Changes in 0.10.0 and 0.10.0.9000
New functionality:
-
abbreviations: Abbreviations are now …
- … matched case-insensitive inside of groups of conneceted lower/upper case sequences.
- … consistently formatted according to the supplied case.
- … formatted exactly as supplied for the cases title, mixed, lower and upper camel.
- … protected from the parsing. This means that
- special characters in abbreviations don’t need to be taken care of anymore. The formatting of digits or special characters like hyphens, colons etc. will be formatted as specified.
-
parsing_option:
- new parsing option 3 parses “SOmeNIceSTUFf”.
- parsing options starting with a minus (-1, -2, -3) interprete non-alphanumeric characters as word boundaries. E.g. “This.text” will stay “This.text” in upper camel case.
-
cases
-
title: New case which should be especially useful for automatic generation of proper labels within graphics or business reports. Builds up on sentence case which is wrapped within
tools::toTitleCase().
-
random: New case, which will randomly convert letters into upper or lower case.
-
numerals: new alignment option
"tight" which allows to suppress all underscores between numerals and letters.
-
sep_out:
sep_out gets proper vectorization. In case of length(sep_out) > 1 differing substrings are connected by the (possibly) differing elements of sep_out. Therefore, sep_out gets accordingly shortened or the last element of sep_out gets recycled to match the number of needed separators for each element of string.
-
UTF8 Conversion: Input is now always converted to UTF-8 and returned as UTF8. Also the
transliterations argument is now aware of non-UTF8 encoded input.
Infrastructure:
-
CII best practices badge: Achieved the criteria for the best practices badge. Current status is under https://bestpractices.coreinfrastructure.org/de/projects/2475
-
R Version: Increase from 3.1 to 3.2 regarding the usage of
tools::toTitleCase() inside to_title_case().
-
Tests: Skip
to_any_case() tests (janitor-pkg-tests, transliterations and complex strings) when platform charset is not UTF-8 to resolve CRAN notification regarding character encoding.
-
Vignettes: The blog article “Introducing the snakecase package” was added as a vignette.
-
Documentation: The readme, the examples and the function documentation were updated according to the new functionality.
-
cases: Title case now builds up on sentence case (instead of parsed_case).
-
abbreviations: Abbreviations are now
- matched case-insensitive
- formatted as they are supplied for title-, upper-camel-, lower-camel- and mixed case. Apart of that abbreviations are still formatted according to the specified case.
- protected from the parsing. This means that
- special characters in abbreviations don’t need to be excluded via a regex in the
sep_in parameter
- the formatting of digits or special characters like hyphens, colons etc. will be formatted as specified.
-
parsing_option:
- new parsing option 3 implemented which parses “SOmeNIceSTUFf”.
- each parsing option can now be prefixed by a minus (-1, -2, -3). In this way
-
numerals: new alignment option
"tight" which allows to suppress all underscores between numerals and letters.
-
improved speed: time of internal parsing could be reduced about a factor of ten, due to some
vapply() replacements.
-
UTF8 Conversion: Input is now always converted to UTF-8 and returned as UTF8. Also the
transliterations argument is now aware of non-UTF8 encoded input.
-
Improved vecorisation of
sep_out: sep_out now behaves better for vector input (length > 1). Instead of returning different strings, the return is now one string, which uses separators according to the supplied order. When length(sep_out) > 1, the last element of sep_out gets recycled.
-
random_case: added
to_random_case(), which will randomly convert letters into upper or lower case.
-
title_case: added
to_title_case(), which is basically to_parsed_case() with sep_out = " " wrapped within tools::toTitleCase() and should be especially useful for proper labels within graphics or business reports.
- increase R Version dependency from 3.1 up to 3.2 regarding the usage of
tools::toTitleCase().
- skip
to_any_case() tests (janitor-pkg-tests, transliterations and complex strings) when platform charset is not UTF-8.
