diff options
| author | Marius <mariausol@gmail.com> | 2012-09-04 19:20:28 +0300 |
|---|---|---|
| committer | Marius <mariausol@gmail.com> | 2012-09-04 19:20:28 +0300 |
| commit | edd2b22c9d92fd77d47b5fb6517a230470ec7038 (patch) | |
| tree | c4d5386d55a53d423cfd6cb4157b6bda2813d8c7 | |
| parent | e0f42793d14b7571ae6221a15e691d28b3f76d8f (diff) | |
| download | context-edd2b22c9d92fd77d47b5fb6517a230470ec7038.tar.gz | |
beta 2012.09.04 18:08
36 files changed, 1377 insertions, 707 deletions
diff --git a/scripts/context/lua/mtxrun.lua b/scripts/context/lua/mtxrun.lua index b2d3fec54..40fe71c0e 100644 --- a/scripts/context/lua/mtxrun.lua +++ b/scripts/context/lua/mtxrun.lua @@ -8,6 +8,11 @@ if not modules then modules = { } end modules ['mtxrun'] = { license = "see context related readme files" } +-- if not lpeg then require("lpeg") end +-- if not md5 then require("md5") end +-- if not lfs then require("lfs") end +-- if not texconfig then texconfig = { } end + -- one can make a stub: -- -- #!/bin/sh @@ -177,7 +182,8 @@ if not modules then modules = { } end modules ['l-table'] = { license = "see context related readme files" } -local type, next, tostring, tonumber, ipairs, table, string = type, next, tostring, tonumber, ipairs, table, string +local type, next, tostring, tonumber, ipairs = type, next, tostring, tonumber, ipairs +local table, string = table, string local concat, sort, insert, remove = table.concat, table.sort, table.insert, table.remove local format, find, gsub, lower, dump, match = string.format, string.find, string.gsub, string.lower, string.dump, string.match local getmetatable, setmetatable = getmetatable, setmetatable @@ -188,6 +194,8 @@ local getinfo = debug.getinfo -- impact on ConTeXt was not that large; the remaining ipairs already -- have been replaced. In a similar fashion we also hardly used pairs. -- +-- Hm, actually ipairs was retained, but we no longer use it anyway. +-- -- Just in case, we provide the fallbacks as discussed in Programming -- in Lua (http://www.lua.org/pil/7.3.html): @@ -571,12 +579,25 @@ local function do_serialize(root,name,depth,level,indexed) end -- we could check for k (index) being number (cardinal) if root and next(root) then - local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- if compact then + -- -- NOT: for k=1,#root do (we need to quit at nil) + -- for k,v in ipairs(root) do -- can we use next? + -- if not first then first = k end + -- last = last + 1 + -- end + -- end + local first, last = nil, 0 if compact then - -- NOT: for k=1,#root do (we need to quit at nil) - for k,v in ipairs(root) do -- can we use next? - if not first then first = k end - last = last + 1 + last = #root + for k=1,last do + if not root[k] then + last = k - 1 + break + end + end + if last > 0 then + first = 1 end end local sk = sortedkeys(root) @@ -3662,6 +3683,24 @@ local isdir = lfs.isdir local isfile = lfs.isfile local currentdir = lfs.currentdir +-- in case we load outside luatex + +if not isdir then + function isdir(name) + local a = attributes(name) + return a and a.mode == "directory" + end + lfs.isdir = isdir +end + +if not isfile then + function isfile(name) + local a = attributes(name) + return a and a.mode == "file" + end + lfs.isfile = isfile +end + -- handy function dir.current() @@ -4083,6 +4122,18 @@ if not modules then modules = { } end modules ['l-unicode'] = { license = "see context related readme files" } +-- this module will be reorganized + +-- todo: utf.sub replacement (used in syst-aux) + +local concat = table.concat +local type = type +local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs +local lpegmatch, patterns = lpeg.match, lpeg.patterns +local utftype = patterns.utftype +local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format +local utfsplitlines = string.utfsplitlines + if not unicode then unicode = { } @@ -4140,12 +4191,156 @@ if not utf.char then end -local concat = table.concat -local utfchar, utfbyte, utfgsub = utf.char, utf.byte, utf.gsub -local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format -local type = type +if not utf.byte then -local utfsplitlines = string.utfsplitlines + local utf8byte = patterns.utf8byte + + function utf.byte(c) + return lpegmatch(utf8byte,c) + end + +end + +local utfchar, utfbyte = utf.char, utf.byte + +-- As we want to get rid of the (unmaintained) utf library we implement our own +-- variants (in due time an independent module): + +function unicode.filetype(data) + return data and lpegmatch(utftype,data) or "unknown" +end + +local toentities = Cs ( + ( + patterns.utf8one + + ( + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end + )^0 +) + +patterns.toentities = toentities + +function utf.toentities(str) + return lpegmatch(toentities,str) +end + + + + +local one = P(1) +local two = C(1) * C(1) +local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) + +-- actually one of them is already utf ... sort of useless this one + +-- function utf.char(n) +-- if n < 0x80 then +-- return char(n) +-- elseif n < 0x800 then +-- return char( +-- 0xC0 + floor(n/0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x10000 then +-- return char( +-- 0xE0 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x40000 then +-- return char( +-- 0xF0 + floor(n/0x40000), +-- 0x80 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- else +-- -- return char( +-- -- 0xF1 + floor(n/0x1000000), +-- -- 0x80 + floor(n/0x40000), +-- -- 0x80 + floor(n/0x1000), +-- -- 0x80 + (floor(n/0x40) % 0x40), +-- -- 0x80 + (n % 0x40) +-- -- ) +-- return "?" +-- end +-- end +-- +-- merge into: + +local pattern = P("\254\255") * Cs( ( + four / function(a,b,c,d) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(a,b) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + + P("\255\254") * Cs( ( + four / function(b,a,d,c) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(b,a) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + +function string.toutf(s) + return lpegmatch(pattern,s) or s -- todo: utf32 +end + +local validatedutf = Cs ( + ( + patterns.utf8one + + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + + P(1) / "�" + )^0 +) + +patterns.validatedutf = validatedutf + +function string.validutf(str) + return lpegmatch(validatedutf,str) +end + + +utf.length = string.utflength +utf.split = string.utfsplit +utf.splitines = string.utfsplitlines +utf.valid = string.validutf + +if not utf.len then + utf.len = utf.length +end + +-- a replacement for simple gsubs: + +local utf8char = patterns.utf8char + +function utf.remapper(mapping) + local pattern = Cs((utf8char/mapping)^0) + return function(str) + if not str or str == "" then + return "" + else + return lpegmatch(pattern,str) + end + end, pattern +end + +-- local remap = utf.remapper { a = 'd', b = "c", c = "b", d = "a" } +-- print(remap("abcd 1234 abcd")) -- 0 EF BB BF UTF-8 -- 1 FF FE UTF-16-little-endian @@ -4338,11 +4533,22 @@ local function big(c) end end +-- function unicode.utf8_to_utf16(str,littleendian) +-- if littleendian then +-- return char(255,254) .. utfgsub(str,".",little) +-- else +-- return char(254,255) .. utfgsub(str,".",big) +-- end +-- end + +local _, l_remap = utf.remapper(little) +local _, b_remap = utf.remapper(big) + function unicode.utf8_to_utf16(str,littleendian) if littleendian then - return char(255,254) .. utfgsub(str,".",little) + return char(255,254) .. lpegmatch(l_remap,str) else - return char(254,255) .. utfgsub(str,".",big) + return char(254,255) .. lpegmatch(b_remap,str) end end @@ -4364,96 +4570,6 @@ function unicode.xstring(s) end -local lpegmatch = lpeg.match -local patterns = lpeg.patterns -local utftype = patterns.utftype - -function unicode.filetype(data) - return data and lpegmatch(utftype,data) or "unknown" -end - -local toentities = lpeg.Cs ( - ( - patterns.utf8one - + ( - patterns.utf8two - + patterns.utf8three - + patterns.utf8four - ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end - )^0 -) - -patterns.toentities = toentities - -function utf.toentities(str) - return lpegmatch(toentities,str) -end - - - - -local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs - -local one = P(1) -local two = C(1) * C(1) -local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) - --- actually one of them is already utf ... sort of useless this one - -local pattern = P("\254\255") * Cs( ( - four / function(a,b,c,d) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(a,b) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - + P("\255\254") * Cs( ( - four / function(b,a,d,c) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(b,a) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - -function string.toutf(s) - return lpegmatch(pattern,s) or s -- todo: utf32 -end - -local validatedutf = Cs ( - ( - patterns.utf8one - + patterns.utf8two - + patterns.utf8three - + patterns.utf8four - + P(1) / "�" - )^0 -) - -patterns.validatedutf = validatedutf - -function string.validutf(str) - return lpegmatch(validatedutf,str) -end - - -utf.length = string.utflength -utf.split = string.utfsplit -utf.splitines = string.utfsplitlines -utf.valid = string.validutf - -if not utf.len then - utf.len = utf.length -end - - end -- of closure do -- create closure to overcome 200 locals limit @@ -4513,8 +4629,8 @@ local tables = utilities.tables local format, gmatch, rep, gsub = string.format, string.gmatch, string.rep, string.gsub local concat, insert, remove = table.concat, table.insert, table.remove local setmetatable, getmetatable, tonumber, tostring = setmetatable, getmetatable, tonumber, tostring -local type, next, rawset, tonumber = type, next, rawset, tonumber -local lpegmatch = lpeg.match +local type, next, rawset, tonumber, loadstring = type, next, rawset, tonumber, loadstring +local lpegmatch, P, Cs = lpeg.match, lpeg.P, lpeg.Cs function tables.definetable(target) -- defines undefined tables local composed, t, n = nil, { }, 0 @@ -4668,19 +4784,27 @@ function tables.encapsulate(core,capsule,protect) end end -local function serialize(t,r) -- no mixes +local escaped = Cs ( ( + P('\\' ) + + P('"' )/'\\"' + + P('\n')/'\\n' + + P('\r')/'\\r' + + 1 +)^0 ) + +local function serialize(t,r,outer) -- no mixes r[#r+1] = "{" local n = #t if n > 0 then for i=1,n do local v = t[i] local tv = type(v) - if tv == "table" then - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("%q,",v) elseif tv == "number" then r[#r+1] = format("%s,",v) + elseif tv == "table" then + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("%s,",tostring(v)) end @@ -4688,27 +4812,46 @@ local function