What is underscore in terms of real or floating point fragments #4649
Replies: 4 comments 10 replies
-
|
To clarify a bit, basically, I want to support the following kinds of use cases: Maybe that's asking too much. i.e. just define an |
Beta Was this translation helpful? Give feedback.
-
|
Having a devil of a time making this work correctly. Trying to get this "simple" DSL working, but it cannot seem to parse beyond the first fragment DIGIT: [0-9];
fragment DOT: '.';
fragment MINUS: '-';
DECIMAL_LITERAL: DIGIT+ ( DOT DIGIT+ )?;
NUMERIC_LITERAL: MINUS? DECIMAL_LITERAL;
SP: ' ';
WS: ( SP | HT | VT )+ -> skip;
location: coordinate ( SP+ coordinate )*;
coordinate: NUMERIC_LITERAL;I've tried different iterations of this, none of them seem to be quite working. For now focused there. Goal is to support the decimal forms above. But like I said, I am willing to consider an Eventually in its slightly more complex axis based form: // Importing the baseline grammar
fragment X: [xX];
fragment Y: [yY];
fragment Z: [zZ];
AXIS: X | Y | Z;
OP_EQ: '=';
coordinate: AXIS SP* OP_EQ SP* DECIMAL_LITERAL; |
Beta Was this translation helpful? Give feedback.
-
|
By comparison, or perhaps contrast at this point, the same problem approached from a regular expression perspective has been proven far more reliable. decimal val;
var re = new Regex(@$"\s*(?<{nameof(val)}>\-?(\d+(\.\d*)?|\d*\.\d+))");
// And subsequent logic parsing through the matches etc. |
Beta Was this translation helpful? Give feedback.
-
|
Ah... I was wondering whether maybe whitespace was being gobbled up when I didn't want it to be. I'll try those suggestoins, thanks... |
Beta Was this translation helpful? Give feedback.
-
|
Following up, I tried adjusting the lexer and rules, still no luck. It is not matching even the first coordinate attribute. // Lexer rules
fragment DIGIT: [0-9];
fragment DOT: '.';
fragment MINUS: '-';
fragment DECIMAL_LITERAL: DIGIT+ ( DOT DIGIT+ )?;
WS: ( ' ' | '\t' | '\f' )+ -> skip;
// Grammar rules
location: coordinate+;
coordinate: numericLiteral;
numericLiteral: MINUS DECIMAL_LITERAL | DECIMAL_LITERAL;And the axis based grammar, "extending" only coordinate. fragment X: [xX];
fragment Y: [yY];
fragment Z: [zZ];
AXIS: X | Y | Z;
OP_EQ: '=';
coordinate: AXIS OP_EQ numericLiteral;I'm a bit confused, when to or not to Under test, it's not even reaching the first coordinate, with breakpoints set in my: public override void ExitCoordinate([NotNull] ScumLocationParser.CoordinateContext context) { ... }So something's wrong, but I'm not sure exactly what. |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Thinking I tried the Put breakpoints in for Enter/Exit Every Rule. Enter, then Exit hit; EOF. It is not even recognizing the first decimal numeric literal. Context showing |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Not sure what you are doing wrong. Here's a zip with driver, grammar, etc., for C#. |
Beta Was this translation helpful? Give feedback.
-
|
Definitely not accustomed to running command line anything, and the scaffold I have is already established. Doing the best I can running through the listeners and such; although perhaps maybe could add more diagnostics there, but that aside. Tried a couple of adjustments, definitely MINUS: '-';
DECIMAL_LITERAL: DIGIT* DOT DIGIT+ | DIGIT+ ( DOT DIGIT* )?;
// ...
numeric: MINUS? DECIMAL_LITERAL;Also reverted to my original decimal definition, as I think this is more reflective of the actual potential use cases. Closer, I think, the results indicate as much. Now all value only coordinates use cases passing. I can troubleshoot the axis based cases; probably something similar going on there, something whether it should be |
Beta Was this translation helpful? Give feedback.
-
|
Okay, and verified axis listeners, now having victory in both areas. 🎆 🍾 And here I was ready to explore the possitibility of a SWIG C++ Boost Spirit X3 and/or Qi based approach, big fan of that technology. However, there's a lot more integration work to do around that possibility extending into a dotnet assembly. 🐑 😬 |
Beta Was this translation helpful? Give feedback.
-
|
C++/SWIG is tough to deal with sometimes. Antlr with the C# target and Antlr4BuildTasks is much easier. Make sure to use antlr.runtime.standard, not the old Antlr4cs. |
Beta Was this translation helpful? Give feedback.
-
|
Yeah I know. In and of itself, I was "ready" with a templated X3 parser. But getting this to work was a huge relief. Good to know re: |
Beta Was this translation helpful? Give feedback.
-
Referencing in particular the CSharp
REAL_LITERALlexicon, perhaps others have modeled similarly:Mainly am interested in a subset portion of the definition:
However, what is the significance of the
'_'in the definition, I am unfamiliar with that syntax.As if to say, could be something like this?
0___0, really? After a bit more digging, yes, truly, I guess so; if I am keeping with that definition.No harm if my particular use cases do not need it? Otherwise, the lexicon seems pretty solid. Better than what I was tinkering with, fewer alternatives that can trip up and foul things up:
Thoughts, suggestions? Other than of course to pare back the CSharp description a bit and just run with it.
Beta Was this translation helpful? Give feedback.
All reactions