An other question about jep (grammar)
guillau...@gmail.com
I follow the integration of your API in my application,
but I meet some new problems in my expressions.
This time I got a conditionnal expression a bit more complicate,
I can't format this one like the parser want, this is my expression :
A_N_E =
if (CT == 1,
(L_c / L_u) -1,
if (CT == 2,
if (TP == 1,
0.46,
if (TP == 2,
(84/(L * PE(81/(L + 0.5)))) -1,
if (TP == 3,
(81/(L * PE(78/(L + 0.5)))) -1,
if (CT == 3,
if (TP == 1 || 2,
0.46,
if (TP == 3,
1- L_c/ (PE(L_B/(L + 0,5)) * L ),
0))))))))
The parser launch me an error :
com.singularsys.jep.ParseException: Encountered "," at line 1, column
409.
Some help needed, thank's you in advance .
Guillaume
Richard Morris
Last but one line has
> (L + 0,5),
You want
> (L + 0.5),
BTW you may find it easier to break the input into multiple lines
that way its easier to use the error message to identify the bug
String eqn = "A_N_E =\n"+
" if (CT == 1,\n"+
" (L_c / L_u) -1,\n"+
" if (CT == 2,\n"+
" if (TP == 1,\n"+
" 0.46,\n"+
" if (TP == 2,\n"+
" (84/(L * PE(81/(L + 0.5)))) -1,\n"+
" if (TP == 3,\n"+
" (81/(L * PE(78/(L + 0.5)))) -1,\n"+
" if (CT == 3,\n"+
" if (TP == 1 || 2,\n"+
" 0.46,\n"+
" if (TP == 3,\n"+
" 1- L_c/ (PE(L_B/(L + 0,5)) * L ),\n"+
" 0))))))))";
jep.parse(eqn);
Which gives me
com.singularsys.jep.ParseException: Encountered "," at line 15, column 47.
Line 15 column 47 is easier to find than line 1, column 409.
Rich
--
Richard Morris
Web: www.singsurf.org www.pfaf.org
Email: ri...@singsurf.org
Tel: (+44) 01208 872963
Post: 1 Lerryn View, Lerryn, Lostwithiel, Cornwall, PL22 0QJ
guillau...@gmail.com
but after i correct the "," by a "."
I got now a new error :
Function "if" illegal number of arguments 2
From where it is come from i use the if formalism that you told me ...
thank's in advance
Guilllaume
On 14 août, 13:30, Richard Morris <r...@singsurf.org> wrote:
guillau...@gmail.com
conditionnal expression,
i explain, for a simple if i do :
if(test, process_if_true, process_if_false).
If i got 2 inside themself if :
if(test_1, process_if_true, if(test_2, process_if_test_2_true,
process_if_test_2_false),process_if_test_1_false,
I am quitte sure that i got a syntax problem with imbricate if
expression, so how can i do ??
Guillaume
Richard Morris
> I think I don't have really understand the way to write multi
> conditionnal expression,
> i explain, for a simple if i do :
> if(test, process_if_true, process_if_false).
>
> If i got 2 inside themself if :
> if(test_1, process_if_true, if(test_2, process_if_test_2_true,
> process_if_test_2_false),process_if_test_1_false,
>
> I am quitte sure that i got a syntax problem with imbricate if
> expression, so how can i do ??
You do have the syntax right, I think its a problem of just counting the
number of brackets right.
>>>> A_N_E =
>>>> if (CT == 1,
>>>> (L_c / L_u) -1,
>>>> if (CT == 2,
>>>> if (TP == 1,
>>>> 0.46,
>>>> if (TP == 2,
>>>> (84/(L * PE(81/(L + 0.5)))) -1,
>>>> if (TP == 3,
>>>> (81/(L * PE(78/(L + 0.5)))) -1,
>>>> if (CT == 3,
>>>> if (TP == 1 || 2,
>>>> 0.46,
>>>> if (TP == 3,
>>>> 1- L_c/ (PE(L_B/(L + 0,5)) * L ),
>>>> 0))))))))
Looking that this a switch statement might make things simpler
public class Switch extends PostfixMathCommand {
private static final long serialVersionUID = 330L;
public Switch() {
this.numberOfParameters = -1;
}
public void run(Stack<Object> stack) throws EvaluationException {
Object vals[] = new Object[this.curNumberOfParameters];
for(int i=this.curNumberOfParameters-1;i>=0;++i)
vals[i] = stack.pop();
int index = (Integer) vals[0];
if(index>= this.curNumberOfParameters)
throw new EvaluationException("Switch: condition out of range
"+index+" max "+(this.curNumberOfParameters-1));
stack.push(vals[index]);
}
}
using the above finction which could be added with
jep.addFunction("switch",new Switch());
you could then transform your code to
>>>> A_N_E =
>>>> switch(CT,
>>>> (L_c / L_u) -1,
>>>> switch(TP,
>>>> 0.46,
>>>> (84/(L * PE(81/(L + 0.5)))) -1,
>>>> (81/(L * PE(78/(L + 0.5)))) -1),
>>>> switch(TP,
>>>> 0.46,
>>>> 0.46,
>>>> 1- L_c/ (PE(L_B/(L + 0.5)) * L )))
Rich
--
Richard Morris
Web: www.singsurf.org www.pfaf.org
Email: ri...@singsurf.org
guillau...@gmail.com
String formule = "A_N_E =\n"+
" 0.46, \n"+
" if (TP == 2, \n"+
" 2),3), \n" +
" if (TP == 1, \n"+
" 4),5), \n"+
" 0.66, \n"+
" 6),7), \n"+
" if (CT == 4,\n"+
" 8),9), \n"+
" 1))))\n";
Ok this formula is wel parsed by jep, but I found something
interresting :
-Always check the number of brackets.
