Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[madk]

Yes. I have it working now. I also added several things to make math and stuff that much less frustrating for the modder to work with, including ditching the confusing [? (...) ] for [# (...) ] as an option, and even better, ( (...) ). [A (...) ] is for performing absolute value on the operation.

Here are the current operators, keeping them as a single character is a lot easier to implement than full keywords, so deal with it.

Standard operators +-*/
Exponential operators ^ (to power) and V (root)
Bitwise logic operators & (and) | (or) X (xor) U (nand) N (nor) and Z (xnor)
Bitshifting operators L (shift left) and R (shift right)
Comparison operations = > and <
Modulo operation M (would have been %, but then the stuff that recognizes variables would screw up.)

I just need to find out why the floating point stuff isn't working (it's insisting on removing the decimals for some reason) and make sure the standard input works OK and I'll give a release. It won't be useful yet, but it'll be better than having no proof of progress save these forum posts.


Oh, and I have yet to find out if my xor and xnor operations work properly. I had to hack them up with not ands, because the language I'm using doesn't have xor readily available as an operation.

[madk]

Oh, lovely. Something about strings is still bugged. 

[madk]

Alright, only two bugs left to tackle. Xor works, and I'm very happy about that.

Code: [Select]

define %test% { "       this is a test." }
define %roar% { "hear me roar." }
define %both% { %test% %roar% }
output %both%


define %one% {1}
define %two% { (8^ (9/3)) }
output %two%


define %zero% { 0 }
output [& %one%%zero% ]
define %ten% { %one%%zero% }
output %ten%
output [& "zero  " + (%one%X%one%) ]
The bugs are:
"       this is a test." becomes "this is a test." and "zero " becomes "zero". This should only be a matter of finding out where I'm removing these quotes prematurely by mistake.
And I still haven't looked into the floating points not working. For example, "define %two% { (8^ (9/3)) }" (512) works beautifully, while "define %two% { (8^ (3/9)) }" (2) does not.



EDIT: Ah, no, the string bug is rather major. I guess I had missed making the arithmatic parser able to recognize that "1+1" is a string as opposed to 1+1 being an equation. This lack of recognition of quotes in this particular section of code is the root of all the problems, though, so fixing this should correct the issues across the board.

[madk]

A fresh look in the morning:

I wrote a general parsing function ages back that I've used in programs innumerable to this day. Its purpose is very simple: if you have a line of text formatted like "command argument,argument //oh hey, comment!" it recognizes it and makes it generally easier to keep the code that processes it a little more consistent. Problem is, I forgot that I wrote it to realize that key chracters inside of quotes don't count. But it gets rid of those quotes in the output, too. There the issue lies, since I need these quotes to properly parse the string equations and things, but I'm still rather confused as to exactly why it's still happening after I thought I had changed it.

And the floating points are only because I fogot to declare one variable as a double instead of an int. Silly me.

Release get, as soon as these two things are fixed. (I hope)

[madk]

Ahar, release!

At the moment, it can't do anything to the raws, I haven't worked on that yet since this rewrite. The code will be easy to transfer from the previous work, though.

Please read the comments in the code file to get an idea of what's going on.

Download

Conditional statements and loops are probably next. Shouldn't take long, I've already got the groundwork implemented in the parsing loop.

[madk]

Ah, I forgot to fix one thing before I provided that link. The 10 example is broken atm.

[madk]

I'm moving on to conditional statements.

Also, I added = to compare strings as equal, and < and > for length comparisons. Also added u[pper]case, l[ower]case, p[roper]case, and i[nvert]case <%variable%> to change the cases of letters within strings. (These are not in the download. Next release will be after conditional statments and loops. Probably functions, too.)

EDIT: Modulo, too, now. It gives you the difference in length.

Anyone thinking of operators I may have missed?

[ilfirin]

Anyone thinking of operators I may have missed?

Can you create pointers yet?   

[madk]

It's less of a language than an interpreter. So no.

I can do something with it if you think it'll actually have a use, though. (Even though half the things in there are only there because it took a couple lines of code, I doubt anybody will ever use some of these features)


EDIT: Hm, I just realized I have no support for arrays. I think I should probably do that.

[bdog]

Anyone thinking of operators I may have missed?

Can you create pointers yet?   

