summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorContext Git Mirror Bot <phg42.2a@gmail.com>2014-10-03 20:15:04 +0200
committerContext Git Mirror Bot <phg42.2a@gmail.com>2014-10-03 20:15:04 +0200
commite891b55d71c2b4139dd25f40cfe344a4899a1ef2 (patch)
tree84525412dc1c62a5cb345d327a2daebd4bbf8405
parentb7b69c20fae6b104e134dae90ac34a35221b82ed (diff)
downloadcontext-e891b55d71c2b4139dd25f40cfe344a4899a1ef2.tar.gz
2014-10-03 19:29:00
-rw-r--r--scripts/context/lua/mtxrun.lua30
-rw-r--r--scripts/context/stubs/mswin/mtxrun.lua30
-rw-r--r--scripts/context/stubs/unix/mtxrun30
-rw-r--r--scripts/context/stubs/win64/mtxrun.lua30
-rw-r--r--tex/context/base/cont-new.mkiv2
-rw-r--r--tex/context/base/context-version.pdfbin4390 -> 4381 bytes
-rw-r--r--tex/context/base/context.mkiv2
-rw-r--r--tex/context/base/l-unicode.lua24
-rw-r--r--tex/context/base/lpdf-epa.lua22
-rw-r--r--tex/context/base/lpdf-epd.lua732
-rw-r--r--tex/context/base/lpdf-ini.lua56
-rw-r--r--tex/context/base/status-files.pdfbin24682 -> 24685 bytes
-rw-r--r--tex/context/base/status-lua.pdfbin332148 -> 332174 bytes
-rw-r--r--tex/context/base/strc-ren.mkiv10
-rw-r--r--tex/generic/context/luatex/luatex-fonts-merged.lua2
15 files changed, 571 insertions, 399 deletions
diff --git a/scripts/context/lua/mtxrun.lua b/scripts/context/lua/mtxrun.lua
index ff54a0927..a6cbfd681 100644
--- a/scripts/context/lua/mtxrun.lua
+++ b/scripts/context/lua/mtxrun.lua
@@ -4394,7 +4394,7 @@ do -- create closure to overcome 200 locals limit
package.loaded["l-unicode"] = package.loaded["l-unicode"] or true
--- original size: 35080, stripped down to: 15712
+-- original size: 35374, stripped down to: 15946
if not modules then modules={} end modules ['l-unicode']={
version=1.001,
@@ -4705,7 +4705,9 @@ local utf_16_be_linesplitter=patterns.utfbom_16_be^-1*lpeg.tsplitat(patterns.utf
local utf_16_le_linesplitter=patterns.utfbom_16_le^-1*lpeg.tsplitat(patterns.utf_16_le_nl)
if bytepairs then
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4732,7 +4734,9 @@ if bytepairs then
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -4759,7 +4763,9 @@ if bytepairs then
return t
end
utf32_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4783,7 +4789,9 @@ if bytepairs then
return t
end
utf32_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4808,7 +4816,9 @@ if bytepairs then
end
else
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4838,7 +4848,9 @@ else
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -17593,8 +17605,8 @@ end -- of closure
-- used libraries : l-lua.lua l-package.lua l-lpeg.lua l-function.lua l-string.lua l-table.lua l-io.lua l-number.lua l-set.lua l-os.lua l-file.lua l-gzip.lua l-md5.lua l-url.lua l-dir.lua l-boolean.lua l-unicode.lua l-math.lua util-str.lua util-tab.lua util-sto.lua util-prs.lua util-fmt.lua trac-set.lua trac-log.lua trac-inf.lua trac-pro.lua util-lua.lua util-deb.lua util-mrg.lua util-tpl.lua util-env.lua luat-env.lua lxml-tab.lua lxml-lpt.lua lxml-mis.lua lxml-aux.lua lxml-xml.lua trac-xml.lua data-ini.lua data-exp.lua data-env.lua data-tmp.lua data-met.lua data-res.lua data-pre.lua data-inp.lua data-out.lua data-fil.lua data-con.lua data-use.lua data-zip.lua data-tre.lua data-sch.lua data-lua.lua data-aux.lua data-tmf.lua data-lst.lua util-lib.lua luat-sta.lua luat-fmt.lua
-- skipped libraries : -
--- original bytes : 724999
--- stripped bytes : 257847
+-- original bytes : 725293
+-- stripped bytes : 257907
-- end library merge
diff --git a/scripts/context/stubs/mswin/mtxrun.lua b/scripts/context/stubs/mswin/mtxrun.lua
index ff54a0927..a6cbfd681 100644
--- a/scripts/context/stubs/mswin/mtxrun.lua
+++ b/scripts/context/stubs/mswin/mtxrun.lua
@@ -4394,7 +4394,7 @@ do -- create closure to overcome 200 locals limit
package.loaded["l-unicode"] = package.loaded["l-unicode"] or true
--- original size: 35080, stripped down to: 15712
+-- original size: 35374, stripped down to: 15946
if not modules then modules={} end modules ['l-unicode']={
version=1.001,
@@ -4705,7 +4705,9 @@ local utf_16_be_linesplitter=patterns.utfbom_16_be^-1*lpeg.tsplitat(patterns.utf
local utf_16_le_linesplitter=patterns.utfbom_16_le^-1*lpeg.tsplitat(patterns.utf_16_le_nl)
if bytepairs then
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4732,7 +4734,9 @@ if bytepairs then
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -4759,7 +4763,9 @@ if bytepairs then
return t
end
utf32_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4783,7 +4789,9 @@ if bytepairs then
return t
end
utf32_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4808,7 +4816,9 @@ if bytepairs then
end
else
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4838,7 +4848,9 @@ else
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -17593,8 +17605,8 @@ end -- of closure
-- used libraries : l-lua.lua l-package.lua l-lpeg.lua l-function.lua l-string.lua l-table.lua l-io.lua l-number.lua l-set.lua l-os.lua l-file.lua l-gzip.lua l-md5.lua l-url.lua l-dir.lua l-boolean.lua l-unicode.lua l-math.lua util-str.lua util-tab.lua util-sto.lua util-prs.lua util-fmt.lua trac-set.lua trac-log.lua trac-inf.lua trac-pro.lua util-lua.lua util-deb.lua util-mrg.lua util-tpl.lua util-env.lua luat-env.lua lxml-tab.lua lxml-lpt.lua lxml-mis.lua lxml-aux.lua lxml-xml.lua trac-xml.lua data-ini.lua data-exp.lua data-env.lua data-tmp.lua data-met.lua data-res.lua data-pre.lua data-inp.lua data-out.lua data-fil.lua data-con.lua data-use.lua data-zip.lua data-tre.lua data-sch.lua data-lua.lua data-aux.lua data-tmf.lua data-lst.lua util-lib.lua luat-sta.lua luat-fmt.lua
