tdmslite 2026.10.31__tar.gz

tdmslite-2026.10.31/.gitattributes

@@ -0,0 +1,2 @@
+# Auto detect text files and perform LF normalization
+* text=auto

tdmslite-2026.10.31/.github/workflows/publish.yml

@@ -0,0 +1,35 @@
+name: Publish to PyPI
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  pypi-publish:
+    name: Build and publish to PyPI
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/p/tdmslite
+    permissions:
+      id-token: write
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install build dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install build
+
+      - name: Build sdist and wheel
+        run: python -m build
+
+      - name: Publish package to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

tdmslite-2026.10.31/.gitignore

@@ -0,0 +1,181 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/latest/usage/project/#working-with-version-control
+.pdm.toml
+.pdm-python
+.pdm-build/
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Cursor  
+#  Cursor is an AI-powered code editor.`.cursorignore` specifies files/directories to 
+#  exclude from AI features like autocomplete and code analysis. Recommended for sensitive data
+#  refer to https://docs.cursor.com/context/ignore-files
+.cursorignore
+.cursorindexingignore

tdmslite-2026.10.31/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 ajbenjam-purdue
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

tdmslite-2026.10.31/PKG-INFO

@@ -0,0 +1,65 @@
+Metadata-Version: 2.5
+Name: tdmslite
+Version: 2026.10.31
+Summary: A lightweight python package to write to and read from National Instruments TDMS files with no external dependencies.
+Project-URL: Homepage, https://github.com/ajben/TDMSlite
+Project-URL: Repository, https://github.com/ajben/TDMSlite
+License-Expression: MIT
+License-File: LICENSE
+Keywords: data,labview,measurement,national-instruments,ni,tdms
+Classifier: Development Status :: 2 - Pre-Alpha
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Programming Language :: Python :: 3.15
+Classifier: Programming Language :: Python :: 3.16
+Classifier: Topic :: File Formats
+Classifier: Topic :: Scientific/Engineering
+Classifier: Typing :: Typed
+Requires-Python: >=3.12
+Description-Content-Type: text/markdown
+
+# TDMSlite
+
+TDMSlite is a lightweight python package to write to and read from National Instruments TDMS files with no external dependencies. TDMSlite is available under an MIT license and was originally written by Aidan Benjamin.
+
+### Usage 
+
+TDMSlite is zero-dependency and allows for platform-agnostic read interface with NI's open standard. Write capability is planned. TDMSlite is hosted on PyPi:
+```cmd
+pip install tdmslite
+```
+
+The core object is `TDMSDocument`, which can be created using a binary stream or with an in-place method (`.from_file`).
+
+```python
+import tdmslite, pathlib
+
+path = pathlib.Path("tdms_example.tdms")
+
+document = tdmslite.TDMSDocument.from_file(path)
+group = document['group name']
+channel = group.channels['channel name']
+raw_data = channel.data # easy!
+```
+
+### Limitations
+
+While TDMS is an open standard, it is strangely not opinionated and allows nonstandard formats like 80-bit floats and fixed point numbers. These types are largely relics of old hardware or byproducts of FPGAs (which cannot write fixed point numbers to the filetype regardless...), but are seldom supported by third parties or NI's own extensions. 
+
+TDMSlite aims to mostly support these niche features. Floating point numbers are all coerced into vanilla python floats, and any integer (aside from a boolean) is coerced into a python integer. Strings and booleans are preserved as independant data types, and complex casting automatically occurs where applicable. 
+
+Data storage is done on a per-channel basis. If a channel's data cannot be embedded in an instance of Python's `array.array`, a costly list is necessary. Very large datasets full of strings will take signficantly more memory than the equivalent space on disk in the TDMS file by virtue of Python's lists and strings. The price we pay for convenience!
+
+### References
+ - [128-bit floating point numbers](https://en.wikipedia.org/wiki/Quadruple-precision_floating-point_format)
+ - [80-bit floating point numbers](https://en.wikipedia.org/wiki/Extended_precision)
+ - [Fixed point numbers](https://en.wikipedia.org/wiki/Fixed-point_arithmetic) (wikipedia)
+ - [Fixed point numbers](https://www.ni.com/docs/en-US/bundle/labview/page/fixed-point-numbers.html) (NI)
+ - [TDMS Versions](https://www.ni.com/docs/en-US/bundle/labview/page/file-format-versions.html)
+ - [TDMS Standard](https://www.ni.com/en/support/documentation/supplemental/07/tdms-file-format-internal-structure.html)
+ - [NI Type Descriptors](https://www.ni.com/docs/en-US/bundle/labview/page/type-descriptors.html)

