summaryrefslogtreecommitdiff
path: root/src/fontloader/misc/fontloader-l-lpeg.lua
diff options
context:
space:
mode:
authorPhilipp Gesang <phg42.2a@gmail.com>2014-12-14 12:29:58 +0100
committerPhilipp Gesang <phg42.2a@gmail.com>2014-12-14 12:29:58 +0100
commitc029158f450cd96f71db2f2a4da72a5c946089c0 (patch)
tree1c5d30d6853f6deb57cc9818b8739106a2d5e03f /src/fontloader/misc/fontloader-l-lpeg.lua
parentea148be27d5a79fd11759856e0ab4033f328ce5a (diff)
parentca673f7b14af906606a188fd98978d3501842f63 (diff)
downloadluaotfload-c029158f450cd96f71db2f2a4da72a5c946089c0.tar.gz
Merge pull request #260 from phi-gamma/master
add import helper, new directory layout, revised letterspacing, fontloader gone nuts
Diffstat (limited to 'src/fontloader/misc/fontloader-l-lpeg.lua')
-rw-r--r--src/fontloader/misc/fontloader-l-lpeg.lua1107
1 files changed, 1107 insertions, 0 deletions
diff --git a/src/fontloader/misc/fontloader-l-lpeg.lua b/src/fontloader/misc/fontloader-l-lpeg.lua
new file mode 100644
index 0000000..192e32f
--- /dev/null
+++ b/src/fontloader/misc/fontloader-l-lpeg.lua
@@ -0,0 +1,1107 @@
+if not modules then modules = { } end modules ['l-lpeg'] = {
+ version = 1.001,
+ comment = "companion to luat-lib.mkiv",
+ author = "Hans Hagen, PRAGMA-ADE, Hasselt NL",
+ copyright = "PRAGMA ADE / ConTeXt Development Team",
+ license = "see context related readme files"
+}
+
+-- lpeg 12 vs lpeg 10: slower compilation, similar parsing speed (i need to check
+-- if i can use new features like capture / 2 and .B (at first sight the xml
+-- parser is some 5% slower)
+
+-- a new lpeg fails on a #(1-P(":")) test and really needs a + P(-1)
+
+-- move utf -> l-unicode
+-- move string -> l-string or keep it here
+
+lpeg = require("lpeg")
+
+-- The latest lpeg doesn't have print any more, and even the new ones are not
+-- available by default (only when debug mode is enabled), which is a pitty as
+-- as it helps nailign down bottlenecks. Performance seems comparable: some 10%
+-- slower pattern compilation, same parsing speed, although,
+--
+-- local p = lpeg.C(lpeg.P(1)^0 * lpeg.P(-1))
+-- local a = string.rep("123",100)
+-- lpeg.match(p,a)
+--
+-- seems slower and is also still suboptimal (i.e. a match that runs from begin
+-- to end, one of the cases where string matchers win).
+
+if not lpeg.print then function lpeg.print(...) print(lpeg.pcode(...)) end end
+
+-- tracing (only used when we encounter a problem in integration of lpeg in luatex)
+
+-- some code will move to unicode and string
+
+-- local lpmatch = lpeg.match
+-- local lpprint = lpeg.print
+-- local lpp = lpeg.P
+-- local lpr = lpeg.R
+-- local lps = lpeg.S
+-- local lpc = lpeg.C
+-- local lpb = lpeg.B
+-- local lpv = lpeg.V
+-- local lpcf = lpeg.Cf
+-- local lpcb = lpeg.Cb
+-- local lpcg = lpeg.Cg
+-- local lpct = lpeg.Ct
+-- local lpcs = lpeg.Cs
+-- local lpcc = lpeg.Cc
+-- local lpcmt = lpeg.Cmt
+-- local lpcarg = lpeg.Carg
+
+-- function lpeg.match(l,...) print("LPEG MATCH") lpprint(l) return lpmatch(l,...) end
+
+-- function lpeg.P (l) local p = lpp (l) print("LPEG P =") lpprint(l) return p end
+-- function lpeg.R (l) local p = lpr (l) print("LPEG R =") lpprint(l) return p end
+-- function lpeg.S (l) local p = lps (l) print("LPEG S =") lpprint(l) return p end
+-- function lpeg.C (l) local p = lpc (l) print("LPEG C =") lpprint(l) return p end
+-- function lpeg.B (l) local p = lpb (l) print("LPEG B =") lpprint(l) return p end
+-- function lpeg.V (l) local p = lpv (l) print("LPEG V =") lpprint(l) return p end
+-- function lpeg.Cf (l) local p = lpcf (l) print("LPEG Cf =") lpprint(l) return p end
+-- function lpeg.Cb (l) local p = lpcb (l) print("LPEG Cb =") lpprint(l) return p end
+-- function lpeg.Cg (l) local p = lpcg (l) print("LPEG Cg =") lpprint(l) return p end
+-- function lpeg.Ct (l) local p = lpct (l) print("LPEG Ct =") lpprint(l) return p end
+-- function lpeg.Cs (l) local p = lpcs (l) print("LPEG Cs =") lpprint(l) return p end
+-- function lpeg.Cc (l) local p = lpcc (l) print("LPEG Cc =") lpprint(l) return p end
+-- function lpeg.Cmt (l) local p = lpcmt (l) print("LPEG Cmt =") lpprint(l) return p end
+-- function lpeg.Carg (l) local p = lpcarg(l) print("LPEG Carg =") lpprint(l) return p end
+
+local type, next, tostring = type, next, tostring
+local byte, char, gmatch, format = string.byte, string.char, string.gmatch, string.format
+----- mod, div = math.mod, math.div
+local floor = math.floor
+
+local P, R, S, V, Ct, C, Cs, Cc, Cp, Cmt = lpeg.P, lpeg.R, lpeg.S, lpeg.V, lpeg.Ct, lpeg.C, lpeg.Cs, lpeg.Cc, lpeg.Cp, lpeg.Cmt
+local lpegtype, lpegmatch, lpegprint = lpeg.type, lpeg.match, lpeg.print
+
+-- let's start with an inspector:
+
+if setinspector then
+ setinspector(function(v) if lpegtype(v) then lpegprint(v) return true end end)
+end
+
+-- Beware, we predefine a bunch of patterns here and one reason for doing so
+-- is that we get consistent behaviour in some of the visualizers.
