- dev-context - NTG Mailing Lists

/MarkInfo refers to empty object
by Pablo Rodriguez 24 Oct '17

24 Oct '17

Dear list, [I accidentally discovered this issue when validating PDF/A documents with veraPDF.] I have the following sample: \nopdfcompression \setuptagging[state=start] \starttext \startTEXpage[offset=1em] \TeX \stopTEXpage \stoptext The resulting PDF document contains a catalog with the entry "/MarkInfo 17 0 R". If compiled with betas up to 2017.10.15 12:29, the PDF object reads: 17 0 obj << /Marked true >> endobj If compiled with betas starting from 2017.10.19 13:50, the PDF object is empty: 17 0 obj << >> endobj Line 153 in lpdf-tag.lua was changed from its original content in beta 2017.10.15 12:29: Marked = pdfboolean(true), to the new content in beta 2017.10.19 13:50: Marked = lpdf.majorversion == 1 and pdfboolean(true) or nil, I wonder whether this may be related to the LuaTeX version. But this is the newest version from the ConTeXt Suite for the Linux 32bit platform (in beta from 17.10.19 13:50): This is LuajitTeX, Version 1.05.0 (TeX Live 2017) Is this also an issue with more recent LuaTeX versions? Sorry for asking this in the developers’ mailing list, but checking it may be too technical for the standard mailing list. Many thanks for your help, Pablo -- http://www.ousia.tk

2 3

type-del.mkiv
by Wolfgang Schuster 23 Oct '17

23 Oct '17

Hi, there are missing \bgroup (and one \aftergroup) on line 793 and 810 in typo-del.mkiv: \def\typo_delimited_attributed {\dontleavehmode \begingroup \dostarttaggedchained\t!delimited\currentdelimitedtext\??delimitedtext \usedelimitedtextstyleandcolor\c!style\c!color \typo_delimited_start_content + \bgroup + \aftergroup\typo_delimited_attributed_e - \typo_delimited_attributed_e \let\next=} \def\typo_delimited_attributed_e {\typo_delimited_stop_content \dostoptagged \typo_delimited_pop \endgroup} \def\typo_delimited_fontdriven {\dontleavehmode \begingroup \dostarttaggedchained\t!delimited\currentdelimitedtext\??delimitedtext \usedlanguageparameter{\c!left\currentparentdelimitedtext}% was: \currentdelimitedtext \usedelimitedtextstyleandcolor\c!style\c!color \typo_delimited_start_content + \bgroup \aftergroup\typo_delimited_fontdriven_e \let\next=} \def\typo_delimited_fontdriven_e {\typo_delimited_stop_content \usedlanguageparameter{\c!right\currentparentdelimitedtext}% was: \currentdelimitedtext \dostoptagged \typo_delimited_pop \endgroup} Wolfgang

1 0

Re: [dev-context] mathpairs
by Aditya Mahajan 06 Aug '17

06 Aug '17

Hi, I tested with the latest beta and everything works well. Can you also add a virtual character for --. Then I can add the mappings for -->, <--. Others that can be mapped are ::, ::=, -: for which I am attaching a patch file. Aditya

1 0

mathpairs
by Aditya Mahajan 06 Aug '17

06 Aug '17

Hi Hans, Currently the following mathpairs are defined: := \colonequals =: \equalscolon <= \le >= \ge != \neq Are these all? (Why is != defined using specials?) Does it make sense to define others? I can think of the following: << \ll >> \gg <<< \lll >>> \ggg == \equiv In addition, we can also add arrows. Because one wants to distinguish between \Leftarrow (<=) and \le (<=), I am defining arrows using two ==. -> \rightarrow <- \leftarrow <-> \leftrightarrow ==> \Rightarrow <== \Leftarrow <=> \Leftrightarrow ===> \Longrightarrow <=== \Longleftarrow <==> \Longleftrightarrow I see this as marginally useful. Does it make sense to add support for math pairs of size three and four? If so, I am attaching a patch for char-def.lua that provides this support (as well as a test file). Everything works except ===>. The reason is that === is not mapped to anything. In the attached file, I tried mapping === (i.e., 0x2261 + 0x3D) to 0x2A76 (I gave it a mathname "eqeqeq" as well; this is consistent with latex's unicode-math package). However, this glyph is missing from LM and cambria, and === to \eqeqeq mapping doesn't work. Any idea on how to fix this or is there another approach without defining a mapping for ===. (that will also help us to define --> where I couldn't figure out what I should map -- to; I can map it to en-dash, but that just feels wrong). Thanks, Aditya PS: Currently, !=== !<= etc doesn't work (neither does \not <=, etc.). It can be fixed, but at that stage I am not sure it worth using math pairs. PPS: Monospace fonts with ligatures: https://www.hanselman.com/blog/MonospacedProgrammingFontsWithLigatures.aspx

