Skip to content

Latest commit

 

History

History
85 lines (51 loc) · 3.22 KB

README.md

File metadata and controls

85 lines (51 loc) · 3.22 KB
Raw

Actions Status

NAME

Terminal::ANSIParser - ANSI/VT stream parser

SYNOPSIS

use Terminal::ANSIParser;

my @parsed;

# Default: Assume UTF-8 decoded inputs, thus working with codepoints
my &parse-codepoint := make-ansi-parser(emit-item => { @parsed.push: $_ });
parse-codepoint($_) for $input-buffer.list;

# Raw bytes: Assume raw byte stream input, for pre-Unicode terminal emulation
my &parse-byte := make-ansi-parser(emit-item => { @parsed.push: $_ },
                                   :raw-bytes);
parse-byte($_) for $input-buffer.list;

# DEC Pedantic: Ignore xterm extensions, forcing pure DEC VT compatibility
#               (implies :raw-bytes as well)
my &parse-pedantic := make-ansi-parser(emit-item => { @parsed.push: $_ },
                                       :dec-pedantic);
parse-pedantic($_) for $input-buffer.list;

DESCRIPTION

Terminal::ANSIParser is a general parser for ANSI/VT escape codes, as defined by the related specs ANSI X3.64, ECMA-48, ISO/IEC 6429, and of course the actual physical DEC VT terminals generally considered the standard for escape code behavior.

The basic make-ansi-parser() routine builds and returns a codepoint-by-codepoint (or byte-by-byte, if the :raw-bytes option is True) table-based binary parsing state machine, based on (and extended from) the error-recovering state machine built from observed DEC VT behavior described at https://vt100.net/emu/dec_ansi_parser.

Each time the parser determines that it has parsed enough input, it emits a token representing the parsed data, which can take one of the following forms:

  • A plain codepoint (or byte if :raw-bytes is True), for passed through data when no escape sequence is active

  • A Terminal::ANSIParser::Sequence object, if an escape sequence is parsed

  • A Terminal::ANSIParser::String object, if a control string is parsed

A few Sequence subclasses exist for separate cases:

  • Ignored: invalid sequences that the parser decides should be ignored

  • Incomplete: sequences that were cut off by the start of another sequence or the end of the input data (see [End of Input](End of Input) below)

  • SimpleEscape: simple escape sequences such as function key codes

  • CSI: parameterized sequences beginning with CSI (ESC [)

Likewise, String has its own subclasses:

  • DCS: Device Control Strings

  • OSC: Operating System Commands

  • SOS: Strings beginning with a general Start Of String indicator

  • PM: Privacy Message (NOTE: NOT A SECURE FUNCTION)

  • APC: Application Program Command

End of Input

End of input can be signaled by parsing an undefined "codepoint"/"byte"; any partial sequence in progress will be flushed as Incomplete, and the undefined marker will be emitted as well, so that downstream consumers are also notified that input is complete.

AUTHOR

Geoffrey Broadwell [email protected]

COPYRIGHT AND LICENSE

Copyright 2021-2024 Geoffrey Broadwell

This library is free software; you can redistribute it and/or modify it under the Artistic License 2.0.