+
+lpeg.patterns = lpeg.patterns or { } -- so that we can share
+local patterns = lpeg.patterns
+
+local anything = P(1)
+local endofstring = P(-1)
+local alwaysmatched = P(true)
+
+patterns.anything = anything
+patterns.endofstring = endofstring
+patterns.beginofstring = alwaysmatched
+patterns.alwaysmatched = alwaysmatched
+
+local sign = S('+-')
+local zero = P('0')
+local digit = R('09')
+local octdigit = R("07")
+local lowercase = R("az")
+local uppercase = R("AZ")
+local underscore = P("_")
+local hexdigit = digit + lowercase + uppercase
+local cr, lf, crlf = P("\r"), P("\n"), P("\r\n")
+----- newline = crlf + S("\r\n") -- cr + lf
+local newline = P("\r") * (P("\n") + P(true)) + P("\n")
+local escaped = P("\\") * anything
+local squote = P("'")
+local dquote = P('"')
+local space = P(" ")
+local period = P(".")
+local comma = P(",")
+
+local utfbom_32_be = P('\000\000\254\255') -- 00 00 FE FF
+local utfbom_32_le = P('\255\254\000\000') -- FF FE 00 00
+local utfbom_16_be = P('\254\255') -- FE FF
+local utfbom_16_le = P('\255\254') -- FF FE
+local utfbom_8 = P('\239\187\191') -- EF BB BF
+local utfbom = utfbom_32_be + utfbom_32_le
+ + utfbom_16_be + utfbom_16_le
+ + utfbom_8
+local utftype = utfbom_32_be * Cc("utf-32-be") + utfbom_32_le * Cc("utf-32-le")
+ + utfbom_16_be * Cc("utf-16-be") + utfbom_16_le * Cc("utf-16-le")
+ + utfbom_8 * Cc("utf-8") + alwaysmatched * Cc("utf-8") -- assume utf8
+local utfstricttype = utfbom_32_be * Cc("utf-32-be") + utfbom_32_le * Cc("utf-32-le")
+ + utfbom_16_be * Cc("utf-16-be") + utfbom_16_le * Cc("utf-16-le")
+ + utfbom_8 * Cc("utf-8")
+local utfoffset = utfbom_32_be * Cc(4) + utfbom_32_le * Cc(4)
+ + utfbom_16_be * Cc(2) + utfbom_16_le * Cc(2)
+ + utfbom_8 * Cc(3) + Cc(0)
+
+local utf8next = R("\128\191")
+
+patterns.utfbom_32_be = utfbom_32_be
+patterns.utfbom_32_le = utfbom_32_le
+patterns.utfbom_16_be = utfbom_16_be
+patterns.utfbom_16_le = utfbom_16_le
+patterns.utfbom_8 = utfbom_8
+
+patterns.utf_16_be_nl = P("\000\r\000\n") + P("\000\r") + P("\000\n") -- P("\000\r") * (P("\000\n") + P(true)) + P("\000\n")
+patterns.utf_16_le_nl = P("\r\000\n\000") + P("\r\000") + P("\n\000") -- P("\r\000") * (P("\n\000") + P(true)) + P("\n\000")
+
+patterns.utf_32_be_nl = P("\000\000\000\r\000\000\000\n") + P("\000\000\000\r") + P("\000\000\000\n")
+patterns.utf_32_le_nl = P("\r\000\000\000\n\000\000\000") + P("\r\000\000\000") + P("\n\000\000\000")
+
+patterns.utf8one = R("\000\127")
+patterns.utf8two = R("\194\223") * utf8next
+patterns.utf8three = R("\224\239") * utf8next * utf8next
+patterns.utf8four = R("\240\244") * utf8next * utf8next * utf8next
+patterns.utfbom = utfbom
+patterns.utftype = utftype
+patterns.utfstricttype = utfstricttype
+patterns.utfoffset = utfoffset
+
+local utf8char = patterns.utf8one + patterns.utf8two + patterns.utf8three + patterns.utf8four
+local validutf8char = utf8char^0 * endofstring * Cc(true) + Cc(false)
+
+local utf8character = P(1) * R("\128\191")^0 -- unchecked but fast
+
+patterns.utf8 = utf8char
+patterns.utf8char = utf8char
+patterns.utf8character = utf8character -- this one can be used in most cases so we might use that one
+patterns.validutf8 = validutf8char
+patterns.validutf8char = validutf8char
+
+local eol = S("\n\r")
+local spacer = S(" \t\f\v") -- + char(0xc2, 0xa0) if we want utf (cf mail roberto)
+local whitespace = eol + spacer
+local nonspacer = 1 - spacer
+local nonwhitespace = 1 - whitespace
+
+patterns.eol = eol
+patterns.spacer = spacer
+patterns.whitespace = whitespace
+patterns.nonspacer = nonspacer
+patterns.nonwhitespace = nonwhitespace
+
+local stripper = spacer ^0 * C((spacer ^0 * nonspacer ^1)^0) -- from example by roberto
+local fullstripper = whitespace^0 * C((whitespace^0 * nonwhitespace^1)^0)
