run e-TeX rather than Knuth's TeX I
get \scantokens, etc.). The point here is that a plain luatex format
will have the new primitves, but no new macros (which only Knuth can
do), surely.
--
Joseph Wright
ust because one thing you want needs LuaTeX doesn't mean that
everything else will have been re-written as well.)
--
Joseph Wright
east for LaTeX). If I'm using a lot of big
packages (say komascript, hyperref, pgf, mhchem, ...), then most of the
code is not about to be re-written. So even if some new package uses
quite a lot of Lua (say a LuaTeX equivalent of fontspec + microtype),
most of the code I'm relying on is still TeX-based.
--
Joseph Wright
ws preferred? I would
> start with vista 32 bit and see what happens.
>
> Again, thanks for your responses.
>
> Patrick
Just a quick question. I assume there is a reason you want to build
luatex yourself, rather than use one of the binary version?
--
Joseph Wright
to do basic parsing of input at the user level, or
does it actually use Lua and just look like TeX, i.e. using Lua as a
pure scripting language to read user data into an internal normalised
format?)
--
Joseph Wright
Taco Hoekwater wrote:
> Joseph Wright wrote:
>> Hello all,
>>
>> A couple of things have come up on the LaTeX-L list (readers of that
>> list will soon spot where) that I'd like to get a feel of in a LuaTeX
>> context. Both are to do with arguments.
>
d think that this issue is a
> luatex's bug.
No, \ifpdfprimitve is meant for dealing with circumstances where an
*available* primitive has been redefined. In your test, \attribute is
not available, so of course you also cannot call the primitive.
--
Joseph Wright
main chunk.
l.27 }
?
! Emergency stop.
l.27 }
What did I get wrong?
--
Joseph Wright
On 16/02/2010 10:55, Taco Hoekwater wrote:
Joseph Wright wrote:
This is LuaTeX, Version beta-0.40.6-2009110118 (Web2C 2009)
(format.ltx
! LuaTeX error wrong number of arguments
stack traceback:
[C]: in function 'enableprimitives'
<\directlua>:1: in
[Leave D on input stream]
end
end
(Also, am I correct in thinking that there is no way to avoid needing
\unexpanded to get the string comparison "as is"?)
--
Joseph Wright
ing comparison "as is"?)
That depends on the actual strings and the current catcode assignments
of course, but \directlua expands its argument, so yes, if you want to
compare control sequence names, you will need some way to make sure
they don't expand automatically.
Just thought I should check: thanks.
--
Joseph Wright
clear! (At the moment, expl3 as a
format will fail with luaTeX as the e-TeX extensions are not around "out
of the box", which is not great.)
--
Joseph Wright
On 16/02/2010 15:43, luigi scarso wrote:
On Tue, Feb 16, 2010 at 1:14 PM, Joseph Wright wrote:
Hello all,
Thanks for the help thus far. I'm now stuck on the next step, I'm afraid.
With a file to build a format something like:
\catcode `\{ = 1 \relax
\catcode `\} = 2 \relax
\catco
x27;s only one function, I think some
awkwardness setting things up is acceptable. AS I'm looking at this from
a LaTeX-like point of view I'd like the format to be self-contained (no
additional files required for it to at least load and function "as
advertised").
--
Joseph Wright
the proper testing :-)
--
Joseph Wright
sly one to bear in mind.
--
Joseph Wright
le and potentially trash someone
else's work?
(The context here is the LaTeX3 \ExplSyntax stack. It would be nice to
take advantage of LuaTeX, when available, and exploit category code
tables. However, first I need to understand how they work.)
--
Joseph Wright
n up eventually. While the lualatex packages provide
the necessary features for now, however, I think it's sensible to call
them the "supported" way to do things.
For l3in2e I'm all for simply requiring existing packages that do the
job, as it keeps down the chance of clashes.
--
Joseph Wright
-related.)
--
Joseph Wright
applied.
