harp-python 0.3.0__tar.gz → 0.4.1__tar.gz

{harp_python-0.3.0 → harp_python-0.4.1}/.github/workflows/build.yml

@@ -16,7 +16,7 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest, windows-latest, macos-latest]
-        python-version: [3.9, 3.11]
+        python-version: [3.9, 3.12]
       fail-fast: false
     steps:
       - name: Checkout

harp_python-0.4.1/.gitignore

@@ -0,0 +1,23 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+
+# Distribution / packaging
+dist/
+_version.py
+*.egg-info/
+*.egg
+
+# IDE
+.vscode/*
+
+# misc
+log*.txt
+scratch/
+scratch*.py
+
+# Test
+.coverage
+
+# Environment
+.venv/
+uv.lock

{harp_python-0.3.0 → harp_python-0.4.1}/PKG-INFO

@@ -1,9 +1,9 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: harp-python
-Version: 0.3.0
+Version: 0.4.1
 Summary: A low-level interface for loading binary Harp protocol data
 Author-email: Goncalo Lopes <g.lopes@neurogears.org>
-License: MIT License
+License-Expression: MIT
 Project-URL: Bug Tracker, https://github.com/harp-tech/harp-python/issues
 Project-URL: Source Code, https://github.com/harp-tech/harp-python
 Classifier: Development Status :: 4 - Beta
@@ -12,7 +12,6 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Operating System :: OS Independent
-Classifier: License :: OSI Approved :: MIT License
 Requires-Python: >=3.9.0
 Description-Content-Type: text/markdown
 License-File: LICENSE
@@ -29,6 +28,7 @@ Requires-Dist: codespell; extra == "dev"
 Provides-Extra: jupyter
 Requires-Dist: ipykernel; extra == "jupyter"
 Requires-Dist: matplotlib; extra == "jupyter"
+Dynamic: license-file
 
 # harp-python
 

harp_python-0.4.1/harp/__init__.py

@@ -0,0 +1,5 @@
+from harp.io import REFERENCE_EPOCH, MessageType, read, to_buffer, to_file
+from harp.reader import create_reader
+from harp.schema import read_schema
+
+__all__ = ["REFERENCE_EPOCH", "MessageType", "read", "to_buffer", "to_file", "create_reader", "read_schema"]