-- skipped libraries : -
--- original bytes : 724999
--- stripped bytes : 257847
+-- original bytes : 725293
+-- stripped bytes : 257907
-- end library merge
diff --git a/scripts/context/stubs/unix/mtxrun b/scripts/context/stubs/unix/mtxrun
index ff54a0927..a6cbfd681 100644
--- a/scripts/context/stubs/unix/mtxrun
+++ b/scripts/context/stubs/unix/mtxrun
@@ -4394,7 +4394,7 @@ do -- create closure to overcome 200 locals limit
package.loaded["l-unicode"] = package.loaded["l-unicode"] or true
--- original size: 35080, stripped down to: 15712
+-- original size: 35374, stripped down to: 15946
if not modules then modules={} end modules ['l-unicode']={
version=1.001,
@@ -4705,7 +4705,9 @@ local utf_16_be_linesplitter=patterns.utfbom_16_be^-1*lpeg.tsplitat(patterns.utf
local utf_16_le_linesplitter=patterns.utfbom_16_le^-1*lpeg.tsplitat(patterns.utf_16_le_nl)
if bytepairs then
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4732,7 +4734,9 @@ if bytepairs then
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -4759,7 +4763,9 @@ if bytepairs then
return t
end
utf32_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4783,7 +4789,9 @@ if bytepairs then
return t
end
utf32_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4808,7 +4816,9 @@ if bytepairs then
end
else
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4838,7 +4848,9 @@ else
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -17593,8 +17605,8 @@ end -- of closure
-- used libraries : l-lua.lua l-package.lua l-lpeg.lua l-function.lua l-string.lua l-table.lua l-io.lua l-number.lua l-set.lua l-os.lua l-file.lua l-gzip.lua l-md5.lua l-url.lua l-dir.lua l-boolean.lua l-unicode.lua l-math.lua util-str.lua util-tab.lua util-sto.lua util-prs.lua util-fmt.lua trac-set.lua trac-log.lua trac-inf.lua trac-pro.lua util-lua.lua util-deb.lua util-mrg.lua util-tpl.lua util-env.lua luat-env.lua lxml-tab.lua lxml-lpt.lua lxml-mis.lua lxml-aux.lua lxml-xml.lua trac-xml.lua data-ini.lua data-exp.lua data-env.lua data-tmp.lua data-met.lua data-res.lua data-pre.lua data-inp.lua data-out.lua data-fil.lua data-con.lua data-use.lua data-zip.lua data-tre.lua data-sch.lua data-lua.lua data-aux.lua data-tmf.lua data-lst.lua util-lib.lua luat-sta.lua luat-fmt.lua
-- skipped libraries : -
--- original bytes : 724999
--- stripped bytes : 257847
+-- original bytes : 725293
+-- stripped bytes : 257907
-- end library merge
diff --git a/scripts/context/stubs/win64/mtxrun.lua b/scripts/context/stubs/win64/mtxrun.lua
index ff54a0927..a6cbfd681 100644
--- a/scripts/context/stubs/win64/mtxrun.lua
+++ b/scripts/context/stubs/win64/mtxrun.lua
@@ -4394,7 +4394,7 @@ do -- create closure to overcome 200 locals limit
package.loaded["l-unicode"] = package.loaded["l-unicode"] or true
--- original size: 35080, stripped down to: 15712
+-- original size: 35374, stripped down to: 15946
if not modules then modules={} end modules ['l-unicode']={
version=1.001,
@@ -4705,7 +4705,9 @@ local utf_16_be_linesplitter=patterns.utfbom_16_be^-1*lpeg.tsplitat(patterns.utf
local utf_16_le_linesplitter=patterns.utfbom_16_le^-1*lpeg.tsplitat(patterns.utf_16_le_nl)
if bytepairs then
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4732,7 +4734,9 @@ if bytepairs then
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -4759,7 +4763,9 @@ if bytepairs then
return t
end
utf32_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4783,7 +4789,9 @@ if bytepairs then
return t
end
utf32_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utflinesplitter,t)
end
local result={}
@@ -4808,7 +4816,9 @@ if bytepairs then
end
else
utf16_to_utf8_be=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_be_linesplitter,t)
end
local result={}
@@ -4838,7 +4848,9 @@ else
return t
end
utf16_to_utf8_le=function(t)
- if type(t)=="string" then
+ if not t then
+ return nil
+ elseif type(t)=="string" then
t=lpegmatch(utf_16_le_linesplitter,t)
end
local result={}
@@ -17593,8 +17605,8 @@ end -- of closure
-- used libraries : l-lua.lua l-package.lua l-lpeg.lua l-function.lua l-string.lua l-table.lua l-io.lua l-number.lua l-set.lua l-os.lua l-file.lua l-gzip.lua l-md5.lua l-url.lua l-dir.lua l-boolean.lua l-unicode.lua l-math.lua util-str.lua util-tab.lua util-sto.lua util-prs.lua util-fmt.lua trac-set.lua trac-log.lua trac-inf.lua trac-pro.lua util-lua.lua util-deb.lua util-mrg.lua util-tpl.lua util-env.lua luat-env.lua lxml-tab.lua lxml-lpt.lua lxml-mis.lua lxml-aux.lua lxml-xml.lua trac-xml.lua data-ini.lua data-exp.lua data-env.lua data-tmp.lua data-met.lua data-res.lua data-pre.lua data-inp.lua data-out.lua data-fil.lua data-con.lua data-use.lua data-zip.lua data-tre.lua data-sch.lua data-lua.lua data-aux.lua data-tmf.lua data-lst.lua util-lib.lua luat-sta.lua luat-fmt.lua
-- skipped libraries : -
--- original bytes : 724999
--- stripped bytes : 257847
+-- original bytes : 725293
+-- stripped bytes : 257907
-- end library merge
diff --git a/tex/context/base/cont-new.mkiv b/tex/context/base/cont-new.mkiv
index 93cd0ad0f..51f9ed8a0 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{2014.10.02 23:07}
+\newcontextversion{2014.10.03 19:27}
%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
index 81a033e83..53e291920 100644
--- a/tex/context/base/context-version.pdf
+++ b/tex/context/base/context-version.pdf
Binary files differ
diff --git a/tex/context/base/context.mkiv b/tex/context/base/context.mkiv
index 6e999630c..0182e23a2 100644
--- a/tex/context/base/context.mkiv
+++ b/tex/context/base/context.mkiv
@@ -28,7 +28,7 @@
%D up and the dependencies are more consistent.