-
cases: added
to_sentence_case() (same as snake, but first letter is uppercase and default sep_out is space).
-
numerals: added
numerals argument to all caseconverters including to_any_case() to format the alignment of digits (middle, left,right, asis). Therefore parsing_option nr 4 might be removed in later releases, as it is the same as parsing_option = 1 and numerals = "asis".
-
transliterations: When named character elements are supplied as part of
transliterations, anything that mathches the name is replaced by the corresponding value.
- attributes are now preserved (not only names as before)
- CRAN release
- Change default
sep_in from NULL to "[^[:alnum:]]". This will make it easier for beginners and in general also faster to modify cases from names like names(iris). Updated the regarding sections in the vignette and readme.
- abbreviations work now also in conversions to lower- and upper camel case.
- replaced
parsing_options 3 and 4 with 5 and 6.
- removed purrr dependency
- removed magrittr dependency
- remove
replace_special_characters, preprocess and postprocess.
- added parsing option 6, which doesn’t surround digits with separators.
- some breaking changes:
- reordering of the arguments of all
to_xxx_case() functions
- renaming
preprocess to sep_in, postprocess to sep_out, replace_special_characters to transliterations.
- changes since last CRAN submission include:
- to_xxx_case shortcuts are now exact wrappers around to_any_case
-
process is deprecated after changing implementation and setting a reasonable default.
- added
abbreviations argument to any_case()
- case none is now a lot more general for formatting
- added
abbreviation specific behaviour for mixed case
- new parsing_option 5, which suppresses conversion after ., @, etc
- renaming of:
- to_small / to_big_camel_case have been renamed to to_lower / to_upper_camel_case. The old names are and will still be supported in to_any_case
-
parsingoption to parsing_option
- introduced rule that parsing_option <= 0 suppresses parsing from now on
- lots of additional tests and smaller bugfixes
- several documentation updates including help, examples, readme and vignette
- all to_xxx_case functions are now exact wrappers of to_any_case
- to_small / to_big_camel_case have been renamed to to_lower / to_upper_camel_case
- minimal vignette update
- more consistency for case none
- bugfix for parsing option 5
- overhaul readme
- renamed
parsingoption argument to parsing_option
-
process argument: changed implementation in to_any_case, set a reasonable default, implemented the behaviour also in to_xxx shortcut functions, deprecated process argument
- make modifications to case none, which allows now more parsing options
- added special behaviour for abbreviations within “mixed_case”
- added abbreviations argument for better mixed case handling
-
improve consistency with stringr pkg regarding special input handling
if(identical(stringr::str_length(string), integer())){return(character())}
- Changes since last update:
-
to_any_case() and shortcuts (to_xxx_case() functions) preserve name attribute now
- R dependency lowered to 3.1
CRAN update
- Changes since last CRAN submission include:
- string transliteration via updated
replace_special_character argument
- some new cases: “none”, “mixed”, “upper_lower”, “lower_upper”
- aliases: “all_caps”, “lower_camel”, “upper_camel”
- different parsing options
- several bugfixes
- improved internal testing
- internal modularisation
-
case == "none" works now with preprocess.
- fixed bug so that
case = upper_lower no also works for NA’s.
Implemented two further parsing options:
- 3: parses the first UPPER letter group like parsing option 2 and the rest like option 1
- 4: parses the first UPPERlowercase letter group like parsingoption 1 and the rest like option 2
bug fix in dev version: protect works now for beginning and end of substrings and not anymore only for complete substrings.
Input of replace special character is now a character vector. It’s entries have to be elements of stringi::stri_trans_list() or names of the transliteration dictionary list, which comes with this package. This update enables users to transliterate strings from different encodings, in a flexible way. For example UTF8 to Ascii, … .
- provided aliases
"all_caps", "lower_camel" and "upper_camel" for "screaming_snake", "small_camel" and "big_camel".
- small bug fix for upper_lower/lower_upper regarding the behaviour for the combination of preprocess and protect (similar to an earlier bug in the camel cases).
- small bug fix for behaviour of upper_lower/lower_upper. Now substrings with without alphabetical characters are ignored.
- new implemented cases: “none”, “mixed”, “upper_lower”, “lower_upper”.
- empty_fill runs now at the beginning of to any case function (after the first parsing) and a second parsing is introduced in case anything is filled.
- fixed bug in camelcases, for letters after protected symbols (usually one wouldn’t protect in these cases anyway…)
- implemented
check_design_rule() (not exported)
- resolved bugs in other design options (also deleted one which is not valid for screaming snake case)
- included several parsingoptions for
to_any_case().
1: “RRRStudio” -> “RRR_Studio” stays as default
2: “RRRStudio -> RRRS_tudio”
any other number will suppress the parsing (only spaces will be converted into "_")
- This is the (stable) development version now. You can find snakecase on cran here
- fixed bug for combination of protect and postprocess. Therefore clarified and rewrote the internal order of
to_any_case()
- resolved all internal build warnings and notes
- fixed bug in replace_special_characters combined with screaming snake case.
- any local special characters are now supported.
- added 2 arguments to
to_any_case(): empty_fill, which allows to fill strings matching to "" with the supplied string. unique_sep adds an integer suffix separated with the supplied string, when not NULL.
- fixed a bug in to_snake_case_internal(). groups of digits are not separated in between anymore.
- digits that are not direct next to each other, will be split via "_" in both camel case versions. This is a meaningful exception. Otherwise information would be lost and also the consistency rules in the readme wouldn’t hold in this case.
- any_case() is now vectorised over pre-/postprocess, pre-/postfix for all case arguments
- introduces protect as (vectorised) argument to to_any_case(). it accepts regular expressions and cleans "_" or whatever the preprocessing did, between matches.
- some more tests, documentation and fixes.
- introduced case = “parsed” in to_any_case()
- introduced to_parsed_case()
- finished vignette (title: Caseconverters) and replaced the usage part in readme with it.
- changed behaviour in snake_case_internal to always treat whitespaces as underscores. for whitespaces in output postprocess = " " is recommended.
- started testing
to_any_case and removed a bug.
- added vignette
- supports behaviour for german umlauts on all platforms
- introduced internal function
to_snake_case_internal() which does the preprocessing and simplifies all other functions (especially to_any_case()) a little bit.
- introduced
to_screaming_snake_case()
- added arguments prefix, postfix and replace_special_characters to
to_any_case().
- completely renewed readme
- updated tests and highly modularized all tests. (just to_any_case lacks some tests now and in general more examples testcases have to be written)
- introduced
to_any_case() (all functionality and documentation provided, but implementation has to be cleaned and also tests have to be written)
- fixed bug:
to_snake_case_dev(c("ssRRss")) returns now ss_r_rss instead of"ss_r_r_ss"
-
to_snake_case() now treats only spaces like underscores now (not dots anymore)
- functionality to treat different stuff like dots will be added via a new function:
to_any_case(), which will wrap the other three and will have pre- and postprocess arguments
- renamed the single input parameters consistent to
string (without deprecating the old name before, since the package was in early dev-stage anyway).
- started a to develop and implement consistent logic (which still has to be better documented in the readme)
- introduced tests for more hard coded examples and the logic behind it (still more hardcoded examples and a third part of the logic have to be tested)
- internal logic has been simplified and modularised a lot, which makes it easier to maintain and introduce more high-level features in the future
- added integrated tests via AppVeyor on windows
- added badges for cran status and code coverage to readme
- Introduced
to_snake_case() which converts strings to snake_case.
- added functions
to_small_camel_case() and to_big_camel_case() which internally build up on to_snake_case() and convert to the appropriate target case.
- added basic hard coded tests
- added integrated tests on linux via travis.ci