The correspondence is correct, but \directlua is expandable
(\immediate\write is not).
Thanks Taco (and Philipp). I will revisit the l3luatex stuff later today
and try to tighten both the expl3 function names and description up.
--
Joseph Wright
On 03/01/2011 14:09, Joseph Wright wrote:
On 03/01/2011 13:56, Taco Hoekwater wrote:
On 01/03/2011 02:39 PM, Philipp Stephani wrote:
Reading the source code, I guess \directlua behaves identically to
\immediate\write, while \latelua corresponds to \write. If that is
true, then the naming
> l.68 \cs_generate_variant:Nn \lua_shipout_x:n
> { x }
> ?
> ) (/home/norbert/tl/2010/texmf-dist/tex/latex/tools/calc.sty))
Sorted in the CTAN upload of expl3 late yesterday. (We've rearranged
some Lua support functions, and I forgot that the luatex
itself, so the usual set up is to ave the IniTeX values to some other
table, then use a stack to switch out table 0 to the IniTeX one to 'do
stuff', then back again when the job is done. See for example Heiko
Oberdiek's 'luatex' package.
--
Joseph Wright
this a bug, or if not what's the logic for the change here?
--
Joseph Wright
, which seems to silently ignore this. I assume
this is a bug?
(I've not tested the other math primitives which are 'unaltered' here,
but the same issue may well apply.)
--
Joseph Wright
eX, which seems to silently ignore this. I assume
>> this is a bug?
>
> No, that is a feature. AFAIK, it is even documented somewhere in the
> math chapter of the reference manual.
Okay, I'd not looked in detail (as I say, this came up as an aside from
the other point). I'll bear this one in mind, then.
--
Joseph Wright
s. Is this correct, is it by design and if so have I
missed it in the manual?
--
Joseph Wright
Herbert
Issue with the ex height in the font, as noted in
http://tex.stackexchange.com/questions/47157/.
--
Joseph Wright
es. Depending on the current value of \escapechar you will see
different things here.
What you need to remember that the way LaTeX works for environments is
that when you type
\begin{foo}
it looks for a macro which could also be used directly as \foo. As you
are seeing the names, and not the control sequences, this means you need
"begin" and "foo".
--
Joseph Wright
idea is to make things
more 'human readable' in the log and so forth. It also means that these
names are 'out of the way' in namespacing terms, as they are quite
distinct from 'normal' LaTeX2e and LaTeX3 names. You see similar 'odd'
naming approaches for example with the NFSS (things like
"<5><6><7><8><9><10><10.95><12><14.4><17.28><20.74><24.88>cmcsc10" and
"OT1/cmr/m/n/7"), or with pgf (e.g. "pgfk@/pgfplots/table/header/has
colnames/.@cmd").
The thing to remember here is that these names are not really meant for
direct access: not everything in the hash table is a user command.
--
Joseph Wright
ed it is
possible for packages to (re)define the same macros. So you cannot
simply go from the document level commands (e.g. \mathbb) to the
packages which define them.
--
Joseph Wright
re might be any help there. But even if I could, I'm guessing
the code may have been modified.)
--
Joseph Wright
On 04/12/2012 09:19, Paul Isambert wrote:
> Le 04/12/2012 09:47, Joseph Wright a écrit :
>> Hello all,
>>
>> I'm currently taking a look at the approaches used for RTL typesetting
>> in pdfTeX/XeTeX/LuaTeX. In pdfTeX/XeTeX, the TeX--XeT approach is used
>&g
ir
- \mathdir
- \pagedir
- \pardir
- \textdir
based on LuaTeX manual 4.13.1.3 'Direction parameters'. However, I still
don't know how this is implemented. (e-TeX says \beginR, etc. are an
abuse of the math on/off system, so there I have some clue.)
--
Joseph Wright
the overview.