As for operators: as i think now "not equal" (others like >= or <= for numerical comparision can be substituted with normal > and < so they not that needed)

For pointers: heck, I think that I had good imagination but still don't know how would you make use of them here but the more the merrier

===
On a side note: yesterday when I had a little free time I made something that you could call bridge between C# and JavaScript (well, its using already working solutions so it was like connecting bricks).
I made it just for fun and its still not quite working with raws (well, working but not as I would like to)
Engine allows you to use .net code IN javascript code (classes, objects etc)

Spoiler (click to show/hide)

Code: [Select]

var raw = new dfjs.DfRaw('reaction_smelter.txt');
writeln(raw); // prints 'dfjs.DfRaw' on console
writeln(raw.tag + ' : ' + raw.value); // prints 'OBJECT : REACTION'
for (var i=0; i<raw.length; i++) {
  writeln(' > ' + raw[i].tag + ' : ' + raw[i].value); // prints bunch of ' > REACTION : name'
  if (raw[i].length > 0) {
    for (var j=0; i<raw[i].length;  j++) {
      writeln(' >  > ' + raw[i][j].tag + ' : ' + raw[i][j].value); // should print other tags etc but its still not working like I would to
    }
  }
}

With JS one could do almost everything on raws. Problem is in reading raw txt and parsing them into arrays (I'm using regex atm to get all tags and a couple of other regexes to sort them somehow) and saving them back.
Question: whitespaces in raw files are there only for readability or are they a must?

[madk]

Readability.

Here's the relevant BlitzMax code for my parsing of the raws file.

Every raw object is stored with an object (the [OBJECT:___] tag), what file it originally came from, and all of the tokens are stored in sequence in a linked list.

Raw object "obj" object.

EDIT: ack, the codebox takes away tabs. You get spoilers instead.

Spoiler (click to show/hide)

Type obj
   Global list:TList=CreateList()
   Field origin$
   Field kind$,name$
   Field tokens:TList=CreateList()
End Type

Here's where most of the action takes place:

Spoiler (click to show/hide)

Global raws:TList=CreateList()

Function Run()
   Local fil:TStream
   
   'store code in memory
   Print "Reading rawscript.."
   fil=ReadFile("RawScript.txt")
   Local lines%=0
   Repeat
      ReadLine(fil)
      lines:+1
   Until Eof(fil)
   SeekStream fil,0
   code=code[..lines+1]
   lines=0
   Repeat
      Local lin$=ReadLine(fil)
      lin=Trim(Replace(lin,Chr(9)," "))
      code[lines]=lin
      lines:+1
   Until Eof(fil)
   codelen=lines
   CloseFile fil
   Print "Rawscript read."
   
   'request DF raws directory
   Print "Please specify the relevant directory.."
   Local dir$=RequestDir("Please select your Dwarf Fortress raw\objects folder.",CurrentDir())
   If FileType(dir)<>2 End
   Print "Directory received."
   
   'get raw files and add them to list raws
   Print "Detecting raw files.."
   ClearList raws
   getraws dir
   Print "Raws found."
   
   'scan the files for [OBJECT:___] tags, because they're very important.
   Print "Scanning for [OBJECT:___] tags.."
   For Local do$=EachIn raws
      Local this:obj=Null
      fil=ReadFile(do)
      Repeat
         Local lin$=ReadLine(fil)
         grabtokens lin

            For Local k$=EachIn tokens
               readtoken k
               If token="OBJECT" Then objects.addlast uarg(1)
            Next

      Until Eof(fil)
      Print "Finished scanning file "+do
   Next
   Print "Tags scanned."
   
