ligature exceptions

ligature exceptions

[Thomas A. Schmitz via ntg-context]

Hi all,

just two quick questions: a couple of weeks ago, there was a long 
discussion on ligature exceptions in, e.g., German. Hans provided an 
elegant new mechanism (\startlanguageoptions). So my two questions:

1. Is the old mechanism (\replaceword[eg][Auflage][Au{fl}age]) now 
obsolete? I tried using it, but unwanted ligatures still occurred. If it 
is indeed deprecated, I'll add a note to the wiki (this would mean that 
the current texlive version does not have the new mechanism and the 
latest lmtx doesn't use the old one, so that's somewhat problematic).

2. Both mechanisms only work for complete words, is that right? So I 
can't just provide a pattern such as "uf|forder," I have to provide a 
full list with all inflected forms:

auf|fordern
auf|fordert
auf|fordernd
Auf|forderung
Auf|forderungen

and for 1-3 also the forms with a capital letter.

Thank you!

Thomas
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Hans Hagen via ntg-context]

[-- Attachment #1: Type: text/plain, Size: 2472 bytes --]

On 7/28/2021 7:24 PM, Thomas A. Schmitz via ntg-context wrote:
> Hi all,
> 
> just two quick questions: a couple of weeks ago, there was a long 
> discussion on ligature exceptions in, e.g., German. Hans provided an 
> elegant new mechanism (\startlanguageoptions). So my two questions:
> 
> 1. Is the old mechanism (\replaceword[eg][Auflage][Au{fl}age]) now 
> obsolete? I tried using it, but unwanted ligatures still occurred. If it 
> is indeed deprecated, I'll add a note to the wiki (this would mean that 
> the current texlive version does not have the new mechanism and the 
> latest lmtx doesn't use the old one, so that's somewhat problematic).

it's not too hard to make it work, see attached, but the new one is 
nicer because it permits more detailed control

in luametatex you can do things like

Au{\norightligaturing f}lage

that is: each character can have a set of options; ok, one can do that 
with attributes but that is less efficients options have been introduced 
for ligatures, kerns, expansion etc (you can diff the attached lmt file 
with the lua file of the same name to see the difference in approach

> 2. Both mechanisms only work for complete words, is that right? So I 
> can't just provide a pattern such as "uf|forder," I have to provide a 
> full list with all inflected forms:
> 
> auf|fordern
> auf|fordert
> auf|fordernd
> Auf|forderung
> Auf|forderungen
> 
> and for 1-3 also the forms with a capital letter.

just lowercase will do

if you look at the example files you can see that there are pre/post 
snippets possibel so basically you can create something

pre [a|b|c] post

which saves some definitions (make we can have pre/post lists which will 
nicely explode into huge lists but memory is not the issue here)

while the replacement works by running over the node list, the second 
mechanism is using a callback at the hyphenation level (so per language) 
where each word is anyway checked against an exception list, so in 
addition that word can now be fed into a function and depending on 
outcome be fed back into the machinery

(there are trackers that give insight in the process)

Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------

[-- Attachment #2: lang-rep.lmt --]
[-- Type: text/plain, Size: 16964 bytes --]

if not modules then modules = { } end modules ['lang-rep'] = {
    version   = 1.001,
    comment   = "companion to lang-rep.mkiv",
    author    = "Hans Hagen, PRAGMA-ADE, Hasselt NL",
    copyright = "PRAGMA ADE / ConTeXt Development Team",
    license   = "see context related readme files"
}

-- A BachoTeX 2013 experiment, probably not that useful. Eventually I used a simpler
-- more generic example. I'm sure no one ever notices of even needs this code.
--
-- As a follow up on a question by Alan about special treatment of dropped caps I wonder
-- if I can make this one more clever (probably in a few more dev steps). For instance
-- injecting nodes or replacing nodes. It's a prelude to a kind of lpeg for nodes,
-- although (given experiences so far) we don't really need that. After all, each problem
-- is somewhat unique.

local type, tonumber, next = type, tonumber, next
local gmatch, gsub = string.gmatch, string.gsub
local utfbyte, utfsplit = utf.byte, utf.split
local P, C, U, Cc, Ct, Cs, lpegmatch = lpeg.P, lpeg.C, lpeg.patterns.utf8character, lpeg.Cc, lpeg.Ct, lpeg.Cs, lpeg.match
local find = string.find

local zwnj     =  0x200C
local grouped  = P("{") * ( Ct((U/utfbyte-P("}"))^1) + Cc(false) ) * P("}")-- grouped
local splitter = Ct((
                    #P("{") * (
                        P("{}") / function() return zwnj end
                      + Ct(Cc("discretionary") * grouped * grouped * grouped)
                      + Ct(Cc("noligature")    * grouped)
                    )
                  + U/utfbyte
                )^1)

local stripper = P("{") * Cs((1-P(-2))^0) * P("}") * P(-1)

local trace_replacements = false  trackers.register("languages.replacements",         function(v) trace_replacements = v end)
local trace_details      = false  trackers.register("languages.replacements.details", function(v) trace_details      = v end)

local report_replacement = logs.reporter("languages","replacements")

local glyph_code         = nodes.nodecodes.glyph
local glue_code          = nodes.nodecodes.glue

local spaceskip_code     = nodes.gluecodes.spaceskip
local xspaceskip_code    = nodes.gluecodes.xspaceskip

local nuts               = nodes.nuts

local getnext            = nuts.getnext
local getprev            = nuts.getprev
local getattr            = nuts.getattr
local getid              = nuts.getid
local getsubtype         = nuts.getsubtype
local getchar            = nuts.getchar
local isglyph            = nuts.isglyph

local setlink            = nuts.setlink
local setnext            = nuts.setnext
local setprev            = nuts.setprev
local setchar            = nuts.setchar
local setattrlist        = nuts.setattrlist
local setoptions         = nuts.setoptions

local glyphoptioncodes       = tex.glyphoptioncodes
local norightligature_option = glyphoptioncodes.norightligature
local noleftligature_option  = glyphoptioncodes.noleftligature

local insertbefore       = nuts.insertbefore
local insertafter        = nuts.insertafter
local remove_node        = nuts.remove
local copy_node          = nuts.copy
local flushlist          = nuts.flushlist

local nodepool           = nuts.pool
local new_disc           = nodepool.disc

local texsetattribute    = tex.setattribute
local unsetvalue         = attributes.unsetvalue

local enableaction       = nodes.tasks.enableaction

local v_reset            = interfaces.variables.reset

local implement          = interfaces.implement

local processors         = typesetters.processors
local splitprocessor     = processors.split

local replacements       = languages.replacements or { }
languages.replacements   = replacements

local a_replacements     = attributes.private("replacements")

local lists = { }
local last  = 0
local trees = { }

table.setmetatableindex(lists,function(lists,name)
    last = last + 1
    local list = { }
    local data = { name = name, list = list, attribute = last }
    lists[last] = data
    lists[name] = data
    trees[last] = list
    return data
end)

lists[v_reset].attribute = unsetvalue -- so we discard 0

-- todo: glue kern attr

local function add(root,word,replacement)
    local processor, replacement = splitprocessor(replacement,true) -- no check
    replacement = lpegmatch(stripper,replacement) or replacement
    local list = utfsplit(word) -- ,true)
    local size = #list
    for i=1,size do
        local l = utfbyte(list[i])
        if not root[l] then
            root[l] = { }
        end
        if i == size then
            local special = find(replacement,"{",1,true)
            local newlist = lpegmatch(splitter,replacement)
            root[l].final = {
                word        = word,
                replacement = replacement,
                processor   = processor,
                oldlength   = size,
                newcodes    = newlist,
                special     = special,
            }
        end
        root = root[l]
    end
end

function replacements.add(category,word,replacement)
    local root = lists[category].list
    if type(word) == "table" then
        for word, replacement in next, word do
            add(root,word,replacement)
        end
    else
        add(root,word,replacement or "")
    end
end

