v0.1.6 — export formats fix (#226)

#226 — Correct spectrum export formats

Previous release (v0.1.5, #217) mislabeled the export formats:

• claimed LSRM .spe but wrote IAEA ASCII ($SPEC_ID: / $DATA: blocks)
• claimed BecqMoni .tka but .tka is a plain counts-per-line file, not BecqMoni's XML

Fixed to the actual reference formats (verified against SpectraVibe writers):

• BecqMoni → XML (.xml, ResultDataFile FormatVersion 120920,
  EnergySpectrum/EnergyCalibration/{PolynomialOrder,Coefficients/Coefficient…},
  Spectrum/DataPoint…). Identity calibration [0,1] fallback if the sample
  has no calibration (BecqMoni requires ≥2 coefficients).
• LSRM SpectraLine → binary .spe (CP-1251 header
  KEY=VALUE\r\n block with SHIFR/NOMER/CONFIGNAME/MEASBEGIN/TLIVE/TREAL/ENERGY/SPECTRSIZE,
  literal SPECTR= marker, then uint32-LE binary counts block).
• InterSpec/ANSI N42.42 → unchanged (was already correct XML).

File dialog filter:
BecqMoni XML (*.xml);;LSRM (*.spe);;InterSpec/ANSI N42 (*.n42).

Verification

Full pytest: 761 passed (+2 new format-verification tests, −1 removed TKA-writer test).

Bundle

waterfall-viewer-0.1.6-win64.zip — PyInstaller onedir, unzip anywhere and run
waterfall-viewer\waterfall-viewer.exe.