+
+----- collapser = Cs(spacer^0/"" * ((spacer^1 * endofstring / "") + (spacer^1/" ") + P(1))^0)
+local collapser = Cs(spacer^0/"" * nonspacer^0 * ((spacer^0/" " * nonspacer^1)^0))
+
+local b_collapser = Cs( whitespace^0 /"" * (nonwhitespace^1 + whitespace^1/" ")^0)
+local e_collapser = Cs((whitespace^1 * P(-1)/"" + nonwhitespace^1 + whitespace^1/" ")^0)
+local m_collapser = Cs( (nonwhitespace^1 + whitespace^1/" ")^0)
+
+local b_stripper = Cs( spacer^0 /"" * (nonspacer^1 + spacer^1/" ")^0)
+local e_stripper = Cs((spacer^1 * P(-1)/"" + nonspacer^1 + spacer^1/" ")^0)
+local m_stripper = Cs( (nonspacer^1 + spacer^1/" ")^0)
+
+patterns.stripper = stripper
+patterns.fullstripper = fullstripper
+patterns.collapser = collapser
+
+patterns.b_collapser = b_collapser
+patterns.m_collapser = m_collapser
+patterns.e_collapser = e_collapser
+
+patterns.b_stripper = b_stripper
+patterns.m_stripper = m_stripper
+patterns.e_stripper = e_stripper
+
+patterns.lowercase = lowercase
+patterns.uppercase = uppercase
+patterns.letter = patterns.lowercase + patterns.uppercase
+patterns.space = space
+patterns.tab = P("\t")
+patterns.spaceortab = patterns.space + patterns.tab
+patterns.newline = newline
+patterns.emptyline = newline^1
+patterns.equal = P("=")
+patterns.comma = comma
+patterns.commaspacer = comma * spacer^0
+patterns.period = period
+patterns.colon = P(":")
+patterns.semicolon = P(";")
+patterns.underscore = underscore
+patterns.escaped = escaped
+patterns.squote = squote
+patterns.dquote = dquote
+patterns.nosquote = (escaped + (1-squote))^0
+patterns.nodquote = (escaped + (1-dquote))^0
+patterns.unsingle = (squote/"") * patterns.nosquote * (squote/"") -- will change to C in the middle
+patterns.undouble = (dquote/"") * patterns.nodquote * (dquote/"") -- will change to C in the middle
+patterns.unquoted = patterns.undouble + patterns.unsingle -- more often undouble
+patterns.unspacer = ((patterns.spacer^1)/"")^0
+
+patterns.singlequoted = squote * patterns.nosquote * squote
+patterns.doublequoted = dquote * patterns.nodquote * dquote
+patterns.quoted = patterns.doublequoted + patterns.singlequoted
+
+patterns.digit = digit
+patterns.octdigit = octdigit
+patterns.hexdigit = hexdigit
+patterns.sign = sign
+patterns.cardinal = digit^1
+patterns.integer = sign^-1 * digit^1
+patterns.unsigned = digit^0 * period * digit^1
+patterns.float = sign^-1 * patterns.unsigned
+patterns.cunsigned = digit^0 * comma * digit^1
+patterns.cpunsigned = digit^0 * (period + comma) * digit^1
+patterns.cfloat = sign^-1 * patterns.cunsigned
+patterns.cpfloat = sign^-1 * patterns.cpunsigned
+patterns.number = patterns.float + patterns.integer
+patterns.cnumber = patterns.cfloat + patterns.integer
+patterns.cpnumber = patterns.cpfloat + patterns.integer
+patterns.oct = zero * octdigit^1
+patterns.octal = patterns.oct
+patterns.HEX = zero * P("X") * (digit+uppercase)^1
+patterns.hex = zero * P("x") * (digit+lowercase)^1
+patterns.hexadecimal = zero * S("xX") * hexdigit^1
+
+patterns.hexafloat = sign^-1
+ * zero * S("xX")
+ * (hexdigit^0 * period * hexdigit^1 + hexdigit^1 * period * hexdigit^0 + hexdigit^1)
+ * (S("pP") * sign^-1 * hexdigit^1)^-1
+patterns.decafloat = sign^-1
+ * (digit^0 * period * digit^1 + digit^1 * period * digit^0 + digit^1)
+ * S("eE") * sign^-1 * digit^1
+
+patterns.propername = (uppercase + lowercase + underscore) * (uppercase + lowercase + underscore + digit)^0 * endofstring
+
+patterns.somecontent = (anything - newline - space)^1 -- (utf8char - newline - space)^1
+patterns.beginline = #(1-newline)
+
+patterns.longtostring = Cs(whitespace^0/"" * ((patterns.quoted + nonwhitespace^1 + whitespace^1/"" * (P(-1) + Cc(" ")))^0))
+
+local function anywhere(pattern) --slightly adapted from website