-- local strip = lpeg.stripper("{}")

function languages.replacements.addlist(category,list)
    local root = lists[category].list
    if type(list) == "string" then
        for new in gmatch(list,"%S+") do
            local old = gsub(new,"[{}]","")
         -- local old = lpegmatch(strip,new)
            add(root,old,new)
        end
    else
        for i=1,#list do
            local new = list[i]
            local old = gsub(new,"[{}]","")
         -- local old = lpegmatch(strip,new)
            add(root,old,new)
        end
    end
end

local function tonodes(list,template)
    local head, current
    for i=1,#list do
        local new = copy_node(template)
        setchar(new,list[i])
        if head then
            head, current = insertafter(head,current,new)
        else
            head, current = new, new
        end
    end
    return head
end

local is_punctuation = characters.is_punctuation

-- We can try to be clever and use the fact that there is no match to skip
-- over to the next word but it is gives fuzzy code so for now I removed
-- that optimization (when I really need a high performance version myself
-- I will look into it (but so far I never used this mechanism myself).
--
-- We used to have the hit checker as function but is got messy when checks
-- for punctuation was added.

local function replace(head,first,last,final,hasspace,overload)
    local current   = first
    local prefirst  = getprev(first) or head
    local postlast  = getnext(last)
    local oldlength = final.oldlength
    local newcodes  = final.newcodes
    local newlength = newcodes and #newcodes or 0
    if trace_replacements then
        report_replacement("replacing word %a by %a",final.word,final.replacement)
    end
    if hasspace or final.special then
        -- It's easier to delete and insert so we do just that. On the todo list is
        -- turn injected spaces into glue but easier might be to let the char break
        -- handler do that ...
        local prev = getprev(current)
        local next = getnext(last)
        local list = current
        setnext(last)
        setlink(prev,next)
        current = prev
        if not current then
            head = nil
        end
        local i = 1
        while i <= newlength do
            local codes = newcodes[i]
            if type(codes) == "table" then
                local method = codes[1]
                if method == "discretionary" then
                    local pre, post, replace = codes[2], codes[3], codes[4]
                    if pre then
                        pre = tonodes(pre,first)
                    end
                    if post then
                        post = tonodes(post,first)
                    end
                    if replace then
                        replace = tonodes(replace,first)
                    end
                    -- todo: also set attr
                    local new = new_disc(pre,post,replace)
                    setattrlist(new,first)
                    head, current = insertafter(head,current,new)
                elseif method == "noligature" then
                    -- not that efficient to copy but ok for testing
                    local list = codes[2]
                    if list then
                        local n = #list
                        for i=1,n do
                            local new = copy_node(first)
                            setchar(new,list[i])
                            if i == 1 then
                                setoptions(new,norightligature_option)
                            elseif i == n then
                                setoptions(new,glyphoptioncodes.noleftligature | norightligature_option)
                            else
                                setoptions(new,glyphoptioncodes.noleftligature)
                            end
                            head, current = insertafter(head,current,new)
                        end
                    else
                     -- local new = copy_node(first)
                     -- setchar(new,zwnj)
                     -- head, current = insertafter(head,current,new)
                        setoptions(current,norightligature_option)
                    end
                else
                    report_replacement("unknown method %a",method or "?")
                end
            else
                local new = copy_node(first)
                setchar(new,codes)
                head, current = insertafter(head,current,new)
            end
            i = i + 1
        end
        flushlist(list)
    elseif newlength == 0 then
        -- we overload
    elseif oldlength == newlength then
        if final.word ~= final.replacement then
            for i=1,newlength do
                setchar(current,newcodes[i])
                current = getnext(current)
            end
        end
        current = getnext(final)
    elseif oldlength < newlength then
        for i=1,newlength-oldlength do
            local n = copy_node(current)
            setchar(n,newcodes[i])
            head, current = insertbefore(head,current,n)
            current = getnext(current)
        end
        for i=newlength-oldlength+1,newlength do
            setchar(current,newcodes[i])
            current = getnext(current)
        end
    else
        for i=1,oldlength-newlength do
            head, current = remove_node(head,current,true)
        end
        for i=1,newlength do
            setchar(current,newcodes[i])
            current = getnext(current)
        end
    end
    if overload then
        overload(final,getnext(prefirst),getprev(postlast))
    end
    return head, postlast
end

-- we handle just one space

function replacements.handler(head)
    local current   = head
    local overload  = attributes.applyoverloads
    local mode      = false -- we're in word or punctuation mode
    local wordstart = false
    local wordend   = false
    local prevend   = false
    local prevfinal = false
    local tree      = false
    local root      = false
    local hasspace  = false
    while current do
        local id = getid(current) -- or use the char getter
        if id == glyph_code then
            local a = getattr(current,a_replacements)
            if a then
                -- we have a run
                tree = trees[a]
                if tree then
                    local char = getchar(current)
                    local punc = is_punctuation[char]
                    if mode == "punc" then
                        if not punc then
                            if root then
                                local final = root.final
                                if final then
                                    head = replace(head,wordstart,wordend,final,hasspace,overload)
                                elseif prevfinal then
                                    head = replace(head,wordstart,prevend,prevfinal,hasspace,overload)
                                end
                                prevfinal = false
                                root = false
                            end
                            mode = "word"
                        end
                    elseif mode == "word" then
                        if punc then
                            if root then
                                local final = root.final
                                if final then
                                    head = replace(head,wordstart,wordend,final,hasspace,overload)
                                elseif prevfinal then
                                    head = replace(head,wordstart,prevend,prevfinal,hasspace,overload)
                                end
                                prevfinal = false
                                root = false
                            end
                            mode = "punc"
                        end
                    else
                        mode = punc and "punc" or "word"
                    end
                    if root then
                        root = root[char]
                        if root then
                            wordend = current
                        end
                    else
                        if prevfinal then
                            head = replace(head,wordstart,prevend,prevfinal,hasspace,overload)
                            prevfinal = false
                        end
                        root = tree[char]
                        if root then
                            wordstart = current
                            wordend   = current
                            prevend   = false
                            hasspace  = false
                        end
                    end
                else
                    root= false
                end
            else
                tree = false
            end
            current = getnext(current)
        elseif root then
            local final = root.final
            if mode == "word" and id == glue_code then
                local s = getsubtype(current)
                if s == spaceskip_code or s == xspaceskip_code then
                    local r = root[32] -- maybe more types
                    if r then
                        if not prevend then
                            local f = root.final
                            if f then
                                prevend   = wordend
                                prevfinal = f
                            end
                        end
                        wordend  = current
                        root     = r
                        hasspace = true
                        goto moveon
                    end
                end
            end
            if final then
                head, current = replace(head,wordstart,wordend,final,hasspace,overload)
            elseif prevfinal then
                head, current = replace(head,wordstart,prevend,prevfinal,hasspace,overload)
            end
            prevfinal = false
            root = false
          ::moveon::
            current = getnext(current)
        else
            current = getnext(current)
        end
    end
    if root then
        local final = root.final
        if final then
            head = replace(head,wordstart,wordend,final,hasspace,overload)
        elseif prevfinal then
            head = replace(head,wordstart,prevend,prevfinal,hasspace,overload)
        end
    end
    return head
end

