IDfsuFile Interface

The numbers in the ElementTable are node numbers, not indices! Each value in the table are between 1 and number-of-nodes.

Setting the StartDateTime for an axis that is not of calendar type, will throw an exception.

Setting the time step for an axis that is not of equidistant type, will throw an exception.

Compared to the Flush method; the Flush will update statistics of all items (if enabled) and all modifications to the header data. This method will only update the time part. The Flush method can be an expensive operations especially for files with many dynamic items. This method is independent of the file at hand.

If the data from one timestep is to be saved for later, the user is responsible to copy the data before using this method again.

Reads the next dynamic item-timestep. First time called it returns the first timestep of the first item. It cycles through each timestep, and each item in the timestep. For a file with 3 items, it returns (itemnumber, timestepIndex) in the following order: (1,0), (2,0), (3,0), (1,1), (2,1), (3,1), (1,2), etc.

If one of the ReadItemTimeStep is called with for example (1,4) the next call to this method will continue from there and return (2,4).

If one of the methods reading/writing static item data is called, the iterator is reset, and the next call to this method again returns the first item-timestep.

This is the most efficient way to iterate through all the items and timesteps in a file, since it iterates exactly as the data is stored on the disk.

Can also be used if it is required to restart reading the static items.

If the item-timestep exists already, data is updated. If it does not exist, the item number and timestep index must match exactly the next item-timestep after the last one in the file.

The size of the data must match the data size of the item that is to be written.

The time value is only relevant for files with non-equidistant time axis. For files with an equidistant time axis, the time value is ignored, and a zero can be used.

Writes the next dynamic item-timestep.

If the file pointer points to the end of the file, this will append a new item-timestep to the file. If the file pointer points to an existing item-timestep, the data of that item-timestep is updated.

Remember that the file pointer position depends on the mode that the file was opened: In edit mode the file pointer points at the first item-timestep. In append mode the file pointer points initially at the end of file.

It iterates over the item-timesteps as the ReadItemTimeStepNext, see there for more details.

This is the most efficient way to iterate through and update/append the items and timesteps in a file, since it iterates exactly as the data is stored on the disk.

The size of the data must match the data size of the item that is to be written.

The time value is only relevant for files with non-equidistant time axis. For files with an equidistant time axis, the time value is ignored, and a zero can be used.

The result can be used by ReadItemTimeStep(IDfsItemData, Int32)

If not all items are of type {T}, an exception will be thrown.

Each column is identified by matching node id numbers. For 3D elements the last half of the node numbers of the bottom element must match the first half of the node numbers in the top element. For 2D vertical elements the order of the node numbers in the bottom element (last half number of nodes) are reversed compared to those in the top element (first half number of nodes).