Describe the bug
When downloading an image or video from a reblogged tumblr post, using either API or SVC as crawler, with "Download reblogged posts" enabled and with a custom filename template containing either "%o", "%q", or "%r", these template items are filled in as if the post weren't a repost at all in the actual resulting filename. This bug doesn't seem to be affected by any other settings.
For example, consider the download of a JPG image. Suppose the image's Tumblr filename is "somefile.jpg", the blog you're downloading from is called "someblog", the image originates from a post by a blog called "someorigin", and you're using "b=%b o=%o q=%q r=%r f=%f" as the filename template. Then the expected filename for the downloaded image is "b=someblog o=someorigin q=someorigin r=repost f=somefile.jpg", but the actual output is "b=someblog o= q=someblog r= f=somefile.jpg". This output is what one would expect for an original post by the "someblog" blog, rather than a repost.
To Reproduce
We'll use the icecreamhdaches blog as an example; it's the first usable example I stumbled across with a limited total storage footprint. The latest post is a repost from jamzdens, which contains a single image, whose Tumblr filename is "4e5f1ed8171f0812b604025fa1e7f3d01ebebae7.pnj". Note that this image will be converted to a JPG upon download.
Steps to reproduce the behavior:
- Open the app and optionally clear the blog list. If the example blog happens to be one already in the blog list, you'll have to remove it and remove the download folder associated with it first before continuing.
- Add the blog to the list of blogs in the app, be it through clipboard recognition or by entering "https://icecreamhdaches.tumblr.com/" in the text box in the bottom left of the app window explicitly.
- Click on the blog in the blog list and go to the Details pane.
- Modify the Filename template to, say, "b=%b o=%o q=%q r=%r f=%f".
- Enqueue the blog by right-clicking it in the blog list and then clicking "Enqueue selection".
- Click "Download" in the action bar at the bottom of the app window.
- Wait for the downloading process to complete, or wait until the post in question has been downloaded and pause the queue (which shouldn't take long as of the time of writing, as it's the most recent post).
- Navigate to the download location in FIle Explorer, which is the "icecreamhdaches" subfolder of the "Blogs" folder within the app installation directory by default.
- Find the file which is called "b=icecreamhdaches o= q=icecreamhdaches r= f=4e5f1ed8171f0812b604025fa1e7f3d01ebebae7.jpg".
- You've encountered the bug: the file should be called "b=icecreamhdaches o=jamzdens q=jamzdens r=repost f=4e5f1ed8171f0812b604025fa1e7f3d01ebebae7.jpg".
Expected behavior
Images and videos downloaded as parts of reposts should follow the filename template's format correctly.
Screenshots
N/A
Desktop:
- TumblThree version: v2.19.2.514
- OS: Windows 11 Home 25H2
- Browser: Brave Browser v1.84.141
(Btw, can't you add the ability to copy this version info in the "About" screen? You can't even select the text atm.)
Additional context
N/A
Describe the bug
When downloading an image or video from a reblogged tumblr post, using either API or SVC as crawler, with "Download reblogged posts" enabled and with a custom filename template containing either "%o", "%q", or "%r", these template items are filled in as if the post weren't a repost at all in the actual resulting filename. This bug doesn't seem to be affected by any other settings.
For example, consider the download of a JPG image. Suppose the image's Tumblr filename is "somefile.jpg", the blog you're downloading from is called "someblog", the image originates from a post by a blog called "someorigin", and you're using "b=%b o=%o q=%q r=%r f=%f" as the filename template. Then the expected filename for the downloaded image is "b=someblog o=someorigin q=someorigin r=repost f=somefile.jpg", but the actual output is "b=someblog o= q=someblog r= f=somefile.jpg". This output is what one would expect for an original post by the "someblog" blog, rather than a repost.
To Reproduce
We'll use the icecreamhdaches blog as an example; it's the first usable example I stumbled across with a limited total storage footprint. The latest post is a repost from jamzdens, which contains a single image, whose Tumblr filename is "4e5f1ed8171f0812b604025fa1e7f3d01ebebae7.pnj". Note that this image will be converted to a JPG upon download.
Steps to reproduce the behavior:
Expected behavior
Images and videos downloaded as parts of reposts should follow the filename template's format correctly.
Screenshots
N/A
Desktop:
(Btw, can't you add the ability to copy this version info in the "About" screen? You can't even select the text atm.)
Additional context
N/A