World: r3wp

ok this is my first attempt for a basic expression parser

stack: copy []
push: func [x] [insert stack x]
pop: does [take stack]
term: [
	pos: paren! :pos into [expr] (push rejoin ["(" pop ")"]) |
	set t string! (push mold t ) | 

 set sign opt ['+ | '-] set t number! (if none? sign [sign: '+] push 
 mold (t * to-integer join sign "1")) | 

 set gp path! (push rejoin [{$('#} first gp {').attr('} second gp 
 {')} ]) 
]
oper: compose ['+ | '- | '* | (to-lit-word "/")]
operrule: [set o oper (push o)]

expr: [term any [operrule term (t2: pop o: pop t1: pop push rejoin 
[t1 o t2])]]


I'm sure it's not optimized nor correct, so any correction is welcome.

it can parse expression like: [- 1 * (a/b + 2)] and it emits a javascript 
compatible string.
(is a part of a comet server written in rebol)

instead of:

set sign opt ['+ | '-] set t number! (if none? sign [sign: '+] push 
mold (t * to-integer join sign "1")) |
I would use something like:
    set t number! (push mold t ) |
 '- set t number! (push mold negate t ) |

(Which is not perfect as well)

because the above will not work for expressions like: [ - (1 + 1)]

also I'm not sure, if the mold is needed

I am having some parse difficulties.  I have the below code...

records: read/lines http://www.geocities.com/[kalef-:-rogers-:-com]/samples.txt
lang-rule: [
	thru "language%3a" copy language to "+%2b" |
	thru "language%3a" copy language to "&resultview" |
	thru "language%3a" copy language to "&"
	;thru "language="   copy language to "&" |
]
rec: 1
foreach record records[
	language: copy ""
	parse record [lang-rule]
	print [rec tab language]
	rec: rec + 1
]

in the parse rule  lang-rule.  It seems that only two of the three 
rules work at any one time with the data I have loaded. (note the 
last rule is commented out)

If I change the order of the rules and place the pipe or bar '|' 
as appropriate, I can't get all three of these rules to work together.

I expect the output to be either the text 'en' or 'fr'  (for english 
or french) for each record, however records 52 & 66 the parse is 
not ending properly for the current setup.  If you change the order 
of the rules, then other records will not work as expected.   Any 
ideas?

lang-rule: [
	thru "language%3a" copy language ["en" | "fr"] to end
]

You could dehex first?  That would make things more consistent...

thanks for the suggestions.  I never thought of using the actual 
text I was expecting as an option like you suggested Oldes.

Okay Oldes, your solution works, but why does my code fail, any ideas? 
 I have other scenarios that follow this same sort of structure but 
do not have a simple two word expected result.  I've been able to 
handle these so far, simply by changing the order and moving the 
'stop' word of the Ampersand to the bottom of the rule options.  
[I'm trying Chris's option now]

I'm here again ! 

Is possible to write a rule that match any word! but those that are 
in a block ?
example:

list: ['for 'while 'show]
parse block [any [word! *NOT IN* list]]

If there's no a direct approach, is there a workaround ?

Brock: it'sbecause in rows 52 and 66 you find "+%2b".It'snot next 
to lang value you want, but it's there so the first rule is true.

You can use this for any lang-id with 2 chars:
lang-chars: charset [#"a" - #"z"]
lang-rule: [
	thru "language%3a" copy language 2 lang-chars to end
]

I would not use dehexas with dehex you parse the data twiceso it 
must be slower.

You can use this if there may be url-encoded data and not encoded 
as well:
lang-chars: charset [#"a" - #"z"]
lang-rule: [
	thru "language" ["%3a" | "="] copy language 2 lang-chars to end
]

btw. you can also  use:
parse "language=xx" lang-rule

Davide.. in R2:
list: ['for | 'while | 'show]
parse [for each while end] [any [list | set w word! (probe w)]]
In R3 it should be better I think.

dehex, agreed -- until the speed trade off becomes worth it.  Which 
may be if you are wanting to get more than just the language from 
the string.

Oldes, thanks for ripping into that with so many options.  To your 
first response - of course, didn't notice that.  Your last response 
is not obvious to me what that does, I'll need to look at that more.

Dehex isn't an issue for me really.  I am only taking  a very small 
percentage of records.  So in the big picture, it's not a significant 
slow-down.  The process this is attached to runs daily on a group 
of text files totalling less than 10 MB in size.

I was pasing something. I got this:
 ** Script Error: Internal limit reached
I still don't know what's wrong. Anybody?

Too many recursions.
Maybe the rule is too complex or you get an infinte loop.
Just show your rule and the problem.
For sure someone will help.

Thanks Oldes, I've tried your hint, but it made my parse rule too 
complex, so I've used an additional word as discriminant.
Hope that R3 will improve this.

sqlab, your are right. There is an infinite loop. I am fixing it. 
It's a C++ parser, so the rule is very complicated.

paul asked: "Question for you regarding parse.  How do you force 
parse to return false immediately without processing the rest of 
the string or block one it evaluates a paren?"

example: 


parse/all s [some ["12345" here: (print "*" here: tail here) :here 
skip]]


basically, in the paren, you assing the tail of the data being parsed, 
and force parse to move to it, then try going beyond....  the skip 
makes it return false, otherwise it returns true.

oops example forgot string to parse... hehehe

>> s: "12345 12345 12345"
== "12345 12345 12345"

>> parse/all s [some ["12345" here: (print "*" here: tail here) :here]]
*
== true

>> parse/all s [some ["12345" here: (print "*" here: tail here) :here 
skip]]
*
== false

Not exactly what I'm talking about.

Well maybe...

parse/all s [some ["12345" (print "*") end skip]
is that not enough ?

Well the skip is  going to produce false unless you know how much 
to skip

the here: tail here: is just the code you assign IF something you 
want to check programatically matched.

now the question, paul, is do you want the WHOLE parse to fail, or 
just that rule?  up to the paren?

or do you want to do, basically, look ahead assertion?

The skip will produce false because it is attempting to skip past 
the end.

The phrase [end skip] will always fail and return false.

What I'm talking about is this:

words: [this that blah]


result: parse [this that bleh blah][some [set w word! (unless find 
words w [do something to return false here)]]

result should evaluate to false in my example

For now, you want a continuation variable. Like this:

result: parse [this that bleh blah][some [set w word! (cont: unless 
find words w [[end skip]]) cont]]

continue: none	;means continue is ok
stop: [end skip]

result: parse [this that bleh blah][some [set w word! (unless find 
words w [continue: stop]) continue]]

ahah, same idea

Brian, I thought I tried that.

checking that one hold on.

I was working on another idea where only THAT rule failed !

for a general failure, the above works for sure.

yeah that will work Brian.  i think I was setting the word that I 
had placed at the end and that is why it failed before.