vtk_fortran_pvd_file

PVD file class: VTK collection of datasets (time series).

A .pvd file lists datasets (VTK XML files written with vtk_file, pvtk_file or vtm_file) with their time step, and optionally part, group and name: readers such as ParaView load the listed files as a time series.

The file is kept valid after each write_dataset: every new dataset is written over the closing tags, which are written again after it. Therefore a run that stops before finalize (crash, killed job) still leaves a readable collection of all the datasets written so far, and a restarted run can continue an existing collection with action='append'.

Source: src/lib/vtk_fortran_pvd_file.f90

Dependencies

Contents

• pvd_file
• write_closing_tags
• initialize
• write_dataset
• finalize
• get_datasets
• upper

Variables

Name	Type	Attributes	Description
closing_tags	character(len=*)	parameter	Closing tags.

Derived Types

pvd_file

PVD file class.

Components

Name	Type	Attributes	Description
unit	integer(kind=I4P)		Logical unit, 0 when the file is not open.
pos_close	integer(kind=I8P)		Stream position of the closing tags, i.e. where the next dataset is written.
xml	type(xml_scanner)		Index of the file read (action='read').
is_reading	logical		The file is open for reading.

Type-Bound Procedures

Name	Attributes	Description
get_datasets	pass(self)	Return the datasets of a file read.
initialize	pass(self)	Initialize (create or append to) the file.
write_dataset	pass(self)	Write one dataset entry.
finalize	pass(self)	Finalize (close) the file.
write_closing_tags	pass(self)	Write the closing tags at the current end of datasets.

Subroutines

write_closing_tags

Write the closing tags at the current end of datasets, drop anything after them and flush the file.

subroutine write_closing_tags(self, error)

Arguments

Name	Type	Intent	Attributes	Description
self	class(pvd_file)	inout		PVD file.
error	integer(kind=I4P)	out		Error status.

Call graph

Functions

initialize

Initialize the file: create a new collection, or reopen an existing one to append datasets.

Example of usage

 type(pvd_file) :: pvd
 ...
 error = pvd%initialize(filename='simulation.pvd')                  ! new collection (an existing file is replaced)
 error = pvd%initialize(filename='simulation.pvd', action='append') ! continue an existing collection (e.g. restart)
 error = pvd%initialize(filename='simulation.pvd', action='read')   ! read the collection, see get_datasets
 ...

@note With action='append' the file must exist and be a VTK collection (<Collection>): the datasets already listed are kept and the new ones are written after them. @note With action='read' the file is indexed, its datasets are returned by get_datasets and finalize frees the index: the error is 1 if the file cannot be read, 2 if it is not a VTK collection.

Returns: integer(kind=I4P)

function initialize(self, filename, action) result(error)

Arguments

Name	Type	Intent	Attributes	Description
self	class(pvd_file)	inout		PVD file.
filename	character(len=*)	in		File name, with the .pvd extension.
action	character(len=*)	in	optional	Action: new (default), append or read.

Call graph

write_dataset

Write one dataset entry: the file name of the dataset and its time step (and optionally part, group and name).

The collection is valid (closing tags included) as soon as this function returns.

Example of usage

 error = pvd%write_dataset(filename='simulation_0010.vtu', timestep=0.1_R8P)
 error = pvd%write_dataset(filename='simulation_0010_rank1.vtu', timestep=0.1_R8P, part=1)

@note filename is written as given: a relative path is relative to the directory of the .pvd file.

Returns: integer(kind=I4P)

function write_dataset(self, filename, timestep, part, group, name) result(error)

Arguments

Name	Type	Intent	Attributes	Description
self	class(pvd_file)	inout		PVD file.
filename	character(len=*)	in		File name of the dataset.
timestep	real(kind=R8P)	in		Time step (time value) of the dataset.
part	integer(kind=I4P)	in	optional	Part (piece) of the dataset at this time step, default 0.
group	character(len=*)	in	optional	Group of the dataset.
name	character(len=*)	in	optional	Name of the dataset.

Call graph

finalize

Finalize the file: close it. The file is already a valid collection, the datasets written so far included.

Returns: integer(kind=I4P)

function finalize(self) result(error)

Arguments

Name	Type	Intent	Attributes	Description
self	class(pvd_file)	inout		PVD file.

Call graph

get_datasets

Return the datasets of a file read (initialize(..., action='read')), in the order of the file.

Missing attributes are returned as 0 (timestep, part) or empty (group, name, file). The strings are blank padded to the longest one; relative files are relative to the directory of the .pvd file.

 type(pvd_file)                :: pvd
 real(R8P),        allocatable :: timestep(:)
 character(len=:), allocatable :: file(:)
 error = pvd%initialize(filename='simulation.pvd', action='read')
 error = pvd%get_datasets(timestep=timestep, file=file)
 error = pvd%finalize()

Returns: integer(kind=I4P)

function get_datasets(self, timestep, part, group, name, file) result(error)

Arguments

Name	Type	Intent	Attributes	Description
self	class(pvd_file)	in		PVD file.
timestep	real(kind=R8P)	out	allocatable, optional	Time step of the datasets.
part	integer(kind=I4P)	out	allocatable, optional	Part of the datasets.
group	character(len=:)	out	allocatable, optional	Group of the datasets.
name	character(len=:)	out	allocatable, optional	Name of the datasets.
file	character(len=:)	out	allocatable, optional	File of the datasets.

Call graph

upper

Return a string in upper case (ASCII letters only).

Attributes: pure

Returns: character(len=len(string))

function upper(string) result(upper_string)

Arguments

Name	Type	Intent	Attributes	Description
string	character(len=*)	in		Input string.

Call graph