Enable decoding of tbm dataset name

pygac/pod_reader.py

@@ -38,20 +38,20 @@
 
 import datetime
 import logging
+
 try:
     from enum import IntFlag
 except ImportError:
     # python version < 3.6, use a simple object without nice representation
     IntFlag = object
 
 import numpy as np
-
-from pyorbital.geoloc_instrument_definitions import avhrr_gac
 from pyorbital.geoloc import compute_pixels, get_lonlatalt
+from pyorbital.geoloc_instrument_definitions import avhrr_gac
 
 from pygac.clock_offsets_converter import get_offsets
 from pygac.correct_tsm_issue import TSM_AFFECTED_INTERVALS_POD, get_tsm_idx
-from pygac.reader import Reader, ReaderError, NoTLEData
+from pygac.reader import DecodingError, NoTLEData, Reader, ReaderError
 from pygac.slerp import slerp
 from pygac.utils import file_opener
 
@@ -322,18 +322,13 @@ def read_header(cls, filename, fileobj=None, header_date="auto"):
                 fd_.read(tbm_header.itemsize),
                 dtype=tbm_header, count=1)
             try:
-                data_set_name = _tbm_head['data_set_name'].decode()
-            except UnicodeDecodeError:
-                data_set_name = '---'
-            allowed_empty = (42*b'\x00' + b'  ')
-            if (cls.data_set_pattern.match(data_set_name)
-                    or (_tbm_head['data_set_name'] == allowed_empty)):
-                tbm_head = _tbm_head.copy()
+                tbm_head = cls._validate_tbm_header(_tbm_head)
                 tbm_offset = tbm_header.itemsize
-            else:
-                fd_.seek(0)
+            except DecodingError:
                 tbm_head = None
                 tbm_offset = 0
+
+            fd_.seek(tbm_offset, 0)
             header = cls.choose_header_based_on_timestamp(header_date, fd_)
             fd_.seek(tbm_offset, 0)
             # need to copy frombuffer to have write access on head
@@ -344,6 +339,18 @@ def read_header(cls, filename, fileobj=None, header_date="auto"):
         cls._validate_header(head)
         return tbm_head, head
 
+    @classmethod
+    def _validate_tbm_header(cls, potential_tbm_header):
+        data_set_name = potential_tbm_header['data_set_name']
+        allowed_empty = (42*b'\x00' + b'  ')
+        if data_set_name == allowed_empty:
+            return potential_tbm_header.copy()
+
+        # This will raise a DecodingError if the data_set_name is not valid.
+        cls._decode_data_set_name(data_set_name)
+        return potential_tbm_header.copy()
+
+
     @classmethod
     def choose_header_based_on_timestamp(cls, header_date, fd_):
         """Choose the header dtype based on the timestamp."""

pygac/reader.py

@@ -209,16 +209,10 @@ def _correct_data_set_name(cls, header, filename):
             filename (str): path to file
         """
         filename = str(filename)
-        for encoding in "utf-8", "cp500":
-            data_set_name = header['data_set_name']
-            try:
-                data_set_name = cls._decode_data_set_name(data_set_name, encoding)
-            except DecodingError as err:
-                LOG.debug(str(err))
-            else:
-                header["data_set_name"] = data_set_name
-                break
-        else:
+        data_set_name = header['data_set_name']
+        try:
+            header["data_set_name"] = cls._decode_data_set_name(data_set_name)
+        except DecodingError:
             LOG.debug(f'The data_set_name in header {header["data_set_name"]} does not match.'
                       ' Use filename instead.')
             match = cls.data_set_pattern.search(filename)
@@ -232,7 +226,19 @@ def _correct_data_set_name(cls, header, filename):
         return header
 
     @classmethod
-    def _decode_data_set_name(cls, data_set_name, encoding):
+    def _decode_data_set_name(cls, data_set_name):
+        for encoding in "utf-8", "cp500":
+            try:
+                data_set_name = cls._decode_data_set_name_for_encoding(data_set_name, encoding)
+            except DecodingError as err:
+                LOG.debug(str(err))
+            else:
+                return data_set_name
+        else:
+            raise DecodingError("Could not reliably decode the dataset name.")
+
+    @classmethod
+    def _decode_data_set_name_for_encoding(cls, data_set_name, encoding):
         data_set_name = data_set_name.decode(encoding, errors='ignore')
         if not cls.data_set_pattern.match(data_set_name):
             raise DecodingError(f'The data_set_name in header {data_set_name} '

pygac/tests/test_reader.py

@@ -24,20 +24,20 @@
 import os
 import sys
 import unittest
-import pytest
-
 from unittest import mock
+
 import numpy as np
 import numpy.testing
-from pygac.gac_reader import GACReader, ReaderError
-from pygac.lac_reader import LACReader
-from pygac.pod_reader import POD_QualityIndicator
+import pytest
+
 from pygac.gac_pod import scanline
-from pygac.reader import NoTLEData
+from pygac.gac_reader import GACReader, ReaderError
 from pygac.lac_pod import LACPODReader
-
-from pygac.pod_reader import tbm_header as tbm_header_dtype, header3
 from pygac.lac_pod import scanline as lacpod_scanline
+from pygac.lac_reader import LACReader
+from pygac.pod_reader import POD_QualityIndicator, header3
+from pygac.pod_reader import tbm_header as tbm_header_dtype
+from pygac.reader import NoTLEData
 
 
 class TestPath(os.PathLike):
@@ -688,7 +688,7 @@ def pod_file_with_tbm_header(tmp_path):
     number_of_scans = 3
 
     tbm_header = np.zeros(1, dtype=tbm_header_dtype)
-    tbm_header["data_set_name"] = b"BRN.HRPT.NJ.D00322.S0334.E0319.B3031919.BL  "
+    tbm_header["data_set_name"] = "BRN.HRPT.NJ.D00322.S0334.E0319.B3031919.BL\x80\x80".encode("cp500")
     tbm_header["select_flag"] = b"S"
     tbm_header["beginning_latitude"] = b"+77"
     tbm_header["ending_latitude"] = b"+22"