mime.lua

   1 #!/usr/bin/lua
   2 --
   3 -- Copyright (c) 2009 Mauro Iazzi
   4 --
   5 -- Permission is hereby granted, free of charge, to any person
   6 -- obtaining a copy of this software and associated documentation
   7 -- files (the "Software"), to deal in the Software without
   8 -- restriction, including without limitation the rights to use,
   9 -- copy, modify, merge, publish, distribute, sublicense, and/or sell
  10 -- copies of the Software, and to permit persons to whom the
  11 -- Software is furnished to do so, subject to the following
  12 -- conditions:
  13 --
  14 -- The above copyright notice and this permission notice shall be
  15 -- included in all copies or substantial portions of the Software.
  16 --
  17 -- THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
  18 -- EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
  19 -- OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
  20 -- NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
  21 -- HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
  22 -- WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  23 -- FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  24 -- OTHER DEALINGS IN THE SOFTWARE.
  25 --/
  26
  27 require 'lpeg'
  28
  29 -- MIME implementation in Lua
  30
  31 -- XXX: at section 3.4
  32
  33
  34 local string = string
  35 local type = type
  36 local tostring = tostring
  37 local select = select
  38 local setfenv = setfenv
  39 local CRLF = '\r\n'
  40 local lpeg = lpeg
  41 local print = print
  42
  43 -- this creates a named and substitution capture for the given pattern
  44 lpeg.Csg = function (patt, name)
  45         return (#lpeg.Cg(patt, name)) * lpeg.Cs(patt)
  46 end
  47 -- this collects both the table of named captures and the
  48 -- substitution capture of the given pattern
  49 lpeg.Cst = function (patt)
  50         return (#lpeg.Cs(patt) * lpeg.Ct(patt)) / function (s, t)
  51                 local ret = {}
  52                 ret [1] = s
  53                 for k, v in pairs(t) do
  54                         if type(k)~='number' then
  55                                 ret[k] = v
  56                         end
  57                 end
  58                 return ret
  59         end
  60 end
  61
  62 -- rfc2822 Section 2.1:
  63 --
  64 -- A message consists of header fields (collectively called "the header
  65 -- of the message") followed, optionally, by a body.  The header is a
  66 -- sequence of lines of characters with special syntax as defined in
  67 -- this standard. The body is simply a sequence of characters that
  68 -- follows the header and is separated from the header by an empty line
  69 -- (i.e., a line with nothing preceding the CRLF).
  70 --
  71 local split_content = function (c)
  72         -- NOTE: there is no guarantee that the CRLF separating header
  73         -- and body is the first one. However no headers admit a blank
  74         -- line so far
  75         local h, b = string.match(c, '^(.-)'..CRLF..CRLF..'(.*)$')
  76         assert(type(h)=='string' and type(b)=='string')
  77         return h, b
  78 end
  79
  80 -- rfc2822 Section 2.2.3
  81 --
  82 -- The process of moving from this folded multiple-line representation
  83 -- of a header field to its single line representation is called
  84 -- "unfolding". Unfolding is accomplished by simply removing any CRLF
  85 -- that is immediately followed by WSP.  Each header field should be
  86 -- treated in its unfolded form for further syntactic and semantic
  87 -- evaluation.
  88 --
  89 local unfold_header = function (h)
  90         return string.gsub(h, CRLF..'[ \t]', ' ')
  91 end
  92
  93 -- rfc2822 Section 2.1
  94 --
  95 -- At the most basic level, a message is a series of characters.  A
  96 -- message that is conformant with this standard is comprised of
  97 -- characters with values in the range 1 through 127 and interpreted as
  98 -- US-ASCII characters [ASCII].  For brevity, this document sometimes
  99 -- refers to this range of characters as simply "US-ASCII characters".
 100 --
 101
 102 -- rfc2822 Section 2.1.1
 103 --
 104 -- There are two limits that this standard places on the number of
 105 -- characters in a line. Each line of characters MUST be no more than
 106 -- 998 characters, and SHOULD be no more than 78 characters, excluding
 107 -- the CRLF.
 108 --
 109 local check_content = function (c)
 110         local err, warn
 111         -- TODO: implement in a portable way
 112         return err, warn
 113 end
 114
 115 -- rfc2822 Section 2.3
 116 --
 117 -- The body of a message is simply lines of US-ASCII characters.  The
 118 -- only two limitations on the body are as follows:
 119 --
 120 -- - CR and LF MUST only occur together as CRLF; they MUST NOT appear
 121 -- independently in the body.
