mime.lua

   1 #!/usr/bin/lua
   2
   3 require 'lpeg'
   4
   5 -- MIME implementation in Lua
   6
   7 -- XXX: at section 3.2.3
   8
   9
  10 local string = string
  11 local type = type
  12 local CRLF = '\r\n'
  13
  14 -- rfc2822 Section 2.1:
  15 --
  16 -- A message consists of header fields (collectively called "the header
  17 -- of the message") followed, optionally, by a body.  The header is a
  18 -- sequence of lines of characters with special syntax as defined in
  19 -- this standard. The body is simply a sequence of characters that
  20 -- follows the header and is separated from the header by an empty line
  21 -- (i.e., a line with nothing preceding the CRLF).
  22 --
  23 local split_content = function (c)
  24         -- TODO: add checks on types
  25         -- NOTE: there is no guarantee that the CRLF separating header
  26         -- and body is the first one. However no headers admit a blank
  27         -- line so far
  28         local h, b = string.match(c, '^(.-)'..CRLF..CRLF..'(.*)$')
  29         assert(type(h)=='string' and type(b)=='string')
  30         return h, b
  31 end
  32
  33 -- rfc2822 Section 2.2.3
  34 --
  35 -- The process of moving from this folded multiple-line representation
  36 -- of a header field to its single line representation is called
  37 -- "unfolding". Unfolding is accomplished by simply removing any CRLF
  38 -- that is immediately followed by WSP.  Each header field should be
  39 -- treated in its unfolded form for further syntactic and semantic
  40 -- evaluation.
  41 --
  42 local unfold_header = function (h)
  43         return string.gsub(h, CRLF..' ', ' ')
  44 end
  45
  46 -- rfc2822 Section 2.1
  47 --
  48 -- At the most basic level, a message is a series of characters.  A
  49 -- message that is conformant with this standard is comprised of
  50 -- characters with values in the range 1 through 127 and interpreted as
  51 -- US-ASCII characters [ASCII].  For brevity, this document sometimes
  52 -- refers to this range of characters as simply "US-ASCII characters".
  53 --
  54
  55 -- rfc2822 Section 2.1.1
  56 --
  57 -- There are two limits that this standard places on the number of
  58 -- characters in a line. Each line of characters MUST be no more than
  59 -- 998 characters, and SHOULD be no more than 78 characters, excluding
  60 -- the CRLF.
  61 --
  62 local check_content = function (c)
  63         local err, warn
  64         -- TODO: implement in a portable way
  65         return err, warn
  66 end
  67
  68 -- rfc2822 Section 2.3
  69 --
  70 -- The body of a message is simply lines of US-ASCII characters.  The
  71 -- only two limitations on the body are as follows:
  72 --
  73 -- - CR and LF MUST only occur together as CRLF; they MUST NOT appear
  74 -- independently in the body.
  75 --
  76 -- - Lines of characters in the body MUST be limited to 998 characters,
  77 --
  78 -- and SHOULD be limited to 78 characters, excluding the CRLF.
  79 --
  80 local check_body = function (b)
  81         local err, warn
  82         -- TODO: implement
  83         return err, warn
  84 end
  85
  86 -- some core values (not implemented literally)
  87 -- see RFC 2234 Section 6.1
  88 local core_values = {
  89         ["ALPHA"] = lpeg.R("AZ", "az"),
  90         ["BIT"] = lpeg.P("0") + lpeg.P("1"),
  91         ["CHAR"] = lpeg.R("\01\127"),
  92         ["CR"] = lpeg.P("\13"),
  93         ["DIGI"] = lpeg.R("09"),
  94         ["CRLF"] = lpeg.P("\13\10"),
  95         ["LF"] = lpeg.P("\10"),
  96 }
  97
  98 -- lexical tokens used in the specification
  99 -- TODO: could be optimized
 100 -- check for non-obfuscated optimizations
 101 -- TODO: write in a less obfuscated way
 102 -- check if parenthesis can be removed
 103 local lex_tokens = {
 104         -- control characters without whitespaces
