polypheny / polypheny-db Goto Github PK

Implement a function that generated valid and unique names for indexes, constraints and foreign keys.

Especially for long running task, it would beneficial to have a visual indication for the user that the UI is currently waiting on a response from Polypheny-DB.

Add a calendar and time element in the UI to easily add and change fields with a date/time type.

A config element which relates to the values of a specified enum.

See also #65

Add an Information element of type table, which allows rendering data in a tabular form.

Allow to export the current schema (or parts of it) as PolySQL-Script. It should be possible to select whether the export should include the currently stored data.

The UI should also support importing PolySQL scripts.

The Sort-Operator in the Plan-Builder of Polypheny-DB UI is broken in Firefox.

• Remove generated query (redundant because this is already in the header of the box)
• Only show "Affected Rows" if there is no result set

Or is there a reason to keep this information?

Add a special information element for durations. This element should already include the required stopwatch code. Furthermore, it should also be possible to simply provide a duration as long.

The element has to support interims and "sub durations". For every interim and sub duration there should be a label assigned. It would be nice if sub durations can also have interims (but this is not required).

`
d = new DurationElement("Execution Time");
d.start();

s1 = d.subDuration("Task 1");
s1.start();
TIME_CONSUMING_TASK_1
s1.end();

TIME_CONSUMING_TASK_2
d.interim("Foo");
TIME_CONSUMING_TASK_3

d.end();
`

The unit in which the duration it is displayed in the UI should be selected automatically. Furthermore, it should be possible to set the unit manually.

It would be nice if the durations can be visualized in the UI.

Simplify the plan builder by using dropdowns instead of free text.

Only suggested columns from tables for which there is a scan operation.

Support monitoring and configuring multiple instances of Polypheny-DB at the same time.

The current implementation requires to provide all data for the graph at ones. This is fine for graphs where you want to visualize a current situation. In case it is required to visualize the change over time, this is very inconvenient.

The information graph element should therefore be extend to allow adding single values. Furthermore, it is required to allow specifying the number of values which should be stored.

The UI should not provide any options for modifying a table which is read-only (data which is e.g. provided by the CSV adapter).The UI should not provide any options for modifiing a table which is read-only (data which is e.g. provided by the CSV adapter).

Requires #66

It is possible to place plan builder elements completely or partially outside the working area.

The update statements generated by the data view should make use of primary keys (if there is one defined).

Instead of a type filed, consider using the simple name of the class as type in the Config Manager and the Information Manager.

Add support for exporting query analyzer sessions to a file. This would allow to send it to a developer in case of a bug.

Allow configuring the log level of Polypheny-DB on a per class or package basis in Polypheny Control.

Instead of either the current solution, which requires either double clicking or typing the name of the element to delete, it would be more consistent to always present a model to the user that asks for approval. Whether approval is required at all for a delete operation should be configurable.

It is not possible to connect columns in the UML view on a tablet. Already existing relations are displayed correctly.

In the Query Plan Builder, it is not possible to grab operators from the left side.

Tested on both, iPadOS and Android.

The UI should check names (schemas, tables, columns, indexes, constraints, foreign keys) before submitting them to Polypheny-DB.

Rules:

• Starts with a letter
• Contains only letters, digits and underscores
• Length <= 100

Regex: ^[a-zA-Z_][a-zA-Z0-9_]*$

Add more descriptive text and explanations to the UI.

• Polypheny-DB Version (hash, build date)
• Polypheny-DB UI Version (hash, build date)
• Developers
• Used libraries and technologies
• License

It should be possible to open the logical query plan from the query analysis in the plan builder. This would allow to easily change complex query plans without needing to rebuild the complete plan.

The number of parameters is not equal for all data types supported by Polypheny-DB. While the current implementation only supports a parameter for VARCHAR and VARBINARY it should be possible to have an arbitrary number of parameters depending on the data type.

It should be possible to select whether the queries executed by the UI should immediately be committed or if the user should trig the commit manually. This allows creating a whole schema and adding data within one transaction using the UI tools.

The transaction/commit mode (auto commit (default) or manual) should be indicated by an icon in the UI. The transaction is bound to one instance (session) of the UI. If Polypheny-DB loses connection to this UI, the changes should be rolled back.

Add constraint, index, adapter and query interface names to the list of suggested names in the auto-completion provided in the SQL query console of the Polypheny-UI.

There is an editor component int the UI (src/app/components/editor). It most probably makes sense to add a new method in the backend in Polypheny-DB (webui/src/main/java/org/polypheny/db/webui/Crud.java) which returns a list of all names to names / identifiers to be considered.

It would be nice to have an Information element of type Text.

Only fetch a configurable number of elements instead of adding an limit clause to every query submitted via the query console.

For the queries executed by the UI it might be benefitial to stick to the limit / offset construt.

The monitor page in Polypheny-DB UI stops working if an insert query has been executed. There reason for this seems to be a wrong query plan entry in the main information manager.

Refactor and correctly integrate the InformationTable element. The HTML code should be part of the UI.

Add an information element which allows to nicely display java code in the UI.

It is currently not possible to update labels of an information element of type graph.

The suggestions should be more context-dependent (only suggest a column name if it is allowed and expected at this place in the query).

When performing a table scan with the plan builder, it is not yet possible to specify the schema of the table that should be scanned.

Add an Information element, which consists of a key and a value. This allows to easily output the value of an option. The rendering of the value should depend on the data type. It should therefore be considered to implement the value in a generic form. The key can always be a String.

Please be aware that this is a rather large issue for first time contributers!

In the current implementation, any changes to the default values are lost after a restart of Polypheny-DB. To change this, I propose the following changes to the configuration system:

• If there are changes to the configuration (e.g. via the UI) these changes are not directly written to the configuration file but are as in the current implementation only available until the next restart.
• To write the configuration to the config file there should be a button in the UI. Its icon should indicate whether there are unwritten configurations. By clicking on the button, the user is presented a diff between the currently active configuration and the configuration in the file.
• It should be possible to select whether the whole configuration or only those different to the default should be written to the file.
• It should also be possible to download the current configuration in the UI. The user should be able to decide whether the exported file should contain all values or only those different to the default values.
• It should be possible to trigger an import of configuration values from a file. If this is done via the UI it would be beneficial to show the changes in the form of a diff before they are applied.
• Independent of the origin of a request to update a config value (UI, file, ...) the checks have to be applied.

The existing configuration system already supports values from a file (application.conf) containing key-value pairs.

Due to foreign key constraints, it would be possible to already prefill the join operator with appropriate data.

Add length field similar to varchar in the UI.

Some of the input fields a framed green (for valid inputs?) while others are not. This should be consistent. Or is there a specific reason why this is handled like this?

A table scan element with public.emps results in a "Table 'public.emps' not found" message.

Use #74 to propose artificially generated index, foreign key and constraint names in the UI. Those names should be set as placeholder in the corresponding input fields of the UI.

Add missing operators in plan builder.