cal: handle folding

1 files changed, 46 insertions, 6 deletions

diff --git a/cal.lua b/cal.lua
index 7fcbbed..6dd2c6b 100755
--- a/cal.lua
+++ b/cal.lua
@@ -1,4 +1,9 @@
 #!/usr/bin/env lua
+--[[--
+
+    RFC2445 parser.
+
+--]]--
 
 local io            = require "io"
 local ioopen        = io.open
@@ -11,16 +16,37 @@ local println       = function (...) print (stringformat (...)) end
 local parse_calendar do
   local lpeg            = require "lpeg"
   local lpegmatch       = lpeg.match
-  local P               = lpeg.P
   local C               = lpeg.C
   local Cp              = lpeg.Cp
+  local Cs              = lpeg.Cs
+  local P               = lpeg.P
+  local S               = lpeg.S
 
+  local p_space         = P" "
   local p_cr            = P"\r"
   local p_lf            = P"\n"
+
+  local p_white_fold    = S" \t"
+  local p_white         = S" \n\r\t\v"
   local p_eol           = p_cr * p_lf + p_lf
   local p_noeol         = P(1) - p_eol
 
-  local p_content_line  = C (p_noeol^1) * p_eol * Cp ()
+  --[[--
+
+    RFC2445: Long content lines SHOULD be split into a multiple line
+    representations using a line "folding" technique. That is, a long line can
+    be split between any two characters by inserting a CRLF immediately
+    followed by a single linear white space character (i.e., SPACE, US-ASCII
+    decimal 32 or HTAB, US-ASCII decimal 9).
+
+  --]]--
+
+  local p_folded_line_1 = p_noeol^1 * (p_eol / "")
+  local p_folded_line_c = p_white_fold/"" * p_folded_line_1
+
+  local p_content_line  = Cs(p_folded_line_1 * p_folded_line_c^0) * Cp()
+
+  local p_skip_line     = p_noeol^0 * p_eol * Cp()
 
   local parse_content_line = function (raw, pos0)
     local res, pos1 = lpegmatch (p_content_line, raw, pos0)
@@ -28,12 +54,26 @@ local parse_calendar do
     return res, pos1
   end
 
-  parse_calendar = function (raw, pos0, consumed)
-    if pos0 == nil then return parse_calendar (raw, 1, 0) end
+  local skip_line = function (raw, pos0)
+    return lpegmatch (p_skip_line, raw, pos0)
+  end
+
+  local errline = function (pos)
+  end
+
+  parse_calendar = function (raw, pos0, consumed, nline, nskipped)
+    if pos0 == nil then return parse_calendar (raw, 1, 0, 1) end
 
     local cline, pos1 = parse_content_line (raw, pos0)
-    println ("»»» [%d–%d] [%s]", pos0, pos1, cline)
-    return parse_calendar (raw, pos1, consumed + pos1 - pos0)
+    if cline == nil then
+      pos1 = skip_line (raw, pos0)
+      println ("[%d–%d] %d bad content line; skipping", pos0, pos1, nline)
+      nskipped = nskipped + 1
+    else
+      println ("[%d–%d] %d [%s]", pos0, pos1, nline, cline)
+    end
+    return parse_calendar (raw, pos1, consumed + pos1 - pos0, nline + 1,
+                           nskipped)
   end
 end