+ return P { P(pattern) + 1 * V(1) }
+end
+
+lpeg.anywhere = anywhere
+
+function lpeg.instringchecker(p)
+ p = anywhere(p)
+ return function(str)
+ return lpegmatch(p,str) and true or false
+ end
+end
+
+function lpeg.splitter(pattern, action)
+ return (((1-P(pattern))^1)/action+1)^0
+end
+
+function lpeg.tsplitter(pattern, action)
+ return Ct((((1-P(pattern))^1)/action+1)^0)
+end
+
+-- probleem: separator can be lpeg and that does not hash too well, but
+-- it's quite okay as the key is then not garbage collected
+
+local splitters_s, splitters_m, splitters_t = { }, { }, { }
+
+local function splitat(separator,single)
+ local splitter = (single and splitters_s[separator]) or splitters_m[separator]
+ if not splitter then
+ separator = P(separator)
+ local other = C((1 - separator)^0)
+ if single then
+ local any = anything
+ splitter = other * (separator * C(any^0) + "") -- ?
+ splitters_s[separator] = splitter
+ else
+ splitter = other * (separator * other)^0
+ splitters_m[separator] = splitter
+ end
+ end
+ return splitter
+end
+
+local function tsplitat(separator)
+ local splitter = splitters_t[separator]
+ if not splitter then
+ splitter = Ct(splitat(separator))
+ splitters_t[separator] = splitter
+ end
+ return splitter
+end
+
+lpeg.splitat = splitat
+lpeg.tsplitat = tsplitat
+
+function string.splitup(str,separator)
+ if not separator then
+ separator = ","
+ end
+ return lpegmatch(splitters_m[separator] or splitat(separator),str)
+end
+
+-- local p = splitat("->",false) print(lpegmatch(p,"oeps->what->more")) -- oeps what more
+-- local p = splitat("->",true) print(lpegmatch(p,"oeps->what->more")) -- oeps what->more
+-- local p = splitat("->",false) print(lpegmatch(p,"oeps")) -- oeps
+-- local p = splitat("->",true) print(lpegmatch(p,"oeps")) -- oeps
+
+local cache = { }
+
+function lpeg.split(separator,str)
+ local c = cache[separator]
+ if not c then
+ c = tsplitat(separator)
+ cache[separator] = c
+ end
+ return lpegmatch(c,str)
+end
+
+function string.split(str,separator)
+ if separator then
+ local c = cache[separator]
+ if not c then
+ c = tsplitat(separator)
+ cache[separator] = c
+ end
+ return lpegmatch(c,str)
+ else
+ return { str }
+ end
+end
+
+local spacing = patterns.spacer^0 * newline -- sort of strip
+local empty = spacing * Cc("")
+local nonempty = Cs((1-spacing)^1) * spacing^-1
+local content = (empty + nonempty)^1
+
+patterns.textline = content
+
+local linesplitter = tsplitat(newline)
+
+patterns.linesplitter = linesplitter
+
+function string.splitlines(str)
+ return lpegmatch(linesplitter,str)
+end
+
+-- lpeg.splitters = cache -- no longer public
+
+local cache = { }
+
+function lpeg.checkedsplit(separator,str)
+ local c = cache[separator]
+ if not c then
+ separator = P(separator)
+ local other = C((1 - separator)^1)
+ c = Ct(separator^0 * other * (separator^1 * other)^0)
+ cache[separator] = c
+ end
+ return lpegmatch(c,str)
+end
+
+function string.checkedsplit(str,separator)
+ local c = cache[separator]
+ if not c then
+ separator = P(separator)
+ local other = C((1 - separator)^1)
+ c = Ct(separator^0 * other * (separator^1 * other)^0)
+ cache[separator] = c
+ end
+ return lpegmatch(c,str)
+end
+
+-- from roberto's site:
+
+local function f2(s) local c1, c2 = byte(s,1,2) return c1 * 64 + c2 - 12416 end
+local function f3(s) local c1, c2, c3 = byte(s,1,3) return (c1 * 64 + c2) * 64 + c3 - 925824 end
+local function f4(s) local c1, c2, c3, c4 = byte(s,1,4) return ((c1 * 64 + c2) * 64 + c3) * 64 + c4 - 63447168 end
+
+local utf8byte = patterns.utf8one/byte + patterns.utf8two/f2 + patterns.utf8three/f3 + patterns.utf8four/f4
+
+patterns.utf8byte = utf8byte
+
+--~ local str = " a b c d "
+
+--~ local s = lpeg.stripper(lpeg.R("az")) print("["..lpegmatch(s,str).."]")