\edef\contextformat {\jobname}
-\edef\contextversion{2014.10.02 23:07}
+\edef\contextversion{2014.10.03 19:27}
\edef\contextkind {beta}
%D For those who want to use this:
diff --git a/tex/context/base/l-unicode.lua b/tex/context/base/l-unicode.lua
index d75779267..85956308a 100644
--- a/tex/context/base/l-unicode.lua
+++ b/tex/context/base/l-unicode.lua
@@ -683,7 +683,9 @@ if bytepairs then
-- with a little bit more code we could include the linesplitter
utf16_to_utf8_be = function(t)
- if type(t) == "string" then
+ if not t then
+ return nil
+ elseif type(t) == "string" then
t = lpegmatch(utf_16_be_linesplitter,t)
end
local result = { } -- we reuse result
@@ -711,7 +713,9 @@ if bytepairs then
end
utf16_to_utf8_le = function(t)
- if type(t) == "string" then
+ if not t then
+ return nil
+ elseif type(t) == "string" then
t = lpegmatch(utf_16_le_linesplitter,t)
end
local result = { } -- we reuse result
@@ -739,7 +743,9 @@ if bytepairs then
end
utf32_to_utf8_be = function(t)
- if type(t) == "string" then
+ if not t then
+ return nil
+ elseif type(t) == "string" then
t = lpegmatch(utflinesplitter,t)
end
local result = { } -- we reuse result
@@ -764,7 +770,9 @@ if bytepairs then
end
utf32_to_utf8_le = function(t)
- if type(t) == "string" then
+ if not t then
+ return nil
+ elseif type(t) == "string" then
t = lpegmatch(utflinesplitter,t)
end
local result = { } -- we reuse result
@@ -791,7 +799,9 @@ if bytepairs then
else
utf16_to_utf8_be = function(t)
- if type(t) == "string" then
+ if not t then
+ return nil
+ elseif type(t) == "string" then
t = lpegmatch(utf_16_be_linesplitter,t)
end
local result = { } -- we reuse result
@@ -822,7 +832,9 @@ else
end
utf16_to_utf8_le = function(t)
- if type(t) == "string" then
+ if not t then
+ return nil
+ elseif type(t) == "string" then
t = lpegmatch(utf_16_le_linesplitter,t)
end
local result = { } -- we reuse result
diff --git a/tex/context/base/lpdf-epa.lua b/tex/context/base/lpdf-epa.lua
index 0440ba2cd..5f6969f45 100644
--- a/tex/context/base/lpdf-epa.lua
+++ b/tex/context/base/lpdf-epa.lua
@@ -149,11 +149,9 @@ function codeinjections.mergereferences(specification)
local size = specification.size or "crop" -- todo
local pagedata = document.pages[pagenumber]
local annotations = pagedata and pagedata.Annots
-local namespace = makenamespace(fullname)
-local reference = namespace .. pagenumber
+ local namespace = makenamespace(fullname)
+ local reference = namespace .. pagenumber
if annotations and annotations.n > 0 then
--- local namespace = makenamespace(fullname)
--- local reference = namespace .. pagenumber
local mediabox = pagedata.MediaBox
local llx = mediabox[1]
local lly = mediabox[2]
@@ -199,14 +197,14 @@ local reference = namespace .. pagenumber
end
end
context.flushlayer { "epdflinks" }
-end
- context.setgvalue("figurereference",reference) -- global
- if trace_links then
- report_link("setting figure reference to %a",reference)
- end
- specification.reference = reference
- return namespace
--- end
+ end
+ -- moved outside previous test
+ context.setgvalue("figurereference",reference) -- global
+ if trace_links then
+ report_link("setting figure reference to %a",reference)
+ end
+ specification.reference = reference
+ return namespace
end
end
return ""-- no namespace, empty, not nil
diff --git a/tex/context/base/lpdf-epd.lua b/tex/context/base/lpdf-epd.lua
index c13f09614..17007cdd1 100644
--- a/tex/context/base/lpdf-epd.lua
+++ b/tex/context/base/lpdf-epd.lua
@@ -10,34 +10,104 @@ if not modules then modules = { } end modules ['lpdf-epd'] = {
-- I want to be independent of the library (which implements a selection of what a file
-- provides) and also because I want an interface closer to Lua's table model while the API
-- stays close to the original xpdf library. Of course, after prototyping a solution, we can
--- optimize it using the low level epdf accessors.
-
+-- optimize it using the low level epdf accessors. However, not all are accessible (this will
+-- be fixed).
+--
-- It will be handy when we have a __length and __next that can trigger the resolve till then
--- we will provide .n as #.
-
+-- we will provide .n as #; maybe in Lua 5.3 or later.
+--
-- As there can be references to the parent we cannot expand a tree. I played with some
--- expansion variants but it does to pay off.
-
--- Maybe we need a close(). In fact, nilling the document root will result in a gc at some
--- point.
-
--- We cannot access all destinations in one run.
-
+-- expansion variants but it does not pay off; adding extra checks is nto worth the trouble.
+--
+-- The document stays over. In order to free memory one has to explicitly onload the loaded
+-- document.
+--
-- We have much more checking then needed in the prepare functions because occasionally
-- we run into bugs in poppler or the epdf interface. It took us a while to realize that
-- there was a long standing gc issue the on long runs with including many pages could
-- crash the analyzer.
+--
+-- - we cannot access all destinations in one run.
+-- - v:getTypeName(), versus types[v:getType()], the last variant is about twice as fast
+--
+-- A potential speedup is to use local function instead of colon accessors. This will be done
+-- in due time. Normally this code is not really speed sensitive but one never knows.