 122 --
 123 -- - Lines of characters in the body MUST be limited to 998 characters,
 124 --
 125 -- and SHOULD be limited to 78 characters, excluding the CRLF.
 126 --
 127 local check_body = function (b)
 128         local err, warn
 129         -- TODO: implement
 130         return err, warn
 131 end
 132
 133 -- some core values (not implemented literally)
 134 -- see RFC 2234 Section 6.1
 135 local core_values = {
 136         ["ALPHA"] = lpeg.R("AZ", "az"),
 137         ["BIT"] = lpeg.P("0") + lpeg.P("1"),
 138         ["CHAR"] = lpeg.R("\01\127"),
 139         ["CR"] = lpeg.P("\13"),
 140         ["CRLF"] = lpeg.P("\13\10"),
 141         ["DIGIT"] = lpeg.R("09"),
 142         ["LF"] = lpeg.P("\10"),
 143         ["WSP"] = lpeg.S("\32\09"),
 144 }
 145
 146 -- lexical tokens used in the specification
 147 -- TODO: could be optimized
 148 -- check for non-obfuscated optimizations
 149 -- TODO: write in a less obfuscated way
 150 -- check if parenthesis can be removed
 151 local lex_tokens = function ()
 152         setfenv(1, lpeg)
 153         return {
 154         -- control characters without whitespaces
 155         ["NO-WS-CTL"] = R("\1\8") +
 156         P("\11") +
 157         P("\12") +
 158         R("\14\13") +
 159         P("\127"), -- RFC 2822 Section 3.2.1
 160         -- a character in a text
 161         ["text"] =      R("\1\9") +
 162         P("\11") +
 163         P("\12") +
 164         R("\14\127"), -- RFC 2822 Section 3.2.1
 165         -- a special character
 166         ["specials"] =  P("(") + P(")") +
 167         P("<") + P(">") +
 168         P("[") + P("]") +
 169         P(":") + P(";") +
 170         P("@") + P("\\") +
 171         P(",") + P(".") +
 172         P("\""), -- RFC 2822 Section 3.2.1
 173         -- a quoted pair should return only the second character
 174         ["quoted-pair"] = (P("\\")/'' * V("text")) + V("obs-qp"), -- RFC 2822 Section 3.2.2
 175         -- a folding white space (a whitespace that can include a CRLF)
 176         -- should be substituted by a single whitespace
 177         ["FWS"] = (((V("WSP")^0 * V("CRLF"))^-1 * V("WSP")^1) + V("obs-FWS")) / " ", -- RFC 2822 Section 3.2.3
 178         -- a text character allowed inside a comment
 179         ["ctext"] =     V("NO-WS-CTL") +
 180         R("\33\39") +
 181         R("\42\91") +
 182         R("\93\126"), -- RFC 2822 Section 3.2.3
 183         -- the content of a comment (comments can nest)
 184         ["ccontent"] =  V("ctext") + V("quoted-pair") + V("comment"), -- RFC 2822 Section 3.2.3
 185         -- an actual comment
 186         -- should be substituted by a single whitespace
 187         ["comment"] =   (P("(") * (V("FWS")^-1 * V("ccontent"))^0 * V("FWS")^-1 * P(")")) / " ", -- RFC 2822 Section 3.2.3
 188         -- a comment or a folding white space
 189         -- should be substituted by a single whitespace
 190         --
 191         -- Folding white spaces should not be placed in a way that
 192         -- creates lines containing only whitespaces.