serialize(t,r) -- no mixes else for k, v in next, t do local tv = type(v) - if tv == "table" then - r[#r+1] = format("[%q]=",k) - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("[%q]=%q,",k,v) elseif tv == "number" then r[#r+1] = format("[%q]=%s,",k,v) + elseif tv == "table" then + r[#r+1] = format("[%q]=",k) + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("[%q]=%s,",k,tostring(v)) end end end - r[#r+1] = "}" + if outer then + r[#r+1] = "}" + else + r[#r+1] = "}," + end return r end function table.fastserialize(t,prefix) - return concat(serialize(t,{ prefix })) + return concat(serialize(t,{ prefix },true)) end --- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7 } }) +-- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7, e = 'f"g\nh' } }) + +function table.load(filename) + if filename then + local t = io.loaddata(filename) + if t and t ~= "" then + t = loadstring(t) + if type(t) == "function" then + t = t() + if type(t) == "table" then + return t + end + end + end + end +end end -- of closure @@ -7502,7 +7645,7 @@ local utf = unicode.utf8 local concat, remove, insert = table.concat, table.remove, table.insert local type, next, setmetatable, getmetatable, tonumber = type, next, setmetatable, getmetatable, tonumber local format, lower, find, match, gsub = string.format, string.lower, string.find, string.match, string.gsub -local utfchar, utfgsub = utf.char, utf.gsub +local utfchar = utf.char local lpegmatch = lpeg.match local P, S, R, C, V, C, Cs = lpeg.P, lpeg.S, lpeg.R, lpeg.C, lpeg.V, lpeg.C, lpeg.Cs @@ -7752,13 +7895,7 @@ local privates_n = { -- keeps track of defined ones } -local function escaped(s) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_u)) - end -end +local escaped = utf.remapper(privates_u) local function unescaped(s) local p = privates_n[s] @@ -7773,13 +7910,7 @@ local function unescaped(s) return p end -local function unprivatized(s,resolve) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_p)) - end -end +local unprivatized = utf.remapper(privates_p) xml.privatetoken = unescaped xml.unprivatized = unprivatized @@ -14612,7 +14743,7 @@ end resolvers.resolve = resolve resolvers.unresolve = unresolve -if os.uname then +if type(os.uname) == "function" then for k, v in next, os.uname() do if not prefixes[k] then @@ -16369,7 +16500,7 @@ function templates.resolve(t,mapping) end -- inspect(utilities.templates.replace("test %one% test", { one = "%two%", two = "two" })) --- inspect(utilities.templates.resolve({ one = "%two%", two = "two" })) +-- inspect(utilities.templates.resolve({ one = "%two%", two = "two", three = "%three%" })) diff --git a/scripts/context/stubs/mswin/mtxrun.lua b/scripts/context/stubs/mswin/mtxrun.lua index b2d3fec54..40fe71c0e 100644 --- a/scripts/context/stubs/mswin/mtxrun.lua +++ b/scripts/context/stubs/mswin/mtxrun.lua @@ -8,6 +8,11 @@ if not modules then modules = { } end modules ['mtxrun'] = { license = "see context related readme files" } +-- if not lpeg then require("lpeg") end +-- if not md5 then require("md5") end +-- if not lfs then require("lfs") end +-- if not texconfig then texconfig = { } end + -- one can make a stub: -- -- #!/bin/sh @@ -177,7 +182,8 @@ if not modules then modules = { } end modules ['l-table'] = { license = "see context related readme files" } -local type, next, tostring, tonumber, ipairs, table, string = type, next, tostring, tonumber, ipairs, table, string +local type, next, tostring, tonumber, ipairs = type, next, tostring, tonumber, ipairs +local table, string = table, string local concat, sort, insert, remove = table.concat, table.sort, table.insert, table.remove local format, find, gsub, lower, dump, match = string.format, string.find, string.gsub, string.lower, string.dump, string.match local getmetatable, setmetatable = getmetatable, setmetatable @@ -188,6 +194,8 @@ local getinfo = debug.getinfo -- impact on ConTeXt was not that large; the remaining ipairs already -- have been replaced. In a similar fashion we also hardly used pairs. -- +-- Hm, actually ipairs was retained, but we no longer use it anyway. +-- -- Just in case, we provide the fallbacks as discussed in Programming -- in Lua (http://www.lua.org/pil/7.3.html): @@ -571,12 +579,25 @@ local function do_serialize(root,name,depth,level,indexed) end -- we could check for k (index) being number (cardinal) if root and next(root) then - local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- if compact then + -- -- NOT: for k=1,#root do (we need to quit at nil) + -- for k,v in ipairs(root) do -- can we use next? + -- if not first then first = k end + -- last = last + 1 + -- end + -- end + local first, last = nil, 0 if compact then - -- NOT: for k=1,#root do (we need to quit at nil) - for k,v in ipairs(root) do -- can we use next? - if not first then first = k end - last = last + 1 + last = #root + for k=1,last do + if not root[k] then + last = k - 1 + break + end + end + if last > 0 then + first = 1 end end local sk = sortedkeys(root) @@ -3662,6 +3683,24 @@ local isdir = lfs.isdir local isfile = lfs.isfile local currentdir = lfs.currentdir +-- in case we load outside luatex + +if not isdir then + function isdir(name) + local a = attributes(name) + return a and a.mode == "directory" + end + lfs.isdir = isdir +end + +if not isfile then + function isfile(name) + local a = attributes(name) + return a and a.mode == "file" + end + lfs.isfile = isfile +end + -- handy function dir.current() @@ -4083,6 +4122,18 @@ if not modules then modules = { } end modules ['l-unicode'] = { license = "see context related readme files" } +-- this module will be reorganized + +-- todo: utf.sub replacement (used in syst-aux) + +local concat = table.concat +local type = type +local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs +local lpegmatch, patterns = lpeg.match, lpeg.patterns +local utftype = patterns.utftype +local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format +local utfsplitlines = string.utfsplitlines + if not unicode then unicode = { } @@ -4140,12 +4191,156 @@ if not utf.char then end -local concat = table.concat -local utfchar, utfbyte, utfgsub = utf.char, utf.byte, utf.gsub -local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format -local type = type +if not utf.byte then -local utfsplitlines = string.utfsplitlines + local utf8byte = patterns.utf8byte + + function utf.byte(c) + return lpegmatch(utf8byte,c) + end + +end + +local utfchar, utfbyte = utf.char, utf.byte + +-- As we want to get rid of the (unmaintained) utf library we implement our own +-- variants (in due time an independent module): + +function unicode.filetype(data) + return data and lpegmatch(utftype,data) or "unknown" +end + +local toentities = Cs ( + ( + patterns.utf8one + + ( + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end + )^0 +) + +patterns.toentities = toentities + +function utf.toentities(str) + return lpegmatch(toentities,str) +end + + + + +local one = P(1) +local two = C(1) * C(1) +local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) + +-- actually one of them is already utf ... sort of useless this one + +-- function utf.char(n) +-- if n < 0x80 then +-- return char(n) +-- elseif n < 0x800 then +-- return char( +-- 0xC0 + floor(n/0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x10000 then +-- return char( +-- 0xE0 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x40000 then +-- return char( +-- 0xF0 + floor(n/0x40000), +-- 0x80 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- else +-- -- return char( +-- -- 0xF1 + floor(n/0x1000000), +-- -- 0x80 + floor(n/0x40000), +-- -- 0x80 + floor(n/0x1000), +-- -- 0x80 + (floor(n/0x40) % 0x40), +-- -- 0x80 + (n % 0x40) +-- -- ) +-- return "?" +-- end +-- end +-- +-- merge into: + +local pattern = P("\254\255") * Cs( ( + four / function(a,b,c,d) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(a,b) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + + P("\255\254") * Cs( ( + four / function(b,a,d,c) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(b,a) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + +function string.toutf(s) + return lpegmatch(pattern,s) or s -- todo: utf32 +end + +local validatedutf = Cs ( + ( + patterns.utf8one + + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + + P(1) / "�" + )^0 +) + +patterns.validatedutf = validatedutf + +function string.validutf(str) + return lpegmatch(validatedutf,str) +end + + +utf.length = string.utflength +utf.split = string.utfsplit +utf.splitines = string.utfsplitlines +utf.valid = string.validutf + +if not utf.len then + utf.len = utf.length +end + +-- a replacement for simple gsubs: + +local utf8char = patterns.utf8char + +function utf.remapper(mapping) + local pattern = Cs((utf8char/mapping)^0) + return function(str) + if not str or str == "" then + return "" + else + return lpegmatch(pattern,str) + end + end, pattern +end + +-- local remap = utf.remapper { a = 'd', b = "c", c = "b", d = "a" } +-- print(remap("abcd 1234 abcd")) -- 0 EF BB BF UTF-8 -- 1 FF FE UTF-16-little-endian @@ -4338,11 +4533,22 @@ local function big(c) end end +-- function unicode.utf8_to_utf16(str,littleendian) +-- if littleendian then +-- return char(255,254) .. utfgsub(str,".",little) +-- else +-- return char(254,255) .. utfgsub(str,".",big) +-- end +-- end + +local _, l_remap = utf.remapper(little) +local _, b_remap = utf.remapper(big) + function unicode.utf8_to_utf16(str,littleendian) if littleendian then - return char(255,254) .. utfgsub(str,".",little) + return char(255,254) .. lpegmatch(l_remap,str) else - return char(254,255) .. utfgsub(str,".",big) + return char(254,255) .. lpegmatch(b_remap,str) end end @@ -4364,96 +4570,6 @@ function unicode.xstring(s) end -local lpegmatch = lpeg.match -local patterns = lpeg.patterns -local utftype = patterns.utftype - -function unicode.filetype(data) - return data and lpegmatch(utftype,data) or "unknown" -end - -local toentities = lpeg.Cs ( - ( - patterns.utf8one - + ( - patterns.utf8two - + patterns.utf8three - + patterns.utf8four - ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end - )^0 -) - -patterns.toentities = toentities - -function utf.toentities(str) - return lpegmatch(toentities,str) -end - - - - -local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs - -local one = P(1) -local two = C(1) * C(1) -local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) - --- actually one of them is already utf ... sort of useless this one - -local pattern = P("\254\255") * Cs( ( - four / function(a,b,c,d) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(a,b) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - + P("\255\254") * Cs( ( - four / function(b,a,d,c) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(b,a) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - -function string.toutf(s) - return lpegmatch(pattern,s) or s -- todo: utf32 -end - -local validatedutf = Cs ( - ( - patterns.utf8one - + patterns.utf8two - + patterns.utf8three - + patterns.utf8four - + P(1) / "�" - )^0 -) - -patterns.validatedutf = validatedutf - -function string.validutf(str) - return lpegmatch(validatedutf,str) -end - - -utf.length = string.utflength -utf.split = string.utfsplit -utf.splitines = string.utfsplitlines -utf.valid = string.validutf - -if not utf.len then - utf.len = utf.length -end - - end -- of closure do -- create closure to overcome 200 locals limit @@ -4513,8 +4629,8 @@ local tables = utilities.tables local format, gmatch, rep, gsub = string.format, string.gmatch, string.rep, string.gsub local concat, insert, remove = table.concat, table.insert, table.remove local setmetatable, getmetatable, tonumber, tostring = setmetatable, getmetatable, tonumber, tostring -local type, next, rawset, tonumber = type, next, rawset, tonumber -local lpegmatch = lpeg.match +local type, next, rawset, tonumber, loadstring = type, next, rawset, tonumber, loadstring +local lpegmatch, P, Cs = lpeg.match, lpeg.P, lpeg.Cs function tables.definetable(target) -- defines undefined tables local composed, t, n = nil, { }, 0 @@ -4668,19 +4784,27 @@ function tables.encapsulate(core,capsule,protect) end end -local function serialize(t,r) -- no mixes +local escaped = Cs ( ( + P('\\' ) + + P('"' )/'\\"' + + P('\n')/'\\n' + + P('\r')/'\\r' + + 1 +)^0 ) + +local function serialize(t,r,outer) -- no mixes r[#r+1] = "{" local n = #t if n > 0 then for i=1,n do local v = t[i] local tv = type(v) - if tv == "table" then - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("%q,",v) elseif tv == "number" then r[#r+1] = format("%s,",v) + elseif tv == "table" then + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("%s,",tostring(v)) end @@ -4688,27 +4812,46 @@ local function serialize(t,r) -- no mixes else for k, v in next, t do local tv = type(v) - if tv == "table" then - r[#r+1] = format("[%q]=",k) - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("[%q]=%q,",k,v) elseif tv == "number" then r[#r+1] = format("[%q]=%s,",k,v) + elseif tv == "table" then + r[#r+1] = format("[%q]=",k) + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("[%q]=%s,",k,tostring(v)) end end end - r[#r+1] = "}" + if outer then + r[#r+1] = "}" + else + r[#r+1] = "}," + end return r end function table.fastserialize(t,prefix) - return concat(serialize(t,{ prefix })) + return concat(serialize(t,{ prefix },true)) end --- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7 } }) +-- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7, e = 'f"g\nh' } }) + +function table.load(filename) + if filename then + local t = io.loaddata(filename) + if t and t ~= "" then + t = loadstring(t) + if type(t) == "function" then + t = t() + if type(t) == "table" then + return t + end + end + end + end +end end -- of closure @@ -7502,7 +7645,7 @@ local utf = unicode.utf8 local concat, remove, insert = table.concat, table.remove, table.insert local type, next, setmetatable, getmetatable, tonumber = type, next, setmetatable, getmetatable, tonumber local format, lower, find, match, gsub = string.format, string.lower, string.find, string.match, string.gsub -local utfchar, utfgsub = utf.char, utf.gsub +local utfchar = utf.char local lpegmatch = lpeg.match local P, S, R, C, V, C, Cs = lpeg.P, lpeg.S, lpeg.R, lpeg.C, lpeg.V, lpeg.C, lpeg.Cs @@ -7752,13 +7895,7 @@ local privates_n = { -- keeps track of defined ones } -local function escaped(s) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_u)) - end -end +local escaped = utf.remapper(privates_u) local function unescaped(s) local p = privates_n[s] @@ -7773,13 +7910,7 @@ local function unescaped(s) return p end -local function unprivatized(s,resolve) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_p)) - end -end +local unprivatized = utf.remapper(privates_p) xml.privatetoken = unescaped xml.unprivatized = unprivatized @@ -14612,7 +14743,7 @@ end resolvers.resolve = resolve resolvers.unresolve = unresolve -if os.uname then +if type(os.uname) == "function" then for k, v in next, os.uname() do if not prefixes[k] then @@ -16369,7 +16500,7 @@ function templates.resolve(t,mapping) end -- inspect(utilities.templates.replace("test %one% test", { one = "%two%", two = "two" })) --- inspect(utilities.templates.resolve({ one = "%two%", two = "two" })) +-- inspect(utilities.templates.resolve({ one = "%two%", two = "two", three = "%three%" })) diff --git a/scripts/context/stubs/unix/mtxrun b/scripts/context/stubs/unix/mtxrun index b2d3fec54..40fe71c0e 100644 --- a/scripts/context/stubs/unix/mtxrun +++ b/scripts/context/stubs/unix/mtxrun @@ -8,6 +8,11 @@ if not modules then modules = { } end modules ['mtxrun'] = { license = "see context related readme files" } +-- if not lpeg then require("lpeg") end +-- if not md5 then require("md5") end +-- if not lfs then require("lfs") end +-- if not texconfig then texconfig = { } end + -- one can make a stub: -- -- #!/bin/sh @@ -177,7 +182,8 @@ if not modules then modules = { } end modules ['l-table'] = { license = "see context related readme files" } -local type, next, tostring, tonumber, ipairs, table, string = type, next, tostring, tonumber, ipairs, table, string +local type, next, tostring, tonumber, ipairs = type, next, tostring, tonumber, ipairs +local table, string = table, string local concat, sort, insert, remove = table.concat, table.sort, table.insert, table.remove local format, find, gsub, lower, dump, match = string.format, string.find, string.gsub, string.lower, string.dump, string.match local getmetatable, setmetatable = getmetatable, setmetatable @@ -188,6 +194,8 @@ local getinfo = debug.getinfo -- impact on ConTeXt was not that large; the remaining ipairs already -- have been replaced. In a similar fashion we also hardly used pairs. -- +-- Hm, actually ipairs was retained, but we no longer use it anyway. +-- -- Just in case, we provide the fallbacks as discussed in Programming -- in Lua (http://www.lua.org/pil/7.3.html): @@ -571,12 +579,25 @@ local function do_serialize(root,name,depth,level,indexed) end -- we could check for k (index) being number (cardinal) if root and next(root) then - local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- if compact then + -- -- NOT: for k=1,#root do (we need to quit at nil) + -- for k,v in ipairs(root) do -- can we use next? + -- if not first then first = k end + -- last = last + 1 + -- end + -- end + local first, last = nil, 0 if compact then - -- NOT: for k=1,#root do (we need to quit at nil) - for k,v in ipairs(root) do -- can we use next? - if not first then first = k end - last = last + 1 + last = #root + for k=1,last do + if not root[k] then + last = k - 1 + break + end + end + if last > 0 then + first = 1 end end local sk = sortedkeys(root) @@ -3662,6 +3683,24 @@ local isdir = lfs.isdir local isfile = lfs.isfile local currentdir = lfs.currentdir +-- in case we load outside luatex + +if not isdir then + function isdir(name) + local a = attributes(name) + return a and a.mode == "directory" + end + lfs.isdir = isdir +end + +if not isfile then + function isfile(name) + local a = attributes(name) + return a and a.mode == "file" + end + lfs.isfile = isfile +end + -- handy function dir.current() @@ -4083,6 +4122,18 @@ if not modules then modules = { } end modules ['l-unicode'] = { license = "see context related readme files" } +-- this module will be reorganized + +-- todo: utf.sub replacement (used in syst-aux) + +local concat = table.concat +local type = type +local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs +local lpegmatch, patterns = lpeg.match, lpeg.patterns +local utftype = patterns.utftype +local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format +local utfsplitlines = string.utfsplitlines + if not unicode then unicode = { } @@ -4140,12 +4191,156 @@ if not utf.char then end -local concat = table.concat -local utfchar, utfbyte, utfgsub = utf.char, utf.byte, utf.gsub -local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format -local type = type +if not utf.byte then -local utfsplitlines = string.utfsplitlines + local utf8byte = patterns.utf8byte + + function utf.byte(c) + return lpegmatch(utf8byte,c) + end + +end + +local utfchar, utfbyte = utf.char, utf.byte + +-- As we want to get rid of the (unmaintained) utf library we implement our own +-- variants (in due time an independent module): + +function unicode.filetype(data) + return data and lpegmatch(utftype,data) or "unknown" +end + +local toentities = Cs ( + ( + patterns.utf8one + + ( + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end + )^0 +) + +patterns.toentities = toentities + +function utf.toentities(str) + return lpegmatch(toentities,str) +end + + + + +local one = P(1) +local two = C(1) * C(1) +local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) + +-- actually one of them is already utf ... sort of useless this one + +-- function utf.char(n) +-- if n < 0x80 then +-- return char(n) +-- elseif n < 0x800 then +-- return char( +-- 0xC0 + floor(n/0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x10000 then +-- return char( +-- 0xE0 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x40000 then +-- return char( +-- 0xF0 + floor(n/0x40000), +-- 0x80 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- else +-- -- return char( +-- -- 0xF1 + floor(n/0x1000000), +-- -- 0x80 + floor(n/0x40000), +-- -- 0x80 + floor(n/0x1000), +-- -- 0x80 + (floor(n/0x40) % 0x40), +-- -- 0x80 + (n % 0x40) +-- -- ) +-- return "?" +-- end +-- end +-- +-- merge into: + +local pattern = P("\254\255") * Cs( ( + four / function(a,b,c,d) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(a,b) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + + P("\255\254") * Cs( ( + four / function(b,a,d,c) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(b,a) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + +function string.toutf(s) + return lpegmatch(pattern,s) or s -- todo: utf32 +end + +local validatedutf = Cs ( + ( + patterns.utf8one + + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + + P(1) / "�" + )^0 +) + +patterns.validatedutf = validatedutf + +function string.validutf(str) + return lpegmatch(validatedutf,str) +end + + +utf.length = string.utflength +utf.split = string.utfsplit +utf.splitines = string.utfsplitlines +utf.valid = string.validutf + +if not utf.len then + utf.len = utf.length +end + +-- a replacement for simple gsubs: + +local utf8char = patterns.utf8char + +function utf.remapper(mapping) + local pattern = Cs((utf8char/mapping)^0) + return function(str) + if not str or str == "" then + return "" + else + return lpegmatch(pattern,str) + end + end, pattern +end + +-- local remap = utf.remapper { a = 'd', b = "c", c = "b", d = "a" } +-- print(remap("abcd 1234 abcd")) -- 0 EF BB BF UTF-8 -- 1 FF FE UTF-16-little-endian @@ -4338,11 +4533,22 @@ local function big(c) end end +-- function unicode.utf8_to_utf16(str,littleendian) +-- if littleendian then +-- return char(255,254) .. utfgsub(str,".",little) +-- else +-- return char(254,255) .. utfgsub(str,".",big) +-- end +-- end + +local _, l_remap = utf.remapper(little) +local _, b_remap = utf.remapper(big) + function unicode.utf8_to_utf16(str,littleendian) if littleendian then - return char(255,254) .. utfgsub(str,".",little) + return char(255,254) .. lpegmatch(l_remap,str) else - return char(254,255) .. utfgsub(str,".",big) + return char(254,255) .. lpegmatch(b_remap,str) end end @@ -4364,96 +4570,6 @@ function unicode.xstring(s) end -local lpegmatch = lpeg.match -local patterns = lpeg.patterns -local utftype = patterns.utftype - -function unicode.filetype(data) - return data and lpegmatch(utftype,data) or "unknown" -end - -local toentities = lpeg.Cs ( - ( - patterns.utf8one - + ( - patterns.utf8two - + patterns.utf8three - + patterns.utf8four - ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end - )^0 -) - -patterns.toentities = toentities - -function utf.toentities(str) - return lpegmatch(toentities,str) -end - - - - -local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs - -local one = P(1) -local two = C(1) * C(1) -local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) - --- actually one of them is already utf ... sort of useless this one - -local pattern = P("\254\255") * Cs( ( - four / function(a,b,c,d) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(a,b) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - + P("\255\254") * Cs( ( - four / function(b,a,d,c) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(b,a) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - -function string.toutf(s) - return lpegmatch(pattern,s) or s -- todo: utf32 -end - -local validatedutf = Cs ( - ( - patterns.utf8one - + patterns.utf8two - + patterns.utf8three - + patterns.utf8four - + P(1) / "�" - )^0 -) - -patterns.validatedutf = validatedutf - -function string.validutf(str) - return lpegmatch(validatedutf,str) -end - - -utf.length = string.utflength -utf.split = string.utfsplit -utf.splitines = string.utfsplitlines -utf.valid = string.validutf - -if not utf.len then - utf.len = utf.length -end - - end -- of closure do -- create closure to overcome 200 locals limit @@ -4513,8 +4629,8 @@ local tables = utilities.tables local format, gmatch, rep, gsub = string.format, string.gmatch, string.rep, string.gsub local concat, insert, remove = table.concat, table.insert, table.remove local setmetatable, getmetatable, tonumber, tostring = setmetatable, getmetatable, tonumber, tostring -local type, next, rawset, tonumber = type, next, rawset, tonumber -local lpegmatch = lpeg.match +local type, next, rawset, tonumber, loadstring = type, next, rawset, tonumber, loadstring +local lpegmatch, P, Cs = lpeg.match, lpeg.P, lpeg.Cs function tables.definetable(target) -- defines undefined tables local composed, t, n = nil, { }, 0 @@ -4668,19 +4784,27 @@ function tables.encapsulate(core,capsule,protect) end end -local function serialize(t,r) -- no mixes +local escaped = Cs ( ( + P('\\' ) + + P('"' )/'\\"' + + P('\n')/'\\n' + + P('\r')/'\\r' + + 1 +)^0 ) + +local function serialize(t,r,outer) -- no mixes r[#r+1] = "{" local n = #t if n > 0 then for i=1,n do local v = t[i] local tv = type(v) - if tv == "table" then - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("%q,",v) elseif tv == "number" then r[#r+1] = format("%s,",v) + elseif tv == "table" then + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("%s,",tostring(v)) end @@ -4688,27 +4812,46 @@ local function serialize(t,r) -- no mixes else for k, v in next, t do local tv = type(v) - if tv == "table" then - r[#r+1] = format("[%q]=",k) - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("[%q]=%q,",k,v) elseif tv == "number" then r[#r+1] = format("[%q]=%s,",k,v) + elseif tv == "table" then + r[#r+1] = format("[%q]=",k) + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("[%q]=%s,",k,tostring(v)) end end end - r[#r+1] = "}" + if outer then + r[#r+1] = "}" + else + r[#r+1] = "}," + end return r end function table.fastserialize(t,prefix) - return concat(serialize(t,{ prefix })) + return concat(serialize(t,{ prefix },true)) end --- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7 } }) +-- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7, e = 'f"g\nh' } }) + +function table.load(filename) + if filename then + local t = io.loaddata(filename) + if t and t ~= "" then + t = loadstring(t) + if type(t) == "function" then + t = t() + if type(t) == "table" then + return t + end + end + end + end +end end -- of closure @@ -7502,7 +7645,7 @@ local utf = unicode.utf8 local concat, remove, insert = table.concat, table.remove, table.insert local type, next, setmetatable, getmetatable, tonumber = type, next, setmetatable, getmetatable, tonumber local format, lower, find, match, gsub = string.format, string.lower, string.find, string.match, string.gsub -local utfchar, utfgsub = utf.char, utf.gsub +local utfchar = utf.char local lpegmatch = lpeg.match local P, S, R, C, V, C, Cs = lpeg.P, lpeg.S, lpeg.R, lpeg.C, lpeg.V, lpeg.C, lpeg.Cs @@ -7752,13 +7895,7 @@ local privates_n = { -- keeps track of defined ones } -local function escaped(s) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_u)) - end -end +local escaped = utf.remapper(privates_u) local function unescaped(s) local p = privates_n[s] @@ -7773,13 +7910,7 @@ local function unescaped(s) return p end -local function unprivatized(s,resolve) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_p)) - end -end +local unprivatized = utf.remapper(privates_p) xml.privatetoken = unescaped xml.unprivatized = unprivatized @@ -14612,7 +14743,7 @@ end resolvers.resolve = resolve resolvers.unresolve = unresolve -if os.uname then +if type(os.uname) == "function" then for k, v in next, os.uname() do if not prefixes[k] then @@ -16369,7 +16500,7 @@ function templates.resolve(t,mapping) end -- inspect(utilities.templates.replace("test %one% test", { one = "%two%", two = "two" })) --- inspect(utilities.templates.resolve({ one = "%two%", two = "two" })) +-- inspect(utilities.templates.resolve({ one = "%two%", two = "two", three = "%three%" })) diff --git a/tex/context/base/back-exp.lua b/tex/context/base/back-exp.lua index f73ee2d55..46ce4f96c 100644 --- a/tex/context/base/back-exp.lua +++ b/tex/context/base/back-exp.lua @@ -22,7 +22,7 @@ local next, type = next, type local format, match, concat, rep, sub, gsub, gmatch, find = string.format, string.match, table.concat, string.rep, string.sub, string.gsub, string.gmatch, string.find local validstring = string.valid local lpegmatch = lpeg.match -local utfchar, utfbyte, utfsub, utfgsub = utf.char, utf.byte, utf.sub, utf.gsub +local utfchar, utfbyte = utf.char, utf.byte local insert, remove = table.insert, table.remove local topoints = number.topoints local utfvalues = string.utfvalues @@ -155,6 +155,8 @@ local somespace = { [0x20] = true, [" "] = true } -- for testing local entities = { ["&"] = "&", [">"] = ">", ["<"] = "<" } local attribentities = { ["&"] = "&", [">"] = ">", ["<"] = "<", ['"'] = "quot;" } +local entityremapper = utf.remapper(entities) + local alignmapping = { flushright = "right", middle = "center", @@ -1352,7 +1354,7 @@ local function begintag(result,element,nature,depth,di,skip) end result[#result+1] = format("%s<metadata>\n",spaces[depth]) for k, v in table.sortedpairs(metadata) do - v = utfgsub(v,".",entities) + v = entityremapper(v) result[#result+1] = format("%s<metavariable name=%q>%s</metavariable>\n",spaces[depth+1],k,v) end result[#result+1] = format("%s</metadata>\n",spaces[depth]) @@ -1410,7 +1412,7 @@ local function flushtree(result,data,nature,depth) -- whatever elseif di.content then -- already has breaks - local content = utfgsub(di.content,".",entities) + local content = entityremapper(di.content) if i == nofdata and sub(content,-1) == "\n" then -- move check -- can be an end of line in par but can also be the last line if trace_spacing then diff --git a/tex/context/base/char-cjk.lua b/tex/context/base/char-cjk.lua index 6232ebcad..3d7de1423 100644 --- a/tex/context/base/char-cjk.lua +++ b/tex/context/base/char-cjk.lua @@ -12,7 +12,8 @@ local floor = math.floor local format = string.format local utfchar = utf.char -local ranges = characters.ranges +local ranges = characters.ranges +local allocate = utilities.storage.allocate -- Hangul Syllable @@ -209,7 +210,7 @@ local remapped = { -- this might be merged into char-def.lua [0x11C2] = 0x314E, -- H } -characters.hangul = { +characters.hangul = allocate { decomposed = decomposed, description = description, leadconsonant = leadconsonant, diff --git a/tex/context/base/char-utf.lua b/tex/context/base/char-utf.lua index cefaca4e3..bd7463522 100644 --- a/tex/context/base/char-utf.lua +++ b/tex/context/base/char-utf.lua @@ -19,7 +19,7 @@ in special kinds of output (for instance <l n='pdf'/>).</p> over a string.