harp_python-0.4.1/harp/io.py

@@ -0,0 +1,274 @@
+from datetime import datetime
+from enum import IntEnum
+from os import PathLike
+from typing import Any, BinaryIO, Optional, Union
+
+import numpy as np
+import numpy.typing as npt
+import pandas as pd
+from pandas._typing import Axes  # pyright: ignore[reportPrivateImportUsage]
+
+from harp.typing import _BufferLike, _FileLike
+
+REFERENCE_EPOCH = datetime(1904, 1, 1)
+"""The reference epoch for UTC harp time."""
+
+
+class MessageType(IntEnum):
+    """Specifies the type of a Harp message."""
+
+    NA = 0
+    READ = 1
+    WRITE = 2
+    EVENT = 3
+
+
+_SECONDS_PER_TICK = 32e-6
+_PAYLOAD_TIMESTAMP_MASK = 0x10
+_messagetypes = [type.name for type in MessageType]
+_dtypefrompayloadtype = {
+    1: np.dtype(np.uint8),
+    2: np.dtype(np.uint16),
+    4: np.dtype(np.uint32),
+    8: np.dtype(np.uint64),
+    129: np.dtype(np.int8),
+    130: np.dtype(np.int16),
+    132: np.dtype(np.int32),
+    136: np.dtype(np.int64),
+    68: np.dtype(np.float32),
+}
+_payloadtypefromdtype = {v: k for k, v in _dtypefrompayloadtype.items()}
+
+
+def read(
+    file_or_buf: Union[_FileLike, _BufferLike],
+    address: Optional[int] = None,
+    dtype: Optional[np.dtype] = None,
+    length: Optional[int] = None,
+    columns: Optional[Axes] = None,
+    epoch: Optional[datetime] = None,
+    keep_type: bool = False,
+):
+    """Read single-register Harp data from the specified file or buffer.
+
+    Parameters
+    ----------
+    file_or_buf
+        File path, open file object, or buffer containing binary data from
+        a single device register.
+    address
+        Expected register address. If specified, the address of
+        the first message is used for validation.
+    dtype
+        Expected data type of the register payload. If specified, the
+        payload type of the first message is used for validation.
+    length
+        Expected number of elements in register payload. If specified, the
+        payload length of the first message is used for validation.
+    columns
+        The optional column labels to use for the data values.
+    epoch
+        Reference datetime at which time zero begins. If specified,
+        the result data frame will have a datetime index.
+    keep_type
+        Specifies whether to include a column with the message type.
+
+    Returns
+    -------
+        A pandas data frame containing message data, sorted by time.
+    """
+    if isinstance(file_or_buf, (str, PathLike, BinaryIO)) or hasattr(file_or_buf, "readinto"):
+        # TODO: in the below we ignore the type as otherwise
+        # we have no way to runtime check _IOProtocol
+        data = np.fromfile(file_or_buf, dtype=np.uint8)  # type: ignore
+    else:
+        data = np.frombuffer(file_or_buf, dtype=np.uint8)
+
+    if len(data) == 0:
+        return pd.DataFrame(
+            columns=columns,
+            index=pd.DatetimeIndex([], name="Time")
+            if epoch
+            else pd.Index([], dtype=np.float64, name="Time"),
+        )
+
+    if address is not None and address != data[2]:
+        raise ValueError(f"expected address {address} but got {data[2]}")
+
+    index = None
+    stride = int(data[1] + 2)
+    nrows = len(data) // stride
+    payloadtype = data[4]
+    payloadoffset = 5
+    if payloadtype & _PAYLOAD_TIMESTAMP_MASK != 0:
+        seconds = np.ndarray(nrows, dtype=np.uint32, buffer=data, offset=payloadoffset, strides=stride)
+        payloadoffset += 4
+        micros = np.ndarray(nrows, dtype=np.uint16, buffer=data, offset=payloadoffset, strides=stride)
+        payloadoffset += 2
+        time = micros * _SECONDS_PER_TICK + seconds
+        payloadtype = payloadtype & ~np.uint8(_PAYLOAD_TIMESTAMP_MASK)
+        if epoch is not None:
+            time = epoch + pd.to_timedelta(time, "s")  # type: ignore
+        index = pd.Series(time)
+        index.name = "Time"
+
+    payloadsize = stride - payloadoffset - 1
+    payloadtype = _dtypefrompayloadtype[payloadtype]
+    if dtype is not None and dtype != payloadtype:
+        raise ValueError(f"expected payload type {dtype} but got {payloadtype}")
+
+    elementsize = payloadtype.itemsize
+    payloadshape = (nrows, payloadsize // elementsize)
+    if length is not None and length != payloadshape[1]:
+        raise ValueError(f"expected payload length {length} but got {payloadshape[1]}")
+
+    payload = np.ndarray(
+        payloadshape,
+        dtype=payloadtype,
+        buffer=data,
+        offset=payloadoffset,
+        strides=(stride, elementsize),
+    )
+
+    result = pd.DataFrame(payload, index=index, columns=columns)
+    if keep_type:
+        msgtype = np.ndarray(nrows, dtype=np.uint8, buffer=data, offset=0, strides=stride)
+        msgtype = pd.Categorical.from_codes(msgtype, categories=_messagetypes)  # type: ignore
+        result[MessageType.__name__] = msgtype
+    return result
+
+
+def to_file(
