diff options
author | Marius <mariausol@gmail.com> | 2012-10-17 20:20:13 +0300 |
---|---|---|
committer | Marius <mariausol@gmail.com> | 2012-10-17 20:20:13 +0300 |
commit | 3d3b6af2a13fb57dd36377954669124ceac73b3d (patch) | |
tree | 07d6438c44c9ba2ab5cebabf0479092dcc56a63c | |
parent | 25454333dc3c49ab6607696ad221d4f9de91e1ef (diff) | |
download | context-3d3b6af2a13fb57dd36377954669124ceac73b3d.tar.gz |
beta 2012.10.17 19:02
-rw-r--r-- | tex/context/base/cont-new.mkii | 2 | ||||
-rw-r--r-- | tex/context/base/cont-new.mkiv | 2 | ||||
-rw-r--r-- | tex/context/base/context-version.pdf | bin | 4147 -> 4144 bytes | |||
-rw-r--r-- | tex/context/base/context-version.png | bin | 106714 -> 105290 bytes | |||
-rw-r--r-- | tex/context/base/context.mkii | 2 | ||||
-rw-r--r-- | tex/context/base/context.mkiv | 2 | ||||
-rw-r--r-- | tex/context/base/font-lib.mkvi | 2 | ||||
-rw-r--r-- | tex/context/base/font-ota.lua | 25 | ||||
-rw-r--r-- | tex/context/base/font-otd.lua | 21 | ||||
-rw-r--r-- | tex/context/base/font-otn.lua | 15 | ||||
-rw-r--r-- | tex/context/base/status-files.pdf | bin | 24589 -> 24605 bytes | |||
-rw-r--r-- | tex/context/base/status-lua.pdf | bin | 195393 -> 195624 bytes | |||
-rw-r--r-- | tex/generic/context/luatex/luatex-fonts-merged.lua | 736 | ||||
-rw-r--r-- | tex/generic/context/luatex/luatex-fonts.lua | 2 |
14 files changed, 399 insertions, 410 deletions
diff --git a/tex/context/base/cont-new.mkii b/tex/context/base/cont-new.mkii index f75a1d7ce..955d1ef04 100644 --- a/tex/context/base/cont-new.mkii +++ b/tex/context/base/cont-new.mkii @@ -11,7 +11,7 @@ %C therefore copyrighted by \PRAGMA. See mreadme.pdf for %C details. -\newcontextversion{2012.10.17 13:30} +\newcontextversion{2012.10.17 19:02} %D This file is loaded at runtime, thereby providing an %D excellent place for hacks, patches, extensions and new diff --git a/tex/context/base/cont-new.mkiv b/tex/context/base/cont-new.mkiv index 4802c250d..1f85ab6d5 100644 --- a/tex/context/base/cont-new.mkiv +++ b/tex/context/base/cont-new.mkiv @@ -11,7 +11,7 @@ %C therefore copyrighted by \PRAGMA. See mreadme.pdf for %C details. -\newcontextversion{2012.10.17 13:30} +\newcontextversion{2012.10.17 19:02} %D This file is loaded at runtime, thereby providing an excellent place for %D hacks, patches, extensions and new features. diff --git a/tex/context/base/context-version.pdf b/tex/context/base/context-version.pdf Binary files differindex c38355795..c9c7b91bd 100644 --- a/tex/context/base/context-version.pdf +++ b/tex/context/base/context-version.pdf diff --git a/tex/context/base/context-version.png b/tex/context/base/context-version.png Binary files differindex cc0edc78f..4243de3c8 100644 --- a/tex/context/base/context-version.png +++ b/tex/context/base/context-version.png diff --git a/tex/context/base/context.mkii b/tex/context/base/context.mkii index 3a1c1f4f8..a98905678 100644 --- a/tex/context/base/context.mkii +++ b/tex/context/base/context.mkii @@ -20,7 +20,7 @@ %D your styles an modules. \edef\contextformat {\jobname} -\edef\contextversion{2012.10.17 13:30} +\edef\contextversion{2012.10.17 19:02} %D For those who want to use this: diff --git a/tex/context/base/context.mkiv b/tex/context/base/context.mkiv index e023bca32..5718eefdb 100644 --- a/tex/context/base/context.mkiv +++ b/tex/context/base/context.mkiv @@ -25,7 +25,7 @@ %D up and the dependencies are more consistent. \edef\contextformat {\jobname} -\edef\contextversion{2012.10.17 13:30} +\edef\contextversion{2012.10.17 19:02} %D For those who want to use this: diff --git a/tex/context/base/font-lib.mkvi b/tex/context/base/font-lib.mkvi index 6b3d76567..4ccb6057c 100644 --- a/tex/context/base/font-lib.mkvi +++ b/tex/context/base/font-lib.mkvi @@ -34,9 +34,9 @@ \registerctxluafile{font-otf}{1.001} % otf main \registerctxluafile{font-otb}{1.001} % otf main base \registerctxluafile{node-inj}{1.001} % we might split it off +\registerctxluafile{font-ota}{1.001} % otf analyzers \registerctxluafile{font-otn}{1.001} % otf main node \registerctxluafile{font-otd}{1.001} % otf dynamics (does an overload) -\registerctxluafile{font-ota}{1.001} % otf analyzers (needs dynamics) \registerctxluafile{font-otp}{1.001} % otf pack \registerctxluafile{font-otc}{1.001} % otf context \registerctxluafile{font-oth}{1.001} % otf helpers diff --git a/tex/context/base/font-ota.lua b/tex/context/base/font-ota.lua index 962f5d4d3..8262f983e 100644 --- a/tex/context/base/font-ota.lua +++ b/tex/context/base/font-ota.lua @@ -48,11 +48,22 @@ local registerotffeature = otffeatures.register process features right.