Datasets encoded in the TDM/TDMS format come along in pairs comprised of a .tdm and .tdx file. While the .tdm file is a human-readable document providing meta information about the dataset, the .tdx is a binary containing the actual data. The .tdm is represented in XML format and basically reveals what data the .tdx contains and how to read it. The XML tree is usually made up of several groups, each containing an arbitrary amount of channels.

Usage

Although the package is built upon a C++ core, which decodes the data, it may be used as a Python module, as well, by interfacing the C++ library with a Cython wrapper.

C++ core

In order to parse the XML tree of the .tdm file, the library employs pugixml: https://pugixml.org/ and https://github.com/zeux/pugixml
The package currently supports the following datatypes:
- eInt8Usi: 8 byte
- eInt16Usi: 16 byte
- eInt32Usi: 32 byte
- eInt64Usi: 64 byte
- eUInt8Usi: 8 byte
- eUInt16Usi: 16 byte
- eUInt32Usi: 32 byte
- eUInt64Usi: 64 byte
- eFloat32Usi: 32 byte
- eFloat64Usi: 64 byte
The core of the library takes care of the decoding by reinterpretation of the binary in the buffer as the required datatype implemented by
```
uint8_t *dfcast = reinterpret_cast<uint8_t*>(&df);

for ( int i = 0; i < (int)sizeof(double); i++ )
{
  dfcast[i] = (int)bych[i];
}
```
where for instance df is the resulting float and bych contains the binary data as an array of chars.
main.cpp contains an example of how the C++ library might be used to provide the channels and groups of the dataset. It is simply build by
```
make
```
extract_all.cpp takes the .tdm, the .tdx file and some output directory as arguments to provide all given information in .csv format without any logging. To build:
```
make extall
```
For instance, the executable accepts the following arguments:
```
./extract_all samples/SineData.tdm samples/SineData.tdx data/
```

Python module

The library may also be used as a Python module and supports the use of group channels in NumPy arrays as shown in example.py .
To extract all available information and data in the TDM files without any further interaction, the use of extract_all.py is recommended. To exhibit the required arguments:
```
python extract_all.py --help
```
The same functionality may be obtained from an existing python script by importing the tdm_ripper module and calling the extract_all function. For instance
```
import tdm_ripper as td

td.extract_all(b"samples/SineData.tdm",b"samples/SineData.tdx",b"data/",b"my_tdm")
```
where the arguments "data/" and "my_tdm" are optional. "data/" specifies the directory where all .csv output is dumped while "my_tdm" represents a name prefix for all csv. files. Note, that all string arguments must be converted to bytes before passing to the argument list by prepending "b".

Languages

C++ 85.5%

Python 5%

Makefile 4.7%

Cython 2.8%

HTML 1.3%

Other 0.7%