Last active
August 30, 2019 10:05
-
-
Save Mellen/ce5d247587dcb7b49e5145a6f49328be to your computer and use it in GitHub Desktop.
Replacing number based html entities with characters, either hex or decimal.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
const NORMAL = 0; | |
const FOUNDAMP = 1; | |
const FOUNDHASH = 2; | |
const FOUNDX = 3; | |
const FOUNDHEX = 4; | |
const GOTENTITY = 5; | |
const FOUNDDEC = 6; | |
function replaceNumericEntities(input) | |
{ | |
const hexdigits = '0123456789abcdefABCDEF'; | |
const digits = '0123456789'; | |
let output = ''; | |
let entity = ''; | |
let isHex = false; | |
let state = NORMAL; | |
for(let c of input+' ') | |
{ | |
switch(state) | |
{ | |
case NORMAL: | |
if(c === '&') | |
{ | |
state = FOUNDAMP; | |
entity += c; | |
} | |
else | |
{ | |
output += c; | |
} | |
break; | |
case FOUNDAMP: | |
if(c === '#') | |
{ | |
state = FOUNDHASH; | |
entity += c | |
} | |
else | |
{ | |
({output, entity, state} = reset(c, entity, output)); | |
} | |
break; | |
case FOUNDHASH: | |
{ | |
if(c === 'x') | |
{ | |
state = FOUNDX; | |
isHex = true; | |
entity += c; | |
} | |
else if(digits.indexOf(c) === -1) | |
{ | |
({output, entity, state} = reset(c, entity, output)); | |
} | |
else | |
{ | |
state = FOUNDDEC; | |
isHex = false; | |
entity += c; | |
} | |
} | |
break; | |
case FOUNDX: | |
{ | |
if(hexdigits.indexOf(c) === -1) | |
{ | |
({output, entity, state} = reset(c, entity, output)); | |
} | |
else | |
{ | |
state = FOUNDHEX; | |
entity += c; | |
} | |
} | |
break; | |
case FOUNDHEX: | |
{ | |
if(c === ';') | |
{ | |
state = GOTENTITY; | |
} | |
else if(hexdigits.indexOf(c) === -1) | |
{ | |
({output, entity, state