This is possible but it's not trivial and you will likely take a performance penalty. The way SlickGrid works, it needs to be able to answer the following two questions rapidly:
- For a given row X, what is the top offset for that row?
- For a given offset Y, what row is at that offset?
When you use a static row height, answering these questions is trivial; however, with dynamic row height, you'll need to maintain a couple of extra data structures.
Here's roughly what I changed in Slick.Grid.js
- Add a new array to track row sizes. Initialize all rows to the default size
- Remove the css rules in createCssRules which set the cell height
- Add some code at the end of renderRows which checks the rendered height of each cell in the row and then sets the height of all cells to the maximum (and stores the height in the row size array). You also need to adjust the top offset based on the sum of heights of rows above the current one.
- Add code to the end of render to resize the canvas to the sum of all row heights.
- Update getViewport to return the top and bottom rows based on the sums of row heights.
- There are a handful of other places where options.rowHeight is used. You can ignore some of them but at the very least, anywhere the canvas is resized needs to be changed.
To make this practical (performant), you'll also need a cache of row top offsets (a cache of sums of row size). This will enable quick computation for the first question and will allow for binary search to answer the second.
I hope that helps.
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…