From patrick%cs.kuleuven.ac.be@princeton.edu Thu Aug 29 01:20:59 1991
Return-Path: <patrick%cs.kuleuven.ac.be@princeton.edu>
Received: from alpha.xerox.com by palain.PARC (4.1/SMI-4.1)
	id AA13556; Thu, 29 Aug 91 01:20:57 PDT
Received: from Minos.PARC.Xerox.xns by alpha.xerox.com via XNS id <11523>; Thu, 29 Aug 1991 01:20:32 PDT
Received: from Princeton.EDU ([128.112.128.1]) by alpha.xerox.com with SMTP id <11523>; Thu, 29 Aug 1991 01:20:05 PDT
Received: from fs.Princeton.EDU by Princeton.EDU (5.65b/2.80/princeton)
	id AA15585; Thu, 29 Aug 91 04:19:59 -0400
Received: from Princeton.EDU (Princeton.EDU.) by fs.Princeton.EDU (4.0/1.105)
	id AA08885; Thu, 29 Aug 91 04:19:57 EDT
Received: from n-kulcs.cs.kuleuven.ac.be by Princeton.EDU (5.65b/2.80/princeton)
	id AA15582; Thu, 29 Aug 91 04:19:49 -0400
Received: from medusa.cs.kuleuven.ac.be by n-kulcs.cs.kuleuven.ac.be (5.64+/n_kulcs1.1)
	id AA02072; Thu, 29 Aug 91 10:20:48 +0200
Received: by medusa.cs.kuleuven.ac.be (5.65b/cs_kuleuven.02)
	id AA08766; Thu, 29 Aug 91 10:21:14 +0200
X-Ns-Transport-Id: 08002008D0FD000FC327
Date: 	Thu, 29 Aug 1991 01:21:14 PDT
From: patrick%cs.kuleuven.ac.be@princeton.edu (Patrick Weemeeuw)
Subject: Spidery Web
To: nr@princeton.edu
Message-Id: <9108290821.AA08766@medusa.cs.kuleuven.ac.be>
Status: RO

Hello,
I'm trying to develop a web system for Prolog.

One of the difficulties I had is recognizing tokens containing
backslashes. I changed master/spider.web such that it now correctly
handles tokens like:
token \\== category ....
generating the C string "\\==" but checking for a string with lenght *3*
instead of 4.

The context diff is included below.

Patrick.
--
Patrick Weemeeuw                   mail: Katholieke Universiteit Leuven
Tel: +32 16 200656 x 3544                Dept. of Computer Science     
e-mail: patrick@cs.kuleuven.ac.be        Celestijnenlaan 200 A         
Fax: +32 16 205308			 B-3001 Leuven (Belgium)       
======================================================================


*** spider.web	Wed Aug 28 19:32:01 1991
--- spider.web.orig	Wed Aug 28 17:47:57 1991
***************
*** 794,800
  	tokenname[$2]="SP_gen_token_" tokennumber
  	tokennumbers[$2]=tokennumber
  	tokennumber--
- 	#<Find real length of |$2| and put it in |tokenlength[$2]|#>
  	## figure out how to recognize the token
  	temp = sprintf( "strncmp(\"%s\",loc-1,%d)==0", $2, tokenlength[$2])
  	tokentest[$2]=temp

--- 794,799 -----
  	tokenname[$2]="SP_gen_token_" tokennumber
  	tokennumbers[$2]=tokennumber
  	tokennumber--
  	## figure out how to recognize the token
  	temp = sprintf( "strncmp(\"%s\",loc-1,%d)==0", $2, length($2))
  	tokentest[$2]=temp
***************
*** 796,802
  	tokennumber--
  	#<Find real length of |$2| and put it in |tokenlength[$2]|#>
  	## figure out how to recognize the token
! 	temp = sprintf( "strncmp(\"%s\",loc-1,%d)==0", $2, tokenlength[$2])
  	tokentest[$2]=temp
  
  # To find the length of a string considered as a C~string, we scan it looking

--- 795,801 -----
  	tokennumbers[$2]=tokennumber
  	tokennumber--
  	## figure out how to recognize the token
! 	temp = sprintf( "strncmp(\"%s\",loc-1,%d)==0", $2, length($2))
  	tokentest[$2]=temp
  	tokenlength[$2]=length($2)
  
***************
*** 798,803
  	## figure out how to recognize the token
  	temp = sprintf( "strncmp(\"%s\",loc-1,%d)==0", $2, tokenlength[$2])
  	tokentest[$2]=temp
  
  # To find the length of a string considered as a C~string, we scan it looking
  for occurrences of the \.{\\} character. We initialize |templ| with

--- 797,803 -----
  	## figure out how to recognize the token
  	temp = sprintf( "strncmp(\"%s\",loc-1,%d)==0", $2, length($2))
  	tokentest[$2]=temp
+ 	tokenlength[$2]=length($2)
  
  
  # The setting of attributes is as for all tokens:
***************
*** 799,827
  	temp = sprintf( "strncmp(\"%s\",loc-1,%d)==0", $2, tokenlength[$2])
  	tokentest[$2]=temp
  
- # To find the length of a string considered as a C~string, we scan it looking
- for occurrences of the \.{\\} character. We initialize |templ| with
- |length($2)|, and each time we find a legitimate escape sequence, we decrement
- |templ|. Remark that, given the current lexical analysis, \.{\\f}, \.{\\n} etc.
- cannot occur in a token.
- #<Find real length of |$2| and put it in |tokenlength[$2]|#>=
- 	templ=length($2)
- 	i=1 #/
- 	while(i<=length($2)) { #/
- 	  if(substr($2,i,1)=="\\") { #/
- 	      if(i<length($2)) {
- 		  tempc=substr($2,i+1,1) #/
- 		  if((tempc=="a") || (tempc=="b") || (tempc=="\\") || (tempc=="?") || (tempc=="'")) {
- 			templ--
- 			i++
- 		  } else { #/
- 			i++ #/
- 		  } #/
- 	      } #/
- 	  } #/
- 	i++ #/
- 	} #/
- 	tokenlength[$2]=templ
  
  # The setting of attributes is as for all tokens:
  #<Set attributes of token |$2|#>=

--- 799,804 -----
  	tokentest[$2]=temp
  	tokenlength[$2]=length($2)
  
  
  # The setting of attributes is as for all tokens:
  #<Set attributes of token |$2|#>=

