Package 'nvdb2osmr'

Get NVDB Column Name Mappings

Description

Returns a named list mapping short ASCII column names (as they appear in GDB/GeoParquet files) to long descriptive Swedish names. This is useful for documentation and understanding what each column represents.

Usage

get_column_mappings()

Details

The column mappings are stored in inst/extdata/column_mappings.yaml and intentionally contain non-ASCII characters (Swedish åäö) for human reference. Package code should always use the short ASCII names.

Value

A named list where names are short column names and values are long descriptive Swedish names.

Examples

## Not run: 
mappings <- get_column_mappings()
mappings$Vagnr_10370  # "Driftbidrag statligt/Vägnr"

# Get all columns related to roads (väg)
mappings[grepl("vag", names(mappings), ignore.case = TRUE)]

## End(Not run)

---

Get Long Column Name

Description

Look up the descriptive (long) name for a given short ASCII column name.

Usage

get_long_name(short_name)

Arguments

short_name

Character string with the short column name (e.g., "Vagnr_10370")

Value

Character string with the long descriptive name, or the input if not found

Examples

## Not run: 
get_long_name("Vagnr_10370")  # "Driftbidrag statligt/Vägnr"
get_long_name("Klass_181")    # "Funktionell vägklass/Klass"

## End(Not run)

---

List Available NVDB Columns

Description

Returns a data frame with all available column mappings, useful for exploring the dataset structure.

Usage

list_columns(pattern = NULL)

Arguments

pattern

Optional regex pattern to filter column names

Value

A data frame with columns short_name and long_name

Examples

## Not run: 
# All columns
list_columns()

# Columns related to speed (hastighet)
list_columns("hastighet")

# Columns with "vag" in the name
list_columns("vag")

## End(Not run)

---

Convert NVDB data to OSM PBF using parallel processing (WKB optimized)

Description

Convert NVDB data to OSM PBF using parallel processing (WKB optimized)

Usage

nvdb_to_pbf(
  input_path,
  output_pbf,
  municipality_codes = NULL,
  county_codes = NULL,
  split_by = c("municipality", "county", "none"),
  use_geoparquet = "auto",
  global_node_prepass = c("auto", "on", "off"),
  simplify_method = "refname",
  presplit = FALSE,
  max_retries = 2,
  duckdb_memory_limit_gb = 4,
  duckdb_threads = 1
)

Arguments

input_path	Path to input file (.gdb, .gpkg, or .geoparquet)
output_pbf	Path to final output .osm.pbf
municipality_codes	Optional vector of 4-digit municipality codes to process (default: all)
county_codes	Optional vector of 2-digit county codes to process (e.g., "01" for Stockholm)
split_by	Strategy for splitting the work: "municipality" (default), "county", or "none" (process whole file in one go).
use_geoparquet	Use GeoParquet for faster processing: "auto", TRUE, or FALSE (default: "auto")
global_node_prepass	Whether to build a global endpoint-node dictionary before split processing. One of "auto" (default), "on", or "off". For split processing (split_by = "municipality" or "county"), "off" is not allowed and raises an error because split mode requires global node prepass for boundary-safe connectivity.
presplit	Logical: whether to pre-split to temp files (default: FALSE)
max_retries	Maximum retries for failed municipalities (default: 2)
duckdb_memory_limit_gb	Memory limit for DuckDB in GB (numeric). Default 4.
duckdb_threads	Number of threads for DuckDB. Default 1 (ideal for parallel runs).

Details

This function supports parallel processing via the mirai package. To run in parallel, you must set up mirai daemons before calling this function, for example using mirai::daemons(4). To shut down daemons after processing, call mirai::daemons(0). If no daemons are configured, processing will happen sequentially.

Splitting by "municipality" is recommended for high-core counts as it provides more granular tasks (~290 tasks). "county" provides ~21 tasks. "none" handles everything in a single process (memory intensive for large areas).

Value

Path to output PBF file (invisibly)

---

Fast NVDB to PBF conversion using ported Rust algorithm (WKB optimized)

Description

Fast NVDB to PBF conversion using ported Rust algorithm (WKB optimized)

Usage

process_nvdb_fast(
  gdb_path,
  output_pbf,
  municipality_code = NULL,
  county_code = NULL,
  simplify_method = "refname",
  node_id_start = 1L,
  way_id_start = 1L,
  duckdb_memory_limit_gb = 4,
  duckdb_threads = 1,
  verbose = TRUE,
  global_node_dict_path = NULL,
  area_code = NULL,
  prepass_rounding = "duckdb_1e7"
)

Arguments

gdb_path	Path to input file (GDB, GPKG, or GeoParquet)
output_pbf	Output PBF file path
municipality_code	4-digit municipality code to process (e.g., '2480')
county_code	2-digit county code to process (e.g., '24'). Used if municipality_code is NULL.
simplify_method	Simplification method (default: "refname")
node_id_start	Starting node ID for this chunk (default: 1)
way_id_start	Starting way ID for this chunk (default: 1)
duckdb_memory_limit_gb	Memory limit for DuckDB in GB (numeric). Default 4.
duckdb_threads	Number of threads for DuckDB. Default 1.
verbose	Print progress messages (default: TRUE)
global_node_dict_path	Optional path to global endpoint-node dictionary parquet. If provided, per-segment start/end global node IDs are joined into the chunk. Required when municipality_code or county_code is supplied.
area_code	Area code for this chunk (municipality or county code). Required when global_node_dict_path is provided.
prepass_rounding	Rounding scheme for global node dictionary matching. Currently only "duckdb_1e7" is supported.

Value

Path to output PBF file (invisibly)

---

Process NVDB data to OSM PBF (WKB optimized)

Description

Optimized function using WKB geometries and direct R property columns. This avoids JSON serialization overhead for significant speedup.

Usage

process_nvdb_wkb(
  wkb_geoms,
  col_names,
  col_data,
  output_path,
  simplify_method = "refname",
  node_id_start = 1L,
  way_id_start = 1L
)

Arguments

wkb_geoms	List of raw WKB byte vectors (one per geometry)
col_names	Character vector of property column names
col_data	List of vectors (one per column), each same length as wkb_geoms
output_path	Path to write the output .osm.pbf file
simplify_method	Simplification method: "refname" (default), "recursive", "linear", "route", or "segment"
node_id_start	Starting ID for nodes (default: 1)
way_id_start	Starting ID for ways (default: 1)

Value

TRUE on success

---