</p> --ldx]]-- -local utfchar, utfbyte, utfgsub = utf.char, utf.byte, utf.gsub +local utfchar, utfbyte = utf.char, utf.byte local concat, gmatch, gsub, find = table.concat, string.gmatch, string.gsub, string.find local utfcharacters, utfvalues = string.utfcharacters, string.utfvalues local allocate = utilities.storage.allocate @@ -210,9 +210,13 @@ end private.set = set -function private.escape (str) return gsub(str,"(.)", escapes) end -function private.replace(str) return utfgsub(str,"(.)", low ) end -function private.revert (str) return utfgsub(str,"(.)", high ) end +-- function private.escape (str) return gsub(str,"(.)", escapes) end +-- function private.replace(str) return utfgsub(str,"(.)", low ) end +-- function private.revert (str) return utfgsub(str,"(.)", high ) end + +private.escape = utf.remapper(escapes) +private.replace = utf.remapper(low) +private.revert = utf.remapper(high) for ch in gmatch(special,".") do set(ch) end diff --git a/tex/context/base/colo-ini.lua b/tex/context/base/colo-ini.lua index 253027238..f2437cd9d 100644 --- a/tex/context/base/colo-ini.lua +++ b/tex/context/base/colo-ini.lua @@ -846,3 +846,21 @@ end -- context.popcatcodes() -- end +-- handy + +local models = storage.allocate { "gray", "rgb", "cmyk", "spot" } + +colors.models = models -- check for usage elsewhere + +function attributes.colors.spec(name) + local t = colorvalues[attributes_list[name]] or colorvalues[attributes_list.black] + return { + model = models[t[1]], + s = t[2], + r = t[3], g = t[4], b = t[5], + c = t[6], m = t[7], y = t[8], k = t[9], + } +end + +-- inspect(attributes.colors.spec("red")) +-- inspect(attributes.colors.spec("red socks")) diff --git a/tex/context/base/cont-new.mkii b/tex/context/base/cont-new.mkii index 5bc49b4d0..8fbb41f87 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{2012.08.28 20:00} +\newcontextversion{2012.09.04 18:08} %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 14fc10ce7..13e895931 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{2012.08.28 20:00} +\newcontextversion{2012.09.04 18:08} %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.mkii b/tex/context/base/context.mkii index 0fd18a46a..efb1028ec 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{2012.08.28 20:00} +\edef\contextversion{2012.09.04 18:08} %D For those who want to use this: diff --git a/tex/context/base/context.mkiv b/tex/context/base/context.mkiv index 1a587fcc4..3905a49c6 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{2012.08.28 20:00} +\edef\contextversion{2012.09.04 18:08} %D For those who want to use this: diff --git a/tex/context/base/data-pre.lua b/tex/context/base/data-pre.lua index 5b25c5f87..4e82b6186 100644 --- a/tex/context/base/data-pre.lua +++ b/tex/context/base/data-pre.lua @@ -146,7 +146,7 @@ end resolvers.resolve = resolve resolvers.unresolve = unresolve -if os.uname then +if type(os.uname) == "function" then for k, v in next, os.uname() do if not prefixes[k] then diff --git a/tex/context/base/font-afk.lua b/tex/context/base/font-afk.lua index 2674cb233..8b65b0631 100644 --- a/tex/context/base/font-afk.lua +++ b/tex/context/base/font-afk.lua @@ -13,9 +13,10 @@ anything larger is encoding dependent. An interesting complication is that a character can be in an encoding twice but is hashed once.</p> --ldx]]-- +local allocate = utilities.storage.allocate fonts.handlers.afm.helpdata = { - ligatures = { -- okay, nowadays we could parse the name but type 1 fonts + ligatures = allocate { -- okay, nowadays we could parse the name but type 1 fonts ['f'] = { -- don't have that many ligatures anyway { 'f', 'ff' }, { 'i', 'fi' }, @@ -37,7 +38,7 @@ fonts.handlers.afm.helpdata = { { 'j', 'ij' } }, }, - texligatures = { + texligatures = allocate { -- ['space'] = { -- { 'L', 'Lslash' }, -- { 'l', 'lslash' } @@ -61,7 +62,7 @@ fonts.handlers.afm.helpdata = { { 'hyphen', 'emdash' } } }, - leftkerned = { + leftkerned = allocate { AEligature = "A", aeligature = "a", OEligature = "O", oeligature = "o", IJligature = "I", ijligature = "i", @@ -70,7 +71,7 @@ fonts.handlers.afm.helpdata = { IJ = "I", ij = "i", Ssharp = "S", ssharp = "s", }, - rightkerned = { + rightkerned = allocate { AEligature = "E", aeligature = "e", OEligature = "E", oeligature = "e", IJligature = "J", ijligature = "j", @@ -79,7 +80,7 @@ fonts.handlers.afm.helpdata = { IJ = "J", ij = "j", Ssharp = "S", ssharp = "s", }, - bothkerned = { + bothkerned = allocate { Acircumflex = "A", acircumflex = "a", Ccircumflex = "C", ccircumflex = "c", Ecircumflex = "E", ecircumflex = "e", diff --git a/tex/context/base/font-gds.lua b/tex/context/base/font-gds.lua index ce26c66c5..b4b073433 100644 --- a/tex/context/base/font-gds.lua +++ b/tex/context/base/font-gds.lua @@ -234,7 +234,7 @@ end -- colorschemes -local colorschemes = fontgoodies.colorschemes or { } +local colorschemes = fontgoodies.colorschemes or allocate { } fontgoodies.colorschemes = colorschemes colorschemes.data = colorschemes.data or { } @@ -603,7 +603,7 @@ end fontgoodies.register("compositions", initialize) -local designsizes = fontgoodies.designsizes or { } +local designsizes = fontgoodies.designsizes or allocate() fontgoodies.designsizes = designsizes local designdata = designsizes.data or allocate() diff --git a/tex/context/base/font-syn.lua b/tex/context/base/font-syn.lua index 31af73d97..d99114998 100644 --- a/tex/context/base/font-syn.lua +++ b/tex/context/base/font-syn.lua @@ -15,7 +15,6 @@ local find, gmatch = string.find, string.gmatch local concat, sort, format = table.concat, table.sort, string.format local serialize = table.serialize local lpegmatch = lpeg.match -local utfgsub, utflower = utf.gsub, utf.lower local unpack = unpack or table.unpack local allocate = utilities.storage.allocate @@ -34,13 +33,13 @@ using a table that has keys filtered from the font related files.</p> fonts = fonts or { } -- also used elsewhere -local names = font.names or { } +local names = font.names or allocate { } fonts.names = names local filters = names.filters or { } names.filters = filters -names.data = names.data or { } +names.data = names.data or allocate { } names.version = 1.110 names.basename = "names" @@ -310,8 +309,6 @@ end local function cleanname(name) return (gsub(lower(name),"[^%a%d]","")) - -- once we can load files with utf names, we can play with the following: - -- return (utfgsub(utfgsub(lower(str),"[^%a%A%d]",""),"%s","")) end local function cleanfilename(fullname,defaultsuffix) diff --git a/tex/context/base/l-dir.lua b/tex/context/base/l-dir.lua index 71de3114e..3deb660ce 100644 --- a/tex/context/base/l-dir.lua +++ b/tex/context/base/l-dir.lua @@ -25,6 +25,24 @@ local isdir = lfs.isdir local isfile = lfs.isfile local currentdir = lfs.currentdir +-- in case we load outside luatex + +if not isdir then + function isdir(name) + local a = attributes(name) + return a and a.mode == "directory" + end + lfs.isdir = isdir +end + +if not isfile then + function isfile(name) + local a = attributes(name) + return a and a.mode == "file" + end + lfs.isfile = isfile +end + -- handy function dir.current() diff --git a/tex/context/base/l-table.lua b/tex/context/base/l-table.lua index cc69af0ef..413d31eb5 100644 --- a/tex/context/base/l-table.lua +++ b/tex/context/base/l-table.lua @@ -6,7 +6,8 @@ if not modules then modules = { } end modules ['l-table'] = { license = "see context related readme files" } -local type, next, tostring, tonumber, ipairs, table, string = type, next, tostring, tonumber, ipairs, table, string +local type, next, tostring, tonumber, ipairs = type, next, tostring, tonumber, ipairs +local table, string = table, string local concat, sort, insert, remove = table.concat, table.sort, table.insert, table.remove local format, find, gsub, lower, dump, match = string.format, string.find, string.gsub, string.lower, string.dump, string.match local getmetatable, setmetatable = getmetatable, setmetatable @@ -17,6 +18,8 @@ local getinfo = debug.getinfo -- impact on ConTeXt was not that large; the remaining ipairs already -- have been replaced. In a similar fashion we also hardly used pairs. -- +-- Hm, actually ipairs was retained, but we no longer use it anyway. +-- -- Just in case, we provide the fallbacks as discussed in Programming -- in Lua (http://www.lua.org/pil/7.3.html): @@ -400,12 +403,25 @@ local function do_serialize(root,name,depth,level,indexed) end -- we could check for k (index) being number (cardinal) if root and next(root) then - local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- if compact then + -- -- NOT: for k=1,#root do (we need to quit at nil) + -- for k,v in ipairs(root) do -- can we use next? + -- if not first then first = k end + -- last = last + 1 + -- end + -- end + local first, last = nil, 0 if compact then - -- NOT: for k=1,#root do (we need to quit at nil) - for k,v in ipairs(root) do -- can we use next? - if not first then first = k end - last = last + 1 + last = #root + for k=1,last do + if not root[k] then + last = k - 1 + break + end + end + if last > 0 then + first = 1 end end local sk = sortedkeys(root) diff --git a/tex/context/base/l-unicode.lua b/tex/context/base/l-unicode.lua index 4e68dc6e1..6447fb040 100644 --- a/tex/context/base/l-unicode.lua +++ b/tex/context/base/l-unicode.lua @@ -6,6 +6,18 @@ if not modules then modules = { } end modules ['l-unicode'] = { license = "see context related readme files" } +-- this module will be reorganized + +-- todo: utf.sub replacement (used in syst-aux) + +local concat = table.concat +local type = type +local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs +local lpegmatch, patterns = lpeg.match, lpeg.patterns +local utftype = patterns.utftype +local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format +local utfsplitlines = string.utfsplitlines + if not unicode then unicode = { } @@ -63,12 +75,180 @@ if not utf.char then end -local concat = table.concat -local utfchar, utfbyte, utfgsub = utf.char, utf.byte, utf.gsub -local char, byte, find, bytepairs, utfvalues, format = string.char, string.byte, string.find, string.bytepairs, string.utfvalues, string.format -local type = type +if not utf.byte then -local utfsplitlines = string.utfsplitlines + local utf8byte = patterns.utf8byte + + function utf.byte(c) + return lpegmatch(utf8byte,c) + end + +end + +local utfchar, utfbyte = utf.char, utf.byte + +-- As we want to get rid of the (unmaintained) utf library we implement our own +-- variants (in due time an independent module): + +function unicode.filetype(data) + return data and lpegmatch(utftype,data) or "unknown" +end + +local toentities = Cs ( + ( + patterns.utf8one + + ( + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end + )^0 +) + +patterns.toentities = toentities + +function utf.toentities(str) + return lpegmatch(toentities,str) +end + +--~ local utfchr = { } -- 60K -> 2.638 M extra mem but currently not called that often (on latin) +--~ +--~ setmetatable(utfchr, { __index = function(t,k) local v = utfchar(k) t[k] = v return v end } ) +--~ +--~ collectgarbage("collect") +--~ local u = collectgarbage("count")*1024 +--~ local t = os.clock() +--~ for i=1,1000 do +--~ for i=1,600 do +--~ local a = utfchr[i] +--~ end +--~ end +--~ print(os.clock()-t,collectgarbage("count")*1024-u) + +--~ collectgarbage("collect") +--~ local t = os.clock() +--~ for i=1,1000 do +--~ for i=1,600 do +--~ local a = utfchar(i) +--~ end +--~ end +--~ print(os.clock()-t,collectgarbage("count")*1024-u) + +--~ local byte = string.byte +--~ local utfchar = utf.char +--~ local lpegmatch = lpeg.match, lpeg.P, lpeg.C, lpeg.R, lpeg.Cs + +local one = P(1) +local two = C(1) * C(1) +local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) + +-- actually one of them is already utf ... sort of useless this one + +-- function utf.char(n) +-- if n < 0x80 then +-- return char(n) +-- elseif n < 0x800 then +-- return char( +-- 0xC0 + floor(n/0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x10000 then +-- return char( +-- 0xE0 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- elseif n < 0x40000 then +-- return char( +-- 0xF0 + floor(n/0x40000), +-- 0x80 + floor(n/0x1000), +-- 0x80 + (floor(n/0x40) % 0x40), +-- 0x80 + (n % 0x40) +-- ) +-- else +-- -- return char( +-- -- 0xF1 + floor(n/0x1000000), +-- -- 0x80 + floor(n/0x40000), +-- -- 0x80 + floor(n/0x1000), +-- -- 0x80 + (floor(n/0x40) % 0x40), +-- -- 0x80 + (n % 0x40) +-- -- ) +-- return "?" +-- end +-- end +-- +-- merge into: + +local pattern = P("\254\255") * Cs( ( + four / function(a,b,c,d) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(a,b) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + + P("\255\254") * Cs( ( + four / function(b,a,d,c) + local ab = 0xFF * byte(a) + byte(b) + local cd = 0xFF * byte(c) + byte(d) + return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) + end + + two / function(b,a) + return utfchar(byte(a)*256 + byte(b)) + end + + one + )^1 ) + +function string.toutf(s) + return lpegmatch(pattern,s) or s -- todo: utf32 +end + +local validatedutf = Cs ( + ( + patterns.utf8one + + patterns.utf8two + + patterns.utf8three + + patterns.utf8four + + P(1) / "�" + )^0 +) + +patterns.validatedutf = validatedutf + +function string.validutf(str) + return lpegmatch(validatedutf,str) +end + + +utf.length = string.utflength +utf.split = string.utfsplit +utf.splitines = string.utfsplitlines +utf.valid = string.validutf + +if not utf.len then + utf.len = utf.length +end + +-- a replacement for simple gsubs: + +local utf8char = patterns.utf8char + +function utf.remapper(mapping) + local pattern = Cs((utf8char/mapping)^0) + return function(str) + if not str or str == "" then + return "" + else + return lpegmatch(pattern,str) + end + end, pattern +end + +-- local remap = utf.remapper { a = 'd', b = "c", c = "b", d = "a" } +-- print(remap("abcd 1234 abcd")) -- 0 EF BB BF UTF-8 -- 1 FF FE UTF-16-little-endian @@ -364,11 +544,22 @@ local function big(c) end end +-- function unicode.utf8_to_utf16(str,littleendian) +-- if littleendian then +-- return char(255,254) .. utfgsub(str,".",little) +-- else +-- return char(254,255) .. utfgsub(str,".",big) +-- end +-- end + +local _, l_remap = utf.remapper(little) +local _, b_remap = utf.remapper(big) + function unicode.utf8_to_utf16(str,littleendian) if littleendian then - return char(255,254) .. utfgsub(str,".",little) + return char(255,254) .. lpegmatch(l_remap,str) else - return char(254,255) .. utfgsub(str,".",big) + return char(254,255) .. lpegmatch(b_remap,str) end end @@ -388,118 +579,3 @@ end function unicode.xstring(s) return format("0x%05X",type(s) == "number" and s or utfbyte(s)) end - ---~ print(unicode.utfcodes(str)) - -local lpegmatch = lpeg.match -local patterns = lpeg.patterns -local utftype = patterns.utftype - -function unicode.filetype(data) - return data and lpegmatch(utftype,data) or "unknown" -end - -local toentities = lpeg.Cs ( - ( - patterns.utf8one - + ( - patterns.utf8two - + patterns.utf8three - + patterns.utf8four - ) / function(s) local b = utfbyte(s) if b < 127 then return s else return format("&#%X;",b) end end - )^0 -) - -patterns.toentities = toentities - -function utf.toentities(str) - return lpegmatch(toentities,str) -end - ---~ local utfchr = { } -- 60K -> 2.638 M extra mem but currently not called that often (on latin) ---~ ---~ setmetatable(utfchr, { __index = function(t,k) local v = utfchar(k) t[k] = v return v end } ) ---~ ---~ collectgarbage("collect") ---~ local u = collectgarbage("count")*1024 ---~ local t = os.clock() ---~ for i=1,1000 do ---~ for i=1,600 do ---~ local a = utfchr[i] ---~ end ---~ end ---~ print(os.clock()-t,collectgarbage("count")*1024-u) - ---~ collectgarbage("collect") ---~ local t = os.clock() ---~ for i=1,1000 do ---~ for i=1,600 do ---~ local a = utfchar(i) ---~ end ---~ end ---~ print(os.clock()-t,collectgarbage("count")*1024-u) - ---~ local byte = string.byte ---~ local utfchar = utf.char ---~ local lpegmatch = lpeg.match, lpeg.P, lpeg.C, lpeg.R, lpeg.Cs - -local P, C, R, Cs = lpeg.P, lpeg.C, lpeg.R, lpeg.Cs - -local one = P(1) -local two = C(1) * C(1) -local four = C(R(utfchar(0xD8),utfchar(0xFF))) * C(1) * C(1) * C(1) - --- actually one of them is already utf ... sort of useless this one - -local pattern = P("\254\255") * Cs( ( - four / function(a,b,c,d) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(a,b) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - + P("\255\254") * Cs( ( - four / function(b,a,d,c) - local ab = 0xFF * byte(a) + byte(b) - local cd = 0xFF * byte(c) + byte(d) - return utfchar((ab-0xD800)*0x400 + (cd-0xDC00) + 0x10000) - end - + two / function(b,a) - return utfchar(byte(a)*256 + byte(b)) - end - + one - )^1 ) - -function string.toutf(s) - return lpegmatch(pattern,s) or s -- todo: utf32 -end - -local validatedutf = Cs ( - ( - patterns.utf8one - + patterns.utf8two - + patterns.utf8three - + patterns.utf8four - + P(1) / "�" - )^0 -) - -patterns.validatedutf = validatedutf - -function string.validutf(str) - return lpegmatch(validatedutf,str) -end - - -utf.length = string.utflength -utf.split = string.utfsplit -utf.splitines = string.utfsplitlines -utf.valid = string.validutf - -if not utf.len then - utf.len = utf.length -end diff --git a/tex/context/base/lang-url.lua b/tex/context/base/lang-url.lua index 1817fc280..3b354216a 100644 --- a/tex/context/base/lang-url.lua +++ b/tex/context/base/lang-url.lua @@ -9,7 +9,7 @@ if not modules then modules = { } end modules ['lang-url'] = { local utf = unicode.utf8 local utfcharacters, utfvalues = string.utfcharacters, string.utfvalues -local utfbyte, utfchar, utfgsub = utf.byte, utf.char, utf.gsub +local utfbyte, utfchar = utf.byte, utf.char context = context @@ -63,7 +63,7 @@ local characters = utilities.storage.allocate { } local mapping = utilities.storage.allocate { ---~ [utfchar(0xA0)] = "~", -- nbsp (catch) + -- [utfchar(0xA0)] = "~", -- nbsp (catch) } hyphenatedurl.characters = characters diff --git a/tex/context/base/lpdf-col.lua b/tex/context/base/lpdf-col.lua index 60d02c1ff..db9d3268b 100644 --- a/tex/context/base/lpdf-col.lua +++ b/tex/context/base/lpdf-col.lua @@ -13,6 +13,8 @@ local round = math.round local backends, lpdf, nodes = backends, lpdf, nodes +local allocate = utilities.storage.allocate + local nodeinjections = backends.pdf.nodeinjections local codeinjections = backends.pdf.codeinjections local registrations = backends.pdf.registrations @@ -44,7 +46,7 @@ local report_color = logs.reporter("colors","backend") -- page groups (might move to lpdf-ini.lua) -local colorspaceconstants = { -- v_none is ignored +local colorspaceconstants = allocate { -- v_none is ignored gray = pdfconstant("DeviceGray"), rgb = pdfconstant("DeviceRGB"), cmyk = pdfconstant("DeviceCMYK"), @@ -694,7 +696,7 @@ end -- this will move to lpdf-spe.lua -backends.pdf.tables.vfspecials = { -- todo: distinguish between glyph and rule color +backends.pdf.tables.vfspecials = allocate { -- todo: distinguish between glyph and rule color red = { "special", 'pdf: 1 0 0 rg 1 0 0 RG' }, green = { "special", 'pdf: 0 1 0 rg 0 1 0 RG' }, diff --git a/tex/context/base/lxml-tab.lua b/tex/context/base/lxml-tab.lua index fecf42634..9a6ac9082 100644 --- a/tex/context/base/lxml-tab.lua +++ b/tex/context/base/lxml-tab.lua @@ -43,7 +43,7 @@ local utf = unicode.utf8 local concat, remove, insert = table.concat, table.remove, table.insert local type, next, setmetatable, getmetatable, tonumber = type, next, setmetatable, getmetatable, tonumber local format, lower, find, match, gsub = string.format, string.lower, string.find, string.match, string.gsub -local utfchar, utfgsub = utf.char, utf.gsub +local utfchar = utf.char local lpegmatch = lpeg.match local P, S, R, C, V, C, Cs = lpeg.P, lpeg.S, lpeg.R, lpeg.C, lpeg.V, lpeg.C, lpeg.Cs @@ -365,13 +365,7 @@ local privates_n = { -- keeps track of defined ones } -local function escaped(s) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_u)) - end -end +local escaped = utf.remapper(privates_u) local function unescaped(s) local p = privates_n[s] @@ -386,13 +380,7 @@ local function unescaped(s) return p end -local function unprivatized(s,resolve) - if s == "" then - return "" - else - return (utfgsub(s,".",privates_p)) - end -end +local unprivatized = utf.remapper(privates_p) xml.privatetoken = unescaped xml.unprivatized = unprivatized diff --git a/tex/context/base/node-res.lua b/tex/context/base/node-res.lua index 4bf4610cd..2933fc86d 100644 --- a/tex/context/base/node-res.lua +++ b/tex/context/base/node-res.lua @@ -33,6 +33,8 @@ local nodecodes = nodes.nodecodes local glyph_code = nodecodes.glyph +local allocate = utilities.storage.allocate + local reserved, nofreserved = { }, 0 local function register_node(n) @@ -325,7 +327,7 @@ end -- local num = userids["my id"] -- local str = userids[num] -local userids = utilities.storage.allocate() pool.userids = userids +local userids = allocate() pool.userids = userids local lastid = 0 setmetatable(userids, { diff --git a/tex/context/base/page-ins.lua b/tex/context/base/page-ins.lua index fd9d9e024..7f870735d 100644 --- a/tex/context/base/page-ins.lua +++ b/tex/context/base/page-ins.lua @@ -18,26 +18,28 @@ if not modules then modules = { } end modules ['page-ins'] = { -- Maybe we should only register in lua and forget about the tex end. -structures = structures or { } -structures.inserts = structures.inserts or { } -local inserts = structures.inserts +structures = structures or { } +structures.inserts = structures.inserts or { } +local inserts = structures.inserts local report_inserts = logs.reporter("inserts") -inserts.stored = inserts.stored or { } -- combining them in one is inefficient in the -inserts.data = inserts.data or { } -- bytecode storage pool +local allocate = utilities.storage.allocate -local variables = interfaces.variables -local v_page = variables.page -local v_columns = variables.columns -local v_firstcolumn = variables.firstcolumn -local v_lastcolumn = variables.lastcolumn -local v_text = variables.text +inserts.stored = inserts.stored or allocate { } -- combining them in one is inefficient in the +inserts.data = inserts.data or allocate { } -- bytecode storage pool + +local variables = interfaces.variables +local v_page = variables.page +local v_columns = variables.columns +local v_firstcolumn = variables.firstcolumn +local v_lastcolumn = variables.lastcolumn +local v_text = variables.text storage.register("structures/inserts/stored", inserts.stored, "structures.inserts.stored") -local data = inserts.data -local stored = inserts.stored +local data = inserts.data +local stored = inserts.stored for name, specification in next, stored do data[specification.number] = specification diff --git a/tex/context/base/regi-ini.lua b/tex/context/base/regi-ini.lua index 7076c3e7a..f9507bd0b 100644 --- a/tex/context/base/regi-ini.lua +++ b/tex/context/base/regi-ini.lua @@ -14,10 +14,11 @@ runtime.