Skip to contents

Validate what would be written where (pre-flight check)

Source: R/hdf5_read-write.R
h5_validate_write.Rd

Checks a named list of values against an open HDF5 file and summarizes, per dataset path, the current dimensions (with a fallback via h5$ls()), the length/column count of the supplied value, and a suggested handling (SCALAR/1D/ND or TS(2-col) for 2-column time series).

Usage

h5_validate_write(h5, values)

Arguments

h5

An open hdf5r::H5File (e.g., opened with mode = "r+").

values

A named list. Names are absolute dataset paths (starting with "/"); values are the R objects to be written (scalar, vector, matrix/array, or a 2-column data.frame/tibble for time series).

Value

A tibble with columns:

  • path – dataset path (after normalizing "//" → "/").

  • cur_dims – detected dims as "2x4" or "SCALAR".

  • val_len – length of the supplied value (for data frames: nrow).

  • df_cols – number of columns (for data frames/tibbles; otherwise NA).

  • decision – heuristic label: "SCALAR", "1D", "ND", or "TS(2-col)".

  • note – additional note (e.g., "not found").

Details

If get_simple_extent_dims() incorrectly reports a dataset as SCALAR, the function falls back to dimensions derived from list_h5_datasets(h5) (i.e., h5$ls(recursive = TRUE)). Names in values are normalized so that multiple leading slashes collapse to a single leading slash.

See also

list_h5_datasets, hdf5r, and a matching writer like h5_write_values().

Examples

if (FALSE) { # \dontrun{
library(hdf5r)
h5 <- H5File$new("input.h5", mode = "r+")

vals <- list(
  "/Parameters/OutputPath" = "C:/temp/out.h5",
  "/Rain/Hyetograph" = tibble::tibble(time = c(0, 10, 20, 30),
                                      value = c(0, 5, 12, 0)),
  "/Measures/.../LayerThickness" = c(150L, 150L)
)

h5_validate_write(h5, vals)
} # }