tdmslite-2026.10.31/README.md

@@ -0,0 +1,40 @@
+# TDMSlite
+
+TDMSlite is a lightweight python package to write to and read from National Instruments TDMS files with no external dependencies. TDMSlite is available under an MIT license and was originally written by Aidan Benjamin.
+
+### Usage 
+
+TDMSlite is zero-dependency and allows for platform-agnostic read interface with NI's open standard. Write capability is planned. TDMSlite is hosted on PyPi:
+```cmd
+pip install tdmslite
+```
+
+The core object is `TDMSDocument`, which can be created using a binary stream or with an in-place method (`.from_file`).
+
+```python
+import tdmslite, pathlib
+
+path = pathlib.Path("tdms_example.tdms")
+
+document = tdmslite.TDMSDocument.from_file(path)
+group = document['group name']
+channel = group.channels['channel name']
+raw_data = channel.data # easy!
+```
+
+### Limitations
+
+While TDMS is an open standard, it is strangely not opinionated and allows nonstandard formats like 80-bit floats and fixed point numbers. These types are largely relics of old hardware or byproducts of FPGAs (which cannot write fixed point numbers to the filetype regardless...), but are seldom supported by third parties or NI's own extensions. 
+
+TDMSlite aims to mostly support these niche features. Floating point numbers are all coerced into vanilla python floats, and any integer (aside from a boolean) is coerced into a python integer. Strings and booleans are preserved as independant data types, and complex casting automatically occurs where applicable. 
+
+Data storage is done on a per-channel basis. If a channel's data cannot be embedded in an instance of Python's `array.array`, a costly list is necessary. Very large datasets full of strings will take signficantly more memory than the equivalent space on disk in the TDMS file by virtue of Python's lists and strings. The price we pay for convenience!
+
+### References
+ - [128-bit floating point numbers](https://en.wikipedia.org/wiki/Quadruple-precision_floating-point_format)
+ - [80-bit floating point numbers](https://en.wikipedia.org/wiki/Extended_precision)
+ - [Fixed point numbers](https://en.wikipedia.org/wiki/Fixed-point_arithmetic) (wikipedia)
+ - [Fixed point numbers](https://www.ni.com/docs/en-US/bundle/labview/page/fixed-point-numbers.html) (NI)
+ - [TDMS Versions](https://www.ni.com/docs/en-US/bundle/labview/page/file-format-versions.html)
+ - [TDMS Standard](https://www.ni.com/en/support/documentation/supplemental/07/tdms-file-format-internal-structure.html)
+ - [NI Type Descriptors](https://www.ni.com/docs/en-US/bundle/labview/page/type-descriptors.html)

tdmslite-2026.10.31/pyproject.toml

@@ -0,0 +1,36 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "tdmslite"
+version = "2026.10.31"
+description = "A lightweight python package to write to and read from National Instruments TDMS files with no external dependencies."
+readme = "README.md"
+requires-python = ">=3.12"
+dependencies = []
+license = "MIT"
+keywords = [
+  "tdms", "labview", 
+  "national-instruments", 
+  "ni", "measurement", 
+  "data"]
+classifiers = [
+  "Development Status :: 2 - Pre-Alpha",
+  "Intended Audience :: Developers",
+  "Intended Audience :: Science/Research",
+  "Topic :: File Formats",
+  "Topic :: Scientific/Engineering",
+  "Operating System :: OS Independent",
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Programming Language :: Python :: 3.14",
+  "Programming Language :: Python :: 3.15",
+  "Programming Language :: Python :: 3.16",
+  "Typing :: Typed"
+]
+
+[project.urls]
+Homepage = "https://github.com/ajben/TDMSlite"
+Repository = "https://github.com/ajben/TDMSlite"

tdmslite-2026.10.31/src/tdmslite/__init__.py

@@ -0,0 +1,8 @@
+from .models import TDMSSegment, \
+  TDMSDocument, TDMSGroup, TDMSChannel
+
+from .structs import TDMSDataTypes
+
+__all__ = ["TDMSSegment", "TDMSDocument", 
+           "TDMSGroup", "TDMSChannel", 
+           "TDMSDataTypes"]

