force segyio header buffer to big

src/mdio/segy/_workers.py

@@ -42,7 +42,6 @@ def header_scan_worker(
         byte_types: Tuple consisting of the data types for the index attributes.
         trace_range: Tuple consisting of the trace ranges to read
         index_names: Tuple of the names for the index attributes
-        segy_endian: Endianness of the input SEG-Y. Rev.2 allows little endian
 
     Returns:
         dictionary with headers:  keys are the index names, values are numpy
@@ -79,7 +78,9 @@ def header_scan_worker(
     # First we create a struct to unpack the 240-byte trace headers.
     # The struct only knows about dimension keys, and their byte offsets.
     # Pads the rest of the data with voids.
-    endian = ByteOrder[segy_endian.upper()]
+    # NOTE: segyio buffer is always big endian. In the future if we use
+    # a different parser, we need to expose this as a parameter.
+    endian = ByteOrder.BIG
 
     # Handle byte offsets
     offsets = [0 if byte_loc is None else byte_loc - 1 for byte_loc in byte_locs]

src/mdio/segy/parsers.py

@@ -123,7 +123,6 @@ def parse_trace_headers(
             repeat(byte_locs),
             repeat(byte_types),
             repeat(index_names),
-            repeat(segy_endian),
             chunksize=2,  # Not array chunks. This is for `multiprocessing`
         )