+    data: pd.DataFrame,
+    file: _FileLike,
+    address: int,
+    dtype: Optional[np.dtype] = None,
+    length: Optional[int] = None,
+    port: Optional[int] = None,
+    epoch: Optional[datetime] = None,
+    message_type: Optional[MessageType] = None,
+):
+    """Write single-register Harp data to the specified file.
+
+    Parameters
+    ----------
+    data
+        Pandas data frame containing message payload.
+    file
+        File path, or open file object in which to store binary data from
+        a single device register.
+    address
+        Register address used to identify all formatted Harp messages.
+    dtype
+        Data type of the register payload. If specified, all data will
+        be converted before formatting the binary payload.
+    length
+        Expected number of elements in register payload. If specified, the
+        number of columns in the input data frame is validated.
+    port
+        Optional port value used for all formatted Harp messages.
+    epoch
+        Reference datetime at which time zero begins. If specified,
+        the input data frame must have a datetime index.
+    message_type
+        Optional message type used for all formatted Harp messages.
+        If not specified, data must contain a MessageType column.
+    """
+    buffer = to_buffer(data, address, dtype, port, length, epoch, message_type)
+    buffer.tofile(file)
+
+
+def to_buffer(
+    data: pd.DataFrame,
+    address: int,
+    dtype: Optional[np.dtype] = None,
+    length: Optional[int] = None,
+    port: Optional[int] = None,
+    epoch: Optional[datetime] = None,
+    message_type: Optional[MessageType] = None,
+) -> npt.NDArray[np.uint8]:
+    """Convert single-register Harp data to a flat binary buffer.
+
+    Parameters
+    ----------
+    data
+        Pandas data frame containing message payload.
+    address
+        Register address used to identify all formatted Harp messages.
+    dtype
+        Data type of the register payload. If specified, all data will
+        be converted before formatting the binary payload.
+    length
+        Expected number of elements in register payload. If specified, the
+        number of columns in the input data frame is validated.
+    port
+        Optional port value used for all formatted Harp messages.
+    epoch
+        Reference datetime at which time zero begins. If specified,
+        the input data frame must have a datetime index.
+    message_type
+        Optional message type used for all formatted Harp messages.
+        If not specified, data must contain a MessageType column.
+
+    Returns
+    -------
+        An array object containing message data formatted according
+        to the Harp binary protocol.
+    """
+    nrows = len(data)
+    if nrows == 0:
+        return np.empty(0, dtype=np.uint8)
+
+    if MessageType.__name__ in data.columns:
+        msgtype = data[MessageType.__name__].cat.codes
+        payload = data[data.columns.drop(MessageType.__name__)].values
+    elif message_type is not None:
+        msgtype = message_type
+        payload = data.values
+    else:
+        raise ValueError(f"message type must be specified either in the data or as argument")
+
+    time = data.index
+    is_timestamped = True
+    if epoch is not None:
+        if not isinstance(time, pd.DatetimeIndex):
+            raise ValueError(f"expected datetime index to encode with epoch but got {time.inferred_type}")
+        time = (time - epoch).total_seconds()
+    elif isinstance(time, pd.RangeIndex):
+        is_timestamped = False
+
+    if dtype is not None:
+        payload = payload.astype(dtype, copy=False)
+
+    ncols = payload.shape[1]
+    if length is not None and ncols != length:
+        raise ValueError(f"expected payload length {length} but got {ncols}")
+
+    if port is None:
+        port = 255
+
+    payloadtype = _payloadtypefromdtype[payload.dtype]
+    payloadlength = ncols * payload.dtype.itemsize
+    stride = payloadlength + 6
+    if is_timestamped:
+        payloadtype |= _PAYLOAD_TIMESTAMP_MASK
+        stride += 6
+
+    buffer = np.empty((nrows, stride), dtype=np.uint8)
+    buffer[:, 0] = msgtype
+    buffer[:, 1:5] = [stride - 2, address, port, payloadtype]
+
+    payloadoffset = 5
+    if is_timestamped:
+        seconds = time.astype(np.uint32)
+        micros = np.around(((time - seconds) / _SECONDS_PER_TICK).values).astype(np.uint16)
+        buffer[:, 5:9] = np.ndarray((nrows, 4), dtype=np.uint8, buffer=seconds.values)
+        buffer[:, 9:11] = np.ndarray((nrows, 2), dtype=np.uint8, buffer=micros)
+        payloadoffset += 6
+
+    payloadstop = payloadoffset + payloadlength
+    buffer[:, payloadoffset:payloadstop] = np.ndarray(
+        (nrows, payloadlength), dtype=np.uint8, buffer=np.ascontiguousarray(payload)
+    )
+    buffer[:, -1] = np.sum(buffer[:, 0:-1], axis=1, dtype=np.uint8)
+    return buffer.reshape(-1)