</p> --ldx]]-- +analyzers.constants = { + init = 1, + medi = 2, + fina = 3, + isol = 4, + -- devanagari + rphf = 5, + half = 6, + pref = 7, + blwf = 8, + pstf = 9, +} + -- todo: analyzers per script/lang, cross font, so we need an font id hash -> script -- e.g. latin -> hyphenate, arab -> 1/2/3 analyze -- its own namespace -local state = attributes.private('state') - function analyzers.setstate(head,font) local useunicodemarks = analyzers.useunicodemarks local tfmdata = fontdata[font] @@ -234,11 +245,11 @@ local isol_fina_medi_init = { -- syriac - [0x0712] = true, [0x0713] = true, [0x0714] = true, [0x071A] = true, - [0x071B] = true, [0x071C] = true, [0x071D] = true, [0x071F] = true, - [0x0720] = true, [0x0721] = true, [0x0722] = true, [0x0723] = true, - [0x0724] = true, [0x0725] = true, [0x0726] = true, [0x0727] = true, - [0x0729] = true, [0x072B] = true, + [0x0712] = true, [0x0713] = true, [0x0714] = true, [0x071A] = true, + [0x071B] = true, [0x071C] = true, [0x071D] = true, [0x071F] = true, + [0x0720] = true, [0x0721] = true, [0x0722] = true, [0x0723] = true, + [0x0724] = true, [0x0725] = true, [0x0726] = true, [0x0727] = true, + [0x0729] = true, [0x072B] = true, -- also diff --git a/tex/context/base/font-otd.lua b/tex/context/base/font-otd.lua index 1d8c4ad43..0bc8225de 100644 --- a/tex/context/base/font-otd.lua +++ b/tex/context/base/font-otd.lua @@ -116,22 +116,11 @@ end -- we reimplement the dataset resolver -local special_attributes = { - init = 1, - medi = 2, - fina = 3, - isol = 4 - -- devanagari - rphf = 5, - half = 6, - pref = 7, - blwf = 8, - pstf = 9, -} +local constants = fonts.analyzers.constants -local resolved = { } -- we only resolve a font,script,language,attribute pair once -local wildcard = "*" -local default = "dflt" +local resolved = { } -- we only resolve a font,script,language,attribute pair once +local wildcard = "*" +local default = "dflt" local function initialize(sequence,script,language,s_enabled,a_enabled,font,attr,dynamic) local features = sequence.features @@ -158,7 +147,7 @@ local function initialize(sequence,script,language,s_enabled,a_enabled,font,attr what = wildcard end if valid then - local attribute = special_attributes[kind] or false + local attribute = constants[kind] or false if a_e and dynamic < 0 then valid = false end diff --git a/tex/context/base/font-otn.lua b/tex/context/base/font-otn.lua index 6ea35f139..2c670825a 100644 --- a/tex/context/base/font-otn.lua +++ b/tex/context/base/font-otn.lua @@ -2097,18 +2097,7 @@ end) -- fonts.hashes.lookups = lookuphashes -local special_attributes = { - init = 1, - medi = 2, - fina = 3, - isol = 4, - -- devanagari - rphf = 5, - half = 6, - pref = 7, - blwf = 8, - pstf = 9, -} +local constants = fonts.analyzers.constants local function initialize(sequence,script,language,enabled) local features = sequence.features @@ -2118,7 +2107,7 @@ local function initialize(sequence,script,language,enabled) if valid then local languages = scripts[script] or scripts[wildcard] if languages and (languages[language] or languages[wildcard]) then - return { valid, special_attributes[kind] or false, sequence.chain or 0, kind, sequence } + return { valid, constants[kind] or false, sequence.chain or 0, kind, sequence } end end end diff --git a/tex/context/base/status-files.pdf b/tex/context/base/status-files.pdf Binary files differindex 5871378be..e59c01282 100644 --- a/tex/context/base/status-files.pdf +++ b/tex/context/base/status-files.pdf diff --git a/tex/context/base/status-lua.pdf b/tex/context/base/status-lua.pdf Binary files differindex fe945f678..aa27720eb 100644 --- a/tex/context/base/status-lua.pdf +++ b/tex/context/base/status-lua.pdf diff --git a/tex/generic/context/luatex/luatex-fonts-merged.lua b/tex/generic/context/luatex/luatex-fonts-merged.lua index 96be83bcc..cd36a3e5b 100644 --- a/tex/generic/context/luatex/luatex-fonts-merged.lua +++ b/tex/generic/context/luatex/luatex-fonts-merged.lua @@ -1,6 +1,6 @@ -- merged file : luatex-fonts-merged.lua -- parent file : luatex-fonts.lua --- merge date : 10/17/12 13:30:56 +-- merge date : 10/17/12 19:02:50 do -- begin