 193         -- This requirement Is not necessarily enforced by this grammar
 194         ["CFWS"] =      ( (V("FWS") + (V("comment")*V("FWS")^-1)) * (V("comment")*V("FWS")^-1)^0 ) / " ", -- RFC 2822 Section 3.2.3
 195         -- character that can appear in an atom
 196         ["atext"] =     V("ALPHA") + V("DIGIT") +
 197         P("!") + P("#") +
 198         P("$") + P("%") +
 199         P("&") + P("'") +
 200         P("*") + P("+") +
 201         P("-") + P("/") +
 202         P("=") + P("?") +
 203         P("^") + P("_") +
 204         P("`") + P("{") +
 205         P("|") + P("}") +
 206         P("~"), -- RFC 2822 Section 3.2.4
 207         -- an atom is equal to the content only discarding comments and whitespace
 208         ["atom"] =      V("CFWS")^-1 * C(V("atext")^1) * V("CFWS")^-1, -- RFC 2822 Section 3.2.4
 209         -- an atom with dots is only the content, discarding CFWSs
 210         ["dot-atom"] =  V("CFWS")^-1 * C(V("dot-atom-text")) * V("CFWS")^-1, -- RFC 2822 Section 3.2.4
 211         -- the content of an atom text with dots
 212         ["dot-atom-text"] =     V("atext")^1 * (P(".") * V("atext")^1)^0, -- RFC 2822 Section 3.2.4
 213         -- character that can appear in a quoted string
 214         ["qtext"] =     V("NO-WS-CTL") +
 215         P("\33") +
 216         R("\35\91") +
 217         R("\93\126"), -- RFC 2822 Section 3.2.5
 218         -- character or quoted pair (both can appear in a quoted string)
 219         -- it is equivalent to the character itself or to the result of the
 220         -- quoted pair
 221         ["qcontent"] =  C(V("qtext")) + V("quoted-pair"),  -- RFC 2822 Section 3.2.5
 222         -- a quoted string is equal to its content
 223         ["quoted-string"] =     V("CFWS")^-1 *
 224         P("\"") * (V("FWS")^-1 * V("qcontent"))^0 * V("FWS")^-1 * P("\"") * V("CFWS")^-1, -- RFC 2822 Section 3.2.5
 225         -- unstructured patterns for unspecified headers
 226         -- what should these be equal to?
 227         --
 228         -- an generic word
 229         ["word"] =      V("atom") + V("quoted-string"), -- RFC 2822 Section 3.2.6
 230         -- an generic phrase
 231         ["phrase"] =    Cs(V("word")^1) + V("obs-phrase"), -- RFC 2822 Section 3.2.6
 232         -- a character for unstructured text
 233         ["utext"] =     V("NO-WS-CTL") + R("\33\126") + V("obs-utext"), -- RFC 2822 Section 3.2.6
 234         -- an unstructured text
 235         ["unstructured"] =      (V("FWS")^-1 * V("utext"))^0 * V("FWS")^-1, -- RFC 2822 Section 3.2.6
 236 }
 237 end
 238
 239 local show = my.show
 240 local deb = function(...) for i=1,select('#',...) do local v=select(i,...)print(type(v), v)end end
 241
 242 local date_time = function ()
 243         setfenv(1, lpeg)
 244         return {
 245         -- date and time specification
 246         -- dates and times should be valid
 247         -- this grammar does not enforce this yet
 248         ["date-time"] = Cst(( V"day-of-week" * P"," )^-1 * V"date" * V"FWS" * V"time" * V"CFWS"^-1), -- RFC 2822 Section 3.3
 249         ["day-of-week"] = ( V"FWS"^-1 * V"day-name" ) + V"obs-day-of-week", -- RFC 2822 Section 3.3
 250         ["day-name"] = Cg(P"Mon" + P"Tue" + P"Wed" + P"Thu" + P"Fri" + P"Sat" + P"Sun", "weekday"), -- RFC 2822 Section 3.3
 251         ["date"] = V"day" * V"month" * V"year", -- RFC 2822 Section 3.3
 252         ["year"] = Cg(C(V"DIGIT"^4), "year") + V"obs-year", -- RFC 2822 Section 3.3
 253         ["month"] = (V"FWS" * V"month-name" * V"FWS") + V"obs-month", -- RFC 2822 Section 3.3
 254         ["month-name"] = Cg(C"Jan" + C"Feb" + C"Mar" + C"Apr" + C"May" + C"Jun" + C"Jul" + C"Aug" + C"Sep" + C"Oct" + C"Nov" + C"Dec", "month"), -- RFC 2822 Section 3.3
 255         ["day"] = (V"FWS"^-1 * Cg(C(V"DIGIT" * V"DIGIT"^-1), "day")) + V"obs-day", -- RFC 2822 Section 3.3
 256         ["time"] = V"time-of-day" * V"FWS" * V"zone", -- RFC 2822 Section 3.3
 257         ["time-of-day"] = V"hour" * P":" * V"minute" * (P":" * V"second")^-1, -- RFC 2822 Section 3.3
 258         ["hour"] = Cg(C(V"DIGIT" * V"DIGIT"), "hour") + V"obs-hour", -- RFC 2822 Section 3.3
 259         ["minute"] = Cg((V"DIGIT" * V"DIGIT"), "minute") + V"obs-minute", -- RFC 2822 Section 3.3
