CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file.

[0.11.0] - 2026-02-27

🚀 Features

• (pc) Added peek_char
• (pc) Added one_char and one text parsers
• (pc) Added many_str
• (pc) Added specific_str, specific_str_ignoring
• (pc) Added FnCtxParser
• (pc) Introducing ManyCtxParser
• (pc) Replaced FnCtxParser with IifParser
• (pc) Introducing MapToUnitParser
• (pc) Introducing err_supplier
• (pc) Support different context type for the inner parser of flatten
• (Makefile) Repair executable permission issues on Samba

🐛 Bug Fixes

• (pc) Implement SetContext for ThenWith
• (ci) Fix build workflow for PRs (#119)
• (ci) Fix pipeline
• (ci) Revert Clippy from Windows pipeline for now
• (pc) Ensure IifCtxParser context is initialized

🚜 Refactor

• (basic) Improve performance of IndexedMap
• (pc) Rename FlatMap to AndThen
• (pc) Removed FlatMapOkNone in favor of AndThenErr
• (parser) Dropped SpecificStr for simplicity
• (pc) Replaced any_char and peek_char with read_p and peek_p
• (pc) Replaced one_char and with one_p
• (pc) Removed FilterPredicate
• (parser) Parse cr and crlf with the same parser
• (parser) Re-introducing comparison operator token types
• (pc) Remove specific_str and specific_str_ignoring
• (parser) Move keyword_ignoring to keyword module
• (parser) Keyword parser should not handle EOF
• (parser) Remove explicit parsers outside pc package (#117)
• (pc) Merge SetContext trait into Parser
• Add dedicated whitespace module (#120)
• (parser) Improve expression parsers regarding whitespace (#121)
• (pc) Pass context by reference
• (pc) Removed ThenWithLeft parser
• (pc) Removed init_context parser
• (parser) Use IifCtxParser in ParamName
• (pc) Introducing MapDecorator (#122)
• Hide Variant type from rusty_parser
• (linter) Refactor types module into smaller modules
• (linter) Moved variable_info to linter (#123)

⚙️ Miscellaneous Tasks

• Align module versions
• (linter) Fix some clippy warnings
• (basic) Fixing some clippy warnings
• Update gitignore
• (basic) Fixed a few clippy issues
• (pc) Correct generic parameter name for MapErrParser
• (pc) Removed unused inc_position methods of InputTrait
• Updated gitignore
• (ci) Fixing clippy warnings and add clippy to CI
• (pc) Removed and_opt parser, adding one_of_p
• Use clippy in Makefile
• (parser) Break apart expression module to smaller files
• (pc) Move no_context to its own module
• (pc) Moved IifParser to its own module and renamed to IifCtxParser

[0.10.0] - 2026-01-22

🚀 Features

• (pc) Implement no_incomplete directly into and parser
• (Makefile) Print the size of the types
• (parser) Implement no_incomplete directly into WithExpectedMessage parser
• (Makefile) Print longest type length
• (pc) Introducing context for parsers
• (pc) Add flatten parser
• (pc) Support contextual parsers
• (pc) Added PeekParser
• (pc) Support custom error in FilterParser
• (parser) Implement specific string parser
• (parser) Moved the dollar sign check for keywords into the any_token parser
• (pc) Combine char with Vec<char> into a String
• (pc) Added Token.text() method to return the ownership of the String
• (parser) Make identifier token include dots
• (pc) Added map_fatal_err, map_non_fatal_err and flat_map_none
• (parser) Support treating EOF as fatal in keyword parser
• (pc) Introducing init_context parser
• (pc) Introducing then_with_left parser
• (pc) Support many without closures
• (pc) Added surround parser
• (parser) Introducing whitespace_ignoring, parse whitespace but ignore the result
• (parser) Support parsing specific string and ignoring the output
• (parser) Support parsing a specific keyword and ignoring the output
• (pc) Start support for text (char and string) parsing
• (pc) Support error messages in filter_or_err
• (pc) Changed the filter predicate into its own trait, embedding error management in it
• (pc) Add is_soft method to error trait

🐛 Bug Fixes

• (parser) Comment separator must start with EOL
• (parser) QBasic names cannot contain underscores
• (parser) comment_separator should read any number of extra EOL or whitespace tokens
• (pc) Fix broken build for release mode