   'read all the objects
   Print "Storing objects.."
   For Local do$=EachIn raws
      Local this:obj=Null
      fil=ReadFile(do)
      Local thisg:group=Null
      Local child:obj=Null
      Repeat
         Local lin$=ReadLine(fil)
         grabtokens lin
            
            For Local k$=EachIn tokens
               readtoken k
               If isnew(token) Then
                  this=New obj;this.origin=StripAll(do);this.kind=token;this.name=uarg(1);obj.list.addlast this
                  If numparams>1
                     For Local a%=2 To numparams
                        this.args[a-2]=args[a]
                     Next
                  EndIf
                  child=Null
               ElseIf this
                  Local newborn%=0
                  For Local g:group=EachIn group.list
                     If g.parent=this.kind And g.child=token Then
                        thisg=g
                        child=New obj
                        child.origin="IS_CHILD"
                        child.kind=token
                        child.name=uarg(1)
                        If numparams>1
                           For Local a%=2 To numparams
                              child.args[a-2]=args[a]
                           Next
                        EndIf
                        this.children.addlast child
                        newborn=1
                        Exit
                     EndIf
                  Next
                  If child
                     If newborn=0
                        For Local g:group=EachIn group.list
                           If g.parent=this.kind And g.child=token Then
                              thisg=g
                              child=New obj
                              child.origin="IS_CHILD"
                              child.kind=token
                              child.name=uarg(1)
                              If numparams>1
                                 For Local a%=2 To numparams
                                    child.args[a-2]=args[a]
                                 Next
                              EndIf
                              this.children.addlast child
                              Exit
                              newborn=1
                           EndIf
                        Next
                     EndIf
                     If newborn=0 child.tokens.addlast k
                  Else
                     this.tokens.addlast k
                  EndIf
               EndIf
            Next
            
      Until Eof(fil)
      Print "Finished reading file "+do
      CloseFile fil
   Next
   Print "Objects stored."
   
   Print "Executing script.."
   Local pos%=0
   get_groups code
   execute code
   Print "Execution complete."
   
   Print "Saving changes to raws.."
   For Local r$=EachIn raws
      fil=WriteFile(r)
      WriteLine fil,StripAll(r)
      Local first%=1
      WriteLine fil,""
      For Local o:obj=EachIn obj.list
         If o.origin=StripAll(r) Then
            If first Then
               first=0
               WriteLine fil,"This file has been processed by RawScript."
               WriteLine fil,"[OBJECT:"+o.kind+"]"
               WriteLine fil,""
            EndIf
            WriteLine fil,"  ["+o.kind+":"+o.name+"]"
            For Local t$=EachIn o.tokens
               WriteLine fil,"    ["+t+"]"
            Next
            For Local c:obj=EachIn o.children
               WriteLine fil,"    ["+c.kind+":"+c.name+c.doargs()+"]"
               For Local t$=EachIn c.tokens
                  WriteLine fil,"      ["+t+"]"
               Next
            Next
            WriteLine fil,""
         EndIf
      Next
      CloseFile fil
      Print "Saved modifications to "+r
   Next
   Print "Finished saving."
   
   Print ""
   Print "Rawscript execution complete. The modification has been installed."
   Input "Press enter to exit."
   
End Function

Here's some miscellaneous token parsing code:

Spoiler (click to show/hide)

Function match%(o:obj,text$)
   text=Upper(text)
   If Not Instr(text,"*") Then 'Easy! No wildcard involved.
      If o.kind+":"+o.name=text Return 1
      If "OBJECT:"+o.kind=text Return 1
      For Local t$=EachIn o.tokens
         readtoken t
         If "TOKEN:"+t=text Return 1
         If "TOKEN:"+token=text Return 1
      Next
   ElseIf text="*" 'Aha. Only wildcard!
      Return 1
   Else 'has a wildcard and normal text. this'll be a little harder, eh?
      If Left(text,1)="*" 'can be in string
         If Right(text,1)="*" 'beginning and end, both
         
         Else
            If Instr(o.kind+":"+o.name,text) Return 1
            For Local t$=EachIn o.tokens
               readtoken t
               If Instr("TOKEN:"+t,text) Return 1
               If Instr("TOKEN:"+token,text) Return 1
            Next
         EndIf
      ElseIf Right(text,1)="*" 'has to start with string
         If Instr(o.kind+":"+o.name,text)=1 Return 1
         For Local t$=EachIn o.tokens
            readtoken t
            If Instr("TOKEN:"+t,text)=1 Return 1
            If Instr("TOKEN:"+token,text)=1 Return 1
         Next
      ElseIf Instr(text,"#") 'the wildcard specifically indicates a missing character
         If wild(o.kind+":"+o.name,text) Return 1
         For Local t$=EachIn o.tokens
            readtoken t
            If wild("TOKEN:"+t,text) Return 1
            If wild("TOKEN:"+token,text) Return 1
         Next
      EndIf
   EndIf
   Return 0
End Function



