Use more meaningful magic bytes again

The new sequence C1 AC C1 B4 is valid under Latin1 to get reasonable
encoding autodetection but is a nonstandard UTF-8 encoding for "lt"
This commit is contained in:
Marcel Krüger 2020-07-12 02:15:08 +02:00
parent eecfc95aed
commit 5262457cd8

View File

@ -166,7 +166,7 @@ local pdfmeta = {
pdfmeta.__index = pdfmeta
local function open(filename)
local file = io.open(filename, 'wb')
file:write"%PDF-X.X\n%\xe4\xf6\xfc\xdf\n"
file:write"%PDF-X.X\n%\xC1\xAC\xC1\xB4\n"
return setmetatable({file = file, version = '1.7', [0] = 0, pages = {}, objstream = {}}, pdfmeta)
end
return {