alex6480 / studycards Goto Github PK

So that it's possible for the user to click anywhere and put the input in focus.

When the study draft is complete a weighing algorithm needs to be implemented in the following cases:
[ ] When drawing a card from a study deck
[ ] When selecting known cards while assembling a study deck

It should be possible to tag cards within a set, in order to study subsets of a set of to find cards easier when editing them.

Only some components should be connected to redux, but these should not be in charge of any rendering logic. These are containers and should be separated from the normal components.

Besides the StudyData that contains information about the current study status (which cards are due when and such) there should be data for previous study sessions.

This data should be used to generate graphs of how the states of the cards progresses and study heatmaps showing activity.

When set meta is updated, the following should be done to prevent conflicts:

1. Set meta is updated and the request is being sent to the server
2. Before the request is complete, the set meta is updated again
3. The first request is cancelled and a new request that is the result of both updates is sent.

That way only one request will be sent at once to prevent race conditions. And the request will include information from previous cancelled requests.

If possible compile the used icons using webpack

StudyCardsApp should not be connected.
The connection should be moved down one level

Right now all errors prevent the file from being loaded even though some errors can be corrected by the parser.
If the error can be corrected by the parser, it should still be logged and shown to the user, but the set should still be imported.

This can be done by adding a severity parameter to errors and only preventing import on severe errors.

Create an overview of all cards that can easily be printed

Study mode is used to actually read the cards.
What determines which cards are shown, in which order and how many?

An algorithm needs to be developed.

Some cards can be studied in both directions. E.g. if they are words in a foreign language.

Casa -> House

The card can also be studied

House -> Casa

This should be possible when studying. Perhaps by adding an option so cards can be studied in reverse (or only cards with certain tags).

Right now the code quality is all over the place. A linter should enforce a strict style.

It should be possible to edit the current card in the study session without interrupting the study session.

Tests need to be created in general, but the parser especially needs them so that modifications won't prevent it from reading older export file formats.

Currently the text input is not big enough for the tag length limit.
The text input should automatically size based on how much is written in it.

Instead of having a setcontainer, the individual set pages just render the necessary components for the header themselves. By encapsulating these components the boilerplate will be minimal.
This setup is much more convenient when combined with React router later on. It also allows the individual segments to reload only the parts that are relevant for them.

Image and occlusion support is a must, but how will it function?

Right now there is one parser per export file version.
Since changes between version are likely going to be small, it seems more logical to break the parser up into smaller parsers that each handle whatever differences there may be from file to file.

It may be difficult for new people to understand exactly what decisions are being made in regard to which cards are selected for studying. Therefore tooltips should be added to explain these concepts.

[ ] X cards are new. Tooltip on new, explaining what a new card is
[ ] X cards are ready for review. Tooltip on review explaining what that means.
[ ] A help orb besides the buttons for selecting how difficult the card is. Clicking it shall explain exactly what each button does
[ ] A help orb in the editor, showing a window with documentation about the rich text editor. Especially features not found in other editors need explaining, like Reveal (and inline images).