-- luatexja/ltj-inputbuf.lua
--
+luatexja.load_module('base'); local ltjb = luatexja.base
luatexja.load_module('charrange'); local ltjc = luatexja.charrange
require("unicode")
local node_free = node.free
local id_glyph = node.id('glyph')
local getcatcode = tex.getcatcode
-local ltjc_is_ucs_in_japanese_char = ltjc.is_ucs_in_japanese_char
+local ltjc_is_japanese_char_curlist = ltjc.is_japanese_char_curlist
local FFFFF = string.char(0xF3,0xBF,0xBF,0xBF)
--- the following function is modified from jafontspec.lua (by K. Maeda).
--- Instead of "%", we use U+FFFFF for suppressing spaces.
+--DEBUG require"socket"
+local time_line = 0
+local start_time_measure, stop_time_measure
+ = ltjb.start_time_measure, ltjb.stop_time_measure
local function add_comment(buffer)
+ start_time_measure('inputbuf')
local i = utflen(buffer)
- while (i>0) and (getcatcode(utfbyte(buffer, i))==1
+ while (i>0) and (getcatcode(utfbyte(buffer, i))==1
or getcatcode(utfbyte(buffer, i))==2) do
i=i-1
end
-- Is the catcode of endline character is 5 (end-of-line)?
-- Is the catcode of U+FFFFF (new comment char) is 14 (comment)?
if ((ct==11) or (ct==12)) and ctl then
- local p = node_new(id_glyph)
- p.char = c
- if ltjc_is_ucs_in_japanese_char(p) then
+ if ltjc_is_japanese_char_curlist(c) then
buffer = buffer .. FFFFF -- U+FFFFF
end
- node_free(p)
end
end
end
+ stop_time_measure('inputbuf')
return buffer
end
-luatexbase.add_to_callback('process_input_buffer',
+luatexbase.add_to_callback('process_input_buffer',
add_comment,'ltj.process_input_buffer')
--EOF