</p> local commands, context = commands, context -local utfchar, utfgsub = utf.char, utf.gsub +local utfchar = utf.char +local lpegmatch = lpeg.match local char, gsub, format = string.char, string.gsub, string.format local next = next -local insert, remove = table.insert, table.remove +local insert, remove, fastcopy = table.insert, table.remove, table.fastcopy local allocate = utilities.storage.allocate local sequencers = utilities.sequencers @@ -139,17 +140,51 @@ local function translate(line,regime) return line end +-- local remappers = { } +-- +-- local function toregime(vector,str,default) -- toregime('8859-1',"abcde Ä","?") +-- local t = backmapping[vector] +-- local remapper = remappers[vector] +-- if not remapper then +-- remapper = utf.remapper(t) +-- remappers[t] = remapper +-- end +-- local m = getmetatable(t) +-- setmetatableindex(t, function(t,k) +-- local v = default or "?" +-- t[k] = v +-- return v +-- end) +-- str = remapper(str) +-- setmetatable(t,m) +-- return str +-- end +-- +-- -- much faster (but only matters when we have > 10K calls + +local cache = { } -- if really needed we can copy vectors and hash defaults + +setmetatableindex(cache, function(t,k) + local v = { remappers = { } } + t[k] = v + return v +end) + local function toregime(vector,str,default) -- toregime('8859-1',"abcde Ä","?") - local t = backmapping[vector] - local m = getmetatable(t) - setmetatableindex(t, function(t,k) - local v = default or "?" - t[k] = v - return v - end) - str = utfgsub(str,".",t) - setmetatable(t,m) - return str + local d = default or "?" + local c = cache[vector].remappers + local r = c[d] + if not r then + local t = fastcopy(backmapping[vector]) + setmetatableindex(t, function(t,k) + local v = d + t[k] = v + return v + end) + r = utf.remapper(t) + c[d] = r + end + return r(str) end local function disable() diff --git a/tex/context/base/s-inf-03.mkiv b/tex/context/base/s-inf-03.mkiv index 1ff4cd0d0..bd4ec5dbc 100644 --- a/tex/context/base/s-inf-03.mkiv +++ b/tex/context/base/s-inf-03.mkiv @@ -1,6 +1,14 @@ % \nopdfcompression -\enablemode[ipad] +\starttext + +There is a loop somewhere ... todo. + +\stoptext + +\enablemode[ipad] % whatever that means + +\setupbodyfont[dejavu] \doifmodeelse {ipad} { @@ -10,16 +18,16 @@ \definefont [TitlePageFont] - [MonoBold at 18pt] + [MonoBold at 16pt] \setupbodyfont - [tt,10pt] + [tt,8pt] } { \definefont [TitlePageFont] - [MonoBold at 24pt] + [MonoBold at 20pt] \setupbodyfont [tt] @@ -207,6 +215,7 @@ local function childtables(key,tab,handler,depth) else handler(s,t,depth) if variant == 3 then +print("UNMARKED",k) childtables(false,v,handler,depth+1) elseif variant == 2 then childtables(k,v,handler,depth+1) @@ -231,7 +240,6 @@ end local function handler(k,t,depth) k = gsub(k,"([~#$%%^&{}\\\|])",cleanup) --- NC() rlap("\\quad\\tx " .. upper(sub(t,1,1)) .. " ".. k) NC() NC() NR() NC() rlap("\\quad\\tx\\kern" .. (depth or 0).. "em" .. upper(sub(t,1,1)) .. " ".. k) NC() NC() NR() end @@ -260,67 +268,66 @@ local function show(title,subtitle,alias,builtin,t,lib,libcolor,glo,glocolor,mar for i=1,#keys do local k = keys[i] local v = t[k] - if k ~= "obsolete" and not skipkeys[k] and (not obsolete or not obsolete[k]) then + if k and k ~= "obsolete" and not skipkeys[k] and (not obsolete or not obsolete[k]) then local inlib = lib and lib[k] local inglo = glo and glo[k] - if k then - local t = type(v) - local kstr, tstr = k, t - local obs = t_obsolete and t_obsolete[k] - if obs then - tstr = function() overstrike(t) end - kstr = function() overstrike(k) end - end - local marked = marked(v) - if marked then - tstr = "data table" - end - if t == "table" then - local m = getmetatable(v) - if m and m.__call then - tstr = "function" - end - end - if not mark then - -- - elseif inlib and tostring(inlib) ~= tostring(v) then - tstr = "overloaded ".. tstr - elseif inglo and tostring(inglo) ~= tostring(v) then - tstr = "overloaded ".. tstr + local t = type(v) + local kstr, tstr = k, t + local obs = t_obsolete and t_obsolete[k] + if obs then + tstr = function() overstrike(t) end + kstr = function() overstrike(k) end + end + local marked = marked(v) + if marked then + tstr = "data table" + end + if t == "table" then + local m = getmetatable(v) + if m and m.__call then + tstr = "function" end - NC() bf() - if inlib then - if not mark and t == "table" then - goto(function() color( { libcolor }, kstr) end, { k } ) - else - color( { libcolor }, kstr) - end - elseif inglo then - if not mark and t == "table" then - goto(function() color( { glocolor }, kstr) end, { k } ) - else - color( { glocolor }, kstr) - end + end + if not mark then + -- + elseif inlib and tostring(inlib) ~= tostring(v) then + tstr = "overloaded ".. tstr + elseif inglo and tostring(inglo) ~= tostring(v) then + tstr = "overloaded ".. tstr + end + NC() bf() + if inlib then + if not mark and t == "table" then + goto(function() color( { libcolor }, kstr) end, { k } ) else - if not mark and t == "table" then - goto(k, { kstr } ) - else - context(kstr) - end + color( { libcolor }, kstr) end - NC() - if inlib then - color( { libcolor }, tstr) - elseif inglo then - color( { glocolor }, tstr) + elseif inglo then + if not mark and t == "table" then + goto(function() color( { glocolor }, kstr) end, { k } ) else - context(tstr) + color( { glocolor }, kstr) end - NC() NR() - if mark and t == "table" and title ~= "libraries" and title ~= "package" and not marked then - childtables(false,v,handler) -- (k,v,handler) + else + if not mark and t == "table" then + goto(k, { kstr } ) + else + context(kstr) end end + NC() + if inlib then + color( { libcolor }, tstr) + elseif inglo then + color( { glocolor }, tstr) + else + context(tstr) + end + NC() NR() + if mark and t == "table" and title ~= "libraries" and title ~= "package" and not marked then +print(title,subtitle,k) + childtables(false,v,handler) -- (k,v,handler) + end end end context.stoptabulate() diff --git a/tex/context/base/status-files.pdf b/tex/context/base/status-files.pdf Binary files differindex 20c7b966c..0b2b9db93 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 37adeca31..7afcfb69b 100644 --- a/tex/context/base/status-lua.pdf +++ b/tex/context/base/status-lua.pdf diff --git a/tex/context/base/tabl-ntb.mkiv b/tex/context/base/tabl-ntb.mkiv index fca724b71..9c7864184 100644 --- a/tex/context/base/tabl-ntb.mkiv +++ b/tex/context/base/tabl-ntb.mkiv @@ -854,11 +854,25 @@ \c_tabl_ntb_maximum_col \zerocount \c_tabl_ntb_maximum_row \zerocount \c_tabl_ntb_maximum_row_span\plusone + \let\currentTABLErow \tabl_ntb_current_row + \let\currentTABLEcolumn\tabl_ntb_current_column + \let\nofTABLErows \tabl_ntb_n_of_rows + \let\nofTABLEcolumns \tabl_ntb_n_of_columns \let\bTR\dobTR \let\bTD\dobTD \let\bTH\dobTH \let\bTN\dobTN} +\def\tabl_ntb_current_row {\m_tabl_ntb_positive_row} +\def\tabl_ntb_current_column{\m_tabl_ntb_positive_col} +\def\tabl_ntb_n_of_rows {\number\c_tabl_ntb_maximum_row} +\def\tabl_ntb_n_of_columns {\number\c_tabl_ntb_maximum_col} + +\let\currentTABLErow \!!zerocount +\let\currentTABLEcolumn\!!zerocount +\let\nofTABLErows \!!zerocount +\let\nofTABLEcolumns \!!zerocount + \unexpanded\def\dobTR{\dosingleempty\tabl_ntb_tr} % also used in tabl-nte \unexpanded\def\dobTD{\dosingleempty\tabl_ntb_td} % also used in tabl-nte \unexpanded\def\dobTH{\dosingleempty\tabl_ntb_th} % also used in tabl-nte diff --git a/tex/context/base/trac-lmx.lua b/tex/context/base/trac-lmx.lua index e1805099e..e83fdce7f 100644 --- a/tex/context/base/trac-lmx.lua +++ b/tex/context/base/trac-lmx.lua @@ -458,7 +458,7 @@ local function lmxresult(self,variables) if trace_variables then -- will become templates report_lmx("converted size: %s",#converted) end - return converted + return converted or lmxerror("no result from converter") else return lmxerror("invalid converter") end diff --git a/tex/context/base/util-jsn.lua b/tex/context/base/util-jsn.lua index 101a75d6b..7493f108d 100644 --- a/tex/context/base/util-jsn.lua +++ b/tex/context/base/util-jsn.lua @@ -52,7 +52,7 @@ local jsonconverter = { "value", object = lbrace * Cf(Ct("") * V("pair") * (comma * V("pair"))^0,rawset) * rbrace, pair = Cg(optionalws * key * optionalws * colon * V("value")), array = Ct(lparent * V("value") * (comma * V("value"))^0 * rparent), - value = optionalws * (jstring + V("object") + V("array") + jtrue + jfalse + jnull + jnumber) * optionalws, + value = optionalws * (jstring + V("object") + V("array") + jtrue + jfalse + jnull + jnumber + #rparent) * optionalws, } -- local jsonconverter = { "value", @@ -93,7 +93,7 @@ local function tojson(value,t) -- we could optimize #t elseif size == 1 then -- we can optimize for non tables t[#t+1] = "[" - tojson(value[i],t) + tojson(value[1],t) t[#t+1] = "]" else for i=1,size do @@ -111,7 +111,7 @@ local function tojson(value,t) -- we could optimize #t t[#t+1] = format("%q",value) elseif kind == "number" then t[#t+1] = value - else + elseif kind == "boolean" then t[#t+1] = tostring(value) end return t diff --git a/tex/context/base/util-sql.lua b/tex/context/base/util-sql.lua index 1e07bfc3c..3daa60d0b 100644 --- a/tex/context/base/util-sql.lua +++ b/tex/context/base/util-sql.lua @@ -10,7 +10,7 @@ if not modules then modules = { } end modules ['util-sql'] = { -- there is a bit of flux in these libraries. Also, we want the data back in -- a way that we like. --- buffer template +-- Todo: buffer templates when files. local format = string.format local rawset, setmetatable, loadstring, type = rawset, setmetatable, loadstring, type @@ -48,7 +48,6 @@ local engine = "mysql" local runners = { -- --defaults-extra-file="%inifile" mysql = [[mysql --user="%username%" --password="%password%" --host="%host%" --port=%port% --database="%database%" < "%queryfile%" > "%resultfile%"]], --- mysql = [[mysql --user="%username%" --password="%password%" --host="%host%" --port=%port% --database="%database%" < "%queryfile%"]], } sql.runners = runners @@ -151,7 +150,7 @@ local function validspecification(specification) specification.queryfile = queryfile specification.resultfile = resultfile if trace_sql then - report_state("template file: %q",templatefile) + report_state("template file: %q",templatefile or "<none>") report_state("query file: %q",queryfile) report_state("result file: %q",resultfile) end @@ -167,10 +166,12 @@ local function dataprepared(specification) end if query then io.savedata(specification.queryfile,query) + os.remove(specification.resultfile) return true else -- maybe push an error os.remove(specification.queryfile) + os.remove(specification.resultfile) end end @@ -183,7 +184,6 @@ local function datafetched(specification) report_state("fetchtime: %.3f sec",osclock()-t) -- not okay under linux else os.execute(command) - -- return os.resultof(command) end return true end @@ -455,28 +455,83 @@ methods.library = { -- -- -- -local e_pattern = lpeg.replacer { { '\\"','\\\\""' }, {'"','""'} } +local e_pattern = lpeg.replacer { { '\\"','\\\\""' }, {'"','""'}, {'\\\n', "\\n" }, {'\\\r', "\\r" }, {'\t', " " } } local u_pattern = lpeg.replacer { { '\\\\','\\' } } +local u_pattern = lpeg.replacer { { '\\\\','\\' }, { "\n","\\n" } } -function sql.escape(str) - return lpegmatch(e_replace,str) +-- library: + +function methods.library.serialize(t) + local str = fastserialize(t,"return") + local escaped = lpegmatch(e_pattern,str) +-- print("LIBRARY PUT STR",str) +-- print("LIBRARY PUT ESC",escaped) + return escaped end -function sql.unescape(str) - return lpegmatch(u_replace,str) +function methods.library.deserialize(str) + local unescaped = lpegmatch(u_pattern,str) +-- print("LIBRARY GET STR",str) +-- print("LIBRARY GET UES",unescaped) + if not unescaped then + return + end + local code = loadstring(unescaped) +-- print("INVALID CODE") + if not code then + return + end + code = code() +-- table.print(code) + if not code then + return + end + return code end -function sql.serialize(t) +-- client + +local e_pattern = lpeg.replacer { { '\\"','\\\\""' }, {'"','""'}, {'\\\n', "\\n" }, {'\\\r', "\\r" } } +local u_pattern = lpeg.replacer { { '\\\\','\\' } } + +function methods.client.serialize(t) return lpegmatch(e_pattern,fastserialize(t,"return")) end -function sql.deserialize(data) - data = lpegmatch(u_pattern,data) - data = data and loadstring(data) - data = data and data() - return data +function methods.client.deserialize(str) + local unescaped = lpegmatch(u_pattern,str) + if not unescaped then + return + end + local code = loadstring(unescaped) + if not code then + return + end + code = code() + if not code then + return + end + return code +end + +sql.serialize = methods.client.serialize +sql.deserialize = methods.client.deserialize + +function sql.escape(str) + return lpegmatch(e_pattern,str) +end + +function sql.unescape(str) + return lpegmatch(u_pattern,str) end +-- local s = sql.serialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7, e = 'f"g\nh' } } +-- local u = sql.unescape(s) +-- local t = sql.deserialize(s) +-- inspect(s) +-- inspect(u) +-- inspect(t) + -- -- -- if tex and tex.systemmodes then diff --git a/tex/context/base/util-str.lua b/tex/context/base/util-str.lua index 75cf443fb..377dd163f 100644 --- a/tex/context/base/util-str.lua +++ b/tex/context/base/util-str.lua @@ -62,11 +62,8 @@ function strings.newrepeater(str,offset) return t end ---~ local dashes = strings.newrepeater("--",-1) - ---~ print(dashes[2]) ---~ print(dashes[3]) ---~ print(dashes[1]) +-- local dashes = strings.newrepeater("--",-1) +-- print(dashes[2],dashes[3],dashes[1]) local extra, tab, start = 0, 0, 4, 0 diff --git a/tex/context/base/util-tab.lua b/tex/context/base/util-tab.lua index 7dc9e09f8..d45b58f6f 100644 --- a/tex/context/base/util-tab.lua +++ b/tex/context/base/util-tab.lua @@ -13,8 +13,8 @@ local tables = utilities.tables local format, gmatch, rep, gsub = string.format, string.gmatch, string.rep, string.gsub local concat, insert, remove = table.concat, table.insert, table.remove local setmetatable, getmetatable, tonumber, tostring = setmetatable, getmetatable, tonumber, tostring -local type, next, rawset, tonumber = type, next, rawset, tonumber -local lpegmatch = lpeg.match +local type, next, rawset, tonumber, loadstring = type, next, rawset, tonumber, loadstring +local lpegmatch, P, Cs = lpeg.match, lpeg.P, lpeg.Cs function tables.definetable(target) -- defines undefined tables local composed, t, n = nil, { }, 0 @@ -168,19 +168,27 @@ function tables.encapsulate(core,capsule,protect) end end -local function serialize(t,r) -- no mixes +local escaped = Cs ( ( + P('\\' ) + + P('"' )/'\\"' + + P('\n')/'\\n' + + P('\r')/'\\r' + + 1 +)^0 ) + +local function serialize(t,r,outer) -- no mixes r[#r+1] = "{" local n = #t if n > 0 then for i=1,n do local v = t[i] local tv = type(v) - if tv == "table" then - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("%q,",v) elseif tv == "number" then r[#r+1] = format("%s,",v) + elseif tv == "table" then + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("%s,",tostring(v)) end @@ -188,24 +196,43 @@ local function serialize(t,r) -- no mixes else for k, v in next, t do local tv = type(v) - if tv == "table" then - r[#r+1] = format("[%q]=",k) - serialize(v,r) - elseif tv == "string" then + if tv == "string" then r[#r+1] = format("[%q]=%q,",k,v) elseif tv == "number" then r[#r+1] = format("[%q]=%s,",k,v) + elseif tv == "table" then + r[#r+1] = format("[%q]=",k) + serialize(v,r) elseif tv == "boolean" then r[#r+1] = format("[%q]=%s,",k,tostring(v)) end end end - r[#r+1] = "}" + if outer then + r[#r+1] = "}" + else + r[#r+1] = "}," + end return r end function table.fastserialize(t,prefix) - return concat(serialize(t,{ prefix })) + return concat(serialize(t,{ prefix },true)) end --- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7 } }) +-- inspect(table.fastserialize { a = 1, b = { 4, { 5, 6 } }, c = { d = 7, e = 'f"g\nh' } }) + +function table.load(filename) + if filename then + local t = io.loaddata(filename) + if t and t ~= "" then + t = loadstring(t) + if type(t) == "function" then + t = t() + if type(t) == "table" then + return t + end + end + end + end +end diff --git a/tex/context/base/util-tpl.lua b/tex/context/base/util-tpl.lua index 61fbbd48e..f97f97d11 100644 --- a/tex/context/base/util-tpl.lua +++ b/tex/context/base/util-tpl.lua @@ -78,5 +78,5 @@ function templates.resolve(t,mapping) end -- inspect(utilities.templates.replace("test %one% test", { one = "%two%", two = "two" })) --- inspect(utilities.templates.resolve({ one = "%two%", two = "two" })) +-- inspect(utilities.templates.resolve({ one = "%two%", two = "two", three = "%three%" })) diff --git a/tex/context/base/x-mathml.lua b/tex/context/base/x-mathml.lua index bf4cae43f..cc7238f94 100644 --- a/tex/context/base/x-mathml.lua +++ b/tex/context/base/x-mathml.lua @@ -12,7 +12,7 @@ local type, next = type, next local utf = unicode.utf8 local format, lower, find, gsub = string.format, string.lower, string.find, string.gsub local strip = string.strip -local utfchar, utfgsub = utf.char, utf.gsub +local utfchar = utf.char local xmlsprint, xmlcprint, xmltext, xmlcontent = xml.sprint, xml.cprint, xml.text, xml.content local getid = lxml.getid local utfcharacters, utfvalues = string.utfcharacters, string.utfvalues @@ -87,6 +87,8 @@ local o_replacements = { -- in main table } +local simpleoperatorremapper = utf.remapper(o_replacements) + --~ languages.data.labels.functions local i_replacements = { @@ -457,8 +459,7 @@ function xml.functions.remapopenmath(e) end function mathml.checked_operator(str) - str = utfgsub(str,".",o_replacements) - context(str) + context(simpleoperatorremapper(str)) end function mathml.stripped(str) @@ -481,10 +482,8 @@ end function mathml.mo(id) local str = xmlcontent(getid(id)) or "" - local rep = gsub(str,"&.-;","") - local rep = utfgsub(rep,".",o_replacements) - context(rep) - -- context.mo(rep) -- fails with \left etc + local rep = gsub(str,"&.-;","") -- todo + context(simpleoperatorremapper(rep)) end function mathml.mi(id) diff --git a/tex/generic/context/luatex/luatex-fonts-merged.lua b/tex/generic/context/luatex/luatex-fonts-merged.lua index afdabf931..9650b1024 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 : 08/28/12 20:00:42 +-- merge date : 09/04/12 18:08:16 do -- begin closure to overcome local limits and interference @@ -144,7 +144,8 @@ if not modules then modules = { } end modules ['l-table'] = { license = "see context related readme files" } -local type, next, tostring, tonumber, ipairs, table, string = type, next, tostring, tonumber, ipairs, table, string +local type, next, tostring, tonumber, ipairs = type, next, tostring, tonumber, ipairs +local table, string = table, string local concat, sort, insert, remove = table.concat, table.sort, table.insert, table.remove local format, find, gsub, lower, dump, match = string.format, string.find, string.gsub, string.lower, string.dump, string.match local getmetatable, setmetatable = getmetatable, setmetatable @@ -155,6 +156,8 @@ local getinfo = debug.getinfo -- impact on ConTeXt was not that large; the remaining ipairs already -- have been replaced. In a similar fashion we also hardly used pairs. -- +-- Hm, actually ipairs was retained, but we no longer use it anyway. +-- -- Just in case, we provide the fallbacks as discussed in Programming -- in Lua (http://www.lua.org/pil/7.3.html): @@ -538,12 +541,25 @@ local function do_serialize(root,name,depth,level,indexed) end -- we could check for k (index) being number (cardinal) if root and next(root) then - local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- local first, last = nil, 0 -- #root cannot be trusted here (will be ok in 5.2 when ipairs is gone) + -- if compact then + -- -- NOT: for k=1,#root do (we need to quit at nil) + -- for k,v in ipairs(root) do -- can we use next? + -- if not first then first = k end + -- last = last + 1 + -- end + -- end + local first, last = nil, 0 if compact then - -- NOT: for k=1,#root do (we need to quit at nil) - for k,v in ipairs(root) do -- can we use next? - if not first then first = k end - last = last + 1 + last = #root + for k=1,last do + if not root[k] then + last = k - 1 + break + end + end + if last > 0 then + first = 1 end end local sk = sortedkeys(root) |