Very useful, thanks. Point about detail understood: at the moment I'm
trying to get my head around the entire area, and to see how the LuaTeX
version contrasts with the pdfTeX/XeTeX approach.
--
Joseph Wright
On 04/12/2012 14:33, Khaled Hosny wrote:
> On Tue, Dec 04, 2012 at 01:52:19PM +0000, Joseph Wright wrote:
>> Very useful, thanks. Point about detail understood: at the moment I'm
>> trying to get my head around the entire area, and to see how the LuaTeX
>> version contr
On 04/12/2012 23:49, Reinhard Kotucha wrote:
> On 2012-12-04 at 11:43:40 +0000, Joseph Wright wrote:
>
> > Tracking changes through from one engine to another is pretty
> > tricky, particularly when the docs for the older engines are not so
> > easy to find.
>
&
rrect, and if so is the expectation that numerals in math mode are
handled using a Lua callback?
--
Joseph Wright
geable person means that
> \mathdir does exactly what Aleph did. I *also* suspect it is wrong, but
> I am not going to mess with it without a person at hand who can tell me
> exactly what *should* happen.
Makes sense: perhaps one of the Arabic typesetting experts might step in
here and give an opinion :-)
--
Joseph Wright
'm still not totally sure of units!). Certainly it's
possible to avoid math mode for many things by doing a bit more macro
programming (e.g. you can vertically centre without \vcenter with a bit
of measurement).
--
Joseph Wright
. Hello world.
Hello world. Hello world. Hello world. Hello world.
\bye
--
Joseph Wright
On 06/12/2012 22:42, Joseph Wright wrote:
> Hello all,
>
> Playing around with RTL primitives, I find something a bit strange, or
> at least unexpected, with the treatment of paragraph parameters. In the
> following, "\pardir TRT" swaps the effect of \leftskip and
On 07/12/2012 07:34, Taco Hoekwater wrote:
> On 12/07/2012 08:09 AM, Joseph Wright wrote:
>> On 06/12/2012 22:42, Joseph Wright wrote:
>>> Hello all,
>>>
>>> Playing around with RTL primitives, I find something a bit strange, or
>>> at least unexpecte
don't have.
I can't imagine paragraphs in vertical typesetting, but I may be wrong.
--
Joseph Wright
viour 'expected', and more
importantly is it useful? (OK, most of the time \pardir and \textdir
will agree on direction, but my reading of the primitive names was that
they are intended to separate out the run of the text and the shape of
the paragraph.)
--
Joseph Wright
On 12/12/2012 09:03, Joseph Wright wrote:
> Hello all,
>
> Continuing my attempts to understand RTL primitives in LuaTeX, I've come
> across what I'd regard as unexpected behaviour in terms of the
> relationship between \textdir and \pardir.
>
> For example, i
the 'shape within a paragraph', in the same way that say
\rightskip is (used to alter what happens *within* any 'outer' \parshape
or \hangindent). However, it's treated as part of the text itself, not
the shaping. May well make sense to others: I'm mainly trying to work
out 'how this is supposed to go' as there is not so much detail in the
docs I can find.
--
Joseph Wright
: how other people see this.
> I find the result of this (in a latex document) much more
> disturbing:
>
> a\\b\\c\par
>
> Only the last line is on the right ;-(.
There's a reason for all of these questions :-) Hopefully the
information will make it's way into l3galley, soon-ish.
--
Joseph Wright
he paragraph,
which alters what \leftskip and \rightskip do (they 'push against
the outline'), and where \parfillskip goes (same idea).
\pagedir seems to be even more tricky: at present, I'm going with 'it
has to agree with \bodydir'.
--
Joseph Wright
..\tenrm D
..\glue 0.0 plus 1.0fill
(Can't imagine this is deliberate.)
--
Joseph Wright
p).
>
> But then, I find the direction stuff utterly confusing, so before I
> commit the change, I would really like a second opinion.
