Interesting idea. It seems this could be done without messing with the parser at all. What would be the benefit of a parsed option over an enhancement to the function?
If extending the functionality of table() is all that is wanted, you might consider something along the lines of newlist (posted in July and appended to this post). Newlist replaces list() in an attempt to get some consistency with set(). There were some discussions and a few design enhancements were left out because they would have been confusing/inconsistent. But what would an enhanced table look like? Why not just add a second argument to table? (I don't think one is used). table([k1,v1,k2,v2]) # only sets the default table(,[k1,v1,k2,v2]) # an obvious choice table(,[[k1,v1],...]) # also obvious, but is it useful table(,table()) # copy ?? error ?? table(,set()) # what should this do? other types? David --- Clint Jeffery <[EMAIL PROTECTED]> wrote: > > > [Wade] came up with the following: > > newtable := [< "key1": "value1", "key2": > "value2", "key3": "value3" >] > > This is not bad... the obvious question is whether > the < and > are needed... > > > Where I was tinkering has the characteristic > that the new literal type is > > down in the Icon code, not the Unicon code. I > never, in fact, thought to > > look in the Unicon layer. Would it need to know? > unigram.y looks like it > > would need to (why?). > > Yes, correct. If you change Icon's grammar, > Unicon's grammar needs to the > same change in order to parse the new construct and > allow it to "pass through". > The opposite is not true; a change to Unicon needs > no change to icont. > > As things stand, you are proposing adding a new VM > instruction, which is > much more serious business than adding some new > syntax or extending a > built-in function (table()) to support that new > syntax. > > Clint > > > ------------------------------------------------------- > This SF.Net email is sponsored by the JBoss Inc. > Get Certified Today > Register for a JBoss Training Course. Free > Certification Exam > for All Training Attendees Through End of 2005. For > more info visit: > http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click > _______________________________________________ > Unicon-ldif mailing list > [EMAIL PROTECTED] > https://lists.sourceforge.net/lists/listinfo/unicon-ldif > ############################################################################ # # File: newlist.icn # # Subject: Unicon extension for built-in procedure list # # Authors: Steve Wampler, David Gamey # # Date: July 2005 # ############################################################################ # # This file is in the public domain. # ############################################################################ # # This procedure overrides and extends the built-in function list(). # # list(X) returns X converted to a list, where X is a table, set, or record # # list(X,x1,...,xn) returns a list with appended elements x1,...,xn # # list(L) returns L (not a copy), where L is a list # # list(L,x1,...,xn) returns a copy of L with appended elements x1,...,xn # # otherwise list behaves exactly as the built-in function # # Notes: # # 1. Returning unmodified lists saves surprises with garbage collections. # It seemed more consistent with other functions. # 2. This now makes list() symmetric functionality with set(). # 3. list(T) converts the values of a table to a list. # To convert the keys of a table, include "link tables" and # use list(keyset(T)). # 4. list(set(...)) is orderless. # 5. list(x) was not extended to handle strings differently than the # built-in function because strings are handled inconsistently and # no really useful and compatible interpretation was obvious. # For example, list("3") returns [,,] and list("a") generates and error. # 6. list() was not extended to csets, because elements of !csets generate # single character strings. # 7. On the surface, this code looks bilingual (Unicon/Icon) but we suspect # it relies on Unicon extensions. # ############################################################################ procedure list(x[]) #: convert structure to list and append local L static LIST initial LIST := proc("list",0) if *x = 0 then return x # we already got one! case type(x[1]) of { "list" : # list type if *x = 1 then return x[1] # . return original if no work else L := copy(get(x)) # . make copy and prepare for work "table" | "set" : # other structures every put(L := [], !get(x)) # . convert elements to list default: if image(x[1]) ? ="record " then # record structure every put(L := [], !get(x)) # . convert else # null, numeric (including numeric string) return LIST!x # . use old built-in behaviour } every put(L,!x) # concatenate any remaining varargs return L end ------------------------------------------------------- This SF.Net email is sponsored by the JBoss Inc. Get Certified Today Register for a JBoss Training Course. Free Certification Exam for All Training Attendees Through End of 2005. For more info visit: http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click _______________________________________________ Unicon-group mailing list Unicon-group@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/unicon-group
