June 2012 - dev-luatex - NTG Mailing Lists

Cron <luatex@boekplan> /var/www/luatex.org/www/bin/luatex-svn
by root＠mail.boekplan.nl 21 Jun '12

21 Jun '12

------------------------------------------------------------------------ r4448 | taco | 2012-06-21 17:07:03 +0200 (Thu, 21 Jun 2012) | 2 lines Changed paths: M /trunk/source/texk/web2c/luatexdir/pdf/pdfdest.w document odd do_dest behaviour ------------------------------------------------------------------------

1 0

Re: [Dev-luatex] \lastnodetype not working as expected in LuaTeX
by Frank Mittelbach 12 Jun '12

12 Jun '12

Hans, my fluency in the linebreak algorithm is a little bit rusty. Are you saying that there are other points and stuff in that space that is not updated or more on the page builder side of the house? Or asked differently is there more that needs updating to make my little example work, and if so could that be already done through a workaround like I did with tex.nest[tex.nest.ptr].tail = node.tail(head) or would it absolutely need fixes in the program itself? thanks frank

2 3

Getting all Symbols
by Sensei 11 Jun '12

11 Jun '12

Hi all, I've posted this question to comp.text.tex, and they've suggested to take a look at LuaTeX. So I forwarded the message to the User mailing list, but probably this one is more appropriate. If not, excuse me! I'm trying to find *at runtime* all the symbols allowed in a latex environment. So, given a tex file, I'd like to access - the complete predefined commands/symbols/definitions; - the complete list of commands/symbols/definitions defined in packages. So, the predefined language list would include \LaTeX, \def, \newcommand, ~, \usepackage, and so on. I've been informed that this is a naive view, since there are lots of other commands such as \@ and similars, and a simple match with a \ would be unrealistic (e.g., anything beginning with a \ is a command). When including a package, for example graphicx, I expect to find a way to have \includegraphics and its siblings. Again, packages may mess up a little bit. The best would be having commands along with their options, but I don't know if this is possible. One thing I'd like to avoid: creating my own parser. Why on earth should I create a parser when latex knows more than me (or everyone else) what symbols and commands it knows? The suggestions from the newsgroup is that *maybe* LuaTeX gives a way to access all TeX's macro hash table. If so, I'd be really happy. Thanks & Cheers!

2 1

show_node_list() of discretionary
by Andreas Matthias 11 Jun '12

11 Jun '12

In luatex, function show_node_list() displays the nodes of a discretionary in a different order than pdftex. Is this intended? I'm especially missing the number of nodes of the replacement (in the example: 2). Example: \showboxbreadth=\maxdimen \showboxdepth=\maxdimen \tracingonline=1 \newbox\mybox \setbox\mybox=\hbox{\discretionary{aa}{bb}{cc}} \showbox\mybox \bye Output of luatex: .\discretionary replacing ..\tenrm c ..\tenrm c ..\tenrm a ..\tenrm a .|\tenrm b .|\tenrm b Output of pdftex: .\discretionary replacing 2 ..\tenrm a ..\tenrm a .|\tenrm b .|\tenrm b .\tenrm c .\tenrm c Ciao Andreas

2 1

\lastnodetype not working as expected in LuaTeX
by Patrick Gundlach 10 Jun '12

10 Jun '12

(More from Frank Mittelbach on this subject - he is not subscribed yet but will be ASAP) I'm now fairly certain this this is a bug in the current implementation of linebreak_filter. What seems to happen is the following: when TeX builds a list (in this case a vertical list) it keeps track of it through a pointer to the head and to the tail of that list. Now when a paragraph is broken into lines those lines get appended to the current vertical list by the linebreaking algorithm and the tail pointer is updated to point to the new end of this list. However, if thelinebreak_filter replaces TeX's algorithm the material from that filter is appended to the list but the tail pointer is not changed. Thus in my example it still points to the node before the paragraph (which happens to be nil as the paragraph starts the vbox). Fortunately this tail pointer is accessible and modifiable from Lua code, so there is a way to fix this. We can't do this in linebreak_filter, however, because the current luatex code that calls the filter expects the tail in precisely this wrong as it does some final checkup and changes. But what we can do is to use the followup file post_linebreak_filter to do the work for us, because by then all these things have happened. So the workaround looks like this: function hpack_paragraph (head) h = node.hpack(head) return h end function fix_nest_tail (head) tex.nest[tex.nest.ptr].tail = node.tail(head) return true end callback.register("linebreak_filter", hpack_paragraph) callback.register("post_linebreak_filter",fix_nest_tail) The second filter simply updates TeX's tail pointer in the current semantic nest to become the tail of the result produced from the line breaking. A bit nasty, but at least it allows to use thelinebreak_filter in the current implementation. If you wonder what this this code does: without anything further it changes the output so that every paragraph is returned as a single line (and indentations like from itemize or similar environments are dropped) and inserts like footnotes or \vadjusts stay within those hboxes. So it needs a little more to become useful, but let your imagination run.

