parcr 0.6.1

Bug fix

• function str_match() did not emulate stringr::str_match exactly. The latter function only uses regular expressions.

• function parser_error_context() caused an error when input had only a few lines and the error occurred on the first few lines.

parcr 0.6.0

New feature

• Improved error messaging.
  • The line where the parser fails is now displayed together with its context
  • The error message mentions expected values. Some of these are automatically inferred, whereas others can be induced by creating named parsers

Bug fix

• The function stringparser() depended on stringr::str_match(). However, the dependency was not listed in DESCRIPTION. This may have led to errors when the stringr package was not installed. The function has been fixed by implementing it with base functions.

parcr 0.5.3

New feature

• Added a constructor function for string parsers using stringr::str_match() because the pattern occurs very often in the construction of string parsers.

parcr 0.5.2

Bug fix

• Parser zero_or_one() was not greedy, meaning that it sometimes did not parse input that it should be able to parse (see issue #12).

New feature

• Added the function Ignore() which reads and discards all elements until the end of a vector.

parcr 0.5.1

• Because of the use of base::isa() this package requires R 4.1 or higher, not R 2.1. This has been corrected in the DESCRIPTION. Running on ubuntu-latest and R 4.1 is checked by CI workflow on Github.
• Minor corrections and additions to both vignettes.

parcr 0.5.0

• All function descriptions and documentation thoroughly reviewed and updated.
• Vignettes reviewed and updated.
• Published on CRAN

parcr 0.4.5

• Extended the fasta file example in the vignette by showing the use of %or% to apply alternative parsers. Added a protein sequence fasta file to the data set.
• Function Parser() was changed and renamed to reporter(). This better reflects its use.

parcr 0.4.4

• Added (started) Details vignette.
• Custom error_parser includes line number and line content fields of the element index on which the parser failed.
• Added function finished() to test for complete consumption of the input by the parser.

parcr 0.4.3

• print.marker now emits ‘[]’ to make explicit that it differs from a simple list().
• Export of function failed() to test failure of a parser. Useful in unit testing.

parcr 0.4.2

• Resolved issue #5

parcr 0.4.1

• Correct example in documentation for Parser() function

parcr 0.4.0

• Marker error messaging implemented and tested. Markers are line numbers (element numbers in a character vector) where the parser fails. In case of alternative parsers like p1 %or% p2 the marker with the largest index value is reported when both parsers fail.

parcr 0.3.3.9000

• Developing marker error messaging

parcr 0.3.3

• Again modified exacly(n,p) because, since repeater functions call the same instance of a function, the counter is not reset when this parser is nested in a repeater parser (zero_or_more, etc.). The reset is now performed in the function itself.

parcr 0.3.2

• Added fastafile example data which is used in the vignette.

parcr 0.3.1

• Redesigned exactly(n,p) to behave as expected with non-emitting parsers like literal("A") %ret% NULL and eof().

• Dependencies on packages stringr and methods were removed.

parcr 0.3.0

• Added function eof() which detects an end of file. With that, the behavior of satisfy(b) and match_s(s) has been modified: they now fail on empty input, regardless of the functions b() and s().

parcr 0.2.5

• Initial Github submission.