Totally agree: I've no idea either what is 'correct'. Frank asked John
Plaice if he could enlighten us (LaTeX team) on the Omega approach. I've
CC'd him on this reply: perhaps he has more insight.
--
Joseph Wright
can be used with tlmgr to
install binary updates (at your own risk):
http://tlcontrib.metatex.org/usage.html.
--
Joseph Wright
h
MiKTeX and TeX Live, so suspect it's an issue/decision at the LuaTeX
end. Is this deliberate, or is it a bug?
--
Joseph Wright
ile is not 'right' at the appropriate
place: adding a marker (undefined) control sequence just after the
\directlua line (or indeed after the plot), the final line of the data
file is
1.96078 1.1218274062469
--
Joseph Wright
there anything more definite
on it available?
--
Joseph Wright
m using luatexbase as this seems a little easier than the rather
callback.register, but if there is a solution with the latter I'll be
interested too.)
--
Joseph Wright
On 07/02/2014 09:02, Khaled Hosny wrote:
> On Fri, Feb 07, 2014 at 08:54:59AM +0000, Joseph Wright wrote:
>> what's the correct approach?
>
> Personally, I’d just work on node lists before font layout is done (so
> that they are char nodes) and use attributes to mark what
?
(It's not restricted to the micro symbol, btw.)
--
Joseph Wright
erate or a bug?
--
Joseph Wright
ng this is not deliberate: know issue?
--
Joseph Wright
nd of string
>> that kills characters behind.
>> * Negative values of \escapechar: The escape char should not
>> be printed at all.
>
>
> Both issues should be fixed in the luatex repository trunk now.
>
> Best wishes,
> Taco
Great, thanks!
--
Joseph Wright
fter\show\csname foo#\endcsname
gives
> \foo##=\relax.
\foo##
with LuaTeX but
> \foo#=\relax.
\foo#
with pdfTeX/XeTeX. The pdfTeX behaviour allows for use of 'odd' catcodes
in csnames, whereas the LuaTeX behaviour means a test would be needed.
--
Joseph Wright
and XeTeX write 79 chars then wrap, as expected from
max_print_line = 79
in texmf.cnf. (Testing with TL2014 pretest.)
Bug?
--
Joseph Wright
ot; so it's not clear that this has happened: again, is this deliberate?
P.S. Thanks for feedback on other issues.
--
Joseph Wright
ment I can live with them (all rather esoteric) :-)
--
Joseph Wright
e reason is \luaescapestring doesn't double # tokens.
\message{"\luaescapestring{#}" "\detokenize{#}"}
Is this deliberate, and if so is there a reliable way of emulating
\pdfstrcmp in it's entirety?
--
Joseph Wright
On 02/06/2014 07:24, Joseph Wright wrote:
> Hello all,
>
> The usual approach to emulating \pdfstrcmp in LuaTeX involves escaping
> input using \luaescapestring and then doing a Lua-based comparison,
> something like:
>
> \catcode`\@=11 %
> \ifdefined\directlua
>
On 02/06/2014 08:52, Taco Hoekwater wrote:
>
> On 02 Jun 2014, at 09:46, Joseph Wright
> wrote:
>
>>>
>>> Is this deliberate, and if so is there a reliable way of emulating
>>> \pdfstrcmp in it's entirety?
>
> \luaescapestring does nothing e
Hello all,
With the TL2014 pretest version of LuaTeX,
\csname\csname\endcsname
leads to the error
! Missing \endcsname inserted.
\^^@csnam\^^@endcsnam
l.1 \csname\csname\endcsname
which looks like some form of 'out by one' issue.
--
Joseph Wright
On 06/06/2014 16:53, Heiko Oberdiek wrote:
> I think, Joseph is referring to the messed up error message in LuaTeX.
Excatly: another issue that shows up doing log file comparisons.
--
Joseph Wright
call on page two "late"
vanishes entirely. The manual does not give a lot away: is there a good
explanation/set of demos somewhere?