tdmslite-2026.10.31/src/tdmslite/models.py

@@ -0,0 +1,281 @@
+from typing import BinaryIO
+from array import array
+from pathlib import Path
+from .structs import TOCFlag, TOCVersion, Number, \
+  TDMSLeadInIntegrity, String, TDMSDataTypes, \
+  TDMSMetadataObject, TDMSRawDataLayout, \
+  TDMSMetadataObjectValue, TDMSDataTypesArrayTypes, \
+  TDMSDataTypesListTypes
+from logging import getLogger
+from sys import byteorder as SYSTEM_BYTEORDER
+
+logger = getLogger(__name__)
+
+class TDMSLeadIn:
+  def __init__(self, stream:BinaryIO) -> None:
+      
+    # Set defaults
+    self.integrity:TDMSLeadInIntegrity = TDMSLeadInIntegrity.OK
+    self.big_endian:bool = False
+    self.version:TOCVersion = TOCVersion.UNKNOWN
+    self.next_segment:int = 0
+    self.raw_data_offset:int = 0
+    
+    # Check header
+    initial_position:int = stream.tell()
+    magic_sequence:bytes = stream.read(4)
+    
+    if magic_sequence == b'':
+      stream.seek(initial_position)
+      self.integrity = TDMSLeadInIntegrity.DONE
+      return
+    if magic_sequence != b'TDSm':
+      stream.seek(initial_position)
+      self.integrity = TDMSLeadInIntegrity.ERROR
+      logger.error(f"Invalid magic sequence (Lead-in started with {magic_sequence})")
+      return
+    
+    # Load data
+    members:list[TOCFlag] = TOCFlag.members(stream.read(4))
+    self.big_endian = TOCFlag.kTocBigEndian in members
+    self.interleaved = TOCFlag.kTocInterleavedData in members
+    self.has_metadata = TOCFlag.kTocMetaData in members
+    self.has_rawdata = TOCFlag.kTocRawData in members
+    self.version = TOCVersion(Number.unpack_uint(stream.read(4), self.big_endian))
+    self.next_segment = Number.unpack_uint(stream.read(8), self.big_endian)
+    self.raw_data_offset = Number.unpack_uint(stream.read(8), self.big_endian)
+    # Not sure on kTocNewObjList or kTocDAQmxRawData
+    
+    # Data checks
+    if self.version == TOCVersion.UNKNOWN:
+      logger.error("Unknown segment version")
+      self.integrity |= TDMSLeadInIntegrity.ERROR # This redundant af
+    elif self.version != TOCVersion.TDMS_2:
+      logger.warning(f"Loading old TDMS data ({self.version.name})")
+      self.integrity |= TDMSLeadInIntegrity.WARN
+    if self.next_segment == 4294967295:
+      logger.warning("Likely corrupted data in segment")
+      self.integrity |= TDMSLeadInIntegrity.WARN
+    
+class TDMSMetadata:
+
+  def __init__(self, stream:BinaryIO, big_endian:bool, channel_states:dict[str,TDMSRawDataLayout]) -> None:
+    
+    n_objects = Number.unpack_uint(stream.read(4), big_endian=big_endian)
+    
+    self.objects:list[TDMSMetadataObject] = []
+    for _ in range(n_objects):
+        
+      object_path = String.unpack(stream, big_endian=big_endian)
+      raw_data_index = Number.unpack_uint(stream.read(4), big_endian=big_endian)
+      
+      # Check the index
+      if raw_data_index == 0xFFFFFFFF: # None attached
+        raw_layout = None
+      elif raw_data_index == 0x00000000: # Reuse
+        raw_layout = channel_states.get(object_path)
+      else:
+        raw_data_type = TDMSDataTypes(
+          Number.unpack_uint(stream.read(4), big_endian=big_endian))
+        dim = Number.unpack_uint(stream.read(4), big_endian=big_endian)
+        n_values = Number.unpack_uint(stream.read(8), big_endian=big_endian)
+        
+        total_size:int = Number.unpack_uint(stream.read(8), big_endian=big_endian) \
+          if raw_data_type == TDMSDataTypes.tdsTypeString else 0
+        
+        raw_layout = TDMSRawDataLayout(raw_data_type, dim, n_values, total_size)
+      
+      if raw_layout: channel_states[object_path] = raw_layout