+--~ local s = lpeg.keeper(lpeg.R("az")) print("["..lpegmatch(s,str).."]")
+--~ local s = lpeg.stripper("ab") print("["..lpegmatch(s,str).."]")
+--~ local s = lpeg.keeper("ab") print("["..lpegmatch(s,str).."]")
+
+local cache = { }
+
+function lpeg.stripper(str)
+ if type(str) == "string" then
+ local s = cache[str]
+ if not s then
+ s = Cs(((S(str)^1)/"" + 1)^0)
+ cache[str] = s
+ end
+ return s
+ else
+ return Cs(((str^1)/"" + 1)^0)
+ end
+end
+
+local cache = { }
+
+function lpeg.keeper(str)
+ if type(str) == "string" then
+ local s = cache[str]
+ if not s then
+ s = Cs((((1-S(str))^1)/"" + 1)^0)
+ cache[str] = s
+ end
+ return s
+ else
+ return Cs((((1-str)^1)/"" + 1)^0)
+ end
+end
+
+function lpeg.frontstripper(str) -- or pattern (yet undocumented)
+ return (P(str) + P(true)) * Cs(anything^0)
+end
+
+function lpeg.endstripper(str) -- or pattern (yet undocumented)
+ return Cs((1 - P(str) * endofstring)^0)
+end
+
+-- Just for fun I looked at the used bytecode and
+-- p = (p and p + pp) or pp gets one more (testset).
+
+-- todo: cache when string
+
+function lpeg.replacer(one,two,makefunction,isutf) -- in principle we should sort the keys
+ local pattern
+ local u = isutf and utf8char or 1
+ if type(one) == "table" then
+ local no = #one
+ local p = P(false)
+ if no == 0 then
+ for k, v in next, one do
+ p = p + P(k) / v
+ end
+ pattern = Cs((p + u)^0)
+ elseif no == 1 then
+ local o = one[1]
+ one, two = P(o[1]), o[2]
+ -- pattern = Cs(((1-one)^1 + one/two)^0)
+ pattern = Cs((one/two + u)^0)
+ else
+ for i=1,no do
+ local o = one[i]
+ p = p + P(o[1]) / o[2]
+ end
+ pattern = Cs((p + u)^0)
+ end
+ else
+ pattern = Cs((P(one)/(two or "") + u)^0)
+ end
+ if makefunction then
+ return function(str)
+ return lpegmatch(pattern,str)
+ end
+ else
+ return pattern
+ end
+end
+
+-- local pattern1 = P(1-P(pattern))^0 * P(pattern) : test for not nil
+-- local pattern2 = (P(pattern) * Cc(true) + P(1))^0 : test for true (could be faster, but not much)
+
+function lpeg.finder(lst,makefunction,isutf) -- beware: slower than find with 'patternless finds'
+ local pattern
+ if type(lst) == "table" then
+ pattern = P(false)
+ if #lst == 0 then
+ for k, v in next, lst do
+ pattern = pattern + P(k) -- ignore key, so we can use a replacer table
+ end
+ else
+ for i=1,#lst do
+ pattern = pattern + P(lst[i])
+ end
+ end
+ else
+ pattern = P(lst)
+ end
+ if isutf then
+ pattern = ((utf8char or 1)-pattern)^0 * pattern
+ else
+ pattern = (1-pattern)^0 * pattern
+ end
+ if makefunction then
+ return function(str)
+ return lpegmatch(pattern,str)
+ end
+ else
+ return pattern
+ end
+end
+
+-- print(lpeg.match(lpeg.replacer("e","a"),"test test"))
+-- print(lpeg.match(lpeg.replacer{{"e","a"}},"test test"))
+-- print(lpeg.match(lpeg.replacer({ e = "a", t = "x" }),"test test"))
+
+local splitters_f, splitters_s = { }, { }
+
+function lpeg.firstofsplit(separator) -- always return value
+ local splitter = splitters_f[separator]
+ if not splitter then
+ local pattern = P(separator)
+ splitter = C((1 - pattern)^0)
+ splitters_f[separator] = splitter
+ end
+ return splitter
+end
+
+function lpeg.secondofsplit(separator) -- nil if not split
+ local splitter = splitters_s[separator]
+ if not splitter then
+ local pattern = P(separator)
+ splitter = (1 - pattern)^0 * pattern * C(anything^0)
+ splitters_s[separator] = splitter
+ end
+ return splitter
+end
+
+local splitters_s, splitters_p = { }, { }
+
+function lpeg.beforesuffix(separator) -- nil if nothing but empty is ok
+ local splitter = splitters_s[separator]
+ if not splitter then
+ local pattern = P(separator)
+ splitter = C((1 - pattern)^0) * pattern * endofstring
+ splitters_s[separator] = splitter
+ end