local enabled = false

function replacements.set(n)
    if n == v_reset then
        n = unsetvalue
    else
        n = lists[n].attribute
        if not enabled then
            enableaction("processors","languages.replacements.handler")
            if trace_replacements then
                report_replacement("enabling replacement handler")
            end
            enabled = true
        end
    end
    texsetattribute(a_replacements,n)
end

-- interface

implement {
    name      = "setreplacements",
    actions   = replacements.set,
    arguments = "string"
}

implement {
    name      = "addreplacements",
    actions   = replacements.add,
    arguments = "3 strings",
}

implement {
    name      = "addreplacementslist",
    actions   = replacements.addlist,
    arguments = "2 strings",
}

[-- Attachment #3: lang-rep.mkxl --]
[-- Type: text/plain, Size: 2278 bytes --]

%D \module
%D   [       file=lang-rep,
%D        version=2013.04.28,
%D          title=\CONTEXT\ Language Macros,
%D       subtitle=Substitution,
%D         author=Hans Hagen,
%D           date=\currentdate,
%D      copyright={PRAGMA ADE \& \CONTEXT\ Development Team}]
%C
%C This module is part of the \CONTEXT\ macro||package and is
%C therefore copyrighted by \PRAGMA. See mreadme.pdf for
%C details.

%D As I needed an example of messing with nodes for the bacho\TEX\ tutorial
%D I cooked up this. In the end I decided to stick to a simpler example and
%D just finished this off in case someone really needs it.

\writestatus{loading}{ConTeXt Language Macros / Replacements}

\unprotect

\registerctxluafile{lang-rep}{autosuffix}

\definesystemattribute[replacements][public,global]

%D \startluacode
%D
%D     -- todo: other nodes (prelude to more experiments with auto dropped caps)
%D
%D     languages.replacements.add("basics", {
%D         ["aap"]   = "monkey",
%D         ["noot"]  = "nut",
%D         ["never"] = "forever",
%D         ["newer"] = "cooler",
%D         ["new"]   = "cool",
%D      -- ["special"] = "veryspe{>>>}{<<<}{=}cial",
%D     })
%D
%D \stopluacode
%D
%D \replaceword[more][this][that]
%D \replaceword[more][crap][support]
%D \replaceword[more][---][—]
%D \replaceword[basics][special][veryspe{>>>}{<<<}{=}cial]
%D
%D \starttyping
%D \start \setreplacements[basics] What the heck, it's now or never, isn't it new? \par \stop
%D \start \setreplacements[more]   Do we --- {\it really} --- need this kind of crap? \par \stop
%D \start \setreplacements[basics] All kinds of special thingies! \par \stop
%D \start \setreplacements[basics] \hsize1mm special \par \stop
%D \stoptyping

\permanent\protected\def\setreplacements[#1]%
  {\clf_setreplacements{#1}}

\permanent\protected\def\resetreplacements
  {\c_attr_replacements\attributeunsetvalue}

\permanent\tolerant\protected\def\replaceword[#1]#*[#2]#*[#3]%
  {\ifarguments\or\or
     \clf_addreplacementslist{#1}{#2}%
   \or
     \clf_addreplacements{#1}{#2}{#3}%
   \fi}

\appendtoks
    \resetreplacements
\to \everyresettypesetting

\appendtoks
    \resetreplacements
\to \everyinitializeverbatim

\protect \endinput

[-- Attachment #4: Type: text/plain, Size: 493 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Pablo Rodriguez]

On 10/08/2017 02:54 PM, Henning Hraban Ramm wrote:
> Am 2017-10-08 um 10:37 schrieb Pablo Rodriguez:
>> On 10/08/2017 12:22 AM, Hans Hagen wrote:
>> Many thanks for your reply, Hans.
>>
>> I almost get it all ;-), but what is the "i" variable?
> 
> It defines the kerning between f and i. I also added "f" and "b".

Many thanks for the replies, both Hans and Hraban.

I didn’t notice any change in my sample because I only tested the fl pair.

Many thanks for your help,

Pablo
-- 
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

Am 2017-10-08 um 10:37 schrieb Pablo Rodriguez <oinos@gmx.es>:

> On 10/08/2017 12:22 AM, Hans Hagen wrote:
>> You need to define features before fonts get defined because we share 
>> font definitions and their features as much as possible
>> 
>> \startluacode
>>     fonts.handlers.otf.addfeature {
>>         name = "kernligatures",
>>         type = "kern",
>>         data = {
>>             f = { i = 50, l = 50 },
>>         }
>>     }
>> \stopluacode
> 
> Many thanks for your reply, Hans.
> 
> I almost get it all ;-), but what is the "i" variable?

It defines the kerning between f and i. I also added "f" and "b".

And after playing with the values and setting up exceptions I understand now why Hans hates the Alegreya font(s):
While the kerning works, the font still uses the long-arced f (that is meant for creating ligatures).

I still like the design, but will consider different fonts for upcoming books.


Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

Am 2017-10-08 um 00:22 schrieb Hans Hagen <pragma@wxs.nl>:

> On 10/7/2017 9:21 PM, Pablo Rodriguez wrote:
> You need to define features before fonts get defined because we share font definitions and their features as much as possible
> 
> \startluacode
>    fonts.handlers.otf.addfeature {
>        name = "kernligatures",
>        type = "kern",
>        data = {
>            f = { i = 50, l = 50 },
>        }
>    }
> \stopluacode
> 
> \blockligatures[u:fl:a]
> 
> Now we can use them
> 
> \definefontfamily[mainfont][rm][Brill]
> \definefontfamily[mainfont][tt][Dejavu]
> 
> \setupbodyfont[mainfont]
> 
> \definefontfeature[default:b][default][blockligatures=yes]
> \definefontfeature[default:k][default][blockligatures=yes,kernligatures=yes]

Thank you very much, that helps! I’m working on a set of exceptions...



Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

Am 2017-10-08 um 00:14 schrieb Hans Hagen <pragma@wxs.nl>:

> On 10/7/2017 9:21 PM, Pablo Rodriguez wrote:
>> On 10/07/2017 01:25 PM, Henning Hraban Ramm wrote:
>>> I rejoiced too early:
>>> While the ligature exception handling works with some of the default
>>> fonts (lm, termes, pagella), it fails with the Alegreya fonts that I’m
>>> using.
> 
> forget about Algreya as it's a useless font ... it has no kerns, no ligatures, no features at all so the f an whatever follows will always touch or overlap due to the design

I can’t agree - it’s beautifully designed, and I can’t complain about its kerning. It also responds to tlig or liga features, even if they’re implemented as single (i.e. not ligature) glyphs (there are e.g. different f glyphs). Maybe it’s set up in a strange way, but I don’t think it’s that bad. Of course I don’t have your level of insight.

My problem with Alegreya’s *design* it just that it’s not very well suited for digital printing, it needs a high resolution to look good.

Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Pablo Rodriguez]

On 10/08/2017 12:22 AM, Hans Hagen wrote:
> You need to define features before fonts get defined because we share 
> font definitions and their features as much as possible
> 
> \startluacode
>      fonts.handlers.otf.addfeature {
>          name = "kernligatures",
>          type = "kern",
>          data = {
>              f = { i = 50, l = 50 },
>          }
>      }
> \stopluacode

Many thanks for your reply, Hans.

I almost get it all ;-), but what is the "i" variable?

Setting different values or even removing doesn’t make any difference
(at least, that I notice).

Many thanks for your help again,

Pablo
-- 
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Pablo Rodriguez]

On 10/08/2017 12:14 AM, Hans Hagen wrote:
> On 10/7/2017 9:21 PM, Pablo Rodriguez wrote:
> [...]
> forget about Algreya as it's a useless font ... it has no kerns, no 
> ligatures, no features at all so the f an whatever follows will always 
> touch or overlap due to the design

In fact, the full version is a commercial typeface:
https://www.myfonts.com/fonts/huertatipografica/alegreya-ht-pro/.

I wonder whether it makes sense to have a free font that only contains
the shapes and just nothing more.

Pablo
-- 
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Hans Hagen]

On 10/7/2017 9:21 PM, Pablo Rodriguez wrote:

> The Brill fonts don’t work either (http://www.brill.com/about/brill-fonts).
> 
>      \definefontfamily[mainfont][rm][Alegreya]
>      \definefontfamily[otherfont][rm][Brill]
>      \setupbodyfont [mainfont,16pt]
> 
>      \starttext
>      \startTEXpage[offset=1em]
>      \startbuffer
>      falsch: Auflage Kampffront
> 
>      richtig: Au\noligature{fl}age Kamp\noligature{ff}ront
> 
>      \setreplacements[eka]
>      richtig: Auflage Kampffront
> 
>      f ft fl fk fb ff
> 
>      f f\/t f\/l f\/k f\/b f\/f
>      \resetreplacements
>      \stopbuffer
> 
>      \getbuffer
> 
>      \setupbodyfont[times]\getbuffer
> 
>      \setupbodyfont[otherfont]\getbuffer
>      \stopTEXpage
>      \stoptext
You need to define features before fonts get defined because we share 
font definitions and their features as much as possible

\startluacode
     fonts.handlers.otf.addfeature {
         name = "kernligatures",
         type = "kern",
         data = {
             f = { i = 50, l = 50 },
         }
     }
\stopluacode

\blockligatures[u:fl:a]

Now we can use them

\definefontfamily[mainfont][rm][Brill]
\definefontfamily[mainfont][tt][Dejavu]

\setupbodyfont[mainfont]

\definefontfeature[default:b][default][blockligatures=yes]
\definefontfeature[default:k][default][blockligatures=yes,kernligatures=yes]

\showfontkerns

\starttext

{\definedfont[Serif*default   @ 11pt]auflage}\par
{\definedfont[Serif*default:b @ 11pt]auflage}\par
{\definedfont[Serif*default:k @ 11pt]auflage}\par

\stoptext

so, you need to add a kern too because brill assumes ligatures to be applied

as usual with tex .. there's always a way out but not always a trivial 
one and of course all this gets forgotten and/or lost in the flood of 
solutions

Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Hans Hagen]

On 10/7/2017 9:21 PM, Pablo Rodriguez wrote:
> On 10/07/2017 01:25 PM, Henning Hraban Ramm wrote:
>> I rejoiced too early:
>> While the ligature exception handling works with some of the default
>> fonts (lm, termes, pagella), it fails with the Alegreya fonts that I’m
>> using.

forget about Algreya as it's a useless font ... it has no kerns, no 
ligatures, no features at all so the f an whatever follows will always 
touch or overlap due to the design

Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

Am 2017-10-07 um 21:21 schrieb Pablo Rodriguez <oinos@gmx.es>:

> Both Alegreya and Brill don’t replace the ligature with a single glyph,
> but replace the first glyph (only, as far as I know).
> 
> I don’t know which is the way to solve it, but this might be the cause.

Oh, yes. Thanks for the analysis. I get the startligature-f also on hyphenation, like in auf-fällig, so that the arc of the first f (that is supposed to touch the second) reaches over the hyphen.


Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Pablo Rodriguez]

On 10/07/2017 01:25 PM, Henning Hraban Ramm wrote:
> I rejoiced too early:
> While the ligature exception handling works with some of the default
> fonts (lm, termes, pagella), it fails with the Alegreya fonts that I’m
> using.

The Brill fonts don’t work either (http://www.brill.com/about/brill-fonts).

    \definefontfamily[mainfont][rm][Alegreya]
    \definefontfamily[otherfont][rm][Brill]
    \setupbodyfont [mainfont,16pt]

    \starttext
    \startTEXpage[offset=1em]
    \startbuffer
    falsch: Auflage Kampffront

    richtig: Au\noligature{fl}age Kamp\noligature{ff}ront

    \setreplacements[eka]
    richtig: Auflage Kampffront

    f ft fl fk fb ff

    f f\/t f\/l f\/k f\/b f\/f
    \resetreplacements
    \stopbuffer

    \getbuffer

    \setupbodyfont[times]\getbuffer

    \setupbodyfont[otherfont]\getbuffer
    \stopTEXpage
    \stoptext

Both Alegreya and Brill don’t replace the ligature with a single glyph,
but replace the first glyph (only, as far as I know).

I don’t know which is the way to solve it, but this might be the cause.

Just in case it helps,

Pablo
-- 
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

[-- Attachment #1: Type: text/plain, Size: 1064 bytes --]

I rejoiced too early:
While the ligature exception handling works with some of the default fonts (lm, termes, pagella), it fails with the Alegreya fonts that I’m using.
I’d be glad if someone (you?) could look into my attached typescript file or even try downloading the fonts* to check what’s going on.