closure to overcome local limits and interference @@ -8817,6 +8817,371 @@ end -- closure do -- begin closure to overcome local limits and interference +if not modules then modules = { } end modules ['font-ota'] = { + version = 1.001, + comment = "companion to font-otf.lua (analysing)", + author = "Hans Hagen, PRAGMA-ADE, Hasselt NL", + copyright = "PRAGMA ADE / ConTeXt Development Team", + license = "see context related readme files" +} + +-- this might become scrp-*.lua + +local type, tostring, match, format, concat = type, tostring, string.match, string.format, table.concat + +if not trackers then trackers = { register = function() end } end + +local trace_analyzing = false trackers.register("otf.analyzing", function(v) trace_analyzing = v end) + +local fonts, nodes, node = fonts, nodes, node + +local allocate = utilities.storage.allocate + +local otf = fonts.handlers.otf + +local analyzers = fonts.analyzers +local initializers = allocate() +local methods = allocate() + +analyzers.initializers = initializers +analyzers.methods = methods +analyzers.useunicodemarks = false + +local nodecodes = nodes.nodecodes +local glyph_code = nodecodes.glyph + +local set_attribute = node.set_attribute +local has_attribute = node.has_attribute +local traverse_id = node.traverse_id +local traverse_node_list = node.traverse + +local fontdata = fonts.hashes.identifiers +local state = attributes.private('state') +local categories = characters and characters.categories or { } -- sorry, only in context + +local otffeatures = fonts.constructors.newfeatures("otf") +local registerotffeature = otffeatures.register + +--[[ldx-- +<p>Analyzers run per script and/or language and are needed in order to +process features right.</p> +--ldx]]-- + +analyzers.constants = { + init = 1, + medi = 2, + fina = 3, + isol = 4, + -- devanagari + rphf = 5, + half = 6, + pref = 7, + blwf = 8, + pstf = 9, +} + +-- todo: analyzers per script/lang, cross font, so we need an font id hash -> script +-- e.g. latin -> hyphenate, arab -> 1/2/3 analyze -- its own namespace + +function analyzers.setstate(head,font) + local useunicodemarks = analyzers.useunicodemarks + local tfmdata = fontdata[font] + local characters = tfmdata.characters + local descriptions = tfmdata.descriptions + local first, last, current, n, done = nil, nil, head, 0, false -- maybe make n boolean + while current do + local id = current.id + if id == glyph_code and current.font == font then + local char = current.char + local d = descriptions[char] + if d then + if d.class == "mark" or (useunicodemarks and categories[char] == "mn") then + done = true + set_attribute(current,state,5) -- mark + elseif n == 0 then + first, last, n = current, current, 1 + set_attribute(current,state,1) -- init + else + last, n = current, n+1 + set_attribute(current,state,2) -- medi + end + else -- finish + if first and first == last then + set_attribute(last,state,4) -- isol + elseif last then + set_attribute(last,state,3) -- fina + end + first, last, n = nil, nil, 0 + end + elseif id == disc_code then + -- always in the middle + set_attribute(current,state,2) -- midi + last = current + else -- finish + if first and first == last then + set_attribute(last,state,4) -- isol + elseif last then + set_attribute(last,state,3) -- fina + end + first, last, n = nil, nil, 0 + end + current = current.next + end + if first and first == last then + set_attribute(last,state,4) -- isol + elseif last then + set_attribute(last,state,3) -- fina + end + return head, done +end + +-- in the future we will use language/script attributes instead of the +-- font related value, but then we also need dynamic features which is +-- somewhat slower; and .. we need a chain of them + +local function analyzeinitializer(tfmdata,value) -- attr + local script, language = otf.scriptandlanguage(tfmdata) -- attr + local action = initializers[script] + if not action then + -- skip + elseif type(action) == "function" then + return action(tfmdata,value) + else + local action = action[language] + if action then + return action(tfmdata,value) + end + end +end + +local function analyzeprocessor(head,font,attr) + local tfmdata = fontdata[font] + local script, language = otf.scriptandlanguage(tfmdata,attr) + local action = methods[script] + if not action then + -- skip + elseif type(action) == "function" then + return