我试图写一个bash脚本测试，需要一个参数，并通过curl发送到网站。我需要url编码的值，以确保特殊字符被正确处理。最好的方法是什么?
这是我到目前为止的基本脚本:
#!/bin/bash
host=${1:?'bad host'}
value=$2
shift
shift
curl -v -d "param=${value}" http://${host}/somepath $@
当前回答
Ruby，为了完整性
value="$(ruby -r cgi -e 'puts CGI.escape(ARGV[0])' "$2")"
其他回答
下面是一个使用Lua的单行转换，类似于blueyed的答案，除了所有的RFC 3986 Unreserved Characters都没有被编码(就像这个答案):
url=$(echo 'print((arg[1]:gsub("([^%w%-%.%_%~])",function(c)return("%%%02X"):format(c:byte())end)))' | lua - "$1")
此外，您可能需要确保字符串中的换行符从LF转换为CRLF，在这种情况下，您可以插入一个gsub("\r? "“\n”，“\r\n”)在百分比编码之前。
下面是一个变体，在application/x-www-form-urlencoded的非标准风格中，它执行换行规范化，并将空格编码为'+'而不是'%20'(可以使用类似的技术将其添加到Perl代码片段中)。
url=$(echo 'print((arg[1]:gsub("\r?\n", "\r\n"):gsub("([^%w%-%.%_%~ ]))",function(c)return("%%%02X"):format(c:byte())end):gsub(" ","+"))' | lua - "$1")
安装php后，我使用这种方式:
URL_ENCODED_DATA=`php -r "echo urlencode('$DATA');"`
Note
这些函数不是用来编码URL的数据，而是用来编码URL。 将url以每行一个的方式放入文件中。
#!/bin/dash
replaceUnicodes () { # $1=input/output file
if ! mv -f "$1" "$1".tmp 2>/dev/null; then return 1; fi
output="$1" awk '
function hexValue(chr) {
if(chr=="0") return 0; if(chr=="1") return 1; if(chr=="2") return 2; if(chr=="3") return 3; if(chr=="4") return 4; if(chr=="5") return 5;
if(chr=="6") return 6; if(chr=="7") return 7; if(chr=="8") return 8; if(chr=="9") return 9; if(chr=="A") return 10;
if(chr=="B") return 11; if(chr=="C") return 12; if(chr=="D") return 13; if(chr=="E") return 14; return 15 }
function hexToDecimal(str, value,i,inc) {
str=toupper(str); value=and(hexValue(substr(str,length(str),1)),15); inc=1;
for(i=length(str)-1;i>0;i--) {
value+=lshift(hexValue(substr(str,i,1)),4*inc++)
} return value }
function toDecimal(str, value,i) {
for(i=1;i<=length(str);i++) {
value=(value*10)+substr(str,i,1)
} return value }
function to32BE(high,low) {
# return 0x10000+((high-0xD800)*0x400)+(low-0xDC00) }
return lshift((high-0xD800),10)+(low-0xDC00)+0x10000 }
function toUTF8(value) {
if(value<0x80) {
return sprintf("%%%02X",value)
} else if(value>0xFFFF) {
return sprintf("%%%02X%%%02X%%%02X%%%02X",or(0xF0,and(rshift(value,18),0x07)),or(0x80,and(rshift(value,12),0x3F)),or(0x80,and(rshift(value,6),0x3F)),or(0x80,and(rshift(value,0),0x3F)))
} else if(value>0x07FF) {
return sprintf("%%%02X%%%02X%%%02X",or(0xE0,and(rshift(value,12),0x0F)),or(0x80,and(rshift(value,6),0x3F)),or(0x80,and(rshift(value,0),0x3F)))
} else { return sprintf("%%%02X%%%02X",or(0xC0,and(rshift(value,6),0x1F)),or(0x80,and(rshift(value,0),0x3F))) }
}
function trap(str) { sub(/^\\+/,"\\",str); return str }
function esc(str) { gsub(/\\/,"\\\\",str); return str }
BEGIN { output=ENVIRON["output"] }
{
finalStr=""; while(match($0,/[\\]+u[0-9a-fA-F]{4}/)) {
p=substr($0,RSTART,RLENGTH); num=hexToDecimal(substr(p,RLENGTH-3,4));
bfrStr=substr($0,1,RSTART-1); $0=substr($0,RSTART+RLENGTH,length($0)-(RSTART+RLENGTH-1));
if(surrogate) {
surrogate=0;
if(RSTART!=1 || num<0xD800 || (num>0xDBFF && num<0xDC00) || num>0xDFFF) {
finalStr=sprintf("%s%s%s%s",finalStr,trap(highP),bfrStr,toUTF8(num))
