seabirdfilehandler 0.4.2__py3-none-any.whl → 0.4.3__py3-none-any.whl

seabirdfilehandler/datatablefiles.py

@@ -10,8 +10,8 @@ from seabirdfilehandler.parameter import Parameters
 from seabirdfilehandler.validation_modules import CnvValidationList
 from seabirdfilehandler.seabirdfiles import SeaBirdFile
 from seabirdfilehandler.dataframe_meta_accessor import (
-    SeriesMetaAccessor,
-    DataFrameMetaAccessor,
+    SeriesMetaAccessor,  # noqa: F401
+    DataFrameMetaAccessor,  # noqa: F401
 )
 
 logger = logging.getLogger(__name__)
@@ -384,7 +384,7 @@ class CnvFile(DataTableFile):
         a list of dictionaries, that organize the table header information
 
         """
-        if header_info is []:
+        if header_info == []:
             header_info = self.data_table_description
         table_header = {}
         duplicate_columns = []

seabirdfilehandler/file_collection.py

@@ -4,8 +4,14 @@ from collections import UserList
 from typing import Type
 import pandas as pd
 import numpy as np
-from seabirdfilehandler import SeaBirdFile, CnvFile, BottleFile, BottleLogFile
+from seabirdfilehandler import (
+    SeaBirdFile,
+    CnvFile,
+    BottleFile,
+    BottleLogFile,
+)
 from seabirdfilehandler.datatablefiles import DataTableFile
+from seabirdfilehandler.utils import get_unique_sensor_data
 
 logger = logging.getLogger(__name__)
 
@@ -49,7 +55,9 @@ class FileCollection(UserList):
                 self.df = self.get_collection_dataframe(self.df_list)
             if self.file_type == CnvFile:
                 self.data_meta_info = self.get_data_table_meta_info()
-            self.sensor_data = self.get_sensor_data()
+            self.sensor_data = get_unique_sensor_data(
+                [file.sensors for file in self.data]
+            )
 
     def __str__(self):
         return "/n".join(self.data)
@@ -247,24 +255,3 @@ class FileCollection(UserList):
     def get_data_table_meta_info(self) -> list[list[dict]]:
         """ """
         return [file.data_header_meta_info for file in self.data]
-
-    def get_sensor_data(self) -> list[tuple[list[dict]]]:
-        """ """
-        unique = []
-        last_unique = None
-        for file in [file for file in self.data]:
-            cast_sensors = file.sensors
-            if last_unique is None:
-                unique.append((file.file_name, cast_sensors))
-            else:
-                differing_dicts = [
-                    current_dict
-                    for last_dict, current_dict in zip(
-                        last_unique, cast_sensors
-                    )
-                    if current_dict != last_dict
-                ]
-                if differing_dicts:
-                    unique.append((file.file_name, differing_dicts))
-            last_unique = cast_sensors
-        return unique

seabirdfilehandler/utils.py

@@ -0,0 +1,53 @@
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+def get_unique_sensor_data(
+    sensor_data: list[list[dict]],
+) -> list[tuple[list[dict]]]:
+    """
+    Returns all the unique sensors and their configuration used in the given
+    collection of sensor data. These will typically be parsed from xml inside
+    .cnv or .xmlcon files.
+    If for example, the first oxygen sensor has been replaced after the 8 cast,
+    then we will see that in the output structure by a seconde tuple, with the
+    number 8 and the individual sensor information for that new oxygen sensor.
+
+    Parameters
+    ----------
+    sensor_data:
+        The structure of xml-parsed dicts inside two organizing lists.
+
+    Returns
+    -------
+    The input structure stripped down to unique sensor data and appended by
+    the index, at which this new sensor appeared the first time.
+
+    """
+    unique = []
+    last_unique = None
+    for index, individual_sensor_data in enumerate(
+        [file for file in sensor_data]
+    ):
+        if last_unique is None:
+            unique.append((index, individual_sensor_data))
+        else:
+            differing_dicts = [
+                current_dict
+                for last_dict, current_dict in zip(
+                    last_unique, individual_sensor_data
+                )
+                if current_dict != last_dict
+            ]
+            if differing_dicts:
+                unique.append((index, differing_dicts))
+        last_unique = individual_sensor_data
+    return unique
+
+
+class UnexpectedFileFormat(Exception):
+    def __init__(self, file_type: str, error: str) -> None:
+        message = f"{file_type} is not formatted as expected: {error}"
+        logger.error(message)
+        super().__init__(message)