{harp_python-0.3.0 → harp_python-0.4.1}/harp/model.py

@@ -7,14 +7,7 @@ from __future__ import annotations
 from enum import Enum
 from typing import Annotated, Dict, List, Optional, Union
 
-from pydantic import (
-    BaseModel,
-    BeforeValidator,
-    ConfigDict,
-    Field,
-    RootModel,
-    field_serializer,
-)
+from pydantic import BaseModel, BeforeValidator, ConfigDict, Field, RootModel, field_serializer
 
 
 class PayloadType(str, Enum):

{harp_python-0.3.0 → harp_python-0.4.1}/harp/reader.py

@@ -6,20 +6,21 @@ from functools import partial
 from math import log2
 from os import PathLike
 from pathlib import Path
-from typing import Any, BinaryIO, Callable, Iterable, Mapping, Optional, Protocol, Union
+from typing import Callable, Iterable, Mapping, Optional, Protocol, Union
 
 from numpy import dtype
 from pandas import DataFrame, Series
-from pandas._typing import Axes
+from pandas._typing import Axes  # pyright: ignore[reportPrivateImportUsage]
 
 from harp.io import MessageType, read
 from harp.model import BitMask, GroupMask, Model, PayloadMember, Register
 from harp.schema import read_schema
+from harp.typing import _BufferLike, _FileLike
 
 
 @dataclass
 class _ReaderParams:
-    path: Path
+    base_path: Path
     epoch: Optional[datetime] = None
     keep_type: bool = False
 
@@ -27,7 +28,7 @@ class _ReaderParams:
 class _ReadRegister(Protocol):
     def __call__(
         self,
-        file: Optional[Union[str, bytes, PathLike[Any], BinaryIO]] = None,
+        file_or_buf: Optional[Union[_FileLike, _BufferLike]] = None,
         epoch: Optional[datetime] = None,
         keep_type: bool = False,
     ) -> DataFrame: ...
@@ -81,12 +82,12 @@ def _compose_parser(
     params: _ReaderParams,
 ) -> Callable[..., DataFrame]:
     def parser(
-        file: Optional[Union[str, bytes, PathLike[Any], BinaryIO]] = None,
+        data: Optional[Union[_FileLike, _BufferLike]] = None,
         columns: Optional[Axes] = None,
         epoch: Optional[datetime] = params.epoch,
         keep_type: bool = params.keep_type,
     ):
-        df = g(file, columns, epoch, keep_type)
+        df = g(data, columns, epoch, keep_type)
         result = f(df)
         type_col = df.get(MessageType.__name__)
         if type_col is not None:
@@ -167,16 +168,16 @@ def _create_payloadmember_parser(device: Model, member: PayloadMember):
 
 def _create_register_reader(register: Register, params: _ReaderParams):
     def reader(
-        file: Optional[Union[str, bytes, PathLike[Any], BinaryIO]] = None,
+        file_or_buf: Optional[Union[_FileLike, _BufferLike]] = None,
         columns: Optional[Axes] = None,
         epoch: Optional[datetime] = params.epoch,
         keep_type: bool = params.keep_type,
     ):
-        if file is None:
-            file = f"{params.path}_{register.address}.bin"
+        if file_or_buf is None:
+            file_or_buf = f"{params.base_path}_{register.address}.bin"
 
         data = read(
-            file,
+            file_or_buf,
             address=register.address,
             dtype=dtype(register.type),
             length=register.length,
@@ -189,7 +190,7 @@ def _create_register_reader(register: Register, params: _ReaderParams):
     return reader
 
 
-def _create_register_parser(device: Model, name: str, params: _ReaderParams):
+def _create_register_handler(device: Model, name: str, params: _ReaderParams):
     register = device.registers[name]
     reader = _create_register_reader(register, params)
 
