Changes in version 0.3.1 (2026-08-19)                  

  - Bugfix (CRAN policy compliance): gui_render_report() could attempt
    to write knitr's intermediate .knit.md file into the installed
    package library directory (the default location rmarkdown::render()
    uses for intermediate files, based on the input template's own
    location), rather than into a writable temporary directory. This is
    invisible on most systems since users typically have write access to
    their own package library, but violates CRAN policy against writing
    outside tempdir(), and fails outright on systems where the library
    is mounted read-only (as CRAN's Debian check systems now do). Fixed
    by explicitly setting intermediates_dir and knit_root_dir to a
    session-specific temporary directory; the function's actual output
    file location (supplied by the caller, e.g. Shiny's own temp path
    during a download, or a user-chosen path) is unaffected by this
    change. Verified by reproducing the exact failure as a non-root user
    against a read-only-permissioned package library, then confirming
    the fix resolves it under the same condition.

                 Changes in version 0.3.0 (2026-07-30)                  

  - New TGA Analysis tab, wrapping biocharkit's new TGA module (requires
    biocharkit >= 0.3.0):
      - Curve analysis: upload a full TGA curve, get a smoothed DTG
        curve with auto-detected decomposition peaks (matched against
        typical moisture/hemicellulose/cellulose/lignin temperature
        ranges), a proximate analysis (moisture/VM/ash/fixed carbon)
        read straight off the curve at adjustable temperature
        breakpoints, a TG/DTG plot, and CSV/600 dpi TIFF downloads.
      - Batch mode (same checkbox pattern as the Isotherms/Kinetics
        tabs): run the proximate analysis separately per sample ID
        across a long-format multi-sample data frame.
      - Kissinger kinetics sub-panel: upload a small heating-rate/peak-
        temperature table (one row per heating rate) and fit
        non-isothermal activation energy and pre-exponential factor.
      - New logic functions gui_tga_curve(), gui_tga_stages(),
        gui_tga_stages_batch(), gui_tga_kissinger() in R/logic.R,
        following the same column-mapping/validation pattern as the rest
        of the app.
      - Combined session report gained matching "TGA Analysis" and "TGA
        — Kissinger Kinetics" sections.

                        Changes in version 0.2.0                        

  - Isotherms tab: added Temkin, Dubinin-Radushkevich, and Sips models
    alongside Langmuir/Freundlich; added batch mode (fit every group in
    a sample-ID column separately, one row of results per group).
  - Kinetics tab: added Elovich and intraparticle diffusion models
    alongside pseudo-first/second-order; added batch mode.
  - New Thermodynamics tab (van't Hoff analysis).
  - New BET Surface Area tab.
  - New Proximate / Ultimate Analysis tab.
  - FTIR tab: added optional baseline correction (linear or
    rolling-minimum) and one-click automatic peak detection.
  - XRD tab: added an option to deconvolve a full 2theta/intensity
    pattern into crystalline peaks and a crystallinity index, instead of
    requiring pre-computed peak areas.
  - Results tables now use DT (sortable, searchable, paginated) instead
    of static tables.
  - New combined session report: every analysis run in the session is
    accumulated and downloadable as a single HTML report from the Home
    tab.
  - Internal: DT, grDevices, and rmarkdown usage moved into real,
    independently-tested R/ functions (gui_datatable(),
    gui_snapshot_png(), gui_render_report()) rather than being used only
    inside the Shiny app script.
  - Fixed an upstream bug in biocharkit::find_ftir_peaks() (see that
    package's NEWS) that affected the new automatic peak detection
    feature.

                        Changes in version 0.1.0                        

  - Initial release.
  - run_biocharkit_gui() launches a local Shiny app wrapping the
    biocharkit package.
  - Tabs: Sample ID Parser, Adsorption Capacity, Isotherms, Kinetics,
    FTIR (functional-group density and peak assignment), XRD
    Crystallinity, Correlation Matrix.
  - Excel import via gui_read_excel() / gui_list_sheets(), with
    interactive column mapping (no coding required) for every analysis.
  - Results downloadable as CSV; isotherm/kinetics/FTIR-spectrum plots
    downloadable as 600 dpi TIFF.