頁面切換模式可生成 .strm 本地媒體庫：

  • .strm 是一堆幾十 bytes 的小文字檔（指向雲端原檔），媒體伺服器掃這些飛快、播放時才去串雲端——不必讓 Emby 一直爬整個網盤（又慢、又容易觸發網盤風控）。
  • 兩台機器看到的路徑不一樣時（OpenAver 這台 vs 媒體伺服器那台），設一組替換規則自動改寫進 .strm；之後改規則，既有的也一併同步更新。

🎬 掃描補資料，現在看得見每一片

掃描頁「補資料」不再是乾巴巴的數字——每一片進度都顯示動畫。

  • 逐片顯示狀態：搜尋中 / 命中 / 補到資料但無封面 / 查無 / 失敗 / 唯讀跳過，一眼看出每片結果。
  • 命中且有封面時，封面從進度卡飛進側邊欄「瀏覽」，落地帶微光；這次補了幾片會有累計標記。
  • 查無只是安靜一閃換下一片，不搶你的注意力。

⚙️ 命名設定圖像化，不再手打 {actor}

設定「資料夾 / 檔名」規則時，不用再手打 {actor}、{num} 這種容易打錯的文字——每個變數變成一顆看得懂的膠囊，用點的就好。

  • 番號、女優、片商、標題… 各是一顆膠囊，從選單一鍵插入，膠囊之間可自由加分隔符（-、_、[ ] 等），一眼看出檔名會長怎樣。
  • 資料夾層級可「新增一層 / 移除此層」，最多三層。

🔀 刮削資訊更完整

同一番號有多個版本時可以自己挑，各來源的評分與簡介也會一起寫進 NFO。

  • JavLibrary 撞號多版本手動切換：一個番號對到多個版本時，列出全部候選（新片優先），用 ‹ › 翻封面挑你要的那一版；搜尋、燈箱換源、結果卡替換三個地方都能切。
  • 評分 / 簡介寫進 NFO：6 個來源刮到的評分與簡介會寫進 NFO，讓 Jellyfin / Emby / Kodi 顯示（OpenAver 自己的介面刻意不顯示——自己逛用不到）。

🔍 搜尋更準、更順

精確番號搜尋改成「照你的來源順序一家一家問」，該先查的來源真的先查。

  • 精確番號不再一次打散所有來源再合併，改成優先序查詢（例如 DMM 排第一，就真的先問 DMM）。
  • 搜尋列右側的控制項不再互相擠壓；在搜尋頁整理入庫時的「封面飛入」動畫也更順。

⚡ 其他改進與修復

  • JavDB 打包版修復：之前 Windows / macOS 安裝版指定 JavDB 一律搜不到（含裝在中文 / 日文路徑的情況），已全部修好——升級重開即可。
  • 番號辨識修正：PARATHD- 這類 7 字母前綴的番號拖檔不再掉開頭字母。
  • 另外也做了多項穩定性與安全性修正。

English

This release bundles the whole v0.11 series since the last published release (v0.10.11). Five highlights: read-only cloud-drive support (with .strm files to plug into your media server), scan progress you can actually watch clip by clip, a point-and-click naming template editor so you don't have to type {actor} by hand, richer scraped metadata, and more accurate, smoother search.

☁️ Read-only cloud drive support

Keep your videos on a mounted cloud drive (115, Google Drive, etc.)? You can now mark a source as "read-only" — OpenAver only reads filenames and scrapes metadata online, writing the NFO, cover, and stills to a local folder you choose. Nothing is ever written to the cloud drive, and the original images are never downloaded onto it.

  • Everything scraped for a read-only source lands in its own folder on your local disk — the cloud folder isn't touched, not even one byte.
  • OpenAver can still browse and play the original files on the cloud drive directly, streaming from the original path stored in the database.

If you also run Emby / Jellyfin / Kodi, flip a mode switch on the Settings page to generate a local .strm library:

  • .strm files are tiny text files — just a few dozen bytes each — that point to the original file on the cloud drive. Your media server scans these in a flash and only streams from the cloud when you hit play, so Emby never has to crawl the whole cloud drive (slow, and a good way to trip the drive's rate limits).
  • If the two machines see different paths for the same drive (the one running OpenAver vs. the one running your media server), set up a path-replacement rule and it's rewritten into every .strm automatically — change the rule later and existing files get updated too.

🎬 Scan progress you can actually see, clip by clip

The "fill in missing data" scan on the Scan page is no longer just a bare number — every single clip's progress now animates on screen.

  • Each clip shows its own status as it goes: searching / matched / matched but no cover / not found / failed / skipped (read-only) — so you can see every result at a glance.
  • When a match comes back with a cover, the cover flies from the progress card into the "Browse" sidebar and lands with a little glow; a running counter tracks how many clips got filled in this pass.
  • A "not found" result just quietly flashes and moves on to the next clip — it won't grab your attention.

⚙️ Naming templates go visual — no more typing {actor} by hand

Setting up folder / filename rules no longer means typing error-prone tokens like {actor} or {num} — every variable is now a readable pill you just click to insert.

  • ID, actor, studio, title, and more are each their own pill — pick one from the menu to insert it, add separators between pills (-, _, [ ], etc.) freely, and see at a glance exactly what the resulting filename will look like.
  • Folder levels can be added or removed with "Add level" / "Remove this level", up to three levels deep.

🔀 Richer scraped metadata

When the same ID has multiple versions, you can now pick which one you want yourself — and ratings and synopses from each source get written into the NFO too.

  • Manual version switching for JavLibrary ID collisions: when one ID maps to multiple versions, all candidates are listed (newest first) — flip through covers with ‹ › to pick the one you want. You can switch versions from search, from the lightbox's source-swap, and from the result card's replace action.
  • Ratings / synopsis written into the NFO: ratings and synopses scraped from 6 sources are now written into the NFO so Jellyfin / Emby / Kodi can display them (OpenAver's own interface deliberately leaves them out — you don't need them while browsing here).

🔍 Search, more accurate and smoother

Exact-ID search now queries your sources one by one, in the order you configured — the source you put first actually gets asked first.

  • Exact-ID lookups no longer fan out to every source at once and merge the results — they now query in priority order (put DMM first, and it really is asked first).
  • Controls on the right side of the search bar no longer crowd each other, and the cover-flying-in animation when organizing search results into your library is smoother too.

⚡ Other improvements & fixes

  • JavDB fixed in packaged builds: previously, choosing JavDB as a source simply returned no results in the installed Windows / macOS app (including when installed under a Chinese / Japanese path) — this is now fully fixed, just update and relaunch.
  • ID-recognition fix: dragging in files with a 7-letter prefix ID like PARATHD- no longer drops the first letter.
  • Plus a number of stability and security fixes.
Assets 5
Loading