} else if(num>0xD7FF && num<0xDC00) {
surrogate=1; high=num; finalStr=sprintf("%s%s",finalStr,trap(highP))
} else { finalStr=sprintf("%s%s",finalStr,toUTF8(to32BE(high,num))) }
} else if(num>0xD7FF && num<0xDC00) {
surrogate=1; highP=p; high=num; finalStr=sprintf("%s%s",finalStr,bfrStr)
} else { finalStr=sprintf("%s%s%s",finalStr,bfrStr,toUTF8(num)) }
} finalStr=sprintf("%s%s",finalStr,$0); $0=finalStr
while(match($0,/[\\]+U[0-9a-fA-F]{8}/)) {
str=substr($0,RSTART,RLENGTH); gsub(esc(str),toUTF8(hexToDecimal(substr(str,RLENGTH-7,8))),$0)
}
while(match($0,/[\\]*&#[xX][0-9a-fA-F]{1,8};/)) {
str=substr($0,RSTART,RLENGTH); idx=index(str,"#");
gsub(esc(str),toUTF8(hexToDecimal(substr(str,idx+2,RLENGTH-idx-2))),$0)
}
while(match($0,/[\\]*&#[0-9]{1,10};/)) {
str=substr($0,RSTART,RLENGTH); idx=index(str,"#");
gsub(esc(str),toUTF8(toDecimal(substr(str,idx+1,RLENGTH-idx-1))),$0)
}
printf("%s\n",$0) > output
}' "$1".tmp
rm -f "$1".tmp
}
replaceHtmlEntities () { # $1=input/output file
if ! mv -f "$1" "$1".tmp 2>/dev/null; then return 1; fi
sed 's/%3[aA]/:/g; s/%2[fF]/\//g; s/"/%22/g; s/</%3C/g; s/>/%3E/g; s/ /%A0/g; s/¢/%A2/g; s/£/%A3/g; s/¥/%A5/g; s/©/%A9/g; s/®/%AE/g; s/&/\&/g; s/\\*\//\//g' "$1".tmp > "$1"
rm -f "$1".tmp
}
# "od -v -A n -t u1 -w99999999"
# "hexdump -v -e \47/1 \42%d \42\47"
# Reminder :: Do not encode (, ), [, and ].
toUTF8Encoded () { # $1=input/output file
if ! mv -f "$1" "$1".tmp 2>/dev/null; then return 1; fi
if [ -s "$1".tmp ]; then
# od -A n -t u1 -w99999999 "$1".tmp | \
hexdump -v -e '/1 "%d "' "$1".tmp | \
output="$1" awk 'function hexDigit(chr) { if((chr>47 && chr<58) || (chr>64 && chr<71) || (chr>96 && chr<103)) return 1; return 0 }
BEGIN { output=ENVIRON["output"] }
{ for(i=1;i<=NF;i++) {
flushed=0; c=$(i);
if(c==13) { if($(i+1)==10) i++; printf("%s\n",url) > output; url=""; flushed=1
} else if(c==10) { printf("%s\n",url) > output; url=""; flushed=1
} else if(c==37) {
if(hexDigit($(i+1)) && hexDigit($(i+2))) {
url=sprintf("%s%%%c%c",url,$(i+1),$(i+2)); i+=2
} else { url=sprintf("%s%%25",url) }
} else if(c>32 && c<127 && c!=34 && c!=39 && c!=96 && c!=60 && c!=62) {
url=sprintf("%s%c",url,c)
} else { url=sprintf("%s%%%02X",url,c) }
} if(!flushed) printf("%s\n",url) > output
}'
fi
rm -f "$1".tmp
}
调用replacecodes () -> replacemlentities () --> toUTF8Encoded()
下面是一个POSIX函数:
url_encode() {
awk 'BEGIN {
for (n = 0; n < 125; n++) {
m[sprintf("%c", n)] = n
}
n = 1
while (1) {
s = substr(ARGV[1], n, 1)
if (s == "") {
break
}
t = s ~ /[[:alnum:]_.!~*\47()-]/ ? t s : t sprintf("%%%02X", m[s])
n++
}
print t
}' "$1"
}
例子:
value=$(url_encode "$2")
你可以在perl中模拟javascript的encodeURIComponent。下面是命令:
perl -pe 's/([^a-zA-Z0-9_.!~*()'\''-])/sprintf("%%%02X", ord($1))/ge'
你可以在.bash_profile中设置它为bash别名:
alias encodeURIComponent='perl -pe '\''s/([^a-zA-Z0-9_.!~*()'\''\'\'''\''-])/sprintf("%%%02X",ord($1))/ge'\'
现在你可以管道到encodeURIComponent:
$ echo -n 'hèllo wôrld!' | encodeURIComponent
h%C3%A8llo%20w%C3%B4rld!