+      n_properties = Number.unpack_uint(stream.read(4), big_endian=big_endian)
+      properties:dict[str, TDMSMetadataObjectValue] = {}
+      
+      for _ in range(n_properties):
+        
+        property_name = String.unpack(stream, big_endian=big_endian)
+        property_type = TDMSDataTypes(Number.unpack_uint(stream.read(4), big_endian=big_endian))
+        
+        property_value = property_type.unpack(stream, big_endian=big_endian)
+        logger.info(f"Read Property(name=\"{property_name}\", type={property_type.name}, val={property_value})")
+        properties[property_name] = property_value
+      
+      self.objects.append(TDMSMetadataObject(object_path, raw_layout, properties))
+
+class TDMSChannel:
+
+  def __init__(self, name:str, path:str, data_type:TDMSDataTypes|None=None) -> None:
+    self.name = name
+    self.path = path
+    self.data_type = data_type
+    self.properties:dict[str,TDMSMetadataObjectValue] = {}
+    self.data:array[TDMSDataTypesArrayTypes]|list[TDMSDataTypesListTypes] = []
+  
+  def add_data(self, data:list[TDMSDataTypesListTypes]|array[TDMSDataTypesArrayTypes]) -> None:
+    if len(self.data) == 0:
+      self.data = data.copy() if isinstance(data, list) else data
+    else:
+      if isinstance(self.data, list) and isinstance(data, list):
+        self.data.extend(data)
+      elif isinstance(self.data, array) and isinstance(data, array):
+        if self.data.typecode == data.typecode:
+          self.data.extend(data)
+  
+  def __len__(self) -> int:
+    return len(self.data)
+  
+  def __repr__(self) -> str:
+    return (f"Channel(name={self.name.replace("\n", "\\n")}, path={self.path.replace("\n", "\\n")}, data_type={self.data_type}, "
+            f"data={"None" if len(self) == 0 else len(self)})")
+
+class TDMSGroup:
+  def __init__(self, name: str) -> None:
+    self.name: str = name
+    self.properties: dict[str,TDMSMetadataObjectValue] = {}
+    self.channels: dict[str, TDMSChannel] = {}
+
+  def __getitem__(self, channel_name: str) -> TDMSChannel:
+    return self.channels[channel_name]
+
+  def __repr__(self) -> str:
+    return f"Group(name={self.name}, " \
+      f"properties={list(self.properties.keys())}, " \
+      f"channels={list(self.channels.keys())})"
+
+class TDMSSegment:
+  
+  def __init__(self, stream: BinaryIO, 
+               channel_states: dict[str, TDMSRawDataLayout], 
+               active_objects: list[TDMSMetadataObject]) -> None:
+    self.stream = stream
+    self.start_pos = stream.tell()
+
+    # Lead-in
+    self.lead_in = TDMSLeadIn(self.stream)
+    if self.lead_in.integrity == TDMSLeadInIntegrity.ERROR or \
+     self.lead_in.integrity == TDMSLeadInIntegrity.DONE : return
+
+    # Metadata
+    if self.lead_in.has_metadata:
+      self.metadata = TDMSMetadata(self.stream, self.lead_in.big_endian, channel_states)
+      self.objects = self.metadata.objects
+    else:
+      self.metadata = None
+      self.objects = active_objects # keep previous active objects
+
+    # Rawdata
+    self.raw_chunks: dict[str, array[TDMSDataTypesArrayTypes] | list[TDMSDataTypesListTypes]] = {}
+    if self.lead_in.has_rawdata:
+      
+      self.stream.seek(self.start_pos + 28 + self.lead_in.raw_data_offset)
+      total_bytes_read:int = self._read_raw_data()
+      logger.info(f"Read {total_bytes_read} bytes of raw data")
+
+  def _read_raw_data(self) -> int:
+    total_read_bytes:int = 0
+    for obj in self.objects:
+      if not obj.raw_layout or obj.raw_layout.n_vals == 0:
+        continue
+
+      layout = obj.raw_layout
+      dtype = layout.data_type
+      arr_state = dtype.array_state()
+      
+      if arr_state:
+        typecode, itemsize = arr_state
+        n_bytes = layout.n_vals * itemsize
+        raw_bytes = self.stream.read(n_bytes)
+        total_read_bytes += n_bytes
+
+        arr:array[TDMSDataTypesArrayTypes] = array(typecode)
+        arr.frombytes(raw_bytes)
+
+        # Endian conversion if needed
+        if self.lead_in.big_endian != (SYSTEM_BYTEORDER == 'big'):
+          arr.byteswap()
+
+        self.raw_chunks[obj.path] = arr
+      else:
+        offsets = [Number.unpack_uint(self.stream.read(4), self.lead_in.big_endian) for _ in range(layout.n_vals)]
+
+        # Read chars
+        data_len = layout.total_bytes - (layout.n_vals * 4)
+        char_bytes = self.stream.read(data_len)
+        total_read_bytes += data_len
+        
+        strings: list[TDMSDataTypesListTypes] = []
+        prev_offset = 0
+        for off in offsets:
+            strings.append(char_bytes[prev_offset:off].decode('utf-8', errors='replace'))
+            prev_offset = off
+
+        self.raw_chunks[obj.path] = strings
+    return total_read_bytes
+
+class TDMSDocument:
+  def __init__(self, stream: BinaryIO) -> None:
+    self.properties: dict[str, TDMSMetadataObjectValue] = {}
+    self.groups: dict[str, TDMSGroup] = {}
+
+    # Persistant state
+    channel_states: dict[str, TDMSRawDataLayout] = {}
+    active_objects: list[TDMSMetadataObject] = []
+
+    while True:
+      seg_start = stream.tell()
+
+      # Delegate parsing
+      segment = TDMSSegment(stream, channel_states, active_objects)
+      if segment.lead_in.integrity == TDMSLeadInIntegrity.ERROR or \
+        segment.lead_in.integrity == TDMSLeadInIntegrity.DONE:
+        break
+
+      active_objects = segment.objects
+
+      # Update Document hierarchy
+      for obj in segment.objects:
+        group_name, chan_name = String.parse_path(obj.path)
+
+        if group_name is None and chan_name is None:
+          self.properties.update(obj.properties)
+        elif group_name and not chan_name:
+          group = self.groups.setdefault(group_name, TDMSGroup(group_name))
+          group.properties.update(obj.properties)
+        elif group_name and chan_name:
+          group = self.groups.setdefault(group_name, TDMSGroup(group_name))
+          dtype = obj.raw_layout.data_type if obj.raw_layout else None
+          if chan_name not in group.channels:
+            group.channels[chan_name] = TDMSChannel(chan_name, obj.path, dtype)
+
+          group.channels[chan_name].properties.update(obj.properties)
+
+      # Add segment's chunks
+      for path, chunk in segment.raw_chunks.items():
+        group_name, chan_name = String.parse_path(path)
+        if group_name and chan_name:
+          self.groups[group_name].channels[chan_name].add_data(chunk)
+
+      # EOF?
+      if segment.lead_in.next_segment == 0xFFFFFFFFFFFFFFFF:
+        break
+      stream.seek(seg_start + 28 + segment.lead_in.next_segment)
+  
+  @classmethod
+  def from_file(cls, path:str|Path) -> TDMSDocument:
+    
+    path = Path(path).resolve()
+    if path.suffix.lower() != '.tdms':
+      raise ValueError(f"Unsupported filetype: {path.suffix} (should be .tdms)")
+    if not path.exists():
+      raise FileNotFoundError(f"File at path {path} does not exist")
+    
+    with open(Path(path).resolve(), 'rb') as f:
+      file = TDMSDocument(f)
+    return file
+
+  def __getitem__(self, key:str) -> TDMSGroup | None:
+    if key in self.groups:
+      return self.groups[key]
+    return None
+
+  def __repr__(self) -> str:
+    return f"TDMSDocument(Properties={self.properties}, " \
+      f"Groups=[{", ".join([repr(group) for group in self.groups.values()])}])"