action(head,font,attr) + else + action = action[language] + if action then + return action(head,font,attr) + end + end + return head, false +end + +registerotffeature { + name = "analyze", + description = "analysis of (for instance) character classes", + default = true, + initializers = { + node = analyzeinitializer, + }, + processors = { + position = 1, + node = analyzeprocessor, + } +} + +-- latin + +methods.latn = analyzers.setstate + +-- this info eventually will go into char-def and we will have a state +-- table for generic then + +local zwnj = 0x200C +local zwj = 0x200D + +local isol = { + [0x0600] = true, [0x0601] = true, [0x0602] = true, [0x0603] = true, + [0x0608] = true, [0x060B] = true, [0x0621] = true, [0x0674] = true, + [0x06DD] = true, [zwnj] = true, +} + +local isol_fina = { + [0x0622] = true, [0x0623] = true, [0x0624] = true, [0x0625] = true, + [0x0627] = true, [0x0629] = true, [0x062F] = true, [0x0630] = true, + [0x0631] = true, [0x0632] = true, [0x0648] = true, [0x0671] = true, + [0x0672] = true, [0x0673] = true, [0x0675] = true, [0x0676] = true, + [0x0677] = true, [0x0688] = true, [0x0689] = true, [0x068A] = true, + [0x068B] = true, [0x068C] = true, [0x068D] = true, [0x068E] = true, + [0x068F] = true, [0x0690] = true, [0x0691] = true, [0x0692] = true, + [0x0693] = true, [0x0694] = true, [0x0695] = true, [0x0696] = true, + [0x0697] = true, [0x0698] = true, [0x0699] = true, [0x06C0] = true, + [0x06C3] = true, [0x06C4] = true, [0x06C5] = true, [0x06C6] = true, + [0x06C7] = true, [0x06C8] = true, [0x06C9] = true, [0x06CA] = true, + [0x06CB] = true, [0x06CD] = true, [0x06CF] = true, [0x06D2] = true, + [0x06D3] = true, [0x06D5] = true, [0x06EE] = true, [0x06EF] = true, + [0x0759] = true, [0x075A] = true, [0x075B] = true, [0x076B] = true, + [0x076C] = true, [0x0771] = true, [0x0773] = true, [0x0774] = true, + [0x0778] = true, [0x0779] = true, [0xFEF5] = true, [0xFEF7] = true, + [0xFEF9] = true, [0xFEFB] = true, + + -- syriac + + [0x0710] = true, [0x0715] = true, [0x0716] = true, [0x0717] = true, + [0x0718] = true, [0x0719] = true, [0x0728] = true, [0x072A] = true, + [0x072C] = true, [0x071E] = true, +} + +local isol_fina_medi_init = { + [0x0626] = true, [0x0628] = true, [0x062A] = true, [0x062B] = true, + [0x062C] = true, [0x062D] = true, [0x062E] = true, [0x0633] = true, + [0x0634] = true, [0x0635] = true, [0x0636] = true, [0x0637] = true, + [0x0638] = true, [0x0639] = true, [0x063A] = true, [0x063B] = true, + [0x063C] = true, [0x063D] = true, [0x063E] = true, [0x063F] = true, + [0x0640] = true, [0x0641] = true, [0x0642] = true, [0x0643] = true, + [0x0644] = true, [0x0645] = true, [0x0646] = true, [0x0647] = true, + [0x0649] = true, [0x064A] = true, [0x066E] = true, [0x066F] = true, + [0x0678] = true, [0x0679] = true, [0x067A] = true, [0x067B] = true, + [0x067C] = true, [0x067D] = true, [0x067E] = true, [0x067F] = true, + [0x0680] = true, [0x0681] = true, [0x0682] = true, [0x0683] = true, + [0x0684] = true, [0x0685] = true, [0x0686] = true, [0x0687] = true, + [0x069A] = true, [0x069B] = true, [0x069C] = true, [0x069D] = true, + [0x069E] = true, [0x069F] = true, [0x06A0] = true, [0x06A1] = true, + [0x06A2] = true, [0x06A3] = true, [0x06A4] = true, [0x06A5] = true, + [0x06A6] = true, [0x06A7] = true, [0x06A8] = true, [0x06A9] = true, + [0x06AA] = true, [0x06AB] = true, [0x06AC] = true, [0x06AD] = true, + [0x06AE] = true, [0x06AF] = true, [0x06B0] = true, [0x06B1] = true, + [0x06B2] = true, [0x06B3] = true, [0x06B4] = true, [0x06B5] = true, + [0x06B6] = true, [0x06B7] = true, [0x06B8] = true, [0x06B9] = true, + [0x06BA] = true, [0x06BB] = true, [0x06BC] = true, [0x06BD] = true, + [0x06BE] = true, [0x06BF] = true, [0x06C1] = true, [0x06C2] = true, + [0x06CC] = true, [0x06CE] = true, [0x06D0] = true, [0x06D1] = true, + [0x06FA] = true, [0x06FB] = true, [0x06FC] = true, [0x06FF] = true, + [0x0750] = true, [0x0751] = true, [0x0752] = true, [0x0753] = true, + [0x0754] = true, [0x0755] = true, [0x0756] = true, [0x0757] = true, + [0x0758] = true, [0x075C] = true, [0x075D] = true, [0x075E] = true, + [0x075F] = true, [0x0760] = true, [0x0761] = true, [0x0762] = true, + [0x0763] = true, [0x0764] = true, [0x0765] = true, [0x0766] = true, + [0x0767] = true, [0x0768] = true, [0x0769] = true, [0x076A] = true, + [0x076D] = true, [0x076E] = true, [0x076F] = true, [0x0770] = true, + [0x0772] = true, [0x0775] = true, [0x0776] = true, [0x0777] = true, + [0x077A] = true, [0x077B] = true, [0x077C] = true, [0x077D] = true, + [0x077E] = true, [0x077F] = true, + + -- syriac + + [0x0712] = true, [0x0713] = true, [0x0714] = true, [0x071A] = true, + [0x071B] = true, [0x071C] = true, [0x071D] = true, [0x071F] = true, + [0x0720] = true, [0x0721] = true, [0x0722] = true, [0x0723] = true, + [0x0724] = true, [0x0725] = true, [0x0726] = true, [0x0727] = true, + [0x0729] = true, [0x072B] = true, + + -- also + + [zwj] = true, +} + +local arab_warned = { } + +-- todo: gref + +local function warning(current,what) + local char = current.char + if not arab_warned[char] then + log.report("analyze","arab: character %s (U+%05X) has no %s class", char, char, what) + arab_warned[char] = true + end +end + +local function finish(first,last) + if last then + if first == last then + local fc = first.char + if isol_fina_medi_init[fc] or isol_fina[fc] then + set_attribute(first,state,4) -- isol + else + warning(first,"isol") + set_attribute(first,state,0) -- error + end + else + local lc = last.char + if isol_fina_medi_init[lc] or isol_fina[lc] then -- why isol here ? + -- if laststate == 1 or laststate == 2 or laststate == 4 then + set_attribute(last,state,3) -- fina + else + warning(last,"fina") + set_attribute(last,state,0) -- error + end + end + first, last = nil, nil + elseif first then + -- first and last are either both set so we never com here + local fc = first.char + if isol_fina_medi_init[fc] or isol_fina[fc] then + set_attribute(first,state,4) -- isol + else + warning(first,"isol") + set_attribute(first,state,0) -- error + end + first = nil + end + return first, last +end + +function methods.arab(head,font,attr) -- maybe make a special version with no trace + local useunicodemarks = analyzers.useunicodemarks + local tfmdata = fontdata[font] + local marks = tfmdata.resources.marks + local first, last, current, done = nil, nil, head, false + while current do + if current.id == glyph_code and current.subtype<256 and current.font == font and not has_attribute(current,state) then + done = true + local char = current.char + if marks[char] or (useunicodemarks and categories[char] == "mn") then + set_attribute(current,state,5) -- mark + elseif isol[char] then -- can be zwj or zwnj too + first, last = finish(first,last) + set_attribute(current,state,4) -- isol + first, last = nil, nil + elseif not first then + if isol_fina_medi_init[char] then + set_attribute(current,state,1) -- init + first, last = first or current, current + elseif isol_fina[char] then + set_attribute(current,state,4) -- isol + first, last = nil, nil + else -- no arab + first, last = finish(first,last) + end + elseif isol_fina_medi_init[char] then + first, last = first or current, current + set_attribute(current,state,2) -- medi + elseif isol_fina[char] then + if not has_attribute(last,state,1) then + -- tricky, we need to check what last may be ! + set_attribute(last,state,2) -- medi + end + set_attribute(current,state,3) -- fina + first, last = nil, nil + elseif char >= 0x0600 and char <= 0x06FF then + set_attribute(current,state,6) -- rest + first, last = finish(first,last) + else --no + first, last = finish(first,last) + end + else + first, last = finish(first,last) + end + current = current.next + end + first, last = finish(first,last) + return head, done +end + +methods.syrc = methods.arab + +directives.register("otf.analyze.useunicodemarks",function(v) + analyzers.useunicodemarks = v +end) + +end -- closure + +do -- begin closure to overcome local limits and interference + if not modules then modules = { } end modules ['font-otn'] = { version = 1.001, comment = "companion to font-ini.mkiv", @@ -10916,18 +11281,7 @@ end) -- fonts.hashes.lookups = lookuphashes -local special_attributes = { - init = 1, - medi = 2, - fina = 3, - isol = 4, - -- devanagari - rphf = 5, - half = 6, - pref = 7, - blwf = 8, - pstf = 9, -} +local constants = fonts.analyzers.constants local function initialize(sequence,script,language,enabled) local features = sequence.features @@ -10937,7 +11291,7 @@ local function initialize(sequence,script,language,enabled) if valid then local languages = scripts[script] or scripts[wildcard] if languages and (languages[language] or languages[wildcard]) then - return { valid, special_attributes[kind] or false, sequence.chain or 0, kind, sequence } + return { valid, constants[kind] or false, sequence.chain or 0, kind, sequence } end end end @@ -12162,360 +12516,6 @@ end -- closure do -- begin closure to overcome local limits and interference -if not modules then modules = { } end modules ['font-ota'] = { - version = 1.001, - comment = "companion to font-otf.lua (analysing)", - author = "Hans Hagen, PRAGMA-ADE, Hasselt NL", - copyright = "PRAGMA ADE / ConTeXt Development Team", - license = "see context related readme files" -} - --- this might become scrp-*.lua - -local type, tostring, match, format, concat = type, tostring, string.match, string.format, table.concat - -if not trackers then trackers = { register = function() end } end - -local trace_analyzing = false trackers.register("otf.analyzing", function(v) trace_analyzing = v end) - -local fonts, nodes, node = fonts, nodes, node - -local allocate = utilities.storage.allocate - -local otf = fonts.handlers.otf - -local analyzers = fonts.analyzers -local initializers = allocate() -local methods = allocate() - -analyzers.initializers = initializers -analyzers.methods = methods -analyzers.useunicodemarks = false - -local nodecodes = nodes.nodecodes -local glyph_code = nodecodes.glyph - -local set_attribute = node.set_attribute -local has_attribute = node.has_attribute -local traverse_id = node.traverse_id -local traverse_node_list = node.traverse - -local fontdata = fonts.hashes.identifiers -local state = attributes.private('state') -local categories = characters and characters.categories or { } -- sorry, only in context - -local otffeatures = fonts.constructors.newfeatures("otf") -local registerotffeature = otffeatures.register - ---[[ldx-- -<p>Analyzers run per script and/or language and are needed in order to -process features right.</p> ---ldx]]-- - --- todo: analyzers per script/lang, cross font, so we need an font id hash -> script --- e.g. latin -> hyphenate, arab -> 1/2/3 analyze -- its own namespace - -local state = attributes.private('state') - -function analyzers.setstate(head,font) - local useunicodemarks = analyzers.useunicodemarks - local tfmdata = fontdata[font] - local characters = tfmdata.characters - local descriptions = tfmdata.descriptions - local first, last, current, n, done = nil, nil, head, 0, false -- maybe make n boolean - while current do - local id = current.id - if id == glyph_code and current.font == font then - local char = current.char - local d = descriptions[char] - if d then - if d.class == "mark" or (useunicodemarks and categories[char] == "mn") then - done = true - set_attribute(current,state,5) -- mark - elseif n == 0 then - first, last, n = current, current, 1 - set_attribute(current,state,1) -- init - else - last, n = current, n+1 - set_attribute(current,state,2) -- medi - end - else -- finish - if first and first == last then - set_attribute(last,state,4) -- isol - elseif last then - set_attribute(last,state,3) -- fina - end - first, last, n = nil, nil, 0 - end - elseif id == disc_code then - -- always in the middle - set_attribute(current,state,2) -- midi - last = current - else -- finish - if first and first == last then - set_attribute(last,state,4) -- isol - elseif last then - set_attribute(last,state,3) -- fina - end - first, last, n = nil, nil, 0 - end - current = current.next - end - if first and first == last then - set_attribute(last,state,4) -- isol - elseif last then - set_attribute(last,state,3) -- fina - end - return head, done -end - --- in the future we will use language/script attributes instead of the --- font related value, but then we also need dynamic features which is --- somewhat slower; and .. we need a chain of them - -local function analyzeinitializer(tfmdata,value) -- attr - local script, language = otf.scriptandlanguage(tfmdata) -- attr - local action = initializers[script] - if not action then - -- skip - elseif type(action) == "function" then - return action(tfmdata,value) - else - local action = action[language] - if action then - return action(tfmdata,value) - end - end -end - -local function analyzeprocessor(head,font,attr) - local tfmdata = fontdata[font] - local script, language = otf.scriptandlanguage(tfmdata,attr) - local action = methods[script] - if not action then - -- skip - elseif type(action) == "function" then - return action(head,font,attr) - else - action = action[language] - if action then - return action(head,font,attr) - end - end - return head, false -end - -registerotffeature { - name = "analyze", - description = "analysis of (for instance) character classes", - default = true, - initializers = { - node = analyzeinitializer, - }, - processors = { - position = 1, - node = analyzeprocessor, - } -} - --- latin - -methods.latn = analyzers.setstate - --- this info eventually will go into char-def and we will have a state --- table for generic then - -local zwnj = 0x200C -local zwj = 0x200D - -local isol = { - [0x0600] = true, [0x0601] = true, [0x0602] = true, [0x0603] = true, - [0x0608] = true, [0x060B] = true, [0x0621] = true, [0x0674] = true, - [0x06DD] = true, [zwnj] = true, -} - -local isol_fina = { - [0x0622] = true, [0x0623] = true, [0x0624] = true, [0x0625] = true, - [0x0627] = true, [0x0629] = true, [0x062F] = true, [0x0630] = true, - [0x0631] = true, [0x0632] = true, [0x0648] = true, [0x0671] = true, - [0x0672] = true, [0x0673] = true, [0x0675] = true, [0x0676] = true, - [0x0677] = true, [0x0688] = true, [0x0689] = true, [0x068A] = true, - [0x068B] = true, [0x068C] = true, [0x068D] = true, [0x068E] = true, - [0x068F] = true, [0x0690] = true, [0x0691] = true, [0x0692] = true, - [0x0693] = true, [0x0694] = true, [0x0695] = true, [0x0696] = true, - [0x0697] = true, [0x0698] = true, [0x0699] = true, [0x06C0] = true, - [0x06C3] = true, [0x06C4] = true, [0x06C5] = true, [0x06C6] = true, - [0x06C7] = true, [0x06C8] = true, [0x06C9] = true, [0x06CA] = true, - [0x06CB] = true, [0x06CD] = true, [0x06CF] = true, [0x06D2] = true, - [0x06D3] = true, [0x06D5] = true, [0x06EE] = true, [0x06EF] = true, - [0x0759] = true, [0x075A] = true, [0x075B] = true, [0x076B] = true, - [0x076C] = true, [0x0771] = true, [0x0773] = true, [0x0774] = true, - [0x0778] = true, [0x0779] = true, [0xFEF5] = true, [0xFEF7] = true, - [0xFEF9] = true, [0xFEFB] = true, - - -- syriac - - [0x0710] = true, [0x0715] = true, [0x0716] = true, [0x0717] = true, - [0x0718] = true, [0x0719] = true, [0x0728] = true, [0x072A] = true, - [0x072C] = true, [0x071E] = true, -} - -local isol_fina_medi_init = { - [0x0626] = true, [0x0628] = true, [0x062A] = true, [0x062B] = true, - [0x062C] = true, [0x062D] = true, [0x062E] = true, [0x0633] = true, - [0x0634] = true, [0x0635] = true, [0x0636] = true, [0x0637] = true, - [0x0638] = true, [0x0639] = true, [0x063A] = true, [0x063B] = true, - [0x063C] = true, [0x063D] = true, [0x063E] = true, [0x063F] = true, - [0x0640] = true, [0x0641] = true, [0x0642] = true, [0x0643] = true, - [0x0644] = true, [0x0645] = true, [0x0646] = true, [0x0647] = true, - [0x0649] = true, [0x064A] = true, [0x066E] = true, [0x066F] = true, - [0x0678] = true, [0x0679] = true, [0x067A] = true, [0x067B] = true, - [0x067C] = true, [0x067D] = true, [0x067E] = true, [0x067F] = true, - [0x0680] = true, [0x0681] = true, [0x0682] = true, [0x0683] = true, - [0x0684] = true, [0x0685] = true, [0x0686] = true, [0x0687] = true, - [0x069A] = true, [0x069B] = true, [0x069C] = true, [0x069D] = true, - [0x069E] = true, [0x069F] = true, [0x06A0] = true, [0x06A1] = true, - [0x06A2] = true, [0x06A3] = true, [0x06A4] = true, [0x06A5] = true, - [0x06A6] = true, [0x06A7] = true, [0x06A8] = true, [0x06A9] = true, - [0x06AA] = true, [0x06AB] = true, [0x06AC] = true, [0x06AD] = true, - [0x06AE] = true, [0x06AF] = true, [0x06B0] = true, [0x06B1] = true, - [0x06B2] = true, [0x06B3] = true, [0x06B4] = true, [0x06B5] = true, - [0x06B6] = true, [0x06B7] = true, [0x06B8] = true, [0x06B9] = true, - [0x06BA] = true, [0x06BB] = true, [0x06BC] = true, [0x06BD] = true, - [0x06BE] = true, [0x06BF] = true, [0x06C1] = true, [0x06C2] = true, - [0x06CC] = true, [0x06CE] = true, [0x06D0] = true, [0x06D1] = true, - [0x06FA] = true, [0x06FB] = true, [0x06FC] = true, [0x06FF] = true, - [0x0750] = true, [0x0751] = true, [0x0752] = true, [0x0753] = true, - [0x0754] = true, [0x0755] = true, [0x0756] = true, [0x0757] = true, - [0x0758] = true, [0x075C] = true, [0x075D] = true, [0x075E] = true, - [0x075F] = true, [0x0760] = true, [0x0761] = true, [0x0762] = true, - [0x0763] = true, [0x0764] = true, [0x0765] = true, [0x0766] = true, - [0x0767] = true, [0x0768] = true, [0x0769] = true, [0x076A] = true, - [0x076D] = true, [0x076E] = true, [0x076F] = true, [0x0770] = true, - [0x0772] = true, [0x0775] = true, [0x0776] = true, [0x0777] = true, - [0x077A] = true, [0x077B] = true, [0x077C] = true, [0x077D] = true, - [0x077E] = true, [0x077F] = true, - - -- syriac - - [0x0712] = true, [0x0713] = true, [0x0714] = true, [0x071A] = true, - [0x071B] = true, [0x071C] = true, [0x071D] = true, [0x071F] = true, - [0x0720] = true, [0x0721] = true, [0x0722] = true, [0x0723] = true, - [0x0724] = true, [0x0725] = true, [0x0726] = true, [0x0727] = true, - [0x0729] = true, [0x072B] = true, - - -- also - - [zwj] = true, -} - -local arab_warned = { } - --- todo: gref - -local function warning(current,what) - local char = current.char - if not arab_warned[char] then - log.report("analyze","arab: character %s (U+%05X) has no %s class", char, char, what) - arab_warned[char] = true - end -end - -local function finish(first,last) - if last then - if first == last then - local fc = first.char - if isol_fina_medi_init[fc] or isol_fina[fc] then - set_attribute(first,state,4) -- isol - else - warning(first,"isol") - set_attribute(first,state,0) -- error - end - else - local lc = last.char - if isol_fina_medi_init[lc] or isol_fina[lc] then -- why isol here ? - -- if laststate == 1 or laststate == 2 or laststate == 4 then - set_attribute(last,state,3) -- fina - else - warning(last,"fina") - set_attribute(last,state,0) -- error - end - end - first, last = nil, nil - elseif first then - -- first and last are either both set so we never com here - local fc = first.char - if isol_fina_medi_init[fc] or isol_fina[fc] then - set_attribute(first,state,4) -- isol - else - warning(first,"isol") - set_attribute(first,state,0) -- error - end - first = nil - end - return first, last -end - -function methods.arab(head,font,attr) -- maybe make a special version with no trace - local useunicodemarks = analyzers.useunicodemarks - local tfmdata = fontdata[font] - local marks = tfmdata.resources.marks - local first, last, current, done = nil, nil, head, false - while current do - if current.id == glyph_code and current.subtype<256 and current.font == font and not has_attribute(current,state) then - done = true - local char = current.char - if marks[char] or (useunicodemarks and categories[char] == "mn") then - set_attribute(current,state,5) -- mark - elseif isol[char] then -- can be zwj or zwnj too - first, last = finish(first,last) - set_attribute(current,state,4) -- isol - first, last = nil, nil - elseif not first then - if isol_fina_medi_init[char] then - set_attribute(current,state,1) -- init - first, last = first or current, current - elseif isol_fina[char] then - set_attribute(current,state,4) -- isol - first, last = nil, nil - else -- no arab - first, last = finish(first,last) - end - elseif isol_fina_medi_init[char] then - first, last = first or current, current - set_attribute(current,state,2) -- medi - elseif isol_fina[char] then - if not has_attribute(last,state,1) then - -- tricky, we need to check what last may be ! - set_attribute(last,state,2) -- medi - end - set_attribute(current,state,3) -- fina - first, last = nil, nil - elseif char >= 0x0600 and char <= 0x06FF then - set_attribute(current,state,6) -- rest - first, last = finish(first,last) - else --no - first, last = finish(first,last) - end - else - first, last = finish(first,last) - end - current = current.next - end - first, last = finish(first,last) - return head, done -end - -methods.syrc = methods.arab - -directives.register("otf.analyze.useunicodemarks",function(v) - analyzers.useunicodemarks = v -end) - -end -- closure - -do -- begin closure to overcome local limits and interference - if not modules then modules = { } end modules ['luatex-fonts-lua'] = { version = 1.001, comment = "companion to luatex-*.tex", diff --git a/tex/generic/context/luatex/luatex-fonts.lua b/tex/generic/context/luatex/luatex-fonts.lua index f5045a4e3..535519db7 100644 --- a/tex/generic/context/luatex/luatex-fonts.lua +++ b/tex/generic/context/luatex/luatex-fonts.lua @@ -176,9 +176,9 @@ else loadmodule('font-otf.lua') loadmodule('font-otb.lua') loadmodule('node-inj.lua') -- will be replaced (luatex >= .70) + loadmodule('font-ota.lua') loadmodule('font-otn.lua') -- loadmodule('luatex-fonts-chr.lua') - loadmodule('font-ota.lua') loadmodule('luatex-fonts-lua.lua') loadmodule('font-def.lua') loadmodule('luatex-fonts-def.lua') |