2 3

fix for \attachment[method=hidden]
by Pablo Rodriguez 07 Jun '17

07 Jun '17

Hans, the following sample shows an issue with attachments: \setupinteraction[state=start] \setupinteractionscreen[option=attachment] \setuppapersize[A5, landscape] \setuppagenumbering[location=] \starttext \input knuth\attachment[file=e.pdf, title=Attachment \letterhash1, author=ConTeXt, subtitle=χαλεπὰ τὰ καλά, method=hidden] \stoptext method=hidden (http://pdf.ousia.tk/attachment-hidden.pdf) doesn’t actually attach the file (http://pdf.ousia.tk/attachment-hidden.png) But method=visible (http://pdf.ousia.tk/attachment-visible.pdf) attachs the file (http://pdf.ousia.tk/attachment-visible.png) I’m afraid that method=hidden was missing subtitle, file and name. I would say that buffer is also missing there, but I don’t know what it is used for. The attached .diff file fixes the issue. I have tested it and it works fine. I hope it helps, Pablo -- http://www.ousia.tk

1 1

translate Dutch sample included in grph-inc.mkiv
by Pablo Rodriguez 07 Jun '17

07 Jun '17

Hans, I have just noticed that grph-inc.mkiv includes a small sample in Dutch. I guess this isn’t intended. The attached patch fixes it. I hope it helps, Pablo -- http://www.ousia.tk

1 0

/UseAttachments and /UseOC for /PageMode
by Pablo Rodriguez 28 May '17

28 May '17

Hans, the attachment contains a patch to be able to set /PageMode to attachments or optional contents. Values are attach and oc. These were the simplest names that came to my mind. I realized these are extremely useful, since vast majority of people (including IT guys) don’t seem to know how to extract attachments from PDF documents. Could you apply it to ConTeXt? Many thanks for your help, Pablo -- http://www.ousia.tk

2 2

Is it possible to change the converted figure cache directory for a build?
by Kevin D. Robbins 18 May '17

18 May '17

By default, converted figures (m_k_i_v_foo.pdf for source figure foo.svg) are written to the same directory as the source figure. I would like all converted figures to be written instead to a cache directory in my home directory. How do I set this path? I asked this question on the ConTeXt users mailing list, but no one had an answer. Is there no user interface (environment variable, command line parameter, etc.) available to change where converted figures are written on disk? Best Regards, Kevin

1 0

Re: [dev-context] setup-en.pdf and mkiv command list
by Wolfgang Schuster 17 May '17

17 May '17

> Idris Samawi Hamid ادريس سماوي حامد <mailto:Idris.Hamid@colostate.edu> > 17. Mai 2017 um 20:21 > On Wed, 17 May 2017 08:15:11 -0600, Wolfgang Schuster > <schuster.wolfgang(a)gmail.com> wrote: > > > This is what I get from running > > context x-setups-generate.mkiv > > <cd:command file="strc-sec.mkiv" generated="yes" name="section" > variant="instance"> > <cd:sequence> > <cd:instance value="section"/> > </cd:sequence> > <cd:arguments> > <cd:keywords list="yes" optional="yes"> > <cd:constant type="cd:reference"/> > </cd:keywords> > <cd:keywords delimiters="braces"> > <cd:constant type="cd:text"/> > </cd:keywords> > </cd:arguments> > <cd:instances> > <cd:constant value="chapter"/> > <cd:constant value="section"/> > <cd:constant value="subsection"/> > <cd:constant value="subsubsection"/> > <cd:constant value="subsubsubsection"/> > <cd:constant value="subsubsubsubsection"/> > <cd:constant value="title"/> > <cd:constant value="subject"/> > <cd:constant value="subsubject"/> > <cd:constant value="subsubsubject"/> > <cd:constant value="subsubsubsubject"/> > <cd:constant value="subsubsubsubsubject"/> > </cd:instances> > </cd:command> > > This is different from what I get from the latest ConTeXt - 2017.05.15 > 21:48. > > Notably, \part is missing from the official context version; it's also > missing in the generated x-setups-generate.pdf and the published > setup-en.pdf. So there are bugs in the official version -) > > Wolfgang: When will your latest sources be ready for distribution or > sharing? Thanks in advance. I can send them in a few days. Wolfgang

1 0

Re: [dev-context] setup-en.pdf and mkiv command list
by Wolfgang Schuster 17 May '17

17 May '17

> Yu,Jason <mailto:Jason.Yu@colostate.edu> > 17. Mai 2017 um 20:27 > > I suppose not all abstract commands has <cd:instances> tag. For > example, \ORNAMENT does not have <cd:instances>, but it is still > recognized as an abstract keyword (slanted). In my python script, I > recognized every abstract keyword by determining if it has a > <cd:sequence> tag, which seems to be right. > Not all commands which let you create instances have predefined ones, e.g. there are no predefined descriptions. Wolfgang

1 0