 260         ["second"] = Cg((V"DIGIT" * V"DIGIT"), "second") + V"obs-second", -- RFC 2822 Section 3.3
 261         ["zone"] = Cg( (P"+" + P"-") * V"DIGIT" * V"DIGIT" * V"DIGIT" * V"DIGIT", "zone" ) + V"obs-zone", -- RFC 2822 Section 3.3
 262 }
 263 end
 264
 265 local address = function ()
 266         setfenv(1, lpeg)
 267         return {
 268         -- address specification
 269         -- dates and times should be valid
 270         -- this grammar does not enforce this yet
 271         ["address"] = V"mailbox" + V"group", -- RFC 2822 Section 3.4
 272         ["mailbox"] = V"name-addr" + V"addr-spec", -- RFC 2822 Section 3.4
 273         ["name-addr"] = V"display-name"^-1 * V"angle-addr", -- RFC 2822 Section 3.4
 274         ["angle-addr"] = (V"CFWS"^-1 * P"<" * V"addr-spec" * P">" * V"CFWS"^-1) + V"obs-angle-addr", -- RFC 2822 Section 3.4
 275         ["group"] = V"display-name" * P":" * (V"mailbox-list" + V"CFWS") * P";" * V"CFWS"^-1, -- RFC 2822 Section 3.4
 276         ["display-name"] = V"phrase", -- RFC 2822 Section 3.4
 277         ["mailbox-list"] = (V"mailbox" * (P"," * V"mailbox")^0) + V"obs-mailbox-list", -- RFC 2822 Section 3.4
 278         ["address-list"] = (V"address" * (P"," * V"address")^0) + V"obs-address-list", -- RFC 2822 Section 3.4
 279         -- address specification (name@host.domain)
 280         ["addr-spec"] = Cst(V"local-part" * P"@" * V"domain"), -- RFC 2822 Section 3.4.1
 281         ["local-part"] = Csg(V"dot-atom" + V"quoted-string" + V"obs-local-part", "box"), -- RFC 2822 Section 3.4.1
 282         ["domain"] = Cg(V"dot-atom" + V"domain-literal" + V"obs-domain", "domain"), -- RFC 2822 Section 3.4.1
 283         ["domain-literal"] = Cs(V"CFWS"^-1 * P"[" * (V"FWS"^-1 * V"dcontent")^0 * V"FWS"^-1 * P"]" * V"CFWS"^-1), -- RFC 2822 Section 3.4.1
 284         ["dcontent"] = V"dtext" + V"quoted-pair", -- RFC 2822 Section 3.4.1
 285         ["dtext"] = V"NO-WS-CTL" + R"\33\90" + R"\94\126", -- RFC 2822 Section 3.4.1
 286 }
 287 end
 288
 289 local overall_message = function ()
 290         setfenv(1, lpeg)
 291         return {
 292         -- overall message specification
 293         --
 294         -- RRC 2822 Section 3.5:
 295         --
 296         -- A message consists of header fields, optionally followed by a message
 297         -- body.  Lines in a message MUST be a maximum of 998 characters
 298         -- excluding the CRLF, but it is RECOMMENDED that lines be limited to 78
 299         -- characters excluding the CRLF.
 300         --
 301         -- this grammar does not enforce the 78 chars limit yet
 302         ["message"] = (V"fields" + V"obs-fields") * (V"CRLF" * V"body")^-1, -- RFC 2822 Section 3.5
 303         ["body"] = (V"text"^-998 * V"CRLF")^0 * V"text"^-998, -- RFC 2822 Section 3.5
 304 }
 305 end
 306
 307 local fields = function ()
 308         setfenv(1, lpeg)
 309         return {
 310         -- spec for all fields in the header
 311         --
 312         -- RRC 2822 Section 3.6
 313         --
 314         --the "fields" rule is incomplete
 315         ["fields"] = -- trace fields missing
 316                         (V"orig-date" + V"from" + V"sender" + V"reply-to" + V"to" + V"cc" + V"bcc"
 317                         + V"message-id" + V"in-reply-to" + V"to" + V"cc" + V"bcc")^0, -- RFC 2822 Section 3.6
 318         -- origination date field: the time at which the author decided
 319         -- the message was ready to be sent (e.g. hit the button)
 320         ["orig-date"] = P"Date:" * V"date-time" * V"CRLF", -- RFC 2822 Section 3.6.1
 321         -- the author of the message
 322         ["from"] = P"From:" * V"mailbox-list" * V"CRLF", -- RFC 2822 Section 3.6.2
 323         -- the actual agent that sent the message
 324         ["sender"] = P"Sender:" * V"mailbox" * V"CRLF", -- RFC 2822 Section 3.6.2
 325         -- address which should be replied to
 326         -- see Section 3.6.3 for forming replies
 327         ["reply-to"] = P"Reply-To:" * V"address-list" * V"CRLF", -- RFC 2822 Section 3.6.2
 328         -- The "To:" field contains the address(es) of the primary recipient(s)
 329         -- of the message.
