Wolfgang Schuster wrote:

this is a bug in the mkiv sorting macros, I posted a similiar message a few days ago but got no answer.

http://archive.contextgarden.net/message/20071125.084213.6c34633f.en.html

no time yet -) Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

2007/12/3, Hans Hagen :

Wolfgang Schuster wrote:

...

2007/12/3, Peter Münster :

...

Hello,

I get unexpected output with this test-file:

% engine=luatex \definesynonyms[test][tests] \test{testOne}{first test} \test{testTwo}{second test} \starttext \testOne \placelistoftests \stoptext

I would expect:

testOne testOne first test

but I get:

testOne test testOne

(Perhaps a problem with parameter numbers #1, #2 etc. ?)

Cheers, Peter

Hi Peter,

this is a bug in the mkiv sorting macros, I posted a similiar message a few days ago but got no answer.

\def\mksavesortedlistentry#1#2#3#4% class key entry meaning

{\immediatewriteutilitytua{ti(js['#1'],{'e','#2',\!!bs#3\!!es,\!!bs#4\!!es})}}

seems to work

this line should be \def\mksavesortedlistentry#1#2#3#4% class key entry meaning {\immediatewriteutilitytua{table.insert(js['#1'],{'e','#1',\!!bs#2\!!es,\!!bs#3\!!es})}} but there is also a big in core-syn.lua the following works for me, the order is "key, value" and not "value, key". I patched this at the moment in my format files and the output of the synonym table is now correct :-) function job.loadsortedlist(class) if job.sortedlists[class] then if not sorters.list.data[class] then sorters.list.data[class] = { language = 'en', entries = { }, flush = function(s) tex.sprint(tex.ctxcatcodes,s) end, sorted = false, class = class } local entries = sorters.list.data[class].entries for k,v in ipairs(job.sortedlists[class]) do if v[1] == 'l' then -- language sorters.list.data[class].language = v[2] else entries[#entries+1] = { v[1], -- kind (e) { { v[2], v[3] } }, -- key, entry v[4] -- optional meaning } end end end sorters.list.process(sorters.list.data[class]) end end Wolfgang

2007/12/4, Peter Münster :

On Tue, Dec 04, 2007 at 04:57:51PM +0100, Wolfgang Schuster wrote:

...

...

...

this is a bug in the mkiv sorting macros, I posted a similiar message a few days ago but got no answer.

Hello Wolfgang,

Indeed, now I remember... (But one long week was time enough for me to forget it.)

...

...

\def\mksavesortedlistentry#1#2#3#4% class key entry meaning

{\immediatewriteutilitytua{ti(js['#1'],{'e','#2',\!!bs#3\!!es,\!!bs#4\!!es})}}

seems to work

this line should be

\def\mksavesortedlistentry#1#2#3#4% class key entry meaning {\immediatewriteutilitytua{table.insert(js['#1'],{'e','#1',\!!bs#2\!!es,\!!bs#3\!!es})}}

Both versions give wrong output with optional arguments. But the following version works for me:

I forgot to mention you have to regenrate the format with my solution, your seems to work but I found a way to stop processing because you the optinal argument can also be used to write a mcro in the first pair of braces and this can produce problems. You can try to add the following synonym and TeX give you a error message. \test[zB]{z.\thinspace B.}{zum Beispiel}

% engine=luatex \unprotect \def\mksavesortedlistentry#1#2#3#4% class key entry meaning {\immediatewriteutilitytua{table.insert(js['#1'],{'e','#3',\!!bs#3\!!es,\!!bs#4\!!es})}} \protect \definesynonyms[test][tests] \test[bbb]{short text}{long description} \test[aaa]{a short text}{a long description} \starttext \bbb~\aaa \placelistoftests \stoptext

Cheers, Peter

Wolfgang

Peter Münster wrote:

On Wed, Dec 05, 2007 at 09:29:08AM +0100, Wolfgang Schuster wrote:

...

...

...

this line should be

\def\mksavesortedlistentry#1#2#3#4% class key entry meaning {\immediatewriteutilitytua{table.insert(js['#1'],{'e','#1',\!!bs#2\!!es,\!!bs#3\!!es})}} Both versions give wrong output with optional arguments. But the following version works for me: I forgot to mention you have to regenrate the format with my solution, your seems to work but I found a way to stop processing because you the optinal argument can also be used to write a mcro in the first pair of braces and this can produce problems.

I've also regenrated the format. But now I understand better: it works well with your patch for core-syn.lua and Hans' patch for core-syn.mkiv: {\immediatewriteutilitytua{table.insert(js['#1'],{'e','#2',\!!bs#3\!!es,\!!bs#4\!!es})}}

can you summarize what i need to change? ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

Wolfgang Schuster wrote:

The order of key and entry is, the same as in the mkii macro

{ { v[2], v[3] } }, -- key, entry

but you wrote

{ { v[3], v[2] } }, -- entry, key

ok, fixed just wondering ... now we move to a mix of tex and lua code, how complex is it to piece together what's happening now inside context Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------