 105         ["NO-WS-CTL"] = lpeg.R("\1\8") +
 106                         lpeg.P("\11") +
 107                         lpeg.P("\12") +
 108                         lpeg.R("\14\13") +
 109                         lpeg.P("\127"), -- RFC 2822 Section 3.2.1
 110         -- a character in a text
 111         ["text"] =      lpeg.R("\1\9") +
 112                         lpeg.P("\11") +
 113                         lpeg.P("\12") +
 114                         lpeg.R("\14\127"), -- RFC 2822 Section 3.2.1
 115         -- a special character
 116         ["specials"] =  lpeg.P("(") + lpeg.P(")") +
 117                         lpeg.P("<") + lpeg.P(">") +
 118                         lpeg.P("[") + lpeg.P("]") +
 119                         lpeg.P(":") + lpeg.P(";") +
 120                         lpeg.P("@") + lpeg.P("\\") +
 121                         lpeg.P(",") + lpeg.P(".") +
 122                         lpeg.P("\""), -- RFC 2822 Section 3.2.1
 123         -- a quoted pair should return only the second character
 124         ["quoted-pair"] = (lpeg.P("\\") * lpeg.C(lpeg.V("text"))) + lpeg.V("obs-qp"), -- RFC 2822 Section 3.2.2
 125         -- a folding white space (a whitespace that can include a CRLF)
 126         -- should be substituted by a single whitespace
 127         ["FWS"] = (((lpeg.V("WSP")^0 * lpeg.V("CRLF"))^-1 * lpeg.V("WSP")^1) + lpeg.V("obs-FWS")) / " ", -- RFC 2822 Section 3.2.3
 128         -- a text character allowed inside a comment
 129         ["ctext"] =     lpeg.V("NO-WS-CTL") +
 130                         lpeg.R("\33\39") +
 131                         lpeg.R("\42\91") +
 132                         lpeg.R("\93\126"), -- RFC 2822 Section 3.2.3
 133         -- the content of a comment (comments can nest)
 134         ["ccontent"] =  lpeg.V("ctext") + lpeg.V("quoted-pair") + lpeg.V("comment"), -- RFC 2822 Section 3.2.3
 135         -- an actual comment
 136         -- should be substituted by a single whitespace
 137         ["comment"] =   (lpeg.P("(") * (lpeg.V("FWS")^-1 * lpeg.V("ccontent"))^0 * lpeg.V("FWS")^-1 * lpeg.P(")")) / " ", -- RFC 2822 Section 3.2.3
 138         -- a comment or a folding white space
 139         -- should be substituted by a single whitespace
 140         --
 141         -- Folding white spaces should not be placed in a way that
 142         -- creates lines containing only whitespaces.
 143         -- This requirement is not enforced by this grammar
 144         ["CFWS"] =      ((lpeg.V("FWS")^-1 * lpeg.V("comment"))^0 * ((lpeg.V("FWS")^-1 * lpeg.V("comment")) + lpeg.P("FWS"))) / " ", -- RFC 2822 Section 3.2.3
 145         -- character that can appear in an atom
 146         ["atext"] =     lpeg.V("ALPHA") + lpeg.V("DIGIT") +
 147                         lpeg.P("!") + lpeg.P("#") +
 148                         lpeg.P("$") + lpeg.P("%") +
 149                         lpeg.P("&") + lpeg.P("'") +
 150                         lpeg.P("*") + lpeg.P("+") +
 151                         lpeg.P("-") + lpeg.P("/") +
 152                         lpeg.P("=") + lpeg.P("?") +
 153                         lpeg.P("^") + lpeg.P("_") +
 154                         lpeg.P("`") + lpeg.P("{") +
 155                         lpeg.P("|") + lpeg.P("}") +
 156                         lpeg.P("~"), -- RFC 2822 Section 3.2.4
 157         -- an atom is equal to the content only discarding comments and whitespace
 158         ["atom"] =      lpeg.V("CFWS")^-1 * lpeg.C(lpeg.V("atext")^1) * lpeg.V("CFWS")^-1, -- RFC 2822 Section 3.2.4
 159         -- an atom with dots is only the content
 160         ["dot-atom"] =  lpeg.V("CFWS")^-1 * lpeg.C(lpeg.V("dot-atom-text")) * lpeg.V("CFWS")^-1, -- RFC 2822 Section 3.2.4
 161         -- the content of an atom text with dots
 162         ["dot-atom-text"] =     lpeg.V("atext")^1 * (lpeg.P(".") * lpeg.V("atext")^1)^0, -- RFC 2822 Section 3.2.4
 163 }
 164
 165