@@ -197,29 +198,34 @@ def _create_register_parser(device: Model, name: str, params: _ReaderParams):
         key = register.maskType.root
         bitMask = None if device.bitMasks is None else device.bitMasks.get(key)
         if bitMask is not None:
-            parser = _create_bitmask_parser(bitMask)
-            reader = _compose_parser(parser, reader, params)
+            bitmask_parser = _create_bitmask_parser(bitMask)
+            reader = _compose_parser(bitmask_parser, reader, params)
             return RegisterReader(register, reader)
 
         groupMask = None if device.groupMasks is None else device.groupMasks.get(key)
         if groupMask is not None:
-            parser = _create_groupmask_parser(name, groupMask)
-            reader = _compose_parser(parser, reader, params)
+            groupmask_parser = _create_groupmask_parser(name, groupMask)
+            reader = _compose_parser(groupmask_parser, reader, params)
             return RegisterReader(register, reader)
 
     if register.payloadSpec is not None:
-        payload_parsers = [
+        member_parsers = [
             (key, _create_payloadmember_parser(device, member))
             for key, member in register.payloadSpec.items()
         ]
 
-        def parser(df: DataFrame):
-            return DataFrame({n: f(df) for n, f in payload_parsers}, index=df.index)
+        def payload_parser(df: DataFrame):
+            return DataFrame({n: f(df) for n, f in member_parsers}, index=df.index)
 
-        reader = _compose_parser(parser, reader, params)
+        reader = _compose_parser(payload_parser, reader, params)
         return RegisterReader(register, reader)
 
-    reader = partial(reader, columns=[name])
+    columns = (
+        [name]
+        if register.length is None or register.length == 1
+        else [f"{name}_{i}" for i in range(register.length)]
+    )
+    reader = partial(reader, columns=columns)
     return RegisterReader(register, reader)
 
 
@@ -265,7 +271,7 @@ def create_reader(
         base_path = path / device.device if is_dir else path.parent / device.device
 
     reg_readers = {
-        name: _create_register_parser(device, name, _ReaderParams(base_path, epoch, keep_type))
+        name: _create_register_handler(device, name, _ReaderParams(base_path, epoch, keep_type))
         for name in device.registers.keys()
     }
     return DeviceReader(device, reg_readers)

{harp_python-0.3.0 → harp_python-0.4.1}/harp/schema.py

@@ -8,6 +8,9 @@ from harp.model import Model, Registers
 
 
 def _read_common_registers() -> Registers:
+    if __package__ is None:
+        raise ValueError("__package__ is None: unable to read common registers")
+
     file = resources.files(__package__) / "common.yml"
     with file.open("r") as fileIO:
         return parse_yaml_raw_as(Registers, fileIO.read())

harp_python-0.4.1/harp/typing.py

@@ -0,0 +1,13 @@
+import mmap
+import sys
+from os import PathLike
+from typing import Any, BinaryIO, Union
+
+from numpy.typing import NDArray
+
+if sys.version_info >= (3, 12):
+    from collections.abc import Buffer as _BufferLike
+else:
+    _BufferLike = Union[bytes, bytearray, memoryview, mmap.mmap, NDArray[Any]]
+
+_FileLike = Union[str, PathLike[str], BinaryIO]

{harp_python-0.3.0 → harp_python-0.4.1}/harp_python.egg-info/PKG-INFO

@@ -1,9 +1,9 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: harp-python
-Version: 0.3.0
+Version: 0.4.1
 Summary: A low-level interface for loading binary Harp protocol data
 Author-email: Goncalo Lopes <g.lopes@neurogears.org>
-License: MIT License
+License-Expression: MIT
 Project-URL: Bug Tracker, https://github.com/harp-tech/harp-python/issues
 Project-URL: Source Code, https://github.com/harp-tech/harp-python
 Classifier: Development Status :: 4 - Beta
@@ -12,7 +12,6 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Operating System :: OS Independent
-Classifier: License :: OSI Approved :: MIT License
 Requires-Python: >=3.9.0
 Description-Content-Type: text/markdown
 License-File: LICENSE
@@ -29,6 +28,7 @@ Requires-Dist: codespell; extra == "dev"
 Provides-Extra: jupyter
 Requires-Dist: ipykernel; extra == "jupyter"
 Requires-Dist: matplotlib; extra == "jupyter"
+Dynamic: license-file
 
 # harp-python
 

{harp_python-0.3.0 → harp_python-0.4.1}/harp_python.egg-info/SOURCES.txt

@@ -11,6 +11,7 @@ harp/io.py
 harp/model.py
 harp/reader.py
 harp/schema.py
+harp/typing.py
 harp_python.egg-info/PKG-INFO
 harp_python.egg-info/SOURCES.txt
 harp_python.egg-info/dependency_links.txt
@@ -23,4 +24,6 @@ tests/test_reader.py
 tests/test_schema.py
 tests/data/device.yml
 tests/data/device_0.bin
+tests/data/device_44.bin
+tests/data/empty_0.bin
 tests/data/write_0.bin