tdmslite-2026.10.31/src/tdmslite/structs.py

@@ -0,0 +1,341 @@
+from enum import Enum, IntEnum
+from typing import Iterable, Self, BinaryIO
+from dataclasses import dataclass, field
+from datetime import datetime, timezone, timedelta
+from math import nan, inf, ldexp
+import struct
+
+type TDMSDataTypesArrayTypes = float | int
+type TDMSDataTypesListTypes  = str | datetime | bool
+type TDMSMetadataObjectValue = int | float | str | bool | complex | datetime
+
+class TOCFlag(Enum):
+  kTocMetaData        = 0x00000002
+  kTocNewObjList      = 0x00000004
+  kTocRawData         = 0x00000008
+  kTocDAQmxRawData    = 0x00000080
+  kTocInterleavedData = 0x00000020
+  kTocBigEndian       = 0x00000040
+
+  @classmethod
+  def members(cls, value:int|bytes) -> list[TOCFlag]:
+    if isinstance(value, bytes) and len(value) != 4:
+      raise ValueError("Incorrect bytes object length")
+    return [flag for flag in TOCFlag if (int.from_bytes(value, 'little') if isinstance(value, bytes) else value) & flag.value]
+  
+  @classmethod
+  def pack(cls, flags:Iterable[TOCFlag]) -> int:
+    return sum([flag.value for flag in flags])
+
+class TOCVersion(IntEnum):
+  TDMS_2 = 4713
+  TDMS_1 = 4712
+  UNKNOWN = 0
+  
+  @classmethod
+  def _missing_(cls, value: object) -> TOCVersion:
+    return TOCVersion.UNKNOWN
+
+class TDMSLeadInIntegrity(Enum):
+  DONE  = 0 # EOF
+  OK    = 1 # good
+  WARN  = 2 # suspicious
+  ERROR = 3 # block loading
+  
+  def __or__(self: Self, value: TDMSLeadInIntegrity) -> TDMSLeadInIntegrity:
+    return self if self.value > value.value else value
+
+class TDMSDataTypes(IntEnum):
+  
+  # Named to match NI's asinine naming schema
+  tdsTypeVoid = 0x00
+  tdsTypeI8 = 0x01
+  tdsTypeI16 = 0x02
+  tdsTypeI32 = 0x03
+  tdsTypeI64 = 0x04
+  tdsTypeU8 = 0x05
+  tdsTypeU16 = 0x06
+  tdsTypeU32 = 0x07
+  tdsTypeU64 = 0x08
+  tdsTypeSingleFloat = 0x09
+  tdsTypeDoubleFloat = 0x0A
+  tdsTypeExtendedFloat = 0x0B
+  tdsTypeSingleFloatWithUnit = 0x19
+  tdsTypeDoubleFloatWithUnit = 0x1A
+  tdsTypeExtendedFloatWithUnit = 0x1B
+  tdsTypeString = 0x20
+  tdsTypeBoolean = 0x21
+  tdsTypeTimeStamp = 0x44
+  tdsTypeFixedPoint = 0x4F
+  tdsTypeComplexSingleFloat = 0x08000C
+  tdsTypeComplexDoubleFloat = 0x10000D
+  tdsTypeDAQmxRawData = 0xFFFFFFFF
+  
+  def unpack(self, stream:BinaryIO, big_endian:bool) -> TDMSMetadataObjectValue:
+    match self:
+      case TDMSDataTypes.tdsTypeString:
+        return(String.unpack(stream, big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeBoolean:
+        return(Boolean.unpack(stream))
+      case TDMSDataTypes.tdsTypeI8:
+        return(Number.unpack_int(stream.read(1), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeU8:
+        return(Number.unpack_uint(stream.read(1), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeI16:
+        return(Number.unpack_int(stream.read(2), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeU16:
+        return(Number.unpack_uint(stream.read(2), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeI32:
+        return(Number.unpack_int(stream.read(4), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeU32:
+        return(Number.unpack_uint(stream.read(4), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeI64:
+        return(Number.unpack_int(stream.read(8), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeU64:
+        return(Number.unpack_uint(stream.read(8), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeSingleFloat | TDMSDataTypes.tdsTypeSingleFloatWithUnit:
+        return(Number.unpack_sfloat(stream.read(4), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeComplexSingleFloat:
+        return(Number.unpack_csfloat(stream.read(8), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeDoubleFloat | TDMSDataTypes.tdsTypeDoubleFloatWithUnit:
+        return(Number.unpack_dfloat(stream.read(8), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeComplexDoubleFloat:
+        return(Number.unpack_cdfloat(stream.read(16), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeExtendedFloat | TDMSDataTypes.tdsTypeExtendedFloatWithUnit:
