NEWS.md

tinycodet 0.5.3

• Behaviour change: the atomic typecasting functions now preserve names, dimensions, and dimnames (instead of all attributes), to be more in line with most of base 'R'.
• Behaviour change: pkgs %installed in% lib.loc will now return NA for "packages" that are part of core 'R' (i.e. 'base', 'stats', etc.).
• Behaviour change: All decimal truth testing operators now always give NA when Inf is compared with Inf, or when -Inf is compared to -Inf.
• Streamlined the internal code of the decimal truth testing operators.
• Added %s><% and %s<>% as aliases for %sget% and %strim%, respectively.
• Removed some superfluous text in the import system documentation.
• Re-written one 'C++' script to pure 'C' code.
• Some of the internal 'C' code now support long vectors, when appropriate.
• Small speed improvement in some of the internal code.
• Added new fake packages to the special tests, to perform more thorough tests on the pversion_ - functions.
• Added even more tests.

tinycodet 0.5.0

• Feature Improvement: help.import() now directly evaluates the arguments under help() if both arguments i and alias are missing.
• Feature Improvement: re_exports in import_as() can now include functions from core R if necessary, except 'base'.
• Performance Improvement: Replaced some of the internal code in the import system with 'C++' code via 'Rcpp' for some performance improvement.
• Bug fix: Fixed a (small) bug, where help.import() sometimes gave an unnecessary error, when searching topics of un-exported objects or non-functions via an alias object instead of searching functions directly.
• Bug fix: Fixed a (small) bug where class names were sometimes inconsistently assigned to functions and infix operators exposed by the 'tinycodet' import system.
• Removed Feature: Removed the form() function, as its use-case is too rare to justify having a whole function for it (with all the tests and maintenance that come with it). I rarely remove functions; this is an exception.
• Fixed some mistakes in the documentation.
• Added more tests.
• help.import() can now also be called without any arguments, just like the original help() function.
• Moved the atomic type casting functions to the "DRY" category.
• Cleaned up the internal code of the import system a bit.
• Added the current year to the LICENSE file.

tinycodet 0.4.6

• Feature improvement: Relaxed the length restriction on argument i in stri_locate_ith(). Also improved the performance of stri_locate_ith().
• Tweaked the error messages of strfind()<- to be a bit more informative.
• Added as_cplx() and as_raw() to the atomic typecasting functions.
• Improved the documentation here and there.
• Tweaked description text.
• Added more tests.

tinycodet 0.4.5

• Slight speed improvement of stri_locate_ith().
• New Features: Added the following functions to the "safer functionality" category: form(), aes_pro(), with_pro(), and safer_partialmatch(). Naturally, added tests for these new functions.

tinycodet 0.4.1

• Added an explanation in the string overview help page, regarding the usage of vector recycling.
• Provided additional clarification in the help file for the %row% and %col~% operators.
• Simplified the internal code of the decimal (in)equality testing operators.
• Added a few more tests.
• Moved is_wholenumber() to the decimal truth testing section.
• Argument change: Added the rt argument to strfind()<-, and moved the i and rt arguments more to the end of the functions. The type argument in strcut_brk() can now also directly accept a list produced by stringi::stri_opts_brkiter().

tinycodet 0.4.0

• Clarified in the documentation that %col~% and %row~% strip attributes.
• Clarified the usage of merge = FALSE for pattern charclass, in the documentation of the stri_locate_ith() function.
• Clarified in the documentation that locked objects are not protected from modification by reference.
• Clarified in the documentation of 'stringi' pattern searching infix operators that the p argument can also be a character vector of length 1.
• Clarified in the stri_locate_ith() documentation that one should not pass the capture_groups argument; also clarified how to capture the ith group using stri_locate_ith(). And clarified that stri_locate_ith() does not support long vectors.
• Normalized the links in the "See Also" sections in the documentation.
• Renamed the "str_truth" page to "str_search", as that makes a little more sense.
• Removed the redundant stringi:: piece in the example code for import_LL().
• The help file for the s_pattern functions is now actually titled "s_pattern".
• The messages returned by import_as() when aliasing packages is now slightly less verbose: removed the line "Methods work like normally", and replaced the line "Importing packages ..." with "Importing packages and registering methods...".
• The strcut_brk() function now includes the tolist argument to return a list. Moreover, the n argument may now also be specified (n = -1L by default).
• More safety checks: Improved safety against malformed conditions in the transform_if() function. stri_locate_ith() now gives a warning when an empty string or pattern is given. The help.import() function now gives an error if neither topic/package nor i/alias is supplied, instead of just silently doing nothing. The %sget% and %strim% operators now give an explicit error message if the arguments do not have proper lengths.
• Internal Re-write: stri_locate_ith() has now been partially re-written, and now includes 'C++' code, making it a bit faster. Also cleaned up the internal code for some other functions here and there.
• Added dependency: 'Rcpp' is now added as a dependency, due to the aforementioned partial re-write in 'C++'.
• New Features: Added %ss% to the collection of string arithmetic operators, and added the strfind()<- method to the string search operators. One can now supply the at argument in the list of the right-hand side for the %s{}% operator. Supplying at = "start" will check if the pattern appears at the start of a string. Supplying at = "end" will check if the pattern appears at the end of a string.
• Optimization: Managed to optimise stri_locate_ith() even further. Optimised the 'subset_if' operators a bit more. Also optimised the %n&%, %sget%, %strim%, and %=strtype% operators a bit more.
• Argument change: The s_pattern functions now wrap additional arguments in a list for the user, in case of using vector arguments, preventing potential confusion.
• Argument change: The type argument in stri_locate_ith_boundaries() is now not a mandatory argument; one can now also supply it through stri_opts_brikiter(), to keep it more consistent with the rest of 'stringi'.
• Tests: Improved the tests for the s_pattern operators. Added tests for stri_locate_ith() when using argument merge = FALSE. Added more error check tests for help.import(). Added tests for unequal vector sizes for the logic operators. Added tests for empty condition subsets for transform_if(). Added tests for the new functionalities introduced in this version. Added tests for malformed 'stringi' pattern searches in all relevant functions.
• Bug fix: There was a small bug where s_coll() worked properly when assigned to an object (as usual), but not when called directly (like nested inside a function). This is now fixed.