{harp_python-0.3.0 → harp_python-0.4.1}/pyproject.toml

@@ -7,7 +7,7 @@ description = "A low-level interface for loading binary Harp protocol data"
 readme = "README.md"
 requires-python = ">=3.9.0"
 dynamic = ["version"]
-license = {text = "MIT License"}
+license = "MIT"
 
 dependencies = [
     "pydantic-yaml",
@@ -20,8 +20,7 @@ classifiers = [
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
-    "Operating System :: OS Independent",
-    "License :: OSI Approved :: MIT License",
+    "Operating System :: OS Independent"
 ]
 
 [project.urls]
@@ -67,6 +66,9 @@ exclude = [
     "reflex-generator"
 ]
 
+[tool.ruff.lint]
+select = ["I"]
+
 [tool.pyright]
 venvPath = "."
 venv = ".venv"

{harp_python-0.3.0 → harp_python-0.4.1}/tests/data/device.yml

@@ -18,6 +18,28 @@ registers:
     type: U8
     maskType: DigitalInputs
     description: Reports the state of the digital input lines.
+  AnalogData:
+    address: 44
+    type: S16
+    length: 3
+    access: Event
+    description: Reports the current values of the analog input lines.
+  AnalogDataPayloadSpec:
+    address: 44
+    type: S16
+    length: 3
+    access: Event
+    description: Reports the current values of the analog input lines.
+    payloadSpec:
+      AnalogInput0:
+        offset: 0
+        description: The voltage at the output of the ADC channel 0.
+      Encoder:
+        offset: 1
+        description: The quadrature counter value on Port 2
+      AnalogInput1:
+        offset: 2
+        description: The voltage at the output of the ADC channel 1.
 bitMasks:
   DigitalInputs:
     description: Specifies the state of the digital input lines.

{harp_python-0.3.0 → harp_python-0.4.1}/tests/params.py

@@ -1,4 +1,5 @@
 from dataclasses import dataclass
+from datetime import datetime
 from os import PathLike
 from pathlib import Path
 from typing import Iterable, Optional, Type, Union
@@ -19,6 +20,8 @@ class DataFileParam:
     expected_dtype: Optional[np.dtype] = None
     expected_length: Optional[int] = None
     expected_error: Optional[Type[BaseException]] = None
+    repeat_data: Optional[int] = None
+    epoch: Optional[datetime] = None
     keep_type: bool = False
 
     def __post_init__(self):

{harp_python-0.3.0 → harp_python-0.4.1}/tests/test_io.py

@@ -1,10 +1,11 @@
 from contextlib import nullcontext
 
 import numpy as np
+import pandas as pd
 import pytest
 from pytest import mark
 
-from harp.io import MessageType, read
+from harp.io import REFERENCE_EPOCH, MessageType, read, to_buffer
 from tests.params import DataFileParam
 
 testdata = [
@@ -29,6 +30,10 @@ testdata = [
     ),
     DataFileParam(path="data/write_0.bin", expected_address=0, expected_rows=4),
     DataFileParam(path="data/write_0.bin", expected_address=0, expected_rows=4, keep_type=True),
+    DataFileParam(path="data/device_0.bin", expected_rows=300, repeat_data=300),
+    DataFileParam(path="data/device_0.bin", expected_rows=1, epoch=REFERENCE_EPOCH),
+    DataFileParam(path="data/empty_0.bin", expected_rows=0, epoch=REFERENCE_EPOCH),
+    DataFileParam(path="data/empty_0.bin", expected_rows=0),
 ]
 