+
+-- __newindex = function(t,k,v)
+-- local tk = rawget(t,k)
+-- if not tk then
+-- local o = epdf.Object()
+-- o:initString(v)
+-- d:add(k,o)
+-- end
+-- rawset(t,k,v)
+-- end,
-local setmetatable, rawset, rawget, tostring, tonumber = setmetatable, rawset, rawget, tostring, tonumber
-local lower, match, char, find, sub = string.lower, string.match, string.char, string.find, string.sub
+local setmetatable, rawset, rawget, type = setmetatable, rawset, rawget, type
+local tostring, tonumber = tostring, tonumber
+local lower, match, char, utfchar = string.lower, string.match, string.char, utf.char
local concat = table.concat
local toutf = string.toutf
-local report_epdf = logs.reporter("epdf")
+local lpegmatch, lpegpatterns = lpeg.match, lpeg.patterns
+local P, C, S, R, Ct, Cc, V = lpeg.P, lpeg.C, lpeg.S, lpeg.R, lpeg.Ct, lpeg.Cc, lpeg.V
+
+local epdf = epdf
+ lpdf = lpdf or { }
+local lpdf = lpdf
+lpdf.epdf = { }
+
+local report_epdf = logs.reporter("epdf")
+
+local getDict, getArray, getReal, getNum, getString, getBool, getName, getRef, getRefNum
+local getType, getTypeName
+local dictGetLength, dictGetVal, dictGetValNF, dictGetKey
+local arrayGetLength, arrayGetNF, arrayGet
+local streamReset, streamGetDict, streamGetChar
+
+do
+ local object = epdf.Object()
+ --
+ getDict = object.getDict
+ getArray = object.getArray
+ getReal = object.getReal
+ getNum = object.getNum
+ getString = object.getString
+ getBool = object.getBool
+ getName = object.getName
+ getRef = object.getRef
+ getRefNum = object.getRefNum
+ --
+ getType = object.getType
+ getTypeName = object.getTypeName
+ --
+ streamReset = object.streamReset
+ streamGetDict = object.streamGetDict
+ streamGetChar = object.streamGetChar
+ --
+end
--- v:getTypeName(), versus types[v:getType()], the last variant is about twice as fast
+local function initialize_methods(xref)
+ local dictionary = epdf.Dict(xref)
+ local array = epdf.Array(xref)
+ --
+ dictGetLength = dictionary.getLength
+ dictGetVal = dictionary.getVal
+ dictGetValNF = dictionary.getValNF
+ dictGetKey = dictionary.getKey
+ --
+ arrayGetLength = array.getLength
+ arrayGetNF = array.getNF
+ arrayGet = array.get
+ --
+ -- report_epdf("initializing lpdf.epdf library")
+ --
+ initialize_methods = function()
+ -- already done
+ end
+end
-local typenames = { [0] =
+local typenames = { [0] =
"boolean",
"integer",
"real",
@@ -55,19 +125,18 @@ local typenames = { [0] =
"integer64",
}
-local typenumbers = table.swapped(typenames)
+local typenumbers = table.swapped(typenames)
-local null_code = typenumbers.null
-local ref_code = typenumbers.ref
+local null_code = typenumbers.null
+local ref_code = typenumbers.ref
local function fatal_error(...)
report_epdf(...)
- -- we exit as we will crash anyway
report_epdf("aborting job in order to avoid crash")
os.exit()
end
-local limited = false -- abit of protection
+local limited = false -- a bit of protection
directives.register("system.inputmode", function(v)
if not limited then
@@ -79,77 +148,43 @@ directives.register("system.inputmode", function(v)
end
end)
---
+-- epdf is the built-in library
function epdf.type(o)
local t = lower(match(tostring(o),"[^ :]+"))
return t or "?"
end
-lpdf = lpdf or { }
-local lpdf = lpdf
-
-lpdf.epdf = { }
-
local checked_access
--- dictionaries
-
--- local function prepare(document,d,t,n,k,mt)
--- for i=1,n do
--- local v = d:getVal(i)
--- local r = d:getValNF(i)
--- local key = d:getKey(i)
--- if r and r:getTypeName() == "ref" then
--- r = r:getRef().num
--- local c = document.cache[r]
--- if c then
--- --
--- else
--- c = checked_access[v:getTypeName()](v,document,r)
--- if c then
--- document.cache[r] = c
--- document.xrefs[c] = r
--- end
--- end
--- t[key] = c
--- elseif v then
--- t[key] = checked_access[v:getTypeName()](v,document)
--- else
--- fatal_error("fatal error: no data for key %s in dictionary",key)
--- end
--- end
--- getmetatable(t).__index = nil -- ?? weird
--- setmetatable(t,mt)
--- return t[k]
--- end
+-- dictionaries (can be optimized: ... resolve and redefine when all locals set)
local function prepare(document,d,t,n,k,mt)
--- print("start prepare dict, requesting key ",k,"out of",n)
for i=1,n do
- local v = d:getVal(i)
+ local v = dictGetVal(d,i)
if v then
- local r = d:getValNF(i)
- local kind = v:getType()
--- print("checking",i,d:getKey(i),v:getTypeName())
+ local r = dictGetValNF(d,i)
+ local kind = getType(v)
if kind == null_code then
- -- report_epdf("warning: null value for key %a in dictionary",key)
+ -- ignore
else
- local key = d:getKey(i)
+ local key = dictGetKey(d,i)
if kind then
- if r and r:getType() == ref_code then
- local objnum = r:getRef().num
- local cached = document.cache[objnum]
+ if r and getType(r) == ref_code then
+ local objnum = getRefNum(r)
+ local cached = document.__cache__[objnum]
if not cached then
cached = checked_access[kind](v,document,objnum)
if c then
- document.cache[objnum] = cached
- document.xrefs[cached] = objnum
+ document.__cache__[objnum] = cached
+ document.__xrefs__[cached] = objnum
end
end
t[key] = cached
+ -- rawset(t,key,cached)
else
t[key] = checked_access[kind](v,document)
+ -- rawset(t,key,checked_access[kind](v,document))
end
else
report_epdf("warning: nil value for key %a in dictionary",key)
@@ -159,67 +194,55 @@ local function prepare(document,d,t,n,k,mt)
fatal_error("error: invalid value at index %a in dictionary of %a",i,document.filename)
end
end
--- print("done")
- getmetatable(t).__index = nil -- ?? weird
setmetatable(t,mt)
return t[k]
end