+ return splitter
+end
+
+function lpeg.afterprefix(separator) -- nil if nothing but empty is ok
+ local splitter = splitters_p[separator]
+ if not splitter then
+ local pattern = P(separator)
+ splitter = pattern * C(anything^0)
+ splitters_p[separator] = splitter
+ end
+ return splitter
+end
+
+function lpeg.balancer(left,right)
+ left, right = P(left), P(right)
+ return P { left * ((1 - left - right) + V(1))^0 * right }
+end
+
+-- print(1,lpegmatch(lpeg.firstofsplit(":"),"bc:de"))
+-- print(2,lpegmatch(lpeg.firstofsplit(":"),":de")) -- empty
+-- print(3,lpegmatch(lpeg.firstofsplit(":"),"bc"))
+-- print(4,lpegmatch(lpeg.secondofsplit(":"),"bc:de"))
+-- print(5,lpegmatch(lpeg.secondofsplit(":"),"bc:")) -- empty
+-- print(6,lpegmatch(lpeg.secondofsplit(":",""),"bc"))
+-- print(7,lpegmatch(lpeg.secondofsplit(":"),"bc"))
+-- print(9,lpegmatch(lpeg.secondofsplit(":","123"),"bc"))
+
+-- -- slower:
+--
+-- function lpeg.counter(pattern)
+-- local n, pattern = 0, (lpeg.P(pattern)/function() n = n + 1 end + lpeg.anything)^0
+-- return function(str) n = 0 ; lpegmatch(pattern,str) ; return n end
+-- end
+
+local nany = utf8char/""
+
+function lpeg.counter(pattern)
+ pattern = Cs((P(pattern)/" " + nany)^0)
+ return function(str)
+ return #lpegmatch(pattern,str)
+ end
+end
+
+-- utf extensies
+
+utf = utf or (unicode and unicode.utf8) or { }
+
+local utfcharacters = utf and utf.characters or string.utfcharacters
+local utfgmatch = utf and utf.gmatch
+local utfchar = utf and utf.char
+
+lpeg.UP = lpeg.P
+
+if utfcharacters then
+
+ function lpeg.US(str)
+ local p = P(false)
+ for uc in utfcharacters(str) do
+ p = p + P(uc)
+ end
+ return p
+ end
+
+
+elseif utfgmatch then
+
+ function lpeg.US(str)
+ local p = P(false)
+ for uc in utfgmatch(str,".") do
+ p = p + P(uc)
+ end
+ return p
+ end
+
+else
+
+ function lpeg.US(str)
+ local p = P(false)
+ local f = function(uc)
+ p = p + P(uc)
+ end
+ lpegmatch((utf8char/f)^0,str)
+ return p
+ end
+
+end
+
+local range = utf8byte * utf8byte + Cc(false) -- utf8byte is already a capture
+
+function lpeg.UR(str,more)
+ local first, last
+ if type(str) == "number" then
+ first = str
+ last = more or first
+ else
+ first, last = lpegmatch(range,str)
+ if not last then
+ return P(str)
+ end
+ end
+ if first == last then
+ return P(str)
+ elseif utfchar and (last - first < 8) then -- a somewhat arbitrary criterium
+ local p = P(false)
+ for i=first,last do
+ p = p + P(utfchar(i))
+ end
+ return p -- nil when invalid range
+ else
+ local f = function(b)
+ return b >= first and b <= last
+ end
+ -- tricky, these nested captures
+ return utf8byte / f -- nil when invalid range
+ end
+end
+
+-- print(lpeg.match(lpeg.Cs((C(lpeg.UR("αω"))/{ ["χ"] = "OEPS" })^0),"αωχαω"))
+
+-- lpeg.print(lpeg.R("ab","cd","gh"))
+-- lpeg.print(lpeg.P("a","b","c"))
+-- lpeg.print(lpeg.S("a","b","c"))
+
+-- print(lpeg.count("äáàa",lpeg.P("á") + lpeg.P("à")))
+-- print(lpeg.count("äáàa",lpeg.UP("áà")))
+-- print(lpeg.count("äáàa",lpeg.US("àá")))
+-- print(lpeg.count("äáàa",lpeg.UR("aá")))
+-- print(lpeg.count("äáàa",lpeg.UR("àá")))
+-- print(lpeg.count("äáàa",lpeg.UR(0x0000,0xFFFF)))
+
+function lpeg.is_lpeg(p)
+ return p and lpegtype(p) == "pattern"
+end
+
+function lpeg.oneof(list,...) -- lpeg.oneof("elseif","else","if","then") -- assume proper order
+ if type(list) ~= "table" then
+ list = { list, ... }
+ end
+ -- table.sort(list) -- longest match first
+ local p = P(list[1])
+ for l=2,#list do
+ p = p + P(list[l])
+ end
+ return p
+end
+
+-- For the moment here, but it might move to utilities. Beware, we need to
+-- have the longest keyword first, so 'aaa' comes beforte 'aa' which is why we
+-- loop back from the end cq. prepend.