2 1

Fwd: [lltx] linebreak_filter problems
by Patrick Gundlach 10 Jun '12

10 Jun '12

(follup-up to the first mail from Frank) Von: Frank Mittelbach <frank.mittelbach(a)latex-project.org> Betreff: Re: [lltx] linebreak_filter problems Datum: 10. Juni 2012 20:49:27 MESZ An: <lualatex-dev(a)tug.org> Am 10.06.2012 16:50, schrieb Frank Mittelbach: > Hi folks, > > can somebody help me to identify if I do something wrong or if there is > a bug with linebreak_filter. > > Th problem is documented at length in > > http://tex.stackexchange.com/questions/59176/lastnodetype-not-working-as-ex… > > > In a nutshell I try to replace (at certain times) TeX's linebreaking > algorithm and replace it with a simple version that just packs up the > hlist into an hbox and returns that. > > If I do that the resulting box appears on the current vertical list (ie > is seen via \showlists) but at the same time > > \lastnotetype returns -1 > \lastbox picks up a void box > > ie to some parts of TeX the list appears to be empty. addendum: I'm not sure I fully understand the implementation yet, but it seams to me that the reason for the problem is that the tail pointer for the current list is not updated if the linebreak_filter acts, i.e., there is code to set up just_box, but in the end cur_list.tail_field remains pointing to the last node before the newly appended list rather than to its tail. In other words somewhere around here if (callback_id > 0) { /* find the correct value for the |just_box| */ halfword box_search = cur_list.tail_field; just_box = null; if (box_search != null) { do { if (type(box_search) == hlist_node) { just_box = box_search; } /* perhaps simply like this */ cur_list.tail_field = box_search /* end mod */ box_search = vlink(box_search); } while (box_search != null); } one could update the tail position. cheers frank

1 0

Fwd: [lltx] linebreak_filter problems
by Patrick Gundlach 10 Jun '12

10 Jun '12

(wrong list, forwarded) Von: Frank Mittelbach <frank.mittelbach(a)latex-project.org> Betreff: [lltx] linebreak_filter problems Datum: 10. Juni 2012 16:50:26 MESZ An: <lualatex-dev(a)tug.org> Hi folks, can somebody help me to identify if I do something wrong or if there is a bug with linebreak_filter. Th problem is documented at length in http://tex.stackexchange.com/questions/59176/lastnodetype-not-working-as-ex… In a nutshell I try to replace (at certain times) TeX's linebreaking algorithm and replace it with a simple version that just packs up the hlist into an hbox and returns that. If I do that the resulting box appears on the current vertical list (ie is seen via \showlists) but at the same time \lastnotetype returns -1 \lastbox picks up a void box ie to some parts of TeX the list appears to be empty. Here is an MWE \def\partest{% \directlua{% function hpackparagraph (head) return node.hpack(head) end callback.register("linebreak_filter", hpackparagraph) }% \par } \setbox0\vbox{ABC\par {\tracingonline1\showboxbreadth\maxdimen\showboxdepth\maxdimen\showlists}% \showthe\lastnodetype } \setbox0\vbox{ABC\partest {\tracingonline1\showboxbreadth\maxdimen\showboxdepth\maxdimen\showlists}% \showthe\lastnodetype } Any ideas what goes wrong or how I should alter things to make this possible? thanks frank

1 0

A mechanism to know which registers are already allocated?
by Paul Isambert 10 Jun '12

10 Jun '12

Hello all, When you write something that isn't tied to any format, the successful allocation of registers depends on the same macros being defined in all formats (e.g. \newcount). (Admittedly, you can do very well without registers by using Lua variable, but only for counts and dimens.) Unfortunately, for attributes, you either have \newattribute in ConTeXt (as far as I can tell from the sources), while in plain and LaTeX you have \newluatexattribute iff you're willing to load luatexbase-attr (which I'm not, although I appreciate the effort). In other words, there is no ``tradition'' here on which one could rely. Now, could it be possible to implement e.g. a "tex.counts" table, whose entries would be \countdef'ed names and the values would be the number of the associated registers? Then one could easily reserve a new register on the fly without having to rely on the same macros, e.g.: \def\FindFreeRegister#1{% \directlua{% local t = {} for _, c in pairs(tex["#1"]) do t[c] = true end for i=22, 2^16-1 do if not t[i] then tex.print(i) break end end}% } \countdef\mycount=\FindFreeRegister{counts} Actually, any other way to make sure a register isn't associated with a command would do; the important thing is that it is implemented in the engine, not in the formats. Best, Paul

1 0

Cron <luatex@boekplan> /var/www/luatex.org/www/bin/luatex-svn
by root＠mail.boekplan.nl 01 Jun '12

01 Jun '12

------------------------------------------------------------------------ r4446 | hhenkel | 2012-05-31 23:15:37 +0200 (Thu, 31 May 2012) | 3 lines Changed paths: M /trunk/source/texk/web2c/luatexdir/luapeg/lpeg.c D /trunk/source/texk/web2c/luatexdir/luapeg/lpeg.h lpeg 0.10.2 (r4433) doesn't yet harmonize with luatex, reverted for now to lpeg 0.9 as it was in use up to r4432. ------------------------------------------------------------------------ r4447 | hhenkel | 2012-05-31 23:33:52 +0200 (Thu, 31 May 2012) | 2 lines Changed paths: M /trunk/source/texk/web2c/Makefile.in M /trunk/source/texk/web2c/luatexdir/am/luamisc.am rm lpeg.h from luamisc.am, ./reautoconf ------------------------------------------------------------------------

1 0