Function wild%(str$,sub$)
   For Local k@=1 To Len(str)
      If Not(Mid(str,k,1)=Mid(sub,k,1) Or Mid(sub,k,1)="#") Then Return 0
      If k>Len(sub) Return 1
   Next
   Return 1
End Function





Global cmd$,arg$,arg2$
Function Parse(line$)
   line=Replace(line,Chr(9)," ")
   cmd="";arg=""
   Local on%=0
   For Local readit%=1 To Len(line)
      Local char$=Mid(line,readit,1)
      If char=" " And on=0
         on=1
      ElseIf char="/" ' comment!
         Exit
      ElseIf char="|"
         on=2
      Else
         If on=0 cmd:+char
         If on=1 arg:+char
         If on=2 arg2:+char
      EndIf
   Next
   cmd=Trim(cmd)
   arg=Trim(arg)
   arg2=Trim(arg2)
End Function






Function getraws(dir$)
   Local files$[]=LoadDir(dir)
   For Local x$=EachIn files$
      If ExtractExt(x)="txt" Then
         If FileType(dir+"\"+x)=1 raws.addlast dir+"\"+x
      EndIf
      If Trim(Replace(x,".",""))<>""
         If FileType(dir+"\"+x)=2 getraws dir+"\"+x
      EndIf
   Next
End Function




Global token$,args$[64],numparams@
Function ReadToken(text$)
   numparams=0
   args[0]=""
   CleanString text
   text=Replace(text,"[","")
   text=Replace(text,"]","")
   Local onparam@=0
   For Local p@@=1 To Len(text)
      Local char$=Mid(text,p,1)
      If char=":"
         onparam:+1
         If onparam<64 args[onparam]=""
      ElseIf onparam<64
         args[onparam]:+char
      EndIf
   Next
   token=args[0]
   numparams=onparam
End Function



Function CleanParams()
   CleanString token
   For Local x@=0 To numparams
      CleanString args

   Next
End Function

Function CleanString(text$ Var)
   text=Trim(Replace(text,Chr(9)," "))
   If Left(text,1)="'" And Right(text,1)="'" text=Mid(text,1,Len(text)-2)
End Function

Global tokens:TList=CreateList(),numtokens@
Function GrabTokens(text$)
   numtokens=0
   ClearList tokens
   CleanString text
   Local ontoken@=0
   Local started@=0
   Local grabbed$=""
   For Local p@@=1 To Len(text)
      Local char$=Mid(text,p,1)
      If char="]"
         tokens.addlast grabbed
         ontoken:+1
      ElseIf char<>"["
         grabbed:+char
      EndIf
   Next
   numtokens=ontoken+1
End Function


Function uarg$(index@)
   Return Upper(args[index])
End Function


Function LeftIs@(str$,sub$)
   If Len(sub)>Len(str) Return 0
   If Instr(str,sub)=0 Return 1
   Return 0
End Function

Maybe it'll be of some help to you. Keep in mind that it's old, mostly untested, and I know there are at least a couple bugs. So don't rely on it too much.

Also, as for a not equal operator, that's something I should have thought of myself. I'll be putting that in, too, then. Soon as I've got the arrays working.

[soul4hdwn]

*patently waits for completion*

[Normandy]

I think you should look into regexes (regular expressions) before continuing any further, they are the standard for text processing: http://en.wikipedia.org/wiki/Regex. There are many libraries for dealing with regexes in a simple fashion, so it's more of a question that you don't design a syntax which interferes with regex syntax (which unfortunately, you currently are) rather than any actual code work.

[madk]

I'm going on a trip until Tuesday now, so if I'm near a computer I'll try and do something, but GameJolt has its indie demake competition this week, so that'll have to be my priority until the 8th.

As for Regex, down with standards, up with working your rear off with equivalent but horribly complxex systems!!
Yeah, my system can basically do all regex can, but not as efficiently.

[Star Weaver]

As for Regex, down with standards, up with working your rear off with equivalent but horribly complxex systems!!
Yeah, my system can basically do all regex can, but not as efficiently.

Cool. Hopefully by the time you get back I'll have had the time and brain power to try converting things to your syntax .