-you have to get a else statement inside each sub conditions
Thank's to richard again you help me a lot :)
Guillaume
On 14 août, 19:04, Richard Morris <r...@singsurf.org> wrote:
guillau...@gmail.com
" 0.46, \n"+
" if (TP == 2, \n"+
" 2),3), \n" +
" if (TP == 1, \n"+
" 4),5), \n"+
" 0.66, \n"+
" 6),7), \n"+
" if (CT == 4,\n"+
" 8),9), \n"+
" 1))))\n";
Ok this formula is wel parsed by jep, but I found something
interresting :
-Always check the number of brackets.
-you have to get a else statement inside each sub conditions
Thank's to richard again you help me a lot :)
Guillaume
Richard Morris
Cann't we have any solution to parse 3,5 as floating point number..without changing COMMA to DOT?
static class CommaNumberTokenMatcher extends NumberTokenMatcher {
private static final long serialVersionUID = 1L;
public CommaNumberTokenMatcher() {
super("(\\d+\\,?\\d*)|(\\,\\d+)");
}
@Override
public Token buildToken(String s) {
String s2=s.replace(',','.'); // convert string changing , to .
return super.buildToken(s2); // create a Token using the string, calls Double.valueof(String);
}
}
@Test
public void testCommaNumber() throws JepException {
ConfigurableParser cp = new ConfigurableParser();
cp.addHashComments();
cp.addSlashComments();
cp.addSingleQuoteStrings();
cp.addDoubleQuoteStrings();
cp.addWhiteSpace();
cp.addTokenMatcher(new CommaNumberTokenMatcher());
cp.addExponentNumbers();
cp.addOperatorTokenMatcher();
cp.addSymbols("(",")","[","]",";"); // changed
cp.setImplicitMultiplicationSymbols("(","[");
cp.addIdentifiers();
cp.addSemiColonTerminator();
cp.addWhiteSpaceCommentFilter();
cp.addBracketMatcher("(",")");
cp.addFunctionMatcher("(",")",";"); // changed
cp.addListMatcher("[","]",";"); // changed
cp.addArrayAccessMatcher("[","]");
Jep jep = new Jep(cp);
jep.parse("24,9");
Object result = jep.evaluate();
assertEquals(24.9,result);
jep.parse("if(2,1 > 3,4 ; 5,6 ; 7,8)");
result = jep.evaluate();
assertEquals(7.8,result);
jep.parse("[2,1;3,4]");
result = jep.evaluate();
assertArrayEquals(new Object[]{2.1,3.4}, ((List<?>) result).toArray());
}
Tikeshwar Sahu
< FLOATING_POINT_LITERAL:
(["0"-"9"])+ "." (["0"-"9"])* (<EXPONENT>)?
| "." (["0"-"9"])+ (<EXPONENT>)?
| (["0"-"9"])+ <EXPONENT>
>
here in place of "."(DOT) i need to use "jep.decimalSymabol()"
< FLOATING_POINT_LITERAL:
(["0"-"9"])+ "jep.decimalSymabol()" (["0"-"9"])* (<EXPONENT>)?
| "jep.decimalSymabol()" (["0"-"9"])+ (<EXPONENT>)?
| (["0"-"9"])+ <EXPONENT>
>
where decimalSymabol() will return either ","
or "." , means based on Locale. But whet it return "," then i ll be getting exception like
Caused by: com.singularsys.jep.parser.ParseException: Encountered " "," ","" at line 1, column 5.
Was expecting one of:
"," ...
">" ...
"<" ...
"=" ...
"<=" ...
">=" ...
"<>" ...
"&&" ...
"||" ...
"+" ...
"-" ...
"*" ...
"." ...
"/" ...
what i am doing is wrong or??
--
You received this message because you are subscribed to the Google Groups "Jep Java Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jep-users+...@googlegroups.com.
To post to this group, send email to jep-...@googlegroups.com.
Visit this group at http://groups.google.com/group/jep-users?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
--
Tikeshwar Sahu
Zoho Corporation Pvt. Ltd
+91-7871565242
Richard Morris
<DEFAULT> TOKEN:
{
<INDENTIFIER1: <LETTER1>(<LETTER1>|<DIGIT1>|".")*>
|
< #LETTER1:
[
"\u0024", // $
"\u0041"-"\u005a", // A - Z
"\u005f", // _
"\u0061"-"\u007a", // a - z
...
]
>
|
< #DIGIT1: ["0"-"9"] >
}
<NO_DOT_IN_IDENTIFIERS> TOKEN:
{
<INDENTIFIER2: <LETTER2>(<LETTER2>|<DIGIT2>)*>
|
< #LETTER2:
[
"\u0024", // $
"\u0041"-"\u005a", // A - Z
"\u005f", // _
"\u0061"-"\u007a", // a - z
...
]
>
|
< #DIGIT2: ["0"-"9"] >
}
Latter in the BNF grammar you need to allow for both types of token
String Identifier() :
{
Token t;
}
{
( t = <INDENTIFIER1> | t = <INDENTIFIER2> ) { return t.image; }
}
You can then call
public void setInitialTokenManagerState(int state)
to set the state, using the constants defined in JccParserConstants.
Good luck, I've found working with the JavaCC grammar quite a tricky thing to get right and would not recommend it.
Richard