v0.6.0

Added

• MS3Record.parse_into() instance method for reuse-based parsing in high-throughput
  loops to eliminating per-record allocation/deallocation overhead.

Changed

• MS3Record.sourceid property now skips an unnecessary NULL check.
• cdata_to_string() no longer wraps .decode() in a redundant str() call.

v0.5.0

Added

• MS3Record.parse() for parsing a single record from a bytes-like buffer.

Changed

• Breaking: MS3Record.from_buffer() is now a generator. The context manager protocol (with from_buffer(...) as reader) and .read() method are no longer available. Plain for iteration is unchanged.
  • Replace with from_buffer(buf) as r: for msr in r: with for msr in from_buffer(buf):
  • Replace with from_buffer(buf) as r: msr = r.read() with msr = MS3Record.parse(buf)
• Optimize MS3Record.from_file() iteration by inlining next, eliminating a method dispatch per record

Removed

• MS3RecordBufferReader class removed. Use MS3Record.from_buffer() (generator) or MS3Record.parse() for single records.

v0.4.0

• MS3RecordValidator.from_file() significantly optimized

v0.3.0

Added

• MS3Record.validate_extra_headers() to return detailed validation errors
• MS3RecordValidator class for validating records in buffers and files

Changed

• Capture libmseed error/warning console output and include in MiniSEEDError exceptions
• Numerous edge-case fixes and usability improvements
• Build wheels for Python 3.14
• libmseed updated to v3.2.4

v0.2.0

Added

• MS3Record.get_extra_header() to get a specified extra header
• MS3Record.set_extra_header() to set a specified extra header
• MS3Record.merge_extra_headers() to apply a JSON Merge Patch to extra headers
• MS3Record.valid_extra_headers() to validate extra headers

Changed

• libmseed updated to v3.2.3

v0.1.0

Added

• MS3Record.generate(), a generator to produce miniSEED records
• MS3TraceList.generate(), a generator to produce miniSEED records

Deprecated

• MS3Record.pack() functionality, use MS3Record.generate() instead
• MS3TraceList.pack() functionality, use MS3TraceList.generate() instead

Changed

• Updated examples and inline docs to reflect current recommended patterns

v0.0.5

Fixed

• MS3Record.samprate consistently returns the rate in samples/second,
  and MS3Record.samprate_raw returns the record's raw value.
• Properly return None when MS3Record.sampletype is not set.

Added

• MS2Record.unpack_data() to decode sample payload after parsing.
• MS3Record.samprate_period_ns to return sample period in nanoseconds.
• MS3Record.samprate_period_seconds to return the period in seconds.

v0.0.4

Fixed

• Track update time of segments in an MS3TraceList consistently

Changed

• Update MS3TraceList.pack() to use new libmseed function for performance when packing with a specified flush_idle_seconds set.

v0.0.3

Added

• MS3TraceList.from_buffer() to create a trace list from miniSEED in a buffer
• MS3TraceList.add_buffer() to add miniSEED data from a buffer
• buffer parameter of MS3TracesList initialization
• system_time() to return the current system time in nanoseconds
• flush_idle_seconds parameter of MS3TraceList.pack() to control flushing of data buffers that have not been updated in a specified number of seconds
• libmseed updated to v3.1.8

Changed

• MS3RecordReader parameter source renamed to buffer for consistency

v0.0.2

Added

• List-like access to trace IDs in MS3TraceList (indexing, slicing, iteration)
• List-like access to segments in MS3TraceID (indexing, slicing, iteration)
• MS3Record.with_datasamples() is a context manager for setting sample buffer, type, counts
  allowing an MS3Record to be used for record packing with zero-copy of data
• MS3Record.from_file() and MS3Record.from_buffer() for convenience
• MS3TraceList.from_file() for consistency and future flexibility
• MS3TraceList.to_file() for writing miniSEED
• MS3TraceSeg.has_same_data() for comparison
• docstring documentation including examples
• Comprehensive repr() and summary str() methods

Changed

• Rename MS3TraceList.read_file() to MS3TraceList.add_file() for clarity

Removed

• MS3TraceList.numtraces in favor of supporting len() directly
• MS3TraceList.traceids() in favor of supporting iteration directly
• MS3TraceID.numsegments in favor of supporting len() directly
• MS3TraceID.segments() in favor of supporting iteration directly
• MS3TraceList.read_files() as unnecessary