v1.8.0

What's Changed

• Fix segfault that occurs when trying to add precedence to a productio… by @mingodad in #24
• Proposal for "noisy" keywords by @phorward in #28
• Integrate more ECMA6 features to JavaScript target by @phorward in #29
• Removing several warnings by @phorward in #30

New Contributors

• @mingodad made their first contribution in #24

Full Changelog: v1.7.0...v1.8.0

1.6.2

• v1.6.2: Provide ERRSTYLE_IMPORTANT-flag to show up important warnings

1.6.0

Released on: May 27, 2019

• Renamed top-level directive #!mode options to scannerless (default)
  and scanner.
• !language-directive allows for identifier as language parameter.
• Removed deprecated directives #parser, #description, #copyright
  and #version.
• Removed old-style !model top-level directive.
• Several internal code revisions.
• Compiled and linked against libphorward 0.24.

1.5.0

• New targets for JavaScript and JSON.
• Targets C and C++ improved to avoid memory leaking and handle malloc/realloc
  calls more securely.
• All targets now supporting the UNICC_SUCCESS and UNICC_ERROR flags that
  can be set as current act (action) to stop parsing.
• Started a grammar and templates test suite for better continous integration.
• Compiled and linked against libphorward 0.23.
• Removed outdated syntax-tree feature from C target.

1.4.0

• Compiled and linked against libphorward 0.22
• C++ target forked from the C target startet
• action_prefix value can be defined in target language templates to
  redefine the @-prefix for action code variables by any other desired
  prefix. This change has been done to make target language processing easier
  for languages where the @-sign is part of the target language syntax.

1.3.0

Released on: December 21, 2017

• Code generator does not emit action code when the action is empty.
• New command-line parameter "-l" or "--language" to specify a target language
  via command-line. This is useful when a grammar only contains AST definitions,
  which allows to render and run it in any target.
• Abstract syntax tree generation support:
  • Drafting an abstract syntax tree definition syntax for the grammar language,
    by using the operators "=" and ":=". See examples/expr.ast.par for an
    example working under both C and Python without any modifications.
  • C and Python parser targets extended to construct abstract syntax tree
    data structures dynamically.
  • These are first changes for an upcoming UniCC v2, where the targets shall
    be re-used with only few improvements.
• Fixed several bugs in the Python target on insensitive mode.
• Additionally allow ":" in case of "->" in grammar definitions.
• Internal code revisions started, but interrupted for now.
• Renamed folder "templates/" into "targets/".

1.2.0

• Started an (yet incomplete) new standard template to provide parser generation
  support for the Python programming language (templates/python.tlt).
• Imported the source code of the C standard template into the UniCC repository
  for further developments, the previous repository gets closed.
  (tempates/c.tlt)
• Fixed bug in the lexical analyzer generation by linking against
  libphorward 0.20 that caused a massive overhead of deterministic finite
  automation states in context-sensitive mode due an arbitrary character
  set order. This bug did not produce invalid parsers, but horribly huge tables.
• Imported the source code of min_lalr1 into the unicc repository, because it
  is the only place where it is used.
• Several internal revisions
  • Changed internal names for files, functions and variables.
  • Replaced old-style function headers by more convenient ones.
  • Removed unused code.
• Renewed manpage
• New README.md and updated CHANGELOG.md