+
+local sort = table.sort
+
+local function copyindexed(old)
+ local new = { }
+ for i=1,#old do
+ new[i] = old
+ end
+ return new
+end
+
+local function sortedkeys(tab)
+ local keys, s = { }, 0
+ for key,_ in next, tab do
+ s = s + 1
+ keys[s] = key
+ end
+ sort(keys)
+ return keys
+end
+
+function lpeg.append(list,pp,delayed,checked)
+ local p = pp
+ if #list > 0 then
+ local keys = copyindexed(list)
+ sort(keys)
+ for i=#keys,1,-1 do
+ local k = keys[i]
+ if p then
+ p = P(k) + p
+ else
+ p = P(k)
+ end
+ end
+ elseif delayed then -- hm, it looks like the lpeg parser resolves anyway
+ local keys = sortedkeys(list)
+ if p then
+ for i=1,#keys,1 do
+ local k = keys[i]
+ local v = list[k]
+ p = P(k)/list + p
+ end
+ else
+ for i=1,#keys do
+ local k = keys[i]
+ local v = list[k]
+ if p then
+ p = P(k) + p
+ else
+ p = P(k)
+ end
+ end
+ if p then
+ p = p / list
+ end
+ end
+ elseif checked then
+ -- problem: substitution gives a capture
+ local keys = sortedkeys(list)
+ for i=1,#keys do
+ local k = keys[i]
+ local v = list[k]
+ if p then
+ if k == v then
+ p = P(k) + p
+ else
+ p = P(k)/v + p
+ end
+ else
+ if k == v then
+ p = P(k)
+ else
+ p = P(k)/v
+ end
+ end
+ end
+ else
+ local keys = sortedkeys(list)
+ for i=1,#keys do
+ local k = keys[i]
+ local v = list[k]
+ if p then
+ p = P(k)/v + p
+ else
+ p = P(k)/v
+ end
+ end
+ end
+ return p
+end
+
+-- inspect(lpeg.append({ a = "1", aa = "1", aaa = "1" } ,nil,true))
+-- inspect(lpeg.append({ ["degree celsius"] = "1", celsius = "1", degree = "1" } ,nil,true))
+
+-- function lpeg.exact_match(words,case_insensitive)
+-- local pattern = concat(words)
+-- if case_insensitive then
+-- local pattern = S(upper(characters)) + S(lower(characters))
+-- local list = { }
+-- for i=1,#words do
+-- list[lower(words[i])] = true
+-- end
+-- return Cmt(pattern^1, function(_,i,s)
+-- return list[lower(s)] and i
+-- end)
+-- else
+-- local pattern = S(concat(words))
+-- local list = { }
+-- for i=1,#words do
+-- list[words[i]] = true
+-- end
+-- return Cmt(pattern^1, function(_,i,s)
+-- return list[s] and i
+-- end)
+-- end
+-- end
+
+-- experiment:
+
+-- local function make(t)
+-- local p
+-- local keys = sortedkeys(t)
+-- for i=1,#keys do
+-- local k = keys[i]
+-- local v = t[k]
+-- if not p then
+-- if next(v) then
+-- p = P(k) * make(v)
+-- else
+-- p = P(k)
+-- end
+-- else
+-- if next(v) then
+-- p = p + P(k) * make(v)
+-- else
+-- p = p + P(k)
+-- end
+-- end
+-- end
+-- return p
+-- end
+
+-- local function make(t)
+-- local p = P(false)
+-- local keys = sortedkeys(t)
+-- for i=1,#keys do
+-- local k = keys[i]
+-- local v = t[k]
+-- if next(v) then
+-- p = p + P(k) * make(v)
+-- else
+-- p = p + P(k)
+-- end
+-- end
+-- return p
+-- end
+
+-- function lpeg.utfchartabletopattern(list) -- goes to util-lpg
+-- local tree = { }
+-- for i=1,#list do
+-- local t = tree
+-- for c in gmatch(list[i],".") do
+-- local tc = t[c]
+-- if not tc then
+-- tc = { }
+-- t[c] = tc
+-- end
+-- t = tc
+-- end
+-- end
+-- return make(tree)
+-- end
+
+local function make(t,hash)
+ local p = P(false)
+ local keys = sortedkeys(t)
+ for i=1,#keys do
+ local k = keys[i]
+ local v = t[k]
+ local h = hash[v]
+ if h then
+ if next(v) then
+ p = p + P(k) * (make(v,hash) + P(true))
+ else
+ p = p + P(k) * P(true)
+ end
+ else
+ if next(v) then
+ p = p + P(k) * make(v,hash)
+ else
+ p = p + P(k)
+ end
+ end
+ end
+ return p
+end
+
+function lpeg.utfchartabletopattern(list) -- goes to util-lpg
+ local tree = { }
+ local hash = { }
+ local n = #list
+ if n == 0 then
+ -- we could always use this branch
+ for s in next, list do
+ local t = tree
+ for c in gmatch(s,".") do
+ local tc = t[c]
+ if not tc then
+ tc = { }
+ t[c] = tc
+ end
+ t = tc
+ end
+ hash[t] = s
+ end
+ else
+ for i=1,n do
+ local t = tree
+ local s = list[i]
+ for c in gmatch(s,".") do
+ local tc = t[c]
+ if not tc then
+ tc = { }
+ t[c] = tc
+ end
+ t = tc
+ end
+ hash[t] = s
+ end
+ end
+ return make(tree,hash)
+end
+
+-- inspect ( lpeg.utfchartabletopattern {
+-- utfchar(0x00A0), -- nbsp
+-- utfchar(0x2000), -- enquad
+-- utfchar(0x2001), -- emquad
+-- utfchar(0x2002), -- enspace
+-- utfchar(0x2003), -- emspace
+-- utfchar(0x2004), -- threeperemspace
+-- utfchar(0x2005), -- fourperemspace
+-- utfchar(0x2006), -- sixperemspace
+-- utfchar(0x2007), -- figurespace
+-- utfchar(0x2008), -- punctuationspace
+-- utfchar(0x2009), -- breakablethinspace
+-- utfchar(0x200A), -- hairspace
+-- utfchar(0x200B), -- zerowidthspace
+-- utfchar(0x202F), -- narrownobreakspace
+-- utfchar(0x205F), -- math thinspace
+-- } )
+
+-- a few handy ones:
+--
+-- faster than find(str,"[\n\r]") when match and # > 7 and always faster when # > 3
+
+patterns.containseol = lpeg.finder(eol) -- (1-eol)^0 * eol
+
+-- The next pattern^n variant is based on an approach suggested
+-- by Roberto: constructing a big repetition in chunks.