local function some_dictionary(d,document,r,mt)
- local n = d and d:getLength() or 0
+ local n = d and dictGetLength(d) or 0
if n > 0 then
local t = { }
- setmetatable(t, { __index = function(t,k) return prepare(document,d,t,n,k,mt) end } )
+ setmetatable(t, {
+ __index = function(t,k)
+ return prepare(document,d,t,n,k,mt)
+ end
+ } )
return t
end
end
--- arrays
-
-local done = { }
-
--- local function prepare(document,a,t,n,k)
--- for i=1,n do
--- local v = a:get(i)
--- local r = a:getNF(i)
--- local kind = v:getTypeName()
--- if kind == "null" then
--- -- TH: weird, but appears possible
--- elseif r:getTypeName() == "ref" then
--- r = r:getRef().num
--- local c = document.cache[r]
--- if c then
--- --
--- else
--- c = checked_access[kind](v,document,r)
--- document.cache[r] = c
--- document.xrefs[c] = r
--- end
--- t[i] = c
--- else
--- t[i] = checked_access[kind](v,document)
--- end
--- end
--- getmetatable(t).__index = nil
--- return t[k]
--- end
+local function get_dictionary(object,document,r,mt)
+ local d = getDict(object)
+ local n = d and dictGetLength(d) or 0
+ if n > 0 then
+ local t = { }
+ setmetatable(t, {
+ __index = function(t,k)
+ return prepare(document,d,t,n,k,mt)
+ end,
+ } )
+ return t
+ end
+end
+
+-- arrays (can be optimized: ... resolve and redefine when all locals set)
local function prepare(document,a,t,n,k)
for i=1,n do
- local v = a:get(i)
+ local v = arrayGet(a,i)
if v then
- local kind = v:getType()
+ local kind = getType(v)
if kind == null_code then
- -- report_epdf("warning: null value for index %a in array",i)
+ -- ignore
elseif kind then
- local r = a:getNF(i)
- if r and r:getType() == ref_code then
- local objnum = r:getRef().num
- local cached = document.cache[objnum]
+ local r = arrayGetNF(a,i)
+ if r and getType(r) == ref_code then
+ local objnum = getRefNum(r)
+ local cached = document.__cache__[objnum]
if not cached then
cached = checked_access[kind](v,document,objnum)
- document.cache[objnum] = cached
- document.xrefs[cached] = objnum
+ document.__cache__[objnum] = cached
+ document.__xrefs__[cached] = objnum
end
t[i] = cached
else
@@ -237,10 +260,28 @@ local function prepare(document,a,t,n,k)
end
local function some_array(a,document,r)
- local n = a and a:getLength() or 0
+ local n = a and arrayGetLength(a) or 0
+ if n > 0 then
+ local t = { n = n }
+ setmetatable(t, {
+ __index = function(t,k)
+ return prepare(document,a,t,n,k)
+ end
+ } )
+ return t
+ end
+end
+
+local function get_array(object,document,r)
+ local a = getArray(object)
+ local n = a and arrayGetLength(a) or 0
if n > 0 then
local t = { n = n }
- setmetatable(t, { __index = function(t,k) return prepare(document,a,t,n,k) end } )
+ setmetatable(t, {
+ __index = function(t,k)
+ return prepare(document,a,t,n,k)
+ end
+ } )
return t
end
end
@@ -248,9 +289,9 @@ end
local function streamaccess(s,_,what)
if not what or what == "all" or what == "*all" then
local t, n = { }, 0
- s:streamReset()
+ streamReset(s)
while true do
- local c = s:streamGetChar()
+ local c = streamGetChar(s)
if c < 0 then
break
else
@@ -262,71 +303,22 @@ local function streamaccess(s,_,what)
end
end
-local function some_stream(d,document,r)
+local function get_stream(d,document,r)
if d then
- d:streamReset()
- local s = some_dictionary(d:streamGetDict(),document,r)
+ streamReset(d)
+ local s = some_dictionary(streamGetDict(d),document,r)
getmetatable(s).__call = function(...) return streamaccess(d,...) end
return s
end
end
--- we need epdf.boolean(v) in addition to v:getBool() [dictionary, array, stream, real, integer, string, boolean, name, ref, null]
-
--- checked_access = {
--- dictionary = function(d,document,r)
--- return some_dictionary(d:getDict(),document,r)
--- end,
--- array = function(a,document,r)
--- return some_array(a:getArray(),document,r)
--- end,
--- stream = function(v,document,r)
--- return some_stream(v,document,r)
--- end,
--- real = function(v)
--- return v:getReal()
--- end,
--- integer = function(v)
--- return v:getNum()
--- end,
--- -- integer64 = function(v)
--- -- return v:getNum()
--- -- end,
--- string = function(v)
--- return toutf(v:getString())
--- end,
--- boolean = function(v)
--- return v:getBool()
--- end,
--- name = function(v)
--- return v:getName()
--- end,
--- ref = function(v)
--- return v:getRef()
--- end,
--- null = function()
--- return nil
--- end,
--- none = function()
--- -- why not null
--- return nil
--- end,
--- -- error = function()
--- -- -- shouldn't happen
--- -- return nil
--- -- end,
--- -- eof = function()
--- -- -- we don't care
--- -- return nil
--- -- end,
--- -- cmd = function()
--- -- -- shouldn't happen
--- -- return nil
--- -- end
--- }
-
--- a bit of a speedup in case we want to play with large pdf's and have millions
--- of access .. it might not be worth the trouble
+local function get_string(v)
+ return toutf(getString(v))
+end
+
+local function get_null()
+ return nil
+end
-- we have dual access: by typenumber and by typename
@@ -345,103 +337,30 @@ checked_access = table.setmetatableindex(function(t,k)
end
end)
+checked_access[typenumbers.boolean] = getBool
+checked_access[typenumbers.integer] = getNum
+checked_access[typenumbers.real] = getReal
+checked_access[typenumbers.string] = get_string
+checked_access[typenumbers.name] = getName
+checked_access[typenumbers.null] = get_null
+checked_access[typenumbers.array] = get_array -- d,document,r
+checked_access[typenumbers.dictionary] = get_dictionary -- d,document,r
+checked_access[typenumbers.stream] = get_stream