tinycodet 0.3.0

• Expanded the import system documentation. Also tweaked the documentation of import_as() to be make it slightly clearer.
• Expanded the documentation of the decimal_truth operators.
• import_as() now does not allow more than 10 packages to be loaded under a single alias, to prevent abusive usage of import_as().
• The import system now has a few more tiny safety checks, including checks for widely known meta-packages (such as "fastverse").
• The pkg_get_deps() function now also has the shared_tidy argument to ignore the shared 'tidyverse' libraries ('rlang', 'lifecycle', 'cli', 'glue', and 'withr'). Also changed the default values of the recom, and rstudioapi arguments.
• Added the pkg_get_deps_minimal() function. Also changed the default arguments of pkg_get_deps().
• Changed the error message for wrong extension specifications: replaced "actual reverse-dependencies" into "actual extensions".
• Changed the usage of the word "load" to "import" in the import system documentation, to avoid confusion with the concept of loading a package.
• Added the pversion_ functions to help in checking package versions.
• Removed the "versions" attribute from alias objects, in favour of the aforementioned pversion_ functions.
• Argument name change: in the import_as() function, changed argument name loadorder to import_order, to avoid confusion with the concept of loading a package.
• Tests: Added more tests.
• Optimization: Minor optimizations for the import_* functions and for transform_if().
• Bug fix: Fixed a bug in pkg_get_deps() where it didn't properly ignore all the recommended R-packages when recom = FALSE.

tinycodet 0.2.2

• Small tweak to the pdf manual.
• Replaced the incorrect uses of the word "class" with "type" in the atomic type casting help file.
• Improved the performance of stri_join_mat(), and produced better examples in both help file and website.
• Fixed incorrect regex usage in one of the examples.

tinycodet 0.2

• Removed or changed "a few functions" in the title, description, introduction help page, and website, as 'tinycodet' actually turned out to be somewhat larger than anticipated.
• Fixed the latex issue in the pdf file generation for users using R version < 4.2.0 with the Mac OS.
• Changed the @name and @rdname parameters of the tinycodet overview help files, such that they get on the top of the pdf file.
• Removed the sentence "Note that only the recommended R-packages actually installed in your system are taken into consideration" in the pkgs help file, as this is no longer applicable (the list of recommended R packages is hard coded).
• Expanded the 'tinycodet_import' help page with more info.

tinycodet 0.1.0.6

• Replaced usage of installed.packages() in the %installed in% operator with find.package(). Also added the sentence "As pkgs %installed in% lib.loc does not even load a package, the user can safely use it without fearing any unwanted side-effects." to the help page.
• Replaced usage of installed.packages() with a manually specified character vector in the R scripts/functions "internal_functions.R", "import_misc.R", "import_inops.R", "x.import.R".
• Replaced "\dontrun" with "\donttest" in help page "tinycodet_import", and set condition for examples.
• Re-implemented the fake packages, but this time preventing false positives using a tip from Duncan Murdoch.

tinycodet 0.1.0.5

• The fake packages will now only be tested on my GitHub page, and not in the package folder itself, to prevent false positives from CRAN checks.

tinycodet 0.1.0.4

• Removed the last sentence in the Description of the DESCRIPTION file, which contained the word "vectorized", because both "vectorised" AND "vectorized" cause CRAN checks to complain.

tinycodet 0.1.0.3

• Changed all the spellings of the word "vectorised", to "vectorized", despite the clear Language field.

tinycodet 0.1.0.2

• Corrected the 'Authors@R' field in the DESCRIPTION file.
• Added 'Language: en-gb' to the DESCRIPTION file.
• Removed the "Maintainer" field, as it can be auto-generated from 'Authors@R'.

tinycodet 0.1.0.1

• Corrected the LICENSE file.

tinycodet 0.1

• Initial release.