🚜 Refactor

• (pc) Implement or_default directly on many parser
• (pc) Merge the types NoIncomplete and OrFail into one parser, MapErr
• (parser) Remove specific module
• (pc) Re-introduce the OrDefault trait
• (pc) Remove all extra methods from Parser trait
• (pc) Improve macros for parsers
• (pc) Support more complex parsers in macros
• (pc) Merge parser1_decl and parser1_impl into one macro parser
• (pc) Re-working macro support
• Remove the lazy_parser macro and add a lazy function
• (parser) Add dedicated struct for any_token_of parser
• (parser) Support padded whitespace in any_token_of
• (parser) Support negative maches in any_token_of
• (pc) Split set_context into a separate trait
• (pc) Make Seq2, Seq3, etc structs private
• (parser) Move token-level logic to a new tokens module
• (parser) Move detect_eof to global_statement
• (parser) Implement the Parser trait directly on the char type
• (parser) Add or_expected as a shortcut for or_syntax_error("Expected: ")
• (parser) Remove single-char token types
• (pc) Improve code readability for single char Token
• (parser) Add AnySymbolParser
• (parser) Strengthen module visibility and rename modules
• (parser) Create dedicated struct parser for keyword
• (parser) Removed keyword_choice and improved keyword_map
• (parser) Improvements to statements parser
• (pc) Align then_with to use RefCell just like then_with_right
• (pc) Implement delimited_by and delimited_by_allow_missing
• (pc) Remove OptZip parser
• (pc) Introduce ParserErrorTrait trait
• (parser) Remove the explicit fatal designation on ParserError
• (pc) Switch to a mutable input
• (common) Deleted CaseInsensitiveStr for simplicity

📚 Documentation

• (pc) Improve Token documentation

⚡ Performance

• (parser) Improve TokenType lookup performance

⚙️ Miscellaneous Tasks

• (Makefile) Show only parser types in print-type-size
• (parser) Add _ws suffix to token_kind_parser functions that are surrounded by optional whitespace
• (parser) Move char_parsers to top-level module
• (parser) Move string_parsers to top-level module
• (parser) Use AnyChar struct directly
• (parser) Use filter_or_err
• (pc) Do not use macro in many
• (pc) Removed TokenList and token_list_to_string
• (pc) Removed unused Clone trait from Token
• (pc) Removed macros
• (parser) Reuse eol_ws_zero_or_more function
• (parser) Simplify implementation of any_token_ws in favor of padded_by_ws
• (pc) Change parse to mut
• (pc) Fix various clippy warnings
• (pc) Strengthen Token invariants
• (parser) Fix some clippy issues
• (pc) Move helper methods into the main Parser trait
• (pc) Make private modules public for better visibility in the docs

◀️ Revert

• Revert the specialized no_incomplete and or_default implementations due to code duplication

[0.9.1] - 2025-12-30

🐛 Bug Fixes

• (ci) Do not run the build workflow for tags
• (ci) Correct artifact paths for release workflow

[0.9.0] - 2025-12-30

🚀 Features

• Add RowColView to map a string to row/col taking newlines into account
• Added RcStringView as the basic input unit that is lightweight and clone-able, and suppors row-col information
• Added Opt parser
• Added surround parser
• Added flat_map_negate_none
• Decouple Parser from ParseError
• Move no_incomplete to the Parser level

🐛 Bug Fixes

• Linter was reporting a valid variable as illegal
• Incorrect usage of features
• Move some code to test cfg
• Fixed build in release mode
• Incorrect upper bound for max length of string
• AndThenOkErr should never re-map errors, only okay and incomplete results
• Use syntax error when entire program can't be parsed
• Use Box dyn for to_option and or_default to fix the build on Mac
• Auto-fixed some issues with clippy
• seq parsers should convert errors to fatal
• Removed macro_export from visitor macros
• Remove the build warning about the resolver