 
@@ -36,17 +41,48 @@ testdata = [
 def test_read(dataFile: DataFileParam):
     context = pytest.raises if dataFile.expected_error else nullcontext
     with context(dataFile.expected_error):  # type: ignore
+        file_or_buf = dataFile.path
+        if dataFile.repeat_data:
+            with open(file_or_buf, "rb") as f:
+                file_or_buf = f.read() * dataFile.repeat_data
+
         data = read(
-            dataFile.path,
+            file_or_buf,
             address=dataFile.expected_address,
             dtype=dataFile.expected_dtype,
             length=dataFile.expected_length,
+            epoch=dataFile.epoch,
             keep_type=dataFile.keep_type,
         )
+
         assert len(data) == dataFile.expected_rows
+        assert isinstance(data.index, pd.DatetimeIndex if dataFile.epoch else pd.Index)
         if dataFile.keep_type:
             assert MessageType.__name__ in data.columns and data[MessageType.__name__].dtype == "category"
 
         if dataFile.expected_cols:
             for col in dataFile.expected_cols:
                 assert col in data.columns
+
+
+writedata = [
+    DataFileParam(path="data/device_0.bin", expected_rows=1, expected_address=0, keep_type=True),
+]
+
+
+@mark.parametrize("dataFile", writedata)
+def test_write(dataFile: DataFileParam):
+    if dataFile.expected_address is None:
+        raise AssertionError("expected address must be defined for all write tests")
+
+    buffer = np.fromfile(dataFile.path, np.uint8)
+    data = read(
+        buffer,
+        address=dataFile.expected_address,
+        dtype=dataFile.expected_dtype,
+        length=dataFile.expected_length,
+        keep_type=dataFile.keep_type,
+    )
+    assert len(data) == dataFile.expected_rows
+    write_buffer = to_buffer(data, address=dataFile.expected_address, length=dataFile.expected_length)
+    assert np.array_equal(buffer, write_buffer)

{harp_python-0.3.0 → harp_python-0.4.1}/tests/test_reader.py

@@ -1,4 +1,4 @@
-import numpy as np
+import pandas as pd
 from pytest import mark
 
 from harp.io import REFERENCE_EPOCH, MessageType
@@ -9,12 +9,17 @@ testdata = [
     DeviceSchemaParam(
         path="data",
         expected_whoAmI=0,
-        expected_registers=["DigitalInputMode"],
+        expected_registers=["AnalogData"],
     ),
     DeviceSchemaParam(
         path="data/device.yml",
         expected_whoAmI=0,
-        expected_registers=["DigitalInputMode"],
+        expected_registers=["AnalogData"],
+    ),
+    DeviceSchemaParam(
+        path="data/device.yml",
+        expected_whoAmI=0,
+        expected_registers=["AnalogDataPayloadSpec"],
     ),
 ]
 
@@ -26,8 +31,13 @@ def test_create_reader(schemaFile: DeviceSchemaParam):
 
     whoAmI = reader.WhoAmI.read()
     assert reader.device.whoAmI == whoAmI.iloc[0, 0]
-    assert whoAmI.index.dtype.type == np.datetime64
+    assert isinstance(whoAmI.index, pd.DatetimeIndex)
 
     whoAmI = reader.WhoAmI.read(epoch=None, keep_type=True)
-    assert whoAmI.index.dtype.type == np.float64
+    assert isinstance(whoAmI.index, pd.Index)
     assert whoAmI.iloc[0, -1] == MessageType.READ.name
+
+    if schemaFile.expected_registers:
+        for register_name in schemaFile.expected_registers:
+            data = reader.registers[register_name].read()
+            assert isinstance(data.index, pd.DatetimeIndex)

harp_python-0.3.0/.gitignore

@@ -1,12 +0,0 @@
-# Environment files
-.vscode
-.venv
-
-# Python temp files
-__pycache__
-*.pyc
-*.egg-info
-dist
-
-# Data files
-/data

harp_python-0.3.0/harp/__init__.py

@@ -1,5 +0,0 @@
-from harp.io import REFERENCE_EPOCH, MessageType, read
-from harp.reader import create_reader
-from harp.schema import read_schema
-
-__all__ = ["REFERENCE_EPOCH", "MessageType", "read", "create_reader", "read_schema"]