*) https://fonts.google.com/?selection.family=Alegreya|Alegreya+SC|Alegreya+Sans|Alegreya+Sans+SC

\mainlanguage[de]

%\usetypescriptfile	[type-alegreya]
%\usetypescript	[alegreya-light]
%\setupbodyfont	[alegreya,rm,16pt]
\setupbodyfont	[termes,rm,16pt]

\definefontfeature[default]
[mode=node,kern=yes,
liga=yes,tlig=yes,
ccmp=yes,language=dflt,
protrusion=quality,
expansion=quality]

\replaceword[eka][Auflage][Au{fl}age]
\replaceword[eka][Kampffront][Kamp{ff}ront]

\starttext
falsch: Auflage Kampffront

richtig: Au\noligature{fl}age Kamp\noligature{ff}ront

\setreplacements[eka]
richtig: Auflage Kampffront
\stoptext



Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

[-- Attachment #2: type-alegreya.tex --]
[-- Type: application/octet-stream, Size: 8005 bytes --]

\definefontfeature
[alefeatures]
[mode=node,kern=yes,
  tlig=yes,liga=yes,
  ccmp=yes,language=dflt,
  protrusion=quality,
  expansion=quality]

\starttypescript [sans] [alegreya,alegreya-light,alegreya-medium]
  %\setups[font:fallback:serif]
  \definefontsynonym[AlegreyaSans-Thin]             [name:AlegreyaSans-Thin]
  \definefontsynonym[AlegreyaSans-ThinItalic]       [name:AlegreyaSans-ThinItalic]
  \definefontsynonym[AlegreyaSans-ThinCaps]         [name:AlegreyaSansSC-Thin]
  \definefontsynonym[AlegreyaSans-ThinItalicCaps]   [name:AlegreyaSansSC-ThinItalic]
  \definefontsynonym[AlegreyaSans-Light]            [name:AlegreyaSans-Light]
  \definefontsynonym[AlegreyaSans-LightItalic]      [name:AlegreyaSans-LightItalic]
  \definefontsynonym[AlegreyaSans-LightCaps]        [name:AlegreyaSansSC-Light]
  \definefontsynonym[AlegreyaSans-LightItalicCaps]  [name:AlegreyaSansSC-LightItalic]
  \definefontsynonym[AlegreyaSans-Regular]          [name:AlegreyaSans-Regular]
  \definefontsynonym[AlegreyaSans-Italic]           [name:AlegreyaSans-Italic]
  \definefontsynonym[AlegreyaSans-Caps]             [name:AlegreyaSansSC-Regular]
  \definefontsynonym[AlegreyaSans-ItalicCaps]       [name:AlegreyaSansSC-Italic]
  \definefontsynonym[AlegreyaSans-Medium]           [name:AlegreyaSans-Medium]
  \definefontsynonym[AlegreyaSans-MediumItalic]     [name:AlegreyaSans-MediumItalic]
  \definefontsynonym[AlegreyaSans-MediumCaps]       [name:AlegreyaSansSC-Medium]
  \definefontsynonym[AlegreyaSans-MediumItalicCaps] [name:AlegreyaSansSC-MediumItalic]
  \definefontsynonym[AlegreyaSans-Bold]             [name:AlegreyaSans-Bold]
  \definefontsynonym[AlegreyaSans-BoldItalic]       [name:AlegreyaSans-BoldItalic]
  \definefontsynonym[AlegreyaSans-BoldCaps]         [name:AlegreyaSansSC-Bold]
  \definefontsynonym[AlegreyaSans-BoldItalicCaps]   [name:AlegreyaSansSC-BoldItalic]
  \definefontsynonym[AlegreyaSans-ExtraBold]        [name:AlegreyaSans-ExtraBold]
  \definefontsynonym[AlegreyaSans-ExtraBoldItalic]  [name:AlegreyaSans-ExtraBoldItalic]
  \definefontsynonym[AlegreyaSans-ExtraBoldCaps]    [name:AlegreyaSansSC-ExtraBold]
  \definefontsynonym[AlegreyaSans-ExtraBoldItalicCaps][name:AlegreyaSansSC-ExtraBoldItalic]
  \definefontsynonym[AlegreyaSans-Black]             [name:AlegreyaSans-Black]
  \definefontsynonym[AlegreyaSans-BlackItalic]       [name:AlegreyaSans-BlackItalic]
  \definefontsynonym[AlegreyaSans-BlackCaps]         [name:AlegreyaSansSC-Black]
  \definefontsynonym[AlegreyaSans-BlackItalicCaps]   [name:AlegreyaSansSC-BlackItalic]
\stoptypescript

\starttypescript [serif] [alegreya]
  \definefontsynonym[AlegreyaSerif-Regular]         [name:Alegreya-Regular]
  \definefontsynonym[AlegreyaSerif-Italic]          [name:Alegreya-Italic]
  \definefontsynonym[AlegreyaSerif-Caps]            [name:AlegreyaSC-Regular]
  \definefontsynonym[AlegreyaSerif-Bold]            [name:Alegreya-Bold]
  \definefontsynonym[AlegreyaSerif-BoldItalic]      [name:Alegreya-BoldItalic]
  \definefontsynonym[AlegreyaSerif-BoldCaps]        [name:AlegreyaSC-Bold]
  \definefontsynonym[AlegreyaSerif-Black]           [name:Alegreya-Black]
  \definefontsynonym[AlegreyaSerif-BlackItalic]     [name:Alegreya-BlackItalic]
  \definefontsynonym[AlegreyaSerif-BlackCaps]       [name:AlegreyaSC-Black]
\stoptypescript

\starttypescript [sans] [alegreya] [name]
  %\setups[font:fallback:sans]
  \definefontsynonym [Sans]           [AlegreyaSans-Regular]     [features=alefeatures]
  \definefontsynonym [SansItalic]     [AlegreyaSans-Italic]      [features=alefeatures]
  \definefontsynonym [SansSlanted]    [AlegreyaSans-Italic]      [features=alefeatures]
  \definefontsynonym [SansBold]       [AlegreyaSans-Bold]        [features=alefeatures]
  \definefontsynonym [SansBoldItalic] [AlegreyaSans-BoldItalic]  [features=alefeatures]
  \definefontsynonym [SansBoldSlanted][AlegreyaSans-BoldItalic]  [features=alefeatures]
  \definefontsynonym [SansCaps]       [AlegreyaSans-Caps]        [features=alefeatures]
  \definefontsynonym [SansThin]         [AlegreyaSans-Thin]      [features=alefeatures]
  \definefontsynonym [SansThinItalic]   [AlegreyaSans-ThinItalic][features=alefeatures]
  \definefontsynonym [SansThinSlanted]  [AlegreyaSans-ThinItalic][features=alefeatures]
  \definefontsynonym [SansThinCaps]     [AlegreyaSans-ThinCaps]  [features=alefeatures]
  \definefontsynonym [SansLight]        [AlegreyaSans-Light]     [features=alefeatures]
  \definefontsynonym [SansLightItalic]  [AlegreyaSans-LightItalic][features=alefeatures]
  \definefontsynonym [SansLightSlanted] [AlegreyaSans-LightItalic][features=alefeatures]
  \definefontsynonym [SansLightCaps]    [AlegreyaSans-LightCaps]  [features=alefeatures]
  \definefontsynonym [SansMedium]       [AlegreyaSans-Medium]     [features=alefeatures]
  \definefontsynonym [SansMediumItalic] [AlegreyaSans-MediumItalic][features=alefeatures]
  \definefontsynonym [SansMediumSlanted][AlegreyaSans-MediumItalic][features=alefeatures]
\stoptypescript