🚜 Refactor

• Split and_pc.rs into smaller files
• Removed _pc suffix from modules
• Reworked csv implementation, added loop_while and opt_zip traits
• Removed unused traits
• Various refactorings (#113)
• Do not use expression to parse param_name.rs
• Unify parsing of dim_name and param_name
• Linter improvements
• Replaced methods of TypeQualifier with new traits IntoTypeQualifier and IntoQualified
• Dropped usage of QualifiedName in Name
• Promoted pre_linter.rs to directory
• Using inner mutability in pre_linter
• Adding convertible.rs for converting function/sub parameters in pre-linter
• Removed lifeline from converter.rs objects for simplicity
• Simplified implicits in linter
• Introducing Stateful trait
• Removed SameTypeConverter trait
• Removed SameTypeConverterInContext trait and ConverterImpl
• Reverted Stateful implementation
• Implemented child state for expression conversion
• Using a different child state for ExpressionNode and Expression
• Promoted files to directory modules
• Reworked dim_rules, removed conversion dance from param_type to dim_type
• Removed custom PartialEq implementation of ParamType
• Trying a structure where small types and traits live in modules named types and traits respectively
• Implemented Convertible for ProgramNode and TopLevelToken
• Remove Rc from Context
• Introducing ResolvedParamType
• Dropped position from the name of UserDefinedType
• Simplify pre-linter
• Drop inner mutability for interpreter Input trait
• Removed all usages of Rc and RefCell
• Split collecting phase of LabelLinter into a LabelCollector
• Dropped top level built_ins module
• Using Cargo workspaces
• Decided on public fields for Locatable and VarName, dropped AsRef implementation and accessor methods
• Removed CaseInsensitiveString's Add operations and StringUtils pad_left
• Added rusty_variant crate
• Moved file_constants.rs to rusty_parser
• Introducing CaseInsensitiveStr
• Moved string_utils.rs to interpreter
• Moved to_str_unchecked to interpreter
• Rename Location and Node
• Separate error type per crate
• Moved error_envelope to interpreter
• Moved indexed_map to interpreter
• Refactor Parser trait to generic, so that it can support dyn
• Use dyn for OrParser
• Use new parsers to build Tokenizer
• Use parser combinator in comment.rs
• Use parser combinator for parsing FileHandle
• Use parser combinator for parsing the main program
• Use parser combinator in statement_separator
• Use parser combinator in keyword_choice and keyword_map
• Use parser combinator in select_case
• Remove negate parser
• Remove peek parser
• Remove token from unread method as it is no longer used
• Removed ParserOnce trait
• Removed NonOptParser marker trait
• Introduce ParseResult for old parsers instead of standard Result
• Introducing ParseResult::None enum member
• Use ParseResult in AndThen mapping functions
• Use ParseResult in AndThenOkErr mapping functions
• Reduce usages of ParseResult::Incomplete
• Replaced MapIncompleteError with WithExpectedMessage parser
• Removed ParseError::Incomplete
• Removed incomplete errors
• Port enum member to new
• Removed method
• Added combiner function to AddWithoutUndo
• Remove GuardPC
• Remove concrete KeepLeft and KeepRight structs
• Renamed AndThen to FlatMap
• Deleted AllowDefault and AllowNone
• Expose the combiner function in allow_opt
• Expose the combiner function in and
• Remove the PC suffix from parsers
• Divide Parser into smaller traits
• Redesigned pc_ng to use the standard Result type and not return the parsed input in the Err
• Use the new ParseResult everywhere
• Use Box dyn for or and seq
• Moved surround method to And trait
• Moved loop_while to its own trait LoopWhile
• Moved error related parsers to
• Move logging method to its own trait
• Move allow_none_if method to its own trait
• Moved one_or_more and zero_or_more to Many trait
• Moved method to its own trait
• Removed deprecated parser in favor of
• Removed deprecated in favor of
• Renamed rusty_parser/types to rusty_parser/specific
• Move Implicits and Names under a new module
• Added new type CompactsInfo
• Added wrapper struct NamesInner
• Delegating implementation to new structs in names module
• Expose traits of names outside module
• Store names of all scopes in Names without dropping them
• Expose linter context to interpreter
• Restructure rusty_linter modules
• Restructure rusty_parser modules
• Flatten public api of rusty_parser crate
• Remove complex nested context objects in linter converter
• Simplify converter for statements
• Not allowing 'zero' Position
• Use VariableInfo from linter Context in generator
• Use a struct for FunctionDeclaration and SubDeclaration
• Introducing new Visitor pattern in linter
• Simplify pre_linter types
• Implement visitor pattern for constant_map
• Converted function validate_string_length into a trait
• Converted user_defined_type_rules to the visitor pattern
• Use new Assignment struct in Statement
• Adapt for_next_counter_match_linter and print_linter to the Visitor pattern
• Add a bi_tuple macro to reduce some boilerplate code
• Changed Statement::SubCall to use a struct instead of two fields
• Changed Statement::BuiltInSubCall to use a struct instead of two fields
• Improve encapsulation of CaseBlock struct
• Make bi_tuple macro more expressive
• Migrated UserDefinedNamesCollector to the new Visitor pattern
• Combined all VarType* traits into one VarType trait
• Made fields of TypedName private
• Introducing ref_to_value_visitor
• Converted Name into a struct with private fields
• Add AsBareName and ToBareName traits
• Remove QualifiedName type
• Renamed chain to then_with
• Removed and_without_undo
• Moved specific types away from pc module
• Use small functions to improve constant_map readability
• Implement ConstLookup trait for names_inner
• Removed Position from Token, made fields private
• Moved the pc framework to the rusty_pc package

