bellinitte / parser-parser Goto Github PK

Expected behavior

Any parsed input should be visualized in the form of a (possibly interactable) tree structure. The nodes of the tree should represent the individual production rules.

Current behavior

Currently there is no visualizations of the inputs whatsoever.

Expected behavior

Every instance of left recursion in production rules should be reported as an error (related indentifiers should be underlined in the editor window).

Current behavior

Currently, any form of left recursion in production rules will cause an infinite loop and overflow the stack, which results in an error in the browser console.

Expected behavior

When writing code in the parser editor, any errors reported by the parser should be marked with an underscore regardless of the line number they are on.

Current behavior

The errors are marked only on the first line of the input. Any error past a newline character is discarded.

Possible solution

The core returns the error position as an offset from the beginning of the source string. The CodeMirror editor then interprets this offset as a column number, however the line number is hard-coded to the first line. The core should calculate the position as a line and a column number directly, instead of the naive offset.

Expected behavior

After creating rhe parser by typing the grammar definition into the parse input, the user should then be able to type into the check input to check if the string is valid within the grammar. For now, the parser should respond with a simple true/false.

Current behavior

No checking functionality is present in either the core or the UI at the time.

Expected behavior

Any references to unknown production rules in other production rules should be detected and reported as an error.

Current behavior

It is asserted that every production rule is valid, which is incorrect, so the checker may crash and display an error in the browser console.

Expected behavior

The parser should be able to process characters (or rather Unicode graphemes) consisting of multiple Unicode code points.

Current behavior

No errors are being reported when the input contains the é code point (the U+0065 "latin small letter e" plus the
U+0301 "combining acute accent").

Expected behavior

After typing a valid definition of a grammar, the user should be able to see every production rule in a dropdown menu list and select one of them as the initial production rule in the checking process.

Current behavior

The user currently cannot see the production rules and cannot select an initial production rule.