\starttypescript [sans] [alegreya-light] [name]
  \definefontsynonym [Sans]           [AlegreyaSans-Light]         [features=alefeatures]
  \definefontsynonym [SansItalic]     [AlegreyaSans-LightItalic]   [features=alefeatures]
  \definefontsynonym [SansSlanted]    [AlegreyaSans-LightItalic]   [features=alefeatures]
  \definefontsynonym [SansCaps]       [AlegreyaSans-LightCaps]     [features=alefeatures]
  \definefontsynonym [SansBold]       [AlegreyaSans-Medium]        [features=alefeatures]
  \definefontsynonym [SansBoldItalic] [AlegreyaSans-MediumItalic]  [features=alefeatures]
  \definefontsynonym [SansBoldSlanted][AlegreyaSans-MediumItalic]  [features=alefeatures]
\stoptypescript

\starttypescript [sans] [alegreya-medium] [name]
  \definefontsynonym [Sans]           [AlegreyaSans-Medium]         [features=alefeatures]
  \definefontsynonym [SansItalic]     [AlegreyaSans-MediumItalic]   [features=alefeatures]
  \definefontsynonym [SansSlanted]    [AlegreyaSans-MediumItalic]   [features=alefeatures]
  \definefontsynonym [SansCaps]       [AlegreyaSans-MediumCaps]     [features=alefeatures]
  \definefontsynonym [SansBold]       [AlegreyaSans-ExtraBold]        [features=alefeatures]
  \definefontsynonym [SansBoldItalic] [AlegreyaSans-ExtraBoldItalic]  [features=alefeatures]
  \definefontsynonym [SansBoldSlanted][AlegreyaSans-ExtraBoldItalic]  [features=alefeatures]
\stoptypescript

\starttypescript [serif] [alegreya] [name]
  \definefontsynonym [Serif]           [AlegreyaSerif-Regular]     [features=alefeatures]
  \definefontsynonym [SerifItalic]     [AlegreyaSerif-Italic]      [features=alefeatures]
  \definefontsynonym [SerifSlanted]    [AlegreyaSerif-Italic]      [features=alefeatures]
  \definefontsynonym [SerifBold]       [AlegreyaSerif-Bold]        [features=alefeatures]
  \definefontsynonym [SerifBoldItalic] [AlegreyaSerif-BoldItalic]  [features=alefeatures]
  \definefontsynonym [SerifBoldSlanted][AlegreyaSerif-BoldItalic]  [features=alefeatures]
  \definefontsynonym [SerifCaps]       [AlegreyaSerif-Caps]        [features=alefeatures]
\stoptypescript

\starttypescript [alegreya]
  \definetypeface [alegreya] [rm] [serif] [alegreya] [default]
  \definetypeface [alegreya] [ss] [sans]  [alegreya]  [default]
  \definetypeface [alegreya] [tt] [mono]  [modern]   [default]
  \definetypeface [alegreya] [mm] [math]  [default]   [default]
\stoptypescript

\starttypescript [alegreya-light]
  \definetypeface [alegreya] [rm] [serif] [alegreya] [default]
  \definetypeface [alegreya] [ss] [sans]  [alegreya-light]  [default]
  \definetypeface [alegreya] [tt] [mono]  [modern]   [default]
  \definetypeface [alegreya] [mm] [math]  [default]   [default]
\stoptypescript

\starttypescript [alegreya-medium]
  \definetypeface [alegreya] [rm] [serif] [alegreya] [default]
  \definetypeface [alegreya] [ss] [sans]  [alegreya-medium]  [default]
  \definetypeface [alegreya] [tt] [mono]  [modern]   [default]
  \definetypeface [alegreya] [mm] [math]  [default]   [default]
\stoptypescript

[-- Attachment #3: Type: text/plain, Size: 492 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Pablo Rodriguez]

On 09/28/2017 03:54 PM, Hans Hagen wrote:
> On 9/28/2017 1:51 PM, Pablo Rodriguez wrote:
>> [...]
>> I overlooked that \blockligatures[u:fl:age, u:fl:eg] only works if there
>> is also a \blockligatures[fl].
>>
>> Hans, is this a bug or is it intended?
> do you indeed use the latest beta? anyway, the \replaceword mechanism is 
> a bit more flexible and efficient

Hans,

after updating to latest beta (from 2017.09.25 19:19 to 2017.09.28 10:07
[I didn’t realize that there was a newer beta yesterday]), it works.

Just in case it may be relevant, \blockligatures[u:fl:age] blocks
ligatures in auflegt and Aufl.

Minimal sample:

    \starttext
    \blockligatures[u:fl:age]

    \definefontfeature[default:b][default][blockligatures=yes]

    \setupbodyfont[times]\showfontkerns

    \startTEXpage[offset=1em]
        fl Auflage auflegt Aufl.

    \definedfont[Serif*default:b]
        fl Auflage auflegt Aufl.
    \stopTEXpage
    \stoptext

Pablo
-- 
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Hans Hagen]

On 9/28/2017 1:51 PM, Pablo Rodriguez wrote:
> On 09/28/2017 09:17 AM, Herbert Voss wrote:
>> Am 28.09.2017 um 08:40 schrieb Pablo Rodriguez:
>>
>>> using latest beta (from 2017.09.25 19:19), the following sample works:
>> [...]
>> How do you catch "Aufl." (the abbreviation of Auflage),
>>    which is written _with_ a ligature?
> 
> Herbert,
> 
> I overlooked that \blockligatures[u:fl:age, u:fl:eg] only works if there
> is also a \blockligatures[fl].
> 
> Hans, is this a bug or is it intended?
do you indeed use the latest beta? anyway, the \replaceword mechanism is 
a bit more flexible and efficient

Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Pablo Rodriguez]

On 09/28/2017 09:17 AM, Herbert Voss wrote:
> Am 28.09.2017 um 08:40 schrieb Pablo Rodriguez:
> 
>> using latest beta (from 2017.09.25 19:19), the following sample works:
> [...]
> How do you catch "Aufl." (the abbreviation of Auflage),
>   which is written _with_ a ligature?

Herbert,

I overlooked that \blockligatures[u:fl:age, u:fl:eg] only works if there
is also a \blockligatures[fl].

Hans, is this a bug or is it intended?

