Desktop Cloud Transform (DCT)

Project Page

https://joshuatz.com/projects/applications/desktop-cloud-transform-dct/

Pre-built release

Find under the releases tab on Github.

What is this?

Desktop Cloud Transform (DCT) is a desktop GUI application written with QT/C++ that allows you to upload image files, and optionally apply advanced transformations on "the cloud", which are then downloaded back to your computer.

The benefit is that you can apply extremely complicated combinations of automated transformations and effects on the cloud instead of your computer, which means you don't need to have Photoshop installed, or deal with complex scripting.

You can also just use it as a drag-and-drop uploader for Cloudinary, so that you can drop a local file onto the bottom upload area and instantly get back a public URL to use anywhere you like.

This Git readme file is really meant to cover the technical parts of the application and how to use it, rather than purpose, so for more info on the background and uses for this project, be sure to check out my project page here.

Disclaimers:

First, this was built mainly for myself, most over the course of a week, and as a way to get up to speed on some QT/C++ stuff. Please don't view it as production-level quality code.

Also, although DCT makes extensive use of Cloudinary and its API, it is not affiliated with Cloudinary in any official capacity. I am not responsible for how you use their services, and it is up to you as a user to know their TOS and be responsible with your usage of their services while using my application.

How to use it

GUI

What are DCT "Configurations"?

A configuration is a set of DCT settings (such as how to save the resulting file) and Cloudinary transformation settings (such as adding overlays, modifying colors, etc.). Once you have created a configuration, you can drag and drop a file into its row to upload your image, have the config settings applied, and the resulting image file downloaded back to your hard drive. There are a few preloaded configs so you can get a feel for how they work. You can delete or modify configs at any time.

Macro support

Macros are case-insensitive; I'm using camelcase below just to make it easier to read. Macros should always be enclosed in braces, with no spaces. Certain transformations are only supported on "outgoing transformations", which are transformations that are applied after an image has been uploaded.

Macro	Incoming Transformation	Outgoing Transformation	What it does
{uploaded}	No	Yes	Inserts public ID as overlay into chain - "l_PUBLICID"
{uploadedId}	No	Yes	Inserts public ID, as-is
{width}	Yes	Yes	Inserts width in pixels of local file being uploaded
{height}	Yes	Yes	Inserts height in pixels of local file being uploaded
{filename}	Yes	Yes	Inserts the filename including the extension
{filenameNoExt} or {basename}	Yes	Yes	Inserts the filename without the extension
{createdStamp}	Yes	Yes	Inserts the timestamp of when the uploaded file was created, in secs since epoch
{cloudinaryCloudName}	Yes	Yes	Inserts your cloudinary cloud name / id
{timestamp}	Yes	Yes	Inserts the current time, as secs since epoch

Building

First, note that this contains git submodules, so you either need to clone with git clone --recursive, or use git submodule update --init after cloning.

Actual building:

So far I've only built for Windows, but there shouldn't be any Win specific dependencies in use. On my todo list is to build for Linux and OSX and see if there are any issues with the file read/writes
Compile tested with MSVC. Should work with MinGW too.
Steps should be as easy as:
1. Clone repo
2. Open in QT Creator
3. Build Release
Make sure you have openSSL in your system path

Deploying

I'm using windeployqt to pull in DLLs to bundle with my exe for release. however, it doesn't pull in OpenSSL although it is needed for the Cloudinary API calls.

License

Licensed under Apache 2.0 See LICENSE file for details.

Notes to self:

Stats logging

Syntax:

Stats::getInstance()->logStat(QString category, QString action,bool usedCloudinary);

Example:

Stats::getInstance()->logStat("application","startup",false);

Storage

Stats are stored with id as primary key (auto-increment) and column with timestamp, as sec since epoch (auto generated by Stats class)

Established combos

Row	Category	Action	Uses Cloudinary (credits)
1	application	upload	false
2	cloudinary	upload	true
3	cloudinary	transform	true
4	cloudinary	store	true
5	application	download	false
6	cloudinary	download	true

Considering Cloudinary Quotas

A single action (uploading an image) can trigger many events, some of which can count against a quota. For example
- Uploading an image, and store original = true, and downloading the final transformed image
  - This would use rows 1 & 5
  - This would also use credit rows 2, 3, 4, & 6
  - Currently, the way I'm doing this it only uses credit row 4 (storing an asset) once, since it stores the original, and then to get the transformed version, it simply fetches the original public URL + transformation string. The transformed version is not actually stored on Cloudinary - it is just a derived asset.

Docs

Pretty much best doc link is for "direct calls to api" - https://cloudinary.com/documentation/upload_images#uploading_with_a_direct_call_to_the_api
List of sections - https://cloudinary.com/documentation/cloudinary_references
Incoming transformations on upload - https://cloudinary.com/documentation/upload_images#incoming_transformations

TODOS

!! - High priority - !!

Add white borders inside icon so it shows up better on dark backgrounds (see taskbar screenshot)
Add support for signed URLs (necessary for downloading via transformation URL, if user has enabled "strict transformations")
Fix help tooltip overflow
Preserve filename on regular uploads (right now, it is generating a new unique ID, regardless of filename)
Add "duplicate" button, to clone existing configs and modify
Work on build script for creating deployment releases, and try to reduce output size
~~Timeout for auto-close of success modal needs to be auto-cancelled if clicked on~~
- For right now, just removed auto-close entirely

Backlog / Should implement

Highlight transformation string on click for easy copy and paste
DB migrations for upgrades to app
Refactor naming of classes, objects, methods - distinction between transformations and configs
Tag items uploaded through this tool to make easier to find online
- https://cloudinary.com/documentation/image_upload_api_reference#tags_method
Internal stats collection and display to user
- Optional: API quota warning configured through globalSettings
Unit tests
Cmd line support
Use richtext in textarea for link - https://doc.qt.io/qt-5/richtext-html-subset.html
Add support for more hosts (Imgur anonymous upload as top priority)

Icebox / optional:

When uploading images, use filename to check if already uploaded, and if so, offer option to "rename" version on cloudinary sever
- Use https://cloudinary.com/documentation/image_upload_api_reference#rename_method
Move some reusable JS logic stuff out of QML files and into reusable JS files

admariner / desktop-cloud-transform Goto Github PK

desktop-cloud-transform's Introduction