dlisio 1.0.4

Python library for working with the well log formats DLIS (RP66v1) and LIS79

Introduction

dlisio is an LGPL licensed library for reading well logs in Digital Log Interchange Standard (DLIS V1), also known as RP66 V1, and Log Information Standard 79 (LIS79).

dlisio is designed as a general purpose library for reading well logs in a simple and easy-to-use manner. Its main focus is making all the data and metadata accessible while putting few assumptions on how the data is to be used. This makes it suitable as a building block for higher level applications as well as for being used directly.

dlisio focuses above all on correctness, performance and robustness. Its core, which does all the heavy lifting, is implemented in C++. Both the C++ core and the python wrappers are backed by an extensive test-suite. It strives to be robust against files that do not strictly adhere to the specifications, which is a widespread issue with both DLIS and LIS files. dlisio tries to account for many of the known specification violations out there, but only when it can do so without compromising correctness. It will not do any guess work on your behalf when such violations pose any ambiguity.

Installation

dlisio supplies pre-built python wheels for a variety of platforms and architectures. The wheels are hosted through the Python Package Index (PyPI) and can be installed with:

pip install dlisio

	macOS ARM	Windows 64bit	Windows 32bit	manylinux x86_64	manylinux aarch64	manylinux i686	musllinux x86_64
CPython 3.10	✅	✅	✅	✅	✅	✅	✅
CPython 3.11	✅	✅	✅	✅	✅	✅	✅
CPython 3.12	✅	✅	✅	✅	✅	✅	✅
CPython 3.13	✅	✅	✅	✅	✅	✅	✅
CPython 3.14	✅	✅	✅	✅	✅	✅	✅

See Build dlisio for building dlisio from source.

Getting started

dlisio's documentation is hosted on readthedocs. Please refer there for proper introduction to dlisio and the file-formats DLIS and LIS. Here is a motivating example showcasing how to read the curve-data from a DLIS-file:

from dlisio import dlis

with dlis.load('myfile.dlis') as files:
    for f in files:
        for frame in f.frames:
            curves = frame.curves()
            # Do something with the curves

and from a LIS-file:

from dlisio import lis

with lis.load('myfile.lis') as files:
    for f in files:
        for format_spec in f.data_format_specs():
            curves = lis.curves(f, format_spec)
            # Do something with the curves

In both cases the curves are returned as structured numpy.ndarray with the curve mnemonics as field names (column names).

Build dlisio

To develop dlisio, or to build a particular revision from source, you need:

• A C++11 compatible compiler (tested on gcc, clang, and msvc 2019)
• CMake version 3.18 or greater
• Python version 3.10 or greater
• fmtlib tested mainly with 7.1.3
• mpark_variant
• pybind11 version 2.6 or greater
• layered-file-protocols
• python packages pytest and numpy

If you do not have pybind11 installed on your system, the easiest way to get a working copy is to pip3 install pybind11 (NP! pybind11, not pybind)

layered-file-protocols has to be installed from source if you don't already have it on your system:

git clone https://github.com/equinor/layered-file-protocols.git
mkdir layered-file-protocols/build
cd layered-file-protocols/build
cmake .. -DCMAKE_BUILD_TYPE=Release -DBUILD_SHARED_LIBS=ON
-DLFP_FMT_HEADER_ONLY=ON
make
make install

To then build and install dlisio:

mkdir dlisio/build
cd dlisio/build
cmake .. -DCMAKE_BUILD_TYPE=Release -DBUILD_SHARED_LIBS=ON
make

dlisio follows common cmake rules and conventions, e.g. to set install prefix use -DCMAKE_INSTALL_PREFIX. To build the python library it is usually a good idea to build shared libraries. To disable python, pass -DBUILD_PYTHON=OFF. By default, the python library is built.

Contributing

We welcome all kinds of contributions, including bug reports, issues, feature requests and documentation. The preferred way of submitting a contribution is to make an issue on github.

Writing DLIS files

dlisio currently has no capability to write files, see issue 396 or issue 417. If writing capability is required, one can use dlisio together with other packages to achieve this. For example, the dliswriter created by Well ID contains some utility functions to create DLIS files from dlisio objects and also provides an example.