 330         ["to"] = P"To:" * V"address-list" * V"CRLF", -- RFC 2822 Section 3.6.3
 331         -- The "Cc:" field contains the addresses of others who are to receive
 332         -- the message, though the content of the message may not be directed
 333         -- at them.
 334         ["cc"] = P"Cc:" * V"address-list" * V"CRLF", -- RFC 2822 Section 3.6.3
 335         -- The "Bcc:" field contains addresses of recipients of the message
 336         -- whose addresses are not to be revealed to other recipients of the
 337         -- message.
 338         ["bcc"] = P"Bcc:" * (V"address-list" + V"CFWS") * V"CRLF", -- RFC 2822 Section 3.6.3
 339         ["message-id"] = P"Message-Id:" * V"msg-id" * V"CRLF", -- RFC 2822 Section 3.6.4
 340         ["in-reply-to"] = P"In-Reply-To:" * V"msg-id"^1 * V"CRLF", -- RFC 2822 Section 3.6.4
 341         ["references"] = P"References:" * V"msg-id"^1 * V"CRLF", -- RFC 2822 Section 3.6.4
 342         ["msg-id"] = V"CFWS"^-1 * P"<" * V"id-left" * P"@" * V"id-right" * P">" * V"CFWS"^-1, -- RFC 2822 Section 3.6.4
 343         ["id-left"] = V"dot-atom-text" + V"no-fold-quote" + V"obs-id-left", -- RFC 2822 Section 3.6.4
 344         ["id-right"] = V"dot-atom-text" + V"no-fold-literal" + V"obs-id-right", -- RFC 2822 Section 3.6.4
 345         ["no-fold-quote"] = P"\"" * (V"qtext" + V"quoted-pair")^0 * P"\"", -- RFC 2822 Section 3.6.4
 346         ["no-fold-literal"] = P"[" * (V"dtext" + V"quoted-pair")^0 * P"]", -- RFC 2822 Section 3.6.4
 347 }
 348 end
 349
 350 local obs_strict = {
 351         ["obs-FWS"] = lpeg.P(false),
 352         ["obs-qp"] = lpeg.P(false),
 353         ["obs-phrase"] = lpeg.P(false),
 354         ["obs-utext"] = lpeg.P(false),
 355         ["obs-day-of-week"] = lpeg.P(false),
 356         ["obs-year"] = lpeg.P(false),
 357         ["obs-month"] = lpeg.P(false),
 358         ["obs-day"] = lpeg.P(false),
 359         ["obs-hour"] = lpeg.P(false),
 360         ["obs-minute"] = lpeg.P(false),
 361         ["obs-second"] = lpeg.P(false),
 362         ["obs-zone"] = lpeg.P(false),
 363         ["obs-angle-addr"] = lpeg.P(false),
 364         ["obs-mailbox-list"] = lpeg.P(false),
 365         ["obs-address-list"] = lpeg.P(false),
 366         ["obs-local-part"] = lpeg.P(false),
 367         ["obs-domain"] = lpeg.P(false),
 368         ["obs-fields"] = lpeg.P(false),
 369         ["obs-id-left"] = lpeg.P(false),
 370         ["obs-id-right"] = lpeg.P(false),
 371 }
 372
 373 local os = setmetatable({}, {__index = function() return lpeg.P(false) end})
 374
 375 local join_set = function (...)
 376         local n = select('#', ...)
 377         local ret = {}
 378         for i = 1, n do
 379                 local t = select(i, ...)
 380                 if type(t)=='table' then
 381                         for k, v in pairs(t) do
 382                                 ret[k] = v
 383                         end
 384                 elseif type(t)=='string' then
 385                         ret[1] = t
 386                 elseif type(t)=='boolean' then
 387                         -- TODO: check no overwrite
 388                 else
 389                         error('join_set: bad argument number '..i..' of type '..type(t))
 390                 end
 391         end
 392         return ret
 393 end
 394
 395 local gr = join_set(core_values, lex_tokens(), date_time(), address(), overall_message(), fields(), obs_strict)
 396
 397 return gr
 398