+checked_access[typenumbers.ref] = getRef
+
for i=0,#typenames do
- checked_access[i] = function()
- return function(v,document)
- invalidaccess(i,document)
+ local checker = checked_access[i]
+ if not checker then
+ checker = function()
+ return function(v,document)
+ invalidaccess(i,document)
+ end
end
+ checked_access[i] = checker
end
+ checked_access[typenames[i]] = checker
end
-checked_access[typenumbers.dictionary] = function(d,document,r)
- local getDict = d.getDict
- local getter = function(d,document,r)
- return some_dictionary(getDict(d),document,r)
- end
- checked_access.dictionary = getter
- checked_access[typenumbers.dictionary] = getter
- return getter(d,document,r)
-end
-
-checked_access[typenumbers.array] = function(a,document,r)
- local getArray = a.getArray
- local getter = function(a,document,r)
- return some_array(getArray(a),document,r)
- end
- checked_access.array = getter
- checked_access[typenumbers.array] = getter
- return getter(a,document,r)
-end
-
-checked_access[typenumbers.stream] = function(v,document,r)
- return some_stream(v,document,r) -- or just an equivalent
-end
-
-checked_access[typenumbers.real] = function(v)
- local getReal = v.getReal
- checked_access.real = getReal
- checked_access[typenumbers.real] = getReal
- return getReal(v)
-end
-
-checked_access[typenumbers.integer] = function(v)
- local getNum = v.getNum
- checked_access.integer = getNum
- checked_access[typenumbers.integer] = getNum
- return getNum(v)
-end
-
-checked_access[typenumbers.string] = function(v)
- local getString = v.getString
- local function getter(v)
- return toutf(getString(v))
- end
- checked_access.string = getter
- checked_access[typenumbers.string] = getter
- return toutf(getString(v))
-end
-
-checked_access[typenumbers.boolean] = function(v)
- local getBool = v.getBool
- checked_access.boolean = getBool
- checked_access[typenumbers.boolean] = getBool
- return getBool(v)
-end
-
-checked_access[typenumbers.name] = function(v)
- local getName = v.getName
- checked_access.name = getName
- checked_access[typenumbers.name] = getName
- return getName(v)
-end
-
-checked_access[typenumbers.ref] = function(v)
- local getRef = v.getRef
- checked_access.ref = getRef
- checked_access[typenumbers.ref] = getRef
- return getRef(v)
-end
-
-checked_access[typenumbers.null] = function()
- return nil
-end
-
-checked_access[typenumbers.none] = function()
- -- is actually an error
- return nil
-end
-
-for i=0,#typenames do
- checked_access[typenames[i]] = checked_access[i]
-end
-
--- checked_access.real = epdf.real
--- checked_access.integer = epdf.integer
--- checked_access.string = epdf.string
--- checked_access.boolean = epdf.boolean
--- checked_access.name = epdf.name
--- checked_access.ref = epdf.ref
-
local function getnames(document,n,target) -- direct
if n then
local Names = n.Names
@@ -498,7 +417,6 @@ local function getlayers(document)
local n = layers.n
for i=1,n do
local layer = layers[i]
- -- print(document.xrefs[layer])
t[i] = layer.Name
end
t.n = n
@@ -513,50 +431,33 @@ local function getstructure(document)
end
local function getpages(document,Catalog)
- local data = document.data
- local xrefs = document.xrefs
- local cache = document.cache
- local cata = data:getCatalog()
- local xref = data:getXRef()
- local pages = { }
- local nofpages = cata:getNumPages()
--- local function getpagestuff(pagenumber,k)
--- if k == "MediaBox" then
--- local pageobj = cata:getPage(pagenumber)
--- local pagebox = pageobj:getMediaBox()
--- return { pagebox.x1, pagebox.y1, pagebox.x2, pagebox.y2 }
--- elseif k == "CropBox" then
--- local pageobj = cata:getPage(pagenumber)
--- local pagebox = pageobj:getMediaBox()
--- return { pagebox.x1, pagebox.y1, pagebox.x2, pagebox.y2 }
--- elseif k == "Resources" then
--- print("todo page resources from parent")
--- -- local pageobj = cata:getPage(pagenumber)
--- -- local resources = pageobj:getResources()
--- end
--- end
--- for pagenumber=1,nofpages do
--- local mt = { __index = function(t,k)
--- local v = getpagestuff(pagenumber,k)
--- if v then
--- t[k] = v
--- end
--- return v
--- end }
- local mt = { __index = Catalog.Pages }
+ local __data__ = document.__data__
+ local __xrefs__ = document.__xrefs__
+ local __cache__ = document.__cache__
+ local __xref__ = document.__xref__
+ --
+ local catalog = __data__:getCatalog()
+ local pages = { }
+ local nofpages = catalog:getNumPages()
+ local metatable = { __index = Catalog.Pages }
+ --
for pagenumber=1,nofpages do
- local pagereference = cata:getPageRef(pagenumber).num
- local pagedata = some_dictionary(xref:fetch(pagereference,0):getDict(),document,pagereference,mt)
+ local pagereference = catalog:getPageRef(pagenumber).num
+ local pageobject = __xref__:fetch(pagereference,0)
+ local pagedata = get_dictionary(pageobject,document,pagereference,metatable)
if pagedata then
- pagedata.number = pagenumber
- pages[pagenumber] = pagedata
- xrefs[pagedata] = pagereference
- cache[pagereference] = pagedata
+ -- rawset(pagedata,"number",pagenumber)
+ pagedata.number = pagenumber
+ pages[pagenumber] = pagedata
+ __xrefs__[pagedata] = pagereference
+ __cache__[pagereference] = pagedata
else
report_epdf("missing pagedata at slot %i",i)
end
end
+ --
pages.n = nofpages
+ --
return pages
end
@@ -583,27 +484,25 @@ function lpdf.epdf.load(filename)
local document = loaded[filename]
if not document then
statistics.starttiming(lpdf.epdf)
- local data = epdf.open(filename) -- maybe resolvers.find_file
- if data then
+ local __data__ = epdf.open(filename) -- maybe resolvers.find_file
