Changes in version 1.0.0 (2026-07-23)                  

                        Changes in version 0.1.9                        

New features

  - read_qview_report() reads the flat report exports Q-View writes next
    to the binary container -- the ..._auto_report and
    ..._auto_all-parameters_report files, as either .csv or .xlsx -- and
    returns the same qview object read_qview() builds. Use it when only
    the exports were kept and the original .Q-View project file is
    unavailable. It differs from read_qview() in two deliberate ways:
      - it captures the plain "Reduced Concentration" point estimate
        (one row per sample) with statistic == "reduced"; and
      - it preserves out-of-range cells instead of dropping them: a
        "< 52.50" cell yields concentration = 52.50 with a new flag
        column set to "<" (">" for upper bound, "incalculable" for
        Incalculable), so limit-of-quantification information survives
        import. A small example-report.csv fixture ships in
        inst/extdata/.

Initial release. Pure-R parser for .Q-View binary project files
(chemiluminescent multiplex ELISA plate imaging and quantification). No
Java runtime, no H2 database driver, no compiled code.

Reliability

  - read_qview() now resolves the superseded MVCC page versions that the
    embedded H2 container retains. A version truncated at a 2048-byte
    page boundary (fewer analytes, or a number cut mid-digit) is no
    longer mistaken for the current reading: for each physical (well,
    replicate, analyte) the value occurring most often across the
    committed page copies wins, breaking ties toward the most complete
    source row. Per-well pixel intensities now match Q-View's own grid
    exports to 1e-6, and each well appears once in plate_layout.
  - read_qview_template() auto-detects the field separator, so
    semicolon-delimited (European-locale) and tab-delimited templates
    parse the same as comma-delimited ones.
  - plot(type = "replicate_scatter") no longer errors when a well-group
    label maps to more than one well; duplicate readings are averaged.

Reader

  - read_qview(path, strip_prefix = FALSE): parses a .Q-View container
    and returns a list of class qview with project metadata, the analyte
    panel (units, LOD / LLOQ / ULOQ, assay-control range), well-group
    sample assignments, per-well replicate pixel intensities, summary
    statistics, optional back-calculated concentrations, curve fits, and
    a plate layout (all tidy tibbles).
  - read_qview_template(): parses the companion well-assignment template
    CSV (NxM layout with Group Name / Group Type / Dilution Factor
    sections).

Helpers

  - strip_qview_prefix(): reverses the producer-side naming convention
    (ICal N -> Cal N, GLow -> Low, HHigh -> High, NFD... / N1234... ->
    original sample ID).
  - well_label(): vectorised plate-coordinate helper.

Methods

  - is_qview(): predicate for the S3 class.
  - print.qview(): compact one-screen summary.
  - summary.qview(): per-analyte mean / SD / CV / min / max grouped by
    well type, returned as a qview_summary tibble with its own print
    method.
  - plot.qview(type = ...): quick-look plate map, per-analyte intensity
    heatmap, and replicate-1-vs-2 scatter; viridis throughout.
  - as_tibble.qview(): long-format pixel-intensity tibble.

Export

  - write_qview_xlsx(), write_qview_csv(), write_qview_rds():
    pipe-friendly writers that return the parsed object invisibly.
    qview_to_xlsx() / qview_to_csv_dir() are kept as
    lifecycle::deprecate_warn() aliases for back-compatibility.

Interactive front-end

  - qview_app(): monochrome bslib Shiny app with built-in dark/light
    toggle, hex-sticker brand, large upload cap (default 512 MB),
    per-table xlsx download, and a publication-ready 2x2 Overview tab
    (plate layout / pixel-intensity distribution / replicate concordance
    / mean PI by well type) with high-DPI PNG and vector PDF export.