Iphone - when to calculate heightForRowAtIndexPath for a tableview when each cell height is dynamic?

I have seen this question asked many times but astoundingly, I have not seen a consistent answer, so I will give it a try myself:

If you have a tableview containing your own custom UITableViewCells that contain UITextViews and UILabels whose height must be determined at runtime, how are you supposed to determine the height for each row in heightForRowAtIndexPath?

The most obvious first idea is to calculate the height for each cell by calculating and then summing the heights of each view inside the cell inside of cellForRowAtIndexPath, and store that final total height for later retrieval.

This will not work however because cellForRowAtIndexPath is called AFTER heightForRowAtIndexPath.

The only thing I can think of is to do all the calculations inside viewDidLoad, create all the UITableViewCells then, calculate the cells height and store that in a custom field inside your UITableViewCell subclass, and put each cell in an NSMutableDictionary with the indexPath as the the key, and then simply retrieve the cell from the dictionary using the indexPath inside cellForRowAtIndexPath and heightForRowAtIndexPath, returning either the custom height value or the cell object itself.

This approach seems wrong though because it does not make use of dequeueReusableCellWithIdentifier, instead I would be loading all the cells at once into a dictionary in my controller, and the delegate methods would be doing nothing more than retrieving the correct cell from the dictionary.

I don't see any other way to do it though. Is this a bad idea - if so, what is the correct way to do this?

The way Apple implements UITableView is not intuitive to everyone and it's easy to misunderstand the role of heightForRowAtIndexPath:. The general intention is that this is a faster and light-on-memory method that can be called for every row in the table quite frequently. This contrasts with cellForRowAtIndexPath: which is often slower and more memory intensive, but is only called for the rows that are actually need to be displayed at any given time.

Why do Apple implement it like this? Part of the reason is that it's almost always cheaper (or can be cheaper if you code it right) to calculate the height of a row than it is to build and populate a whole cell. Given that in many tables the height of every cell will be identical, it is often vastly cheaper. And another part of the reason is because iOS needs to know the size of the whole table: this allows it to create the scroll bars and set it up on a scroll view etc.

So, unless every cell height is the same, then when a UITableView is created and whenever you send it a reloadData message, the datasource is sent one heightForRowAtIndexPath message for each cell. So if your table has 30 cells, that message gets sent 30 times. Say only six of those 30 cells are visible on screen. In that case, when created and when you send it a reloadData message, the UITableView will send one cellForRowAtIndexPath message per visible row, i.e. that message gets sent six times.

Some people are sometimes puzzled about how to calculate a cell height without creating the views themselves. But usually this is easy to do.

For example, if your row heights vary in size because they hold varying amounts of text, you can use one of the sizeWithFont: methods on the relevant string to do the calculations. This is quicker than building a view and then measuring the result. Note, that if you change the height of a cell, you will need to either reload the whole table (with reloadData - this will ask the delegate for every height, but only ask for visible cells) OR selectively reload the rows where the size has changed (which, last time I checked, also calls heightForRowAtIndexPath: on ever row but also does some scrolling work for good measure).

See this question and perhaps also this one.

So, I think you can do this without having to create your cells all at once (which, as you suggest, is wasteful and also probably impractical for a large number of cells).

UIKit adds a couple of methods to NSString, you may have missed them as they're not part of the main NSString documentation. The ones of interest to you begin:

- (CGSize)sizeWithFont...

Here is the link to the Apple docs.

In theory, these NSString additions exist for this exact problem: to figure out the size that a block of text will take up without needing to load the view itself. You presumably already have access to the text for each cell as part of your table view datasource.

I say 'in theory' because if you're doing formatting in your UITextView your mileage may vary with this solution. But I'm hoping it will get you at least part way there. There's an example of this on Cocoa is My Girlfriend.

An approach I have used in the past is to create a class variable to hold a single instance of the cell you are going to be using in the table (I call it a prototype cell). Then in the custom cell class I have a method to populate the data and determine the height the cell needs to be. Note that it can be a simpler variant of the method to really populate the data - instead of actually resizing a UILabel in a cell for example, it can just use the NSString height methods to determine how tall the UILabel would be in the final cell and then use the total cell height (plus a border on the bottom) and UILabel placement to determine the real height. YOu use the prototype cell just to get an idea of where elements are placed so you know what it means when a label is going to be 44 units high.

In heightForRow: I then call that method to return the height.

In cellForRow: I use the method that actually populates labels and resizes them (you never resize the UITableView cell yourself).

If you want to get fancy, you can also cache the height for each cell based on the data you pass in (for instance it could just be on one NSString if that's all that determines height). If you have a lot of data that's often the same it may make sense to have a permanent cache instead of just in-memory.

You can also try estimating line count based on character or word count, but in my experience that never works - and when it goes wrong it usually messes up a cell and all the cells below it.