harp_python-0.3.0/harp/io.py

@@ -1,122 +0,0 @@
-from datetime import datetime
-from enum import IntEnum
-from os import PathLike
-from typing import Any, BinaryIO, Optional, Union
-
-import numpy as np
-import pandas as pd
-from pandas._typing import Axes
-
-REFERENCE_EPOCH = datetime(1904, 1, 1)
-"""The reference epoch for UTC harp time."""
-
-
-class MessageType(IntEnum):
-    """Specifies the type of a Harp message."""
-
-    NA = 0
-    READ = 1
-    WRITE = 2
-    EVENT = 3
-
-
-_SECONDS_PER_TICK = 32e-6
-_messagetypes = [type.name for type in MessageType]
-_payloadtypes = {
-    1: np.dtype(np.uint8),
-    2: np.dtype(np.uint16),
-    4: np.dtype(np.uint32),
-    8: np.dtype(np.uint64),
-    129: np.dtype(np.int8),
-    130: np.dtype(np.int16),
-    132: np.dtype(np.int32),
-    136: np.dtype(np.int64),
-    68: np.dtype(np.float32),
-}
-
-
-def read(
-    file: Union[str, bytes, PathLike[Any], BinaryIO],
-    address: Optional[int] = None,
-    dtype: Optional[np.dtype] = None,
-    length: Optional[int] = None,
-    columns: Optional[Axes] = None,
-    epoch: Optional[datetime] = None,
-    keep_type: bool = False,
-):
-    """Read single-register Harp data from the specified file.
-
-    Parameters
-    ----------
-    file
-        Open file object or filename containing binary data from
-        a single device register.
-    address
-        Expected register address. If specified, the address of
-        the first message in the file is used for validation.
-    dtype
-        Expected data type of the register payload. If specified, the
-        payload type of the first message in the file is used for validation.
-    length
-        Expected number of elements in register payload. If specified, the
-        payload length of the first message in the file is used for validation.
-    columns
-        The optional column labels to use for the data values.
-    epoch
-        Reference datetime at which time zero begins. If specified,
-        the result data frame will have a datetime index.
-    keep_type
-        Specifies whether to include a column with the message type.
-
-    Returns
-    -------
-        A pandas data frame containing message data, sorted by time.
-    """
-    data = np.fromfile(file, dtype=np.uint8)
-    if len(data) == 0:
-        return pd.DataFrame(columns=columns, index=pd.Index([], dtype=np.float64, name="Time"))
-
-    if address is not None and address != data[2]:
-        raise ValueError(f"expected address {address} but got {data[2]}")
-
-    index = None
-    stride = int(data[1] + 2)
-    nrows = len(data) // stride
-    payloadtype = data[4]
-    payloadoffset = 5
-    if payloadtype & 0x10 != 0:
-        seconds = np.ndarray(nrows, dtype=np.uint32, buffer=data, offset=payloadoffset, strides=stride)
-        payloadoffset += 4
-        micros = np.ndarray(nrows, dtype=np.uint16, buffer=data, offset=payloadoffset, strides=stride)
-        payloadoffset += 2
-        time = micros * _SECONDS_PER_TICK + seconds
-        payloadtype = payloadtype & ~np.uint8(0x10)
-        if epoch is not None:
-            time = epoch + pd.to_timedelta(time, "s")  # type: ignore
-        index = pd.Series(time)
-        index.name = "Time"
-
-    payloadsize = stride - payloadoffset - 1
-    payloadtype = _payloadtypes[payloadtype]
-    if dtype is not None and dtype != payloadtype:
-        raise ValueError(f"expected payload type {dtype} but got {payloadtype}")
-
-    elementsize = payloadtype.itemsize
-    payloadshape = (nrows, payloadsize // elementsize)
-    if length is not None and length != payloadshape[1]:
-        raise ValueError(f"expected payload length {length} but got {payloadshape[1]}")
-
-    payload = np.ndarray(
-        payloadshape,
-        dtype=payloadtype,
-        buffer=data,
-        offset=payloadoffset,
-        strides=(stride, elementsize),
-    )
-
-    result = pd.DataFrame(payload, index=index, columns=columns)
-    if keep_type:
-        msgtype = np.ndarray(nrows, dtype=np.uint8, buffer=data, offset=0, strides=stride)
-        msgtype = pd.Categorical.from_codes(msgtype, categories=_messagetypes)  # type: ignore
-        result[MessageType.__name__] = msgtype
-    return result