Koneko

Changelog

* Interactive helper UI to assist configuring lscat numbers (`lscat_app.py`).
- Configuration assistant
- image_width
- image_height
- images_x_spacing
- images_y_spacing
- image_thumbnail_size
- page_spacing
- gallery_print_spacing
- users_print_name_xcoord
- Display 'testgallery' and 'testuser' paths (for internal and debugging use, just copy over the files)
- Browse the cache and display a dir
- Display any path (but all names must be left-padded with a three digit number)

* Remove `users_page_spacing`: it's just `gallery_page_spacing - 3`
* Rename `gallery_page_spacing` setting to `page_spacing`
* Show frequently searched IDs/strings with an option to select them with pick.
- On main screen:
- prompt for a mode or
- select all.
- On prompt loop, allow '!freq' to launch pick that goes directly to the mode.

Fixed

* Fixed a bug where downloading a single image crashes
* Fixed a bug where some images or mode might not update outdated images
* `first_img` attribute no longer used
* If an invalid input is given in a loop prompt, it will ask for another, rather than going back to `main()`
* Fixed giving invalid cli args crashing by re-prompting the user. However, leaving (including with ctrl+c) is impossible.
* Fixed one-letter-two-digit-sequence not working
* Fixed gallery -> image crashing

Code maintenance

* Rename `action_before_prefetch()` to `maybe_join_thread()`
* Extract out messy condition to bool variables in `prompt.py`
* Use `shutil.rmtree` instead of `rm` via `os.system`
* Improved attribute and method names in Loop classes
* Removed useless `prompted` boolean variable
* Remove pipey dependency in favour of `flow` from `results`
* Use `funcy.once` instead of boolean guards in `api.py`
* Simplified `generate_orders()` to 2 lines
* Wrap some test in a try-finally block so resources are always restored

Changelog

* lscat now slightly faster as the image number is calculated only when it finishes downloading, not on every recursion.
* Improvement for lscat: the screen will be filled with the most rows as possible
* HACKING.md split from CONTRIBUTING.md to better assist other developers
* Image view can now preview the next four images in multi-image posts (need to enable its experimental flag)

Fixed
* Fixed opening the manual in user modes not working
* Fixed bug in lscat: number of columns now determines when to print page spacings

Code maintenance
* Extract out common behaviour from Gallery and Users ui classes into AbstractUI
* Gallery classes is now structued like Users. The most significant difference is the use of `offset` to prefetch and navigate pages.
* Unified API requests for each mode
* Simplified code everywhere
* Reduced nesting with early returns
* Removed dead code
* Split up functions in:
* `config.py`
* `main.py`
* `cli.py`
* `download.py`
* Renamed data classes (eg GalleryJson to GalleryData)
* Simplify `ui.display_page()`
* Extract common code in `prompt.py`; follow a common structure
* Remove unused cytoolz dependency
* More functional style
* Use `placeholder` to make lambdas look better
* Use `pipey` for functional-style postfix pipes
* Using `returns.result` Result container in `config.py` and `splitpoint`
* Make tests more professional
* Move functions around modules to better reflect (im)purity

Changelog

Features
* Users can now configure lscat numbers, including:
* Image width
* Image height
* Image thumbnail size (For pixcat; the above two are more accurate)
* x and y spacing/padding between images
* Spacing between printed text
* Spacing between pages
* See MANUAL.md and example_config.ini for more details
* Boolean settings in the config now accepts ('1', 'yes', 'true', or 'on') for True, and anything else as False.
* Warn user for invalid credentials
* Image mode can now jump to specified page number, in a multi-image post
* Image mode now prefetches the next image in another thread
* Fixed spinner not propagating exceptions
* On reloading and clearing cache, the directory to be deleted will be shown
* Fixed clearing the cache not working
* Added -v/--version option
* Added --help option
* noprint option renamed to print_info (in the affirmative), and True by default (if no key or setting found)

Internal code
* Added lots of tests
* CI now tries to launch the application to make sure it doesn't crash on start and I didn't make any silly mistakes when building
* Updated requirements.txt
* Moved stuff like begin prompt and show man loop from `utils.py` to `screens.py`
* Refactored cli.py and prompt.py
* Move config functions in lscat and utils to config.py

Changelog
* **Depreciated** legacy lscat (the shell script) and lsix support.
* Fixed import errors in v0.6 (uploaded on PyPI)

Features
* If there are cached images in the users modes (3 & 4), it will be shown immediately before requesting and parsing.
* Gallery and User modes (all modes except mode 2) now prefetch the next page in another thread/in the background, freeing up the prompt to respond to user key presses.
* If the cache dir has images, it will display the images before waiting for login, making it much faster.
* Improved experience of invalid input: it will no longer reload to main after 2 seconds.
* Fixed bug in Image mode where the first picture of a multi-picture post prints the current page indicator before the picture is displayed

Terminal reliability
* lscat now calculates params using terminal size
* Add option to turn off printing messages in lscat

Code
* Removed the `raw` attribute from data classes
* Moved view_post_mode() from main.py to ui.py
* Renamed img_post_page_num to page_num
* Updated unit tests and added more tests
* Revamped lscat to use only the generators to display the images, rather than for loops, heavy classes, and fragile inheritance
* Users view actually got faster, because two yield statements are apparently faster than two nested for loops
* No significant speed difference for Galley views.
* Continuous integration tested on linux and macos

Changelog

Hotfix:
* Fixed bug where mode 5 won't launch

Changelog

Mostly code structural improvements and refactoring:
* No more global variables (API_QUEUE, API_THREAD, API)
* Use APIHandler() class for all api requests, in `api.py`
* Separating mode UI from `main.py`
* Move download functions to `download.py`
* Move docopt stuff to `cli.py`
* Move JSON manipulation to `data.py`
* KONEKODIR now in `__init__.py`
* Merge GalleryLikeMode into AbstractGallery; but pull out download and requests
* IllustFollowModeLoop is now a function
* Use pathlib instead of strings

Features:
* Colored multi-image indicator in gallery view (mode 1, 5)
* Image mode (mode 2) can now display the image in full resolution (useful for wide pictures)

Bug fixes:
* Fixed the multi-image indicator for posts in gallery view (mode 1, 5) disappearing after going next or previous page.