Changelog • yyjsonr

Fix outstanding rchk errors

Prepare for CRAN

Fix some compilation warnings.
Update documentation.
Update to yyjson v0.9.0

Bug fix for serializing NA in factors

Re-introduce GeoJSON support

Re-introduce NDJSON support
Add read_ndjson_str()

call normalizePath() on all file paths
clarify authorship and copyright

Increase maximum number of allowed data.frame columns (during parsing) to 2048

Read JSON from ‘.gz’ files in read_json_file()

New fast_numerics flag when writing.
- Default FALSE
- If TRUE the user is guaranteeing that there are no NA, NaN or Inf values in the numeric and integer vectors, and thus a faster method for writing these vectors to JSON can be used.
Changed writing of raw vectors to always use the fast_numerics method, as raw R vectors (by definition) will not have NA, NaN or Inf values.

Fixes for CRAN
- Adjust pointer arithmetic when calling output_verbose_error() to avoid overflow of size_t

Fixes for CRAN
- Fixed warnings when building with MAKEVARS PKG_CFLAGS = -Wconversion
- Fixed C error where raw buffer was passed to strlen() leading to a check failure on cran with clang-ASAN

Fixes for CRAN
- DESCRIPTION fix: Write C as 'C'
- Add link to original yyjson library
- Fix name in LICENSE

Fixes for CRAN
- Simplify example to remove checkRd NOTE
- Platform specific handling of error location format string to fix WARNING

Add int64 = "double" option to opts_read_json()
Preparations for CRAN

Remove NDJSON and GeoJSON code to simplify preparation for CRAN. Will re-introduce this code in future releases.
Updated to YYJSON v0.8.0

Fix an off-by-one error when reporting line numbers in NDJSON handling.
Increase buffer size when reading lines from NDJSON files.
- MAX_LINE_LENGTH now 131072 (was 10000)

Writing to JSON objects now supports a digits argument for rounding floating point values to the specified number of significant digits
- digits = -1 means don’t do any rounding
- digits = 0 rounds floating point values to integers (and writes the values as JSON integers)

Refactored options for simplification to data.frame
- removed vectors_to_df
- replaced with obj_of_arrs_to_df
- added arr_of_obs_to_df
length-1 vectors marked with class AsIs (using a call to I()) will never be unboxed i.e. will always be serialized as a JSON []-array with one element.
Added parse option length1_array_asis. If TRUE then automatically add the class AsIs to the object object.

Added pre-calculation and caching of data.frame column types for faster serialization when outputting row-by-row.

Added path.expand() when handling filenames. Thanks to
- https://github.com/shikokuchuo
- https://github.com/hrbrmstr
Added support for data.frames without column names to match behaviour of jsonlite (when working within the plotly package)

Unifed naming scheme

FEATURE: Added promote_num_to_string in opts_read_json() to enable forced promotion of numerics to string
BUGFIX: fixes for handling of geometry collection when reading and writing.
TESTING: More tests included for output to geojson
TESTING: Refactored testing of ‘sf’ objects

Bug fix for checking attributes on list columns. This surfaced more errors to do with setting XYZ geometry type.
More extensive testing of write_geojson_str()

Initial alpha GeoJSON support:
- read_geojson_str(), read_geojson_file()
  - Needs more testing for corner cases
- write_geojson_str(), write_geojson_file()
  - At a very alpha stage.

Added validate_json_file() and validate_json_str()

Added write_ndjson_file() and write_ndjson_str()

Update to option setting to remove overhead
Added read_json_raw() to parse JSON directly from a raw vector (which contains UTF-8 string data)

Initial release