+ if __data__ then
+ local __xref__ = __data__:getXRef()
document = {
- filename = filename,
- cache = { },
- xrefs = { },
- data = data,
+ filename = filename,
+ __cache__ = { },
+ __xrefs__ = { },
+ __fonts__ = { },
+ __data__ = __data__,
+ __xref__ = __xref__,
}
- -- table.setmetatablenewindex(document.cache,function(t,k,v)
- -- if rawget(t,k) then
- -- report_epdf("updating object %a in cache",k)
- -- else
- -- report_epdf("storing object %a in cache",k)
- -- end
- -- rawset(t,k,v)
- -- end)
- local Catalog = some_dictionary(data:getXRef():getCatalog():getDict(),document)
- local Info = some_dictionary(data:getXRef():getDocInfo():getDict(),document)
- document.Catalog = Catalog
- document.Info = Info
- -- document.catalog = Catalog
+ --
+ initialize_methods(__xref__)
+ --
+ local Catalog = some_dictionary(__xref__:getCatalog():getDict(),document)
+ local Info = some_dictionary(__xref__:getDocInfo():getDict(),document)
+ --
+ document.Catalog = Catalog
+ document.Info = Info
-- a few handy helper tables
document.pages = delayed(document,"pages", function() return getpages(document,Catalog) end)
document.destinations = delayed(document,"destinations", function() return getnames(document,Catalog.Names and Catalog.Names.Dests) end)
@@ -633,16 +532,199 @@ end
-- for k, v in next, expand(t) do
-function lpdf.epdf.expand(t)
+local function expand(t)
if type(t) == "table" then
local dummy = t.dummy
end
return t
end
+-- for k, v in expanded(t) do
+
+local function expanded(t)
+ if type(t) == "table" then
+ local dummy = t.dummy
+ end
+ return next, t
+end
+
+lpdf.epdf.expand = expand
+lpdf.epdf.expanded = expanded
+
+-- experiment .. will be finished when there is a real need
+
+local hexdigit = R("09","AF")
+local hexword = hexdigit*hexdigit*hexdigit*hexdigit / function(s) return tonumber(s,16) end
+local numchar = ( P("\\") * ( (R("09")^3/tonumber) + C(1) ) ) + C(1)
+local number = lpegpatterns.number / tonumber
+local spaces = lpegpatterns.whitespace^1
+local keyword = P("/") * C(R("AZ","az","09")^1)
+local operator = C((R("AZ","az")+P("'")+P('"'))^1)
+
+local grammar = P { "start",
+ start = (keyword + number + V("dictionary") + V("unicode") + V("string") + V("unicode")+ V("array") + spaces)^1,
+ array = P("[") * Ct(V("start")^1) * P("]"),
+ dictionary = P("<<") * Ct(V("start")^1) * P(">>"),
+ unicode = P("<") * Ct(hexword^1) * P(">"),
+ string = P("(") * Ct((V("string")+numchar)^1) * P(")"), -- untested
+}
+
+local operation = Ct(grammar^1 * operator)
+local parser = Ct((operation + P(1))^1)
+
+-- beginbfrange : <start> <stop> <firstcode>
+-- <start> <stop> [ <firstsequence> <firstsequence> <firstsequence> ]
+-- beginbfchar : <code> <newcode>
+
+-- todo: utf16 -> 8
+-- we could make range more efficient but it's seldom seen anyway
+
+local optionals = spaces^0
+local whatever = optionals * P("<") * hexword * P(">")
+local hexstring = optionals * P("<") * C(hexdigit^1) * P(">")
+local bfchar = Cc(1) * whatever * whatever
+local bfrange = Cc(2) * whatever * whatever * whatever
+ + Cc(3) * whatever * whatever * optionals * P("[") * hexstring^1 * optionals * P("]")
+local fromunicode = Ct ( (
+ P("beginbfchar" ) * Ct(bfchar )^1 * optionals * P("endbfchar" ) +
+ P("beginbfrange") * Ct(bfrange)^1 * optionals * P("endbfrange") +
+ spaces +
+ P(1)
+)^1 )
+
+local utf16_to_utf8_be = utf.utf16_to_utf8_be
+local utfchar = utfchar
+
+local function analyzefonts(document,resources) -- unfinished
+ local fonts = document.__fonts__
+ if resources then
+ local fontlist = resources.Font
+ if fontlist then
+ for id, data in expanded(fontlist) do
+ if not fonts[id] then
+ -- a quck hack ... I will look into it more detail if I find a real
+ -- -application for it
+ local tounicode = data.ToUnicode()
+ if tounicode then
+ tounicode = lpegmatch(fromunicode,tounicode)
+ end
+ if type(tounicode) == "table" then
+ local t = { }
+ for i=1,#tounicode do
+ local u = tounicode[i]
+ local w = u[1]
+ if w == 1 then
+ t[u[2]] = utfchar(u[3])
+ elseif w == 2 then
+ local m = u[4]
+ for i=u[2],u[3] do
+ t[i] = utfchar(m)
+ m = m + 1
+ end
+ elseif w == 3 then
+ local m = 4
+ for i=u[2],u[3] do
+ t[i] = utf16_to_utf8_be(u[m])
+ m = m + 1
+ end
+ end
+ end
+ fonts[id] = {
+ tounicode = t
+ }
+ else
+ fonts[id] = {
+ tounicode = { }
+ }
+ end
+ end
+ end
+ end
+ end
+ return fonts
+end
+
+function lpdf.epdf.getpagecontent(document,pagenumber)
+
+ local page = document.pages[pagenumber]
+
+ if not page then
+ return
+ end
+
+ local fonts = analyzefonts(document,page.Resources)
+
+ local content = page.Contents() or ""
+ local list = lpegmatch(parser,content)
+ local font = nil
+ local unic = nil
+
+ for i=1,#list do
+ local entry = list[i]
+ local size = #entry
+ local operator = entry[size]
+ if operator == "Tf" then
+ font = fonts[entry[1]]
+ unic = font.tounicode
+ elseif operator == "TJ" then -- { array, TJ }
+ local list = entry[1]
+ for i=1,#list do
+ local li = list[i]
+ if type(li) == "table" then
+ for i=1,#li do
+ local c = li[i]
+ local u = unic[c]
+ li[i] = u or utfchar(c)
+ end
+ list[i] = concat(li)
+ end
+ end
+ elseif operator == "Tj" or operator == "'" or operator == '"' then -- { string, Tj } { string, ' } { n, m, string, " }
+ local li = entry[size-1]
+ for i=1,#li do
+ local c = li[i]
+ local u = unic[c]
+ li[i] = utfchar(u or c)
+ end
+ entry[1] = concat(li)
+ end
+ end
+
+ -- for i=1,#list do
+ -- local entry = list[i]
+ -- local size = #entry
+ -- local operator = entry[size]
+ -- if operator == "TJ" then -- { array, TJ }