Version data:
This is LuaTeX, Version beta-0.79.1 (TeX Live 2014) (rev 4971)
--
Joseph Wright
and indeed I'm running exactly
the same file (encoded UTF-8). The issue here is not just that the
results are different but that I get a different number of tokens: in
the real situation that's a big issue. I've not tracked down all of the
codepoints this problem applies to, but there seem to be several, so
it's not just this char.
Any clues one what is up?
--
Joseph Wright
On 04/08/2014 15:50, Joseph Wright wrote:
> Hello all,
>
> Not sure which list to raise this one on, but hopefully there will be
> enough people knowledgeable here to help!
>
> Trying to read some UTF-8 data, I find that a few chars seem to give
> unexpected results with C
up tokenization for me as I
never have two to start with).
> I haven’t checked for a while but back then it proved difficult to
> disable this preprocessing from within a document because they were
> being applied before the document was read.
I was wondering about that: whether there
s not the case with LuaTeX: no catcode logging. Is
this deliberate or a bug? If it's deliberate, what's the reasoning: it's
tricky for doing a diff against pdfTeX log.
--
Joseph Wright
stfont=select font nullfont}
\test ->cmsy10 at10.0pt
{\show}
> \testfont=select font cmsy10.
l.4 \show\testfont
Clearly the assignment happens: bug?
--
Joseph Wright
On 05/12/2014 10:48, Hans Hagen wrote:
> On 12/4/2014 10:56 PM, Joseph Wright wrote:
>
>> this deliberate or a bug? If it's deliberate, what's the reasoning: it's
>> tricky for doing a diff against pdfTeX log.
>
> btw, you cannot expect luatex to be byte-b
On 05/12/2014 10:46, Hans Hagen wrote:
> On 12/4/2014 10:56 PM, Joseph Wright wrote:
>> Hello all,
>>
>> With minimal example
>>
>> {\tracingassigns=1\begingroup\catcode`\+=11\endgroup}
>> \bye
>>
>> pdfTeX writes the expected
interaction of the
size of the mu unit with math mode style.
Is this a documented change or a bug?
--
Joseph Wright
, whilst pdfTeX and XeTeX break both. Is this a
deliberate change of behaviour?
--
Joseph Wright
ow differing outcomes.
- --
Joseph Wright
-BEGIN PGP SIGNATURE-
Version: GnuPG/MacGPG2 v2
Comment: GPGTools - https://gpgtools.org
iQEcBAEBCgAGBQJVApIYAAoJEFn7VsXH86MMPusH/jq60w6+2RKTGNM0FSv0OaZN
vDXxRdswKAfQ7cESCer82AAbAui8JJCZJrUle3yLuTN7CW0p2TUFBMGnVq4SOyXp
UHgWsssuuz744yyObTlcM+
et
>
> a
> b
> c^^Md
>
> which looks ok to me from the perspective of luatex
Isn't that a very different thing: ConTeXt has \catcode`\^=12 here (so
the above fails with plain).
--
Joseph Wright
is change deliberate?
--
Joseph Wright
On 15/04/2015 12:47, Joseph Wright wrote:
> Hello all,
>
> With example
>
> \tracingoutput=1 %
> \showboxdepth=\maxdimen
> \showboxbreadth=\maxdimen
> 1-world
> \bye
>
> and LuaTeX 0.79 I get in the log the expected
>
> ...\tenr
On 15/04/2015 16:29, Ulrike Fischer wrote:
> Am Wed, 15 Apr 2015 13:00:48 +0100 schrieb Joseph Wright:
>
>> It's been pointed out to me that classical TeX does have this, so
>> presumably it is a deliberate change :-)
>
> See also
> http://article.gmane.org/gmane
data file. This is a complex area and we have at present
done little other than copy the current logic.
--
Joseph Wright
On 06/05/2015 15:09, Jonathan Kew wrote:
> On 6/5/15 14:14, Joseph Wright wrote:
>
>> Based on the current files, we have a block to set \XeTeXcharclass,
>> which only applies to XeTeX. The logic followed in that code is that
>> characters in the file LineBreak.
o what 'GREEK CAPITAL LETTER EPSILON WITH PSILI' represents
if it's not the upper case of 'GREEK SMALL LETTER EPSILON WITH PSILI': I
notice in xgreek you map U+1F18 to U+0395 for upper casing and U+1F10
for lower casing.)
--
Joseph Wright
> them at this level.
>
> I think that unicode-letters.def makes most sense as a
> fully automated representation of the UCD data files in TeX syntax.
>
> That way everyone knows what data is in there.
>
> Individual language packages have far fewer characters to worry about
> and can over-ride
> the base settings where appropriate.
Indeed: provided hyphenation is correct then we are OK. (LuaTeX of
course is rather more flexible there than XeTeX.)
--
Joseph Wright
nt penalty.
What's the current position?
--
Joseph Wright
On 17/06/2015 15:09, Hans Hagen wrote:
> On 6/17/2015 9:59 AM, Joseph Wright wrote:
>> Hello all,
>>
>> In Heiko Oberdiek's luatex.dtx there is a note about catcode table
>> allocation
>>
>> % \item Large catcode table numbers should be avoided
esumably this is not a deliberate change?
--
Joseph Wright
- char_exists
- find_cidmap_file
- pdf_stream_filter_callback
- read_cidmap_file
What is the status of these callbacks?
--
Joseph Wright
x" and "luatex"
sets. Moreover, the 'new' versions still point to the older ones, for
example
> \primitive=\pdfprimitive.
l.1 \show\primitive
Is this deliberate?
--
Joseph Wright
On 24/06/2015 09:34, Joseph Wright wrote:
> Hello all,
>
> I'm currently looking over the list of primitive names as given by e.g.
> tex.extraprimitives("luatex"). I notice a couple of oddities.
>
> Primitive \luatexbanner is down in the "pdftex" se
On 06/10/2015 16:44, Hans Hagen wrote:
> On 10/6/2015 5:26 PM, David Carlisle wrote:
>
>> so either we need to update the format as you suggested with a \let
>> aliasing the old name, or we need to start a round of contacting authors
>> and getting updates in place for the transition.
>>
>> Either
On 06/10/2015 15:55, Hans Hagen wrote:
> % promoted (from omega)
>
> \pagedir
> \bodydir
> \pardir
> \textdir
> \mathdir
>
> % promoted (from aleph)
>
> \boxdir
>
> % promoted (from pdftex)
>
> \lpcode
> \rpcode
> \efcode
> \leftmarginkern
> \rightmarginkern
>
> \tagcode
> \letterspacefont
H
On 06/10/2015 22:50, Reinhard Kotucha wrote:
> On 2015-10-06 at 16:54:57 +0100, Joseph Wright wrote:
>
> > For TeX Live users, binary updates (essentially) only happen once
> > per year at release time, so probably something like April/May
> > before any change. (Someo
Hello all,
Trying to use e.g. tex.dimen for accessing primitives seems to lead to
very odd errors:
\directlua{print(tex.dimen["pagewidth"])}
! LuaTeX error [\directlua]:1: incorrect index value -11 for
tex.getdimen()
stack traceback:
[C]: in function '__index'
I'm assu
On 14/10/2015 01:06, Hironori KITAGAWA wrote:
> Hello,
>
> I found that the following source
>
> \documentclass{minimal}
> \usepackage{luatexbase}% 2015/10/04 v1.3
> \usepackage{expl3}% 2015/09/27 v6111
> \ExplSyntaxOn
> \xetex_suppressfontnotfounderror:D 0
> \ExplSyntaxOff
> \begin{documen
1 - 100 of 167 matches
Mail list logo