+--
+-- Being sparse is not needed, and only complicate matters and
+-- the number of redundant entries is not that large.
+
+local function nextstep(n,step,result)
+ local m = n % step -- mod(n,step)
+ local d = floor(n/step) -- div(n,step)
+ if d > 0 then
+ local v = V(tostring(step))
+ local s = result.start
+ for i=1,d do
+ if s then
+ s = v * s
+ else
+ s = v
+ end
+ end
+ result.start = s
+ end
+ if step > 1 and result.start then
+ local v = V(tostring(step/2))
+ result[tostring(step)] = v * v
+ end
+ if step > 0 then
+ return nextstep(m,step/2,result)
+ else
+ return result
+ end
+end
+
+function lpeg.times(pattern,n)
+ return P(nextstep(n,2^16,{ "start", ["1"] = pattern }))
+end
+
+-- local p = lpeg.Cs((1 - lpeg.times(lpeg.P("AB"),25))^1)
+-- local s = "12" .. string.rep("AB",20) .. "34" .. string.rep("AB",30) .. "56"
+-- inspect(p)
+-- print(lpeg.match(p,s))
+
+-- moved here (before util-str)
+
+----- digit = R("09")
+----- period = P(".")
+----- zero = P("0")
+local trailingzeros = zero^0 * -digit -- suggested by Roberto R
+local case_1 = period * trailingzeros / ""
+local case_2 = period * (digit - trailingzeros)^1 * (trailingzeros / "")
+local number = digit^1 * (case_1 + case_2)
+local stripper = Cs((number + 1)^0)
+
+lpeg.patterns.stripzeros = stripper
+
+-- local sample = "bla 11.00 bla 11 bla 0.1100 bla 1.00100 bla 0.00 bla 0.001 bla 1.1100 bla 0.100100100 bla 0.00100100100"
+-- collectgarbage("collect")
+-- str = string.rep(sample,10000)
+-- local ts = os.clock()
+-- lpegmatch(stripper,str)
+-- print(#str, os.clock()-ts, lpegmatch(stripper,sample))
+
+-- for practical reasone we keep this here:
+
+local byte_to_HEX = { }
+local byte_to_hex = { }
+local byte_to_dec = { } -- for md5
+local hex_to_byte = { }
+
+for i=0,255 do
+ local H = format("%02X",i)
+ local h = format("%02x",i)
+ local d = format("%03i",i)
+ local c = char(i)
+ byte_to_HEX[c] = H
+ byte_to_hex[c] = h
+ byte_to_dec[c] = d
+ hex_to_byte[h] = c
+ hex_to_byte[H] = c
+end
+
+local hextobyte = P(2)/hex_to_byte
+local bytetoHEX = P(1)/byte_to_HEX
+local bytetohex = P(1)/byte_to_hex
+local bytetodec = P(1)/byte_to_dec
+local hextobytes = Cs(hextobyte^0)
+local bytestoHEX = Cs(bytetoHEX^0)
+local bytestohex = Cs(bytetohex^0)
+local bytestodec = Cs(bytetodec^0)
+
+patterns.hextobyte = hextobyte
+patterns.bytetoHEX = bytetoHEX
+patterns.bytetohex = bytetohex
+patterns.bytetodec = bytetodec
+patterns.hextobytes = hextobytes
+patterns.bytestoHEX = bytestoHEX
+patterns.bytestohex = bytestohex
+patterns.bytestodec = bytestodec
+
+function string.toHEX(s)
+ if not s or s == "" then
+ return s
+ else
+ return lpegmatch(bytestoHEX,s)
+ end
+end
+
+function string.tohex(s)
+ if not s or s == "" then
+ return s
+ else
+ return lpegmatch(bytestohex,s)
+ end
+end
+
+function string.todec(s)
+ if not s or s == "" then
+ return s
+ else
+ return lpegmatch(bytestodec,s)
+ end
+end
+
+function string.tobytes(s)
+ if not s or s == "" then
+ return s
+ else
+ return lpegmatch(hextobytes,s)
+ end
+end
+
+-- local h = "ADFE0345"
+-- local b = lpegmatch(patterns.hextobytes,h)
+-- print(h,b,string.tohex(b),string.toHEX(b))