seabirdfilehandler/xmlfiles.py

@@ -4,6 +4,8 @@ import xml.etree.ElementTree as ET
 import json
 import xmltodict
 
+from seabirdfilehandler.utils import UnexpectedFileFormat
+
 
 class XMLFile(UserDict):
     """
@@ -78,6 +80,58 @@ class XMLCONFile(XMLFile):
 
     def __init__(self, path_to_file):
         super().__init__(path_to_file)
+        self.sensor_info = self.get_sensor_info()
+
+    def get_sensor_info(self) -> list[dict]:
+        """
+        Creates a multilevel dictionary, dropping the first four dictionaries,
+        to retrieve pure sensor information.
+
+        Returns
+        -------
+        A list of all the individual sensor information, stored in dictionaries
+
+        """
+        try:
+            sensors = self.data["SBE_InstrumentConfiguration"]["Instrument"][
+                "SensorArray"
+            ]["Sensor"]
+        except KeyError as error:
+            raise UnexpectedFileFormat("XMLCON", error)
+        else:
+            # create a tidied version of the xml-parsed sensor dict
+            sensor_names = []
+            tidied_sensor_list = []
+            for entry in sensors:
+                sensor_key = list(entry.keys())[-1]
+                if not sensor_key.endswith(("Sensor", "Meter")):
+                    continue
+                sensor_name = sensor_key.removesuffix("Sensor")
+                # the wetlab sensors feature a suffix _Sensor
+                sensor_name = sensor_name.removesuffix("_")
+                # assuming, that the first sensor in the xmlcon is also on the
+                # first sensor strand, the second occurence of the name is
+                # suffixed with '2'
+                if sensor_name in sensor_names:
+                    sensor_name += "2"
+                sensor_names.append(sensor_name)
+                # move the calibration info one dictionary level up
+                calibration_info = entry[sensor_key]
+                # build the new dictionary
+                try:
+                    new_dict = {
+                        "Channel": str(int(entry["@index"]) + 1),
+                        "SensorName": sensor_name,
+                        **calibration_info,
+                    }
+                except TypeError:
+                    new_dict = {
+                        "Channel": entry["@Channel"],
+                        "SensorName": sensor_name,
+                        "Info": calibration_info,
+                    }
+                tidied_sensor_list.append(new_dict)
+            return tidied_sensor_list
 
 
 class PsaFile(XMLFile):

{seabirdfilehandler-0.4.2.dist-info → seabirdfilehandler-0.4.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: seabirdfilehandler
-Version: 0.4.2
+Version: 0.4.3
 Summary: Library of parsers to interact with SeaBird CTD files.
 Keywords: CTD,parser,seabird,data
 Author: Emil Michels

seabirdfilehandler-0.4.3.dist-info/RECORD

@@ -0,0 +1,14 @@
+seabirdfilehandler/__init__.py,sha256=8Vk2TURWXv2_NG_U_fR0fIVDFymkFkBipvmlS5ucB3M,147
+seabirdfilehandler/dataframe_meta_accessor.py,sha256=x4mSEN49us6Ezzjdt41fl5Ry8IJR09ORrZ1roOIJbyc,6439
+seabirdfilehandler/datatablefiles.py,sha256=yzTAzsWcdrHYaa2QaR6OpdEs-cu2py8V5o79s2Uz7MM,31646
+seabirdfilehandler/file_collection.py,sha256=nWyi5FToCV9-e_zcaLhRb4oOt8KAmyHC-SBGLJO9KQ4,6909
+seabirdfilehandler/logging.yaml,sha256=mXxbhJPio3OGaukTpc3rLGA8Ywq1DNqp0Vn5YCbH6jY,459
+seabirdfilehandler/parameter.py,sha256=UyKb_HGQ57pETdhSfR5FbJ60aOj8_d3_Tgw_akth0TY,13283
+seabirdfilehandler/seabirdfiles.py,sha256=BKLyk5gUMkt1CG4ljDXlCqcr5zej0-9PjPS0sX2E4n8,7449
+seabirdfilehandler/utils.py,sha256=5KXdB8Hdv65dv5tPyXxNMct1mCEOyA3S8XP54AFAnx0,1745
+seabirdfilehandler/validation_modules.py,sha256=eZ6x0giftUtlxnRMOnK_vCkgccdwUXPrDjajFa-E6n0,4698
+seabirdfilehandler/xmlfiles.py,sha256=L_puQf8eg0ojv85AyEMID4jnwkOlV_fgZP3W5yeSUBY,4668
+seabirdfilehandler-0.4.3.dist-info/LICENSE,sha256=Ifd1VPmYv32oJd2QVh3wIQP9X05vYJlcY6kONz360ws,34603
+seabirdfilehandler-0.4.3.dist-info/METADATA,sha256=ODlAzixojiAr_KtwyXVmYlWDn0UJTzFlp_RBAWzLxug,1289
+seabirdfilehandler-0.4.3.dist-info/WHEEL,sha256=XbeZDeTWKc1w7CSIyre5aMDU_-PohRwTQceYnisIYYY,88
+seabirdfilehandler-0.4.3.dist-info/RECORD,,

seabirdfilehandler-0.4.2.dist-info/RECORD

@@ -1,13 +0,0 @@
-seabirdfilehandler/__init__.py,sha256=8Vk2TURWXv2_NG_U_fR0fIVDFymkFkBipvmlS5ucB3M,147
-seabirdfilehandler/dataframe_meta_accessor.py,sha256=x4mSEN49us6Ezzjdt41fl5Ry8IJR09ORrZ1roOIJbyc,6439
-seabirdfilehandler/datatablefiles.py,sha256=LwHAh_2dUSERHvs-ql_wBl2a5jcHabOAD5i33fASfs0,31618
-seabirdfilehandler/file_collection.py,sha256=eFJqKnKMwAGC37D27FsPUi7ceTU3iUC0Y_wG2HzxwmU,7518
-seabirdfilehandler/logging.yaml,sha256=mXxbhJPio3OGaukTpc3rLGA8Ywq1DNqp0Vn5YCbH6jY,459
-seabirdfilehandler/parameter.py,sha256=UyKb_HGQ57pETdhSfR5FbJ60aOj8_d3_Tgw_akth0TY,13283
-seabirdfilehandler/seabirdfiles.py,sha256=BKLyk5gUMkt1CG4ljDXlCqcr5zej0-9PjPS0sX2E4n8,7449
-seabirdfilehandler/validation_modules.py,sha256=eZ6x0giftUtlxnRMOnK_vCkgccdwUXPrDjajFa-E6n0,4698
-seabirdfilehandler/xmlfiles.py,sha256=Es0pUCoB0af-meCH-75h-s-r0mEfrthl-8QDjqNBWPk,2441
-seabirdfilehandler-0.4.2.dist-info/LICENSE,sha256=Ifd1VPmYv32oJd2QVh3wIQP9X05vYJlcY6kONz360ws,34603
-seabirdfilehandler-0.4.2.dist-info/METADATA,sha256=9HcKSVWHN90OnsS86UftA7plWh6ho_BUV1sbf5eHESo,1289
-seabirdfilehandler-0.4.2.dist-info/WHEEL,sha256=XbeZDeTWKc1w7CSIyre5aMDU_-PohRwTQceYnisIYYY,88
-seabirdfilehandler-0.4.2.dist-info/RECORD,,