+        return(Number.unpack_efloat(stream.read(10), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeTimeStamp:
+        return(Timestamp.from_bytes(stream.read(16), big_endian=big_endian))
+      case TDMSDataTypes.tdsTypeFixedPoint:
+        raise NotImplementedError("Fixed point values are not supported by NI in TDMS")
+      case _:
+        raise NotImplementedError(f"Type {self} has not yet been added")
+
+  def array_state(self) -> tuple[str, int] | None:
+    match self:
+      case TDMSDataTypes.tdsTypeI8:
+        return ('b', 1)
+      case TDMSDataTypes.tdsTypeU8:
+        return ('B', 1)
+      case TDMSDataTypes.tdsTypeI16:
+        return ('h', 2)
+      case TDMSDataTypes.tdsTypeU16:
+        return ('H', 2)
+      case TDMSDataTypes.tdsTypeI32:
+        return ('i', 4)
+      case TDMSDataTypes.tdsTypeU32:
+        return ('I', 4)
+      case TDMSDataTypes.tdsTypeI64:
+        return ('q', 8)
+      case TDMSDataTypes.tdsTypeU64:
+        return ('Q', 8)
+      case TDMSDataTypes.tdsTypeSingleFloat:
+        return ('f', 4)
+      case TDMSDataTypes.tdsTypeSingleFloatWithUnit:
+        return ('f', 4)
+      case TDMSDataTypes.tdsTypeDoubleFloat:
+        return ('d', 8)
+      case TDMSDataTypes.tdsTypeDoubleFloatWithUnit:
+        return ('d', 8)
+      case TDMSDataTypes.tdsTypeBoolean:
+        return ('b', 1)
+      case _:
+        return None
+
+@dataclass
+class TDMSRawDataLayout:
+  data_type:  TDMSDataTypes
+  dim:        int
+  n_vals:     int
+  total_bytes:int
+
+@dataclass
+class TDMSMetadataObject:
+  path:       str
+  raw_layout: TDMSRawDataLayout | None = None
+  properties: dict[str, TDMSMetadataObjectValue] = \
+    field(default_factory=dict[str, TDMSMetadataObjectValue])
+
+class String:
+  
+  @classmethod
+  def pack(cls, content:str, big_endian:bool=False) -> bytes:
+    
+    # per https://www.ni.com/en/support/documentation/supplemental/07/tdms-file-format-internal-structure.html
+    content_stripped:str = content.rstrip()
+    return len(content_stripped).to_bytes(length=4, byteorder='big' if big_endian else 'little') + content_stripped.encode('utf-8')
+  
+  @classmethod
+  def unpack(cls, stream:BinaryIO, big_endian:bool=False) -> str:
+    
+    # see String.pack
+    str_length = Number.unpack_uint(stream.read(4), big_endian=big_endian)
+    return stream.read(str_length).decode('utf-8')
+  
+  @classmethod
+  def parse_path(cls, path:str) -> tuple[str|None, str|None]:
+    if path == "/": return (None, None)
+    
+    escaped, skip = False, False
+    current:list[str] = []
+    parts:list[str] = []
+    
+    for i, char in enumerate(path):
+      
+      if skip: 
+        skip = False
+        continue
+      
+      if char == "'": # Escape char
+        if escaped and i + 1 < len(path) and path[i+1] == "'":
+          current.append("'") # Doubled quotation to embed a '
+          skip = True
+        
+        else:
+          escaped = not escaped
+          if not escaped:
+            parts.append("".join(current))
+            current = []
+            
+      elif escaped: current.append(char) # Valid char
+    
+    return (parts[0] if len(parts) > 0 else None, parts[1] if len(parts) > 1 else None)      
+
+class Number:
+  
+  max_u32_bytes:bytes = b'\xFF\xFF\xFF\xFF'
+  max_u32:int = int.from_bytes(max_u32_bytes)
+  max_u64_bytes:bytes = b'\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF'
+  max_u64:int = int.from_bytes(max_u64_bytes)
+  
+  @classmethod
+  def pack_u32(cls, content:int, big_endian:bool=False) -> bytes:
+    return content.to_bytes(byteorder='big' if big_endian else 'little', signed=False, length=4)
+  
+  @classmethod
+  def pack_i32(cls, content:int, big_endian:bool=False) -> bytes:
+    return content.to_bytes(byteorder='big' if big_endian else 'little', signed=True, length=4)
+  
+  @classmethod
+  def pack_u64(cls, content:int, big_endian:bool=False) -> bytes:
+    return content.to_bytes(byteorder='big' if big_endian else 'little', signed=False, length=8)
+  
+  @classmethod
+  def pack_i64(cls, content:int, big_endian:bool=False) -> bytes:
+    return content.to_bytes(byteorder='big' if big_endian else 'little', signed=True, length=8)
+  
+  @classmethod
+  def unpack_uint(cls, content:bytes, big_endian:bool=False) -> int:
+    return int.from_bytes(content, 'big' if big_endian else 'little', signed=False)
+  
+  @classmethod
+  def unpack_int(cls, content:bytes, big_endian:bool=False) -> int:
+    return int.from_bytes(content, 'big' if big_endian else 'little', signed=True)
+  
+  @classmethod
+  def unpack_hfloat(cls, content:bytes, big_endian:bool=False) -> float:
+    return struct.unpack(('>' if big_endian else '<')+'e', content)[0]
+  
+  @classmethod
+  def unpack_sfloat(cls, content:bytes, big_endian:bool=False) -> float:
+    return struct.unpack(('>' if big_endian else '<')+'f', content)[0]
+  
+  @classmethod
+  def unpack_dfloat(cls, content:bytes, big_endian:bool=False) -> float:
+    return struct.unpack(('>' if big_endian else '<')+'d', content)[0]
+  
+  @classmethod
+  def unpack_efloat(cls, content:bytes, big_endian:bool=False) -> float:
+    
+    # https://en.wikipedia.org/wiki/Extended_precision
+    
+    if len(content) != 10:
+      raise ValueError(f"Incorrect number of bytes ({len(content)}) provided to unpack_efloat (reqd: 10)")
+    
+    if big_endian:
+      biased_exp, mantissa = struct.unpack('>HQ', content)
+    else:
+      mantissa, biased_exp = struct.unpack('<QH', content)
+      
+    sign = -1.0 if biased_exp & 0x8000 else 1.0
+    exponent = biased_exp & 0x7FFF
+    
+    # inf/nan
+    if exponent == 0x7FFF:
+      
+      # lower 63 bits are set or bit 63 is 0
+      if (mantissa & 0x7FFFFFFFFFFFFFFF) or not (mantissa & 0x8000000000000000):
+        return nan
+      
+      return inf if sign > 0 else -inf
+    
+    # Zero check
+    if exponent == 0 and mantissa == 0: 
+      return sign * 0.0
+    
+    if exponent == 0:
+      # Subnormal numbers
+      exponent = -16382
+    else:
+      # Normal numbers
+      exponent -= 16383
+    
+    frac = mantissa / (1 << 63)
+    try:
+      return ldexp(sign * frac, exponent)
+    except:
+      return inf if sign > 0 else -inf
+
+  @classmethod
+  def unpack_fixed(cls, content:bytes, word_length:int, int_word_length:int, signed:bool, big_endian:bool=False) -> float:
+    integer = cls.unpack_int(content=content, big_endian=big_endian)
+    
+    fraction = word_length - int_word_length
+    if signed and (integer & (1 << (word_length - 1))):
+      integer -= (1 << word_length)
+    
+    return integer / (2 ** fraction)
+
+  @classmethod
+  def unpack_csfloat(cls, content:bytes, big_endian:bool=False) -> complex:
+    real = cls.unpack_sfloat(content[0:4], big_endian=big_endian)
+    imag = cls.unpack_sfloat(content[4:8], big_endian=big_endian)
+    
+    return real + imag * 1j
+  
+  @classmethod
+  def unpack_cdfloat(cls, content:bytes, big_endian:bool=False) -> complex:
+    real = cls.unpack_dfloat(content[0:8], big_endian=big_endian)
+    imag = cls.unpack_dfloat(content[8:16], big_endian=big_endian)
+    
+    return real + imag * 1j
+
+class Boolean:
+  
+  @classmethod
+  def unpack(cls, stream:BinaryIO) -> bool:
+    return stream.read(1) == b'\x01'
+
+class Timestamp:
+
+  EPOCH = datetime(1904, 1, 1, tzinfo=timezone.utc)
+  TWO_POW_64 = 1 << 64
+
+  @classmethod
+  def from_bytes(cls, content: bytes, big_endian: bool) -> datetime:
+    fraction = Number.unpack_uint(content[0:8], big_endian=big_endian)
+    seconds = Number.unpack_int(content[8:16], big_endian=big_endian)
+    microseconds = round(fraction * 1_000_000 / cls.TWO_POW_64)
+
+    return cls.EPOCH + timedelta(seconds=seconds, microseconds=microseconds)
+
+  @classmethod
+  def to_bytes(cls, dt: datetime, big_endian: bool) -> bytes:
+    delta = dt.astimezone(timezone.utc) - cls.EPOCH
+    
+    # int seconds elapsed
+    seconds = delta.days * 86400 + delta.seconds
+    fraction = (delta.microseconds * cls.TWO_POW_64) // 1_000_000
+
+    content_0_8 = Number.pack_u64(fraction, big_endian=big_endian)
+    content_8_16 = Number.pack_i64(seconds, big_endian=big_endian)
+    return content_0_8 + content_8_16