Pablo
-- 
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

Am 2017-09-27 um 23:02 schrieb Herbert Voss <Herbert.Voss@FU-Berlin.DE>:

>> Thank you for the suggestion.
>> I tried
>> \replaceword[eka][Auflage][Au{fl}age]
>> \setreplacements[eka]
>> and even converted the german wordlist from the selnolig package, but it doesn’t help.
>> If I try
>> \replaceword[eka][Auflage][Au\noligature{fl}age]
>> I get a literal "Au noligature fl age" in my text.
>> Simply Au\noligature{fl}age doesn’t help either.
>> My active features are:
>> \definefontfeature[default]
>>     [mode=node,liga=yes,kern=yes,tlig=yes,
>>     ccmp=yes,language=dflt,
>>     protrusion=quality,
>>     expansion=quality]
> 
> \definefontfeature[default]
> [mode=node,liga=yes,kern=yes,tlig=yes,
> ccmp=yes,language=dflt,
> protrusion=quality,
> expansion=quality]
> 
> \replaceword[eka][Auflage][Au{fl}age]
> 
> \starttext
> Auflage
> Au\noligature{fl}age
> \setreplacements[eka]
> Auflage
> \stoptext
> 
> 
> no problem here, first with ligature the other two without

Yes, today it works. Thank you! Didn’t change anything. Strange.

Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Herbert Voss]

Am 28.09.2017 um 08:40 schrieb Pablo Rodriguez:

> using latest beta (from 2017.09.25 19:19), the following sample works:
> 
>      \starttext
>      \blockligatures[fi,ff] \blockligatures[fl]
>      \blockligatures[u:fl:age, u:fl:eg]
> 
>      \definefontfeature[default:b][default][blockligatures=yes]
> 
>      \setupbodyfont[times]\showfontkerns
> 
>      \startTEXpage[offset=1em]
>          fi ff fl Auflage auflegt
> 
>      \definedfont[Serif*default:b]
>          fi ff fl Auflage auflegt

How do you catch "Aufl." (the abbreviation of Auflage),
  which is written _with_ a ligature?

Herbert



>      \stopTEXpage
>      \stoptext
> 
> Just in case it helps,
> 
> Pablo
> 

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Pablo Rodriguez]

On 09/27/2017 10:27 PM, Henning Hraban Ramm wrote:
> Am 2017-09-27 um 18:13 schrieb Hans Hagen <pragma@wxs.nl>:
>> [...]
>> The \blockligatures[u:fl:age] is new and something of a joke
>> experiment ... I can probably come up with a real advanced and robust
>> mechanism if motivated enough
> 
> Thank you, I tried it, and while blocking ligatures generally works
> (\blockligatures[fl]), the selective blocking
> (\blockligatures[u:fl:age]) doesn’t change anything.
> Do you have this only in your local version?

Hi Hraban,

using latest beta (from 2017.09.25 19:19), the following sample works:

    \starttext
    \blockligatures[fi,ff] \blockligatures[fl]
    \blockligatures[u:fl:age, u:fl:eg]

    \definefontfeature[default:b][default][blockligatures=yes]

    \setupbodyfont[times]\showfontkerns

    \startTEXpage[offset=1em]
        fi ff fl Auflage auflegt

    \definedfont[Serif*default:b]
        fi ff fl Auflage auflegt
    \stopTEXpage
    \stoptext

Just in case it helps,

Pablo
-- 
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Hans Hagen]

On 9/27/2017 11:08 PM, Thomas A. Schmitz wrote:
> On 09/27/2017 10:25 PM, Henning Hraban Ramm wrote:
>> Thank you for the suggestion.
>> I tried
>>
>> \replaceword[eka][Auflage][Au{fl}age]
>> \setreplacements[eka]
>>
>> and even converted the german wordlist from the selnolig package, but 
>> it doesn’t help.
> 
> 
> Works here. You'll have to make a real example.
I'll add an extra (simple) definition format:

\replaceword [eka] [Au{fl}age Shiff{f}ahrt]

I leave it to others to collect (reasonable) lists of words. We can 
these to the distribution then.

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Thomas A. Schmitz]

On 09/27/2017 10:25 PM, Henning Hraban Ramm wrote:
> Thank you for the suggestion.
> I tried
> 
> \replaceword[eka][Auflage][Au{fl}age]
> \setreplacements[eka]
> 
> and even converted the german wordlist from the selnolig package, but it doesn’t help.


Works here. You'll have to make a real example.

Thomas
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Herbert Voss]

Am 27.09.2017 um 22:25 schrieb Henning Hraban Ramm:
> Am 2017-09-27 um 15:39 schrieb Thomas A. Schmitz <thomas.schmitz@uni-bonn.de>:
> 
>> On 09/27/2017 02:16 PM, Thomas Widmann wrote:
>>> That's a nice idea.  If something like that existed, it could be used
>>> to change 's' to 'ſ' in relevant places, too.
>>
>> This exists, see here: https://www.mail-archive.com/ntg-context@ntg.nl/msg79381.html
> 
> Thank you for the suggestion.
> I tried
> 
> \replaceword[eka][Auflage][Au{fl}age]
> \setreplacements[eka]
> 
> and even converted the german wordlist from the selnolig package, but it doesn’t help.
> If I try
> \replaceword[eka][Auflage][Au\noligature{fl}age]
> I get a literal "Au noligature fl age" in my text.
> 
> Simply Au\noligature{fl}age doesn’t help either.
> 
> My active features are:
> \definefontfeature[default]
>      [mode=node,liga=yes,kern=yes,tlig=yes,
>      ccmp=yes,language=dflt,
>      protrusion=quality,
>      expansion=quality]
> 

\definefontfeature[default]
[mode=node,liga=yes,kern=yes,tlig=yes,
ccmp=yes,language=dflt,
protrusion=quality,
expansion=quality]

\replaceword[eka][Auflage][Au{fl}age]

\starttext
Auflage
Au\noligature{fl}age
\setreplacements[eka]
Auflage
\stoptext


no problem here, first with ligature the other two without

Herbert



> 
> Greetlings, Hraban
> ---
> http://www.fiee.net
> http://wiki.contextgarden.net
> GPG Key ID 1C9B22FD
> 
> 
> ___________________________________________________________________________________
> If your question is of interest to others as well, please add an entry to the Wiki!
> 
> maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
> webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
> archive  : https://bitbucket.org/phg/context-mirror/commits/
> wiki     : http://contextgarden.net
> ___________________________________________________________________________________
> 

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

Am 2017-09-27 um 18:13 schrieb Hans Hagen <pragma@wxs.nl>:

> Anyway, in addition to Thomas answer ... we already have this
> 
> \blockligatures[fi,ff] \blockligatures[fl]
> \blockligatures[u:fl:age]
> 
> \definefontfeature[default:nolig][default][blockligatures=yes]
> 
> \startTEXpage[offset=1em]
>    \definedfont[Serif*default:nolig]
>    fi ff fl auflage
>    \definedfont[Serif*default]
>    fi ff fl auflage
> \stopTEXpage
> 
> The \blockligatures[u:fl:age] is new and something of a joke experiment ... I can probably come up with a real advanced and robust mechanism if motivated enough

Thank you, I tried it, and while blocking ligatures generally works (\blockligatures[fl]), the selective blocking (\blockligatures[u:fl:age]) doesn’t change anything.
Do you have this only in your local version?

Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Henning Hraban Ramm]

Am 2017-09-27 um 15:39 schrieb Thomas A. Schmitz <thomas.schmitz@uni-bonn.de>:

> On 09/27/2017 02:16 PM, Thomas Widmann wrote:
>> That's a nice idea.  If something like that existed, it could be used
>> to change 's' to 'ſ' in relevant places, too.
> 
> This exists, see here: https://www.mail-archive.com/ntg-context@ntg.nl/msg79381.html

Thank you for the suggestion.
I tried

\replaceword[eka][Auflage][Au{fl}age]
\setreplacements[eka]

and even converted the german wordlist from the selnolig package, but it doesn’t help.
If I try
\replaceword[eka][Auflage][Au\noligature{fl}age]
I get a literal "Au noligature fl age" in my text.

Simply Au\noligature{fl}age doesn’t help either.

My active features are:
\definefontfeature[default]
    [mode=node,liga=yes,kern=yes,tlig=yes,
    ccmp=yes,language=dflt,
    protrusion=quality,
    expansion=quality]


Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD


___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Herbert Voss]

Am 27.09.2017 um 14:09 schrieb Henning Hraban Ramm:

> 
> I know I can break ligatures with \/, like Auf\/lage, but is there a general dictionary approach, like \hyphenation{} ?

https://ctan.org/tex-archive/macros/luatex/latex/selnolig

is a selnolig.lua which could be used for a solution with context.
A list of all german words where a ligature should be suppressed
is also there.

Herbert


> 
> 
> Greetlings, Hraban
> ---
> http://www.fiee.net
> http://wiki.contextgarden.net
> GPG Key ID 1C9B22FD
> 
> ___________________________________________________________________________________
> If your question is of interest to others as well, please add an entry to the Wiki!
> 
> maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
> webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
> archive  : https://bitbucket.org/phg/context-mirror/commits/
> wiki     : http://contextgarden.net
> ___________________________________________________________________________________
> 

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Hans Hagen]

On 9/27/2017 5:53 PM, Ulrike Fischer wrote:
> Am Wed, 27 Sep 2017 15:10:33 +0200 schrieb Tomas Hala:
> 
>>  From this point of view, Auf{}lage seems better.
> 
> This doesn't prevent the ligature:
> 
> \starttext
> 
> Auflage Auf{}lage Auf\/lage
> 
> \stoptext
> 
> This also documented: "The most important change is that adding a
> brace group in the middle of a word (like in of{}fice) does not
> prevent ligature creation."
> 
> With pdftex it does avoid ligatures, but not reliably. There are
> cases where the brace group disappear when pdftex tries out
> hyphenation points.
in traditional tex hyphenation is integrated in the par builder which 
comes after ligaturing and kerning and tex only hyphenated where it 
'makes sense' and in the process (multi-pass, going into disc nodes) it 
will consult (deconstruct / reconstruct) ligatures and at that point 
it's not known that the input had a {}. Tricks like inserting something 
in between is also unreliable when done at the wrong time and it can 
also prevent for instance kerns of hyphenation.

Anyway, in addition to Thomas answer ... we already have this

\blockligatures[fi,ff] \blockligatures[fl]
\blockligatures[u:fl:age]

\definefontfeature[default:nolig][default][blockligatures=yes]

\startTEXpage[offset=1em]
     \definedfont[Serif*default:nolig]
     fi ff fl auflage
     \definedfont[Serif*default]
     fi ff fl auflage
\stopTEXpage

The \blockligatures[u:fl:age] is new and something of a joke experiment 
... I can probably come up with a real advanced and robust mechanism if 
motivated enough

Hans


-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Ulrike Fischer]

Am Wed, 27 Sep 2017 15:10:33 +0200 schrieb Tomas Hala:

> From this point of view, Auf{}lage seems better.

This doesn't prevent the ligature:

\starttext

Auflage Auf{}lage Auf\/lage 

\stoptext

This also documented: "The most important change is that adding a
brace group in the middle of a word (like in of{}fice) does not
prevent ligature creation."

With pdftex it does avoid ligatures, but not reliably. There are
cases where the brace group disappear when pdftex tries out
hyphenation points. 



-- 
Ulrike Fischer 
http://www.troubleshooting-tex.de/

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Florian Grammel]

[-- Attachment #1.1: Type: text/plain, Size: 342 bytes --]

There is a LuaLaTeX-package "selnolig" to achieve this goal. Its wordlists and rules could probably be adapted. There is also a thorough discussion of this challenge in the documentation.
https://www.ctan.org/pkg/selnolig

Best regards,
Florian

____________________________________________

Florian Grammel

Copenhagen, Denmark


[-- Attachment #1.2: Type: text/html, Size: 754 bytes --]

[-- Attachment #2: Type: text/plain, Size: 492 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Thomas A. Schmitz]

On 09/27/2017 02:16 PM, Thomas Widmann wrote:
> That's a nice idea.  If something like that existed, it could be used
> to change 's' to 'ſ' in relevant places, too.


This exists, see here: 
https://www.mail-archive.com/ntg-context@ntg.nl/msg79381.html
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Tomas Hala]

Hi Hraban,

I am afraid that \/ will break not only the ligature but moreover
it will erase the possible breakpoint between f and l. 
From this point of view, Auf{}lage seems better.

General approach might be connected with font some features as Hans
shown at ConTeXt Meeting this year. 

I recorded this case into the list of things we have to keep in mind 
for testing of free fonts.

Best wishes,

Tomas

Wed, Sep 27, 2017 ve 02:09:50PM +0200 Henning Hraban Ramm napsal(a):
# Hi all,
# 
# I know I can break ligatures with \/, like Auf\/lage, but is there a general dictionary approach, like \hyphenation{} ?
# 
# 
# Greetlings, Hraban
# ---
# http://www.fiee.net
# http://wiki.contextgarden.net
# GPG Key ID 1C9B22FD
# 
# ___________________________________________________________________________________
# If your question is of interest to others as well, please add an entry to the Wiki!
# 
# maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
# webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
# archive  : https://bitbucket.org/phg/context-mirror/commits/
# wiki     : http://contextgarden.net
# ___________________________________________________________________________________

                                         Tomáš Hála
--------------------------------------------------------------------
Mendelova univerzita, Provozně ekonomická fakulta, ústav informatiky
Zemědělská 1, CZ-613 00 Brno,  tel. +420 545 13 22 28
--------------------------------------------------------------------
http://akela.mendelu.cz/~thala
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

Re: ligature exceptions

[Thomas Widmann]

On 27 September 2017 at 13:09, Henning Hraban Ramm <texml@fiee.net> wrote:
>
> I know I can break ligatures with \/, like Auf\/lage, but is there a general
> dictionary approach, like \hyphenation{} ?

That's a nice idea.  If something like that existed, it could be used
to change 's' to 'ſ' in relevant places, too.

Thomas
-- 
Thomas Widmann
Director
Complexli Limited
+44 789 444 3009
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________

ligature exceptions

[Henning Hraban Ramm]

Hi all,

I know I can break ligatures with \/, like Auf\/lage, but is there a general dictionary approach, like \hyphenation{} ?


Greetlings, Hraban
---
http://www.fiee.net
http://wiki.contextgarden.net
GPG Key ID 1C9B22FD

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
archive  : https://bitbucket.org/phg/context-mirror/commits/
wiki     : http://contextgarden.net
___________________________________________________________________________________