+ -- local list = entry[1]
+ -- for i=1,#list do
+ -- local li = list[i]
+ -- if type(li) == "string" then
+ -- --
+ -- elseif li < -50 then
+ -- list[i] = " "
+ -- else
+ -- list[i] = ""
+ -- end
+ -- end
+ -- entry[1] = concat(list)
+ -- elseif operator == "Tf" then
+ -- -- already concat
+ -- elseif operator == "cm" then
+ -- local e = entry[1]
+ -- local sx, rx, ry, sy, tx, ty = e[1], e[2], e[3], e[4], e[5], e[6]
+ -- -- if dy ... newline
+ -- end
+ -- end
+
+ return list
+
+end
+
+-- document.Catalog.StructTreeRoot.ParentTree.Nums[2][1].A.P[1])
+
-- helpers
-- function lpdf.epdf.getdestinationpage(document,name)
--- local destination = document.data:findDest(name)
+-- local destination = document.__data__:findDest(name)
-- return destination and destination.number
-- end
diff --git a/tex/context/base/lpdf-ini.lua b/tex/context/base/lpdf-ini.lua
index b58008e7f..a4725c30e 100644
--- a/tex/context/base/lpdf-ini.lua
+++ b/tex/context/base/lpdf-ini.lua
@@ -8,7 +8,7 @@ if not modules then modules = { } end modules ['lpdf-ini'] = {
local setmetatable, getmetatable, type, next, tostring, tonumber, rawset = setmetatable, getmetatable, type, next, tostring, tonumber, rawset
local char, byte, format, gsub, concat, match, sub, gmatch = string.char, string.byte, string.format, string.gsub, table.concat, string.match, string.sub, string.gmatch
-local utfchar, utfvalues = utf.char, utf.values
+local utfchar, utfbyte, utfvalues = utf.char, utf.byte, utf.values
local sind, cosd, floor, max, min = math.sind, math.cosd, math.floor, math.max, math.min
local lpegmatch, P, C, R, S, Cc, Cs = lpeg.match, lpeg.P, lpeg.C, lpeg.R, lpeg.S, lpeg.Cc, lpeg.Cs
local formatters = string.formatters
@@ -193,29 +193,51 @@ function lpdf.rectangle(width,height,depth)
end
end
---
+-- we could use a hash of predefined unicodes
+
+-- local function tosixteen(str) -- an lpeg might be faster (no table)
+-- if not str or str == "" then
+-- return "<feff>" -- not () as we want an indication that it's unicode
+-- else
+-- local r, n = { "<feff" }, 1
+-- for b in utfvalues(str) do
+-- n = n + 1
+-- if b < 0x10000 then
+-- r[n] = format("%04x",b)
+-- else
+-- -- r[n] = format("%04x%04x",b/1024+0xD800,b%1024+0xDC00)
+-- r[n] = format("%04x%04x",floor(b/1024),b%1024+0xDC00) --bit32.rshift(b,10)
+-- end
+-- end
+-- n = n + 1
+-- r[n] = ">"
+-- return concat(r)
+-- end
+-- end
+
+local cache = table.setmetatableindex(function(t,k) -- can be made weak
+ local v = utfbyte(k)
+ if v < 0x10000 then
+ v = format("%04x",v)
+ else
+ -- v = format("%04x%04x",v/1024+0xD800,v%1024+0xDC00)
+ v = format("%04x%04x",floor(v/1024),v%1024+0xDC00)
+ end
+ t[k] = v
+ return v
+end)
+
+local p = Cs(Cc("<feff") * (lpeg.patterns.utf8character/cache)^1 * Cc(">"))
local function tosixteen(str) -- an lpeg might be faster (no table)
if not str or str == "" then
return "<feff>" -- not () as we want an indication that it's unicode
else
- local r, n = { "<feff" }, 1
- for b in utfvalues(str) do
- n = n + 1
- if b < 0x10000 then
- r[n] = format("%04x",b)
- else
- -- r[n] = format("%04x%04x",b/1024+0xD800,b%1024+0xDC00)
- r[n] = format("%04x%04x",floor(b/1024),b%1024+0xDC00)
- end
- end
- n = n + 1
- r[n] = ">"
- return concat(r)
+ return lpegmatch(p,str)
end
end
-lpdf.tosixteen = tosixteen
+lpdf.tosixteen = tosixteen
-- lpeg is some 5 times faster than gsub (in test) on escaping
@@ -432,7 +454,7 @@ local function pdfstring(str,default)
end
local function pdfunicode(str,default)
- return setmetatable({ str or default or "" },mt_u)
+ return setmetatable({ str or default or "" },mt_u) -- could be a string
end
local cache = { } -- can be weak
diff --git a/tex/context/base/status-files.pdf b/tex/context/base/status-files.pdf
index 5da0e7556..51d01e48d 100644
--- a/tex/context/base/status-files.pdf
+++ b/tex/context/base/status-files.pdf
Binary files differ
diff --git a/tex/context/base/status-lua.pdf b/tex/context/base/status-lua.pdf
index 09c5ba3ad..f4d5e92d4 100644
--- a/tex/context/base/status-lua.pdf
+++ b/tex/context/base/status-lua.pdf
Binary files differ
diff --git a/tex/context/base/strc-ren.mkiv b/tex/context/base/strc-ren.mkiv
index fdf8fb7f4..baeb3ab04 100644
--- a/tex/context/base/strc-ren.mkiv
+++ b/tex/context/base/strc-ren.mkiv
@@ -516,6 +516,11 @@
\fi
\endgroup}
+\def\fakedheadnumber{\vphantom{0}} % needed for mathplus
+
+\unexpanded\def\fakeheadnumbercontent
+ {\hbox to \zeropoint{\let\getheadnumber\fakedheadnumber\headnumbercontent}}
+
\unexpanded\def\strc_rendering_inject_number_and_text
{\edef\p_command{\headparameter\c!command}% assumes \unexpanded definition
\ifx\p_command\empty
@@ -608,6 +613,7 @@
\hsize\headtextwidth
\fi
\noindent
+ \fakeheadnumbercontent % will also be done in the other ones (force consistency with numbered)
\fi
\headtextcontent
}
@@ -637,6 +643,8 @@
\hskip\dimexpr\d_strc_rendering_local_leftoffset+\scratchdistance\relax
}
}
+ \else
+ \fakeheadnumbercontent % will also be done in the other ones (force consistency with numbered)
\fi
\headtextcontent
}
@@ -665,6 +673,8 @@
\strut
\headnumbercontent
\par
+ else
+ \fakeheadnumbercontent % will also be done in the other ones (force consistency with numbered)
\fi
\begstrut
\headtextcontent
diff --git a/tex/generic/context/luatex/luatex-fonts-merged.lua b/tex/generic/context/luatex/luatex-fonts-merged.lua
index 42f85340e..15241bacc 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 : 10/02/14 23:07:34
+-- merge date : 10/03/14 19:27:20
do -- begin closure to overcome local limits and interference