2cyr-try-charset-find

#!/bin/bash

: ${1?"Not enough arguments: `basename $0` "some text to guess charset encoding" [optional (30 by default): amount of recoding percolations]"}

# How much results show
TOP_N="${2:-30}"


#TEXT=тест
#TEXT=$( echo -n тест | iconv -t koi8r )
TEXT="$1"

echo '###############################################'
echo "##### Text for recoding: [$TEXT] #####"
echo "##### Show TOP $TOP_N results of recoding #####"
echo '###############################################'

# Simple:
#http --form http://2cyr.com/decode/ text=тест
# But it does not work with non-unicode!
#http --form https://2cyr.com/decode/ text="$( echo -n тест | iconv -t koi8r )" \

#curl -X POST -d text=тест -H 'Content-Type: application/x-www-form-urlencoded; charset=utf-8' http://2cyr.com/decode/

# Sed by https://stackoverflow.com/questions/17988756/how-to-select-lines-between-two-marker-patterns-which-may-occur-multiple-times-w/17990436#17990436
curl -sS -X POST -d "text=$TEXT" -H 'User-Agent: c_u_r_l' `# User-agent required! Curl blocked!` \
		http://2cyr.com/decode/ \
			| sed '/Select the successful sample/,/:UTF-7:SJIS/{//!b};d' \
				| sed -r 's@<option value="(.+)">|</option>@\1#@g' \
					| while read l; `# drop what incorrect initially` \
						do (echo "$l" | iconv -t utf8 &>/dev/null) && echo "[$l]";
					done \

#						| head -$TOP_N