Add SEG-Y Revision 2.1 to supported formats

README.md

@@ -105,8 +105,10 @@ are still in progress and not all validation logic is implemented yet:
 
 - ✅ Rev 0 (1975)
 - ✅ Rev 1 (2002)
-- ✅ Rev 2 (2017)
-- 🔲 Rev 2.1 (2023)
+- ✅ Rev 2 (2017)\*\*
+- ✅ Rev 2.1 (2023)\*\*
+
+\*\* The XML stanzas and extended trace headers are not currently supported.
 
 ### Custom SEG-Y Standards
 

src/segy/standards/__init__.py

@@ -6,10 +6,12 @@
 from segy.standards.spec import REV0
 from segy.standards.spec import REV1
 from segy.standards.spec import REV2
+from segy.standards.spec import REV21
 
 register_segy_standard(version_or_name=0.0, spec=REV0)
 register_segy_standard(version_or_name=1.0, spec=REV1)
 register_segy_standard(version_or_name=2.0, spec=REV2)
+register_segy_standard(version_or_name=2.1, spec=REV21)
 
 
 __all__ = ["get_segy_standard", "SegyStandard"]

src/segy/standards/fields/binary.py

@@ -57,10 +57,16 @@ class Rev2(SegStandardEnum):
     BYTE_ORDER                        = (97, "int32")
     SEGY_REVISION_MAJOR               = (301, "uint8")
     SEGY_REVISION_MINOR               = (302, "uint8")
-    MAX_EXTENDED_TRACE_HEADERS        = (307, "int16")
-    SURVEY_TYPE                       = (309, "int16")
+    MAX_EXTENDED_TRACE_HEADERS        = (307, "int32")
     TIME_BASIS_CODE                   = (311, "int16")
     NUM_TRACES                        = (313, "uint64")
     BYTE_OFFSET_FIRST_TRACE           = (321, "uint64")
     NUM_DATA_TRAILER_STANZAS          = (329, "int32")
+
+
+class Rev21(SegStandardEnum):
+    """Definition of SEG-Y Rev2 binary headers."""
+
+    MAX_EXTENDED_TRACE_HEADERS        = (307, "int16")
+    SURVEY_TYPE                       = (309, "int16")
 # fmt:on

src/segy/standards/spec.py

@@ -1,6 +1,5 @@
 """SEG-Y Revision 0 Specification."""
 
-
 from segy.schema.base import Endianness
 from segy.schema.format import ScalarType
 from segy.schema.format import TextHeaderEncoding
@@ -15,16 +14,25 @@
 from segy.standards.fields import trace
 
 BIN_HDR_FIELDS_REV0 = [field.model for field in binary.Rev0]
+TRC_HDR_FIELDS_REV0 = [field.model for field in trace.Rev0]
+
 BIN_HDR_FIELDS_REV1 = BIN_HDR_FIELDS_REV0 + [field.model for field in binary.Rev1]
-BIN_HDR_FIELDS_REV2 = BIN_HDR_FIELDS_REV1 + [field.model for field in binary.Rev2]
+TRC_HDR_FIELDS_REV1 = TRC_HDR_FIELDS_REV0 + [field.model for field in trace.Rev1]
 
+# Rev2 removes segy_revision and splits it into major/minor revision fields
+BIN_HDR_FIELDS_REV2 = BIN_HDR_FIELDS_REV1 + [field.model for field in binary.Rev2]
 BIN_HDR_FIELDS_REV2.remove(binary.Rev1.SEGY_REVISION.model)
 BIN_HDR_FIELDS_REV2 = sorted(BIN_HDR_FIELDS_REV2, key=lambda f: f.byte)
-
-TRC_HDR_FIELDS_REV0 = [field.model for field in trace.Rev0]
-TRC_HDR_FIELDS_REV1 = TRC_HDR_FIELDS_REV0 + [field.model for field in trace.Rev1]
 TRC_HDR_FIELDS_REV2 = TRC_HDR_FIELDS_REV1 + [field.model for field in trace.Rev2]
 
+# Rev2.1 splits maximum extended trace headers into two 16-bit.
+# First reduces it from 32-bit to 16-bit
+# Second gets new "survey type" field
+BIN_HDR_FIELDS_REV21 = BIN_HDR_FIELDS_REV2 + [field.model for field in binary.Rev21]
+BIN_HDR_FIELDS_REV21.remove(binary.Rev2.MAX_EXTENDED_TRACE_HEADERS.model)
+BIN_HDR_FIELDS_REV21 = sorted(BIN_HDR_FIELDS_REV21, key=lambda f: f.byte)
+TRC_HDR_FIELDS_REV21 = TRC_HDR_FIELDS_REV2  # no change
+
 
 text_header_ebcdic = TextHeaderSpec(
     rows=40,
@@ -68,3 +76,15 @@
         header=HeaderSpec(fields=TRC_HDR_FIELDS_REV2, item_size=240),
     ),
 )
+
+REV21 = SegySpec(
+    segy_standard=SegyStandard.REV21,
+    endianness=Endianness.BIG,
+    text_header=text_header_ebcdic,
+    binary_header=HeaderSpec(fields=BIN_HDR_FIELDS_REV21, item_size=400, offset=3200),
+    ext_text_header=ext_text_header_ebcdic_3200,
+    trace=TraceSpec(
+        data=TraceDataSpec(format=ScalarType.IBM32),
+        header=HeaderSpec(fields=TRC_HDR_FIELDS_REV21, item_size=240),
+    ),
+)

tests/test_segy_file.py

@@ -131,7 +131,8 @@ class TestSegyFile:
     """Test the usage of SegyFile class."""
 
     @pytest.mark.parametrize(
-        "standard", [SegyStandard.REV0, SegyStandard.REV1, SegyStandard.REV2]
+        "standard",
+        [SegyStandard.REV0, SegyStandard.REV1, SegyStandard.REV2, SegyStandard.REV21],
     )
     @pytest.mark.parametrize("endianness", [Endianness.BIG, Endianness.LITTLE])
     @pytest.mark.parametrize("sample_format", [ScalarType.IBM32, ScalarType.FLOAT32])