🎨 Styling

• Styling changelog according to default options

⚙️ Miscellaneous Tasks

• Ensure no code lives in mod.rs files
• Applied auto-fixes from cargo clippy
• Implementing cargo clippy suggestions
• Removed unnecessary boxing of &dyn Printer
• Implementing cargo clippy suggestions
• Implementing cargo clippy suggestions
• Reduced module visibility
• Extracted rusty_parser crate
• Extracted rusty_linter crate
• Removed funding file
• Update dependencies
• Extract test function in name parser's unit test
• Deprecated some error related methods
• Fix warning about lifeline
• Uniform coding style of parser implementation
• Reverted deprecation of or_fail and no_incomplete
• Removed method Variant::is_array
• Renamed idx variable to index for more clarity
• Removed deprecated method in MockInterpreterTrait
• Applied a few clippy suggestions
• Fixed use_self clippy rule
• Fixed clippy finding branches_sharing_code
• Derive default with annotation in linter converter types
• MainContext of PreLinter does not need to be public
• Marked function as test only
• Renamed implicits to implicit_vars
• Moved NameInfo to its own module
• Improved docs
• Use single field in Names for all one-level data
• Removed mut methods of Position
• Add comments
• Instantiate pre_linter context with Default
• Removed unwrap custom functions in favor of From trait
• Change ConstantMap type into a struct for encapsulation
• Move bi_tuple macro to a new module
• Implement AsRef for TypedName<T>
• Removed From<Token> implementation of BareName
• Removed unused code in type_qualifier
• Improve keyword_enum macro
• Removed AccumulateParser
• Removed AllowNoneIf parser
• Removed obscure ExtractExpression trait
• Code readability improvements to opt_second_expression
• Improve naming of ConstValueResolver for clarity
• Reformat code with nightly rustfmt
• Upgrade packages edition to 2024
• Upgraded rusty_parser to 2024 edition
• Re-arrange packages in rusty_parser
• Update README
• Split Windows build on a separate GitHub workflow
• Migrating GitHub workflows to dtolnay/rust-toolchain
• Unifiy build and build-windows GitHub actions
• Trying a windows-latest GitHub runner instead of cross compilation
• Experimenting with sharing artifacts between jobs
• Align build and release GitHub workflows

[0.8.0] - 2022-09-25

🚜 Refactor

• Using tokenizers (#110)

⚙️ Miscellaneous Tasks

• Updated copyright year in LICENSE
• Use cliff.toml from instarepo
• (changelog) Updated changelog
• (changelog) Updated changelog

[0.7.0] - 2021-12-11

🚀 Features

• [breaking] Use standard PATH_TRANSLATED variable for Apache cgi-script instead of BLR_PROGRAM
• Introducing git-cliff and changelog

[0.5.0] - 2020-11-27

Bugfix

• Arguments appearing out of order due to enumerating over HashMap

Refactoring

• Changed Name::Qualified to use QualifiedName as it used to in the past
• Removed HasQualifier trait
• Renamed TypeDefinition to ExpressionType

[0.3.0] - 2020-07-29

🚜 Refactor

• Moved numeric casts to separate module

Bugfix

• Strings cannot be used as an if condition