# Add extractor of type Data/auto

**URL:** <https://discourse.paraview.org/t/add-extractor-of-type-data-auto/7609>\
**Category:** Feature Requests\
**Created:** [July 14, 2021, 9:17am UTC](https://discourse.paraview.org/t/add-extractor-of-type-data-auto/7609 "2021-07-14T09:17:50Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Alexandre\_Minot](https://discourse.paraview.org/user_avatar/discourse.paraview.org/alexandre_minot/32/7132_2.png) [@Alexandre\_Minot](https://discourse.paraview.org/u/Alexandre_Minot)\
**Post date:** [July 14, 2021, 9:17am UTC](https://discourse.paraview.org/t/add-extractor-of-type-data-auto/7609/1 "2021-07-14T09:17:50Z")

</div>

A data extractor that would choose the correct writer type based on input data type would be a welcomed addition. This could take the form of an extractor of type Data/auto. Such a feature would make in-situ post-processing with ParaView-Catalyst smoother.

Let me give a couple examples:

I generally setup my in-situ post-processing pipeline by loading a previous result file into ParaView. With the disk reader I am using, the data in loaded into ParaView as a MultiBlockDataSet. If I setup a slice and an extractor for that slice, ParaView invites me to use an extractor of type VTM (in fact, other data extractors are grayed-out as ParaView detects my data type). However, Catalyst 2.0 will send data of type vtkPartionedDataSet or vtkPartionedDataSetCollection. I therefore have to change the writer in the Catalyst script to a VTPC.

Say I now want to use a merge block on my slice. Knowing the syntax, I can add it directly in the python script, without opening ParaView again. I now also need to remember to change the writer for the slice to a VTU writer.

Having a data writer of type AUTO would make this smoother, especially for users not familiar with the different data types in ParaView.

---

<div class="post-metadata">

**Author:** ![utkarsh.ayachit](https://discourse.paraview.org/user_avatar/discourse.paraview.org/utkarsh.ayachit/32/39_2.png) [@utkarsh.ayachit](https://discourse.paraview.org/u/utkarsh.ayachit)\
**Post date:** [July 14, 2021, 12:00pm UTC](https://discourse.paraview.org/t/add-extractor-of-type-data-auto/7609/2 "2021-07-14T12:00:44Z")

</div>

I couldn’t agree more. Personally, I find having separate extensions based in the data type for VTK xml format quite annoying. Similar to `.vtk`, it would have been easier if there was simply a `.vtx` file extension for storing all VTK datasets to a VTK-based XML file format ([ref](https://discourse.vtk.org/t/vtk-xml-format-and-data-type-specific-variants/3623)).