mzML export

The to_mzml example writes a minimal HUPO mzML 1.1 document covering the scans in a RAW file. It is also exposed as a method on the Python RawFile:

import opentfraw

raw = opentfraw.RawFile("run.raw")
raw.to_mzml("run.mzML")

The output is intentionally minimal: spectrum index, scan headers, m/z + intensity arrays (base64 + zlib), and the filter string. It validates against the PSI-MS mzML 1.1 schema and is suitable as a bridge into existing mzML-based pipelines. It does not aim to populate every optional controlled-vocabulary annotation the schema permits; it covers the core scan and peak data plus the vendor metadata OpenTFRaw decodes directly from the binary.

Each spectrum also carries the scan's acquisition event id and its opentfraw.* extra values (resolution, AGC, lock mass, instrument-status readings and so on; opentfraw.extra_field_keys() lists them) as <userParam> elements, and references an instrumentConfiguration for its mass analyzer. The extras add roughly 10-14% to the file on small-spectrum data. To choose them:

raw.to_mzml("run.mzML", extra_fields=[])  # none
raw.to_mzml("run.mzML", extra_fields=["opentfraw.resolution"])  # only these
raw.to_mzml("run.mzML", exclude_extra_fields=["opentfraw.sps_masses"])  # all but these

From Rust, pass ExtraFields to OpenTfRawSource::with_extra_fields.

If a scan's data cannot be decoded, the export fails with that error instead of leaving the scan out. To write the readable scans instead:

skipped = raw.to_mzml("run.mzML", skip_unreadable=True)

This emits a UserWarning with the count and returns it. From Rust, an OpenTfRawSource written with openmassspec_core::write_mzml_with_options and RecordErrorPolicy::Skip does the same and reports the count in the returned WriteSummary.

OpenTFRaw's goal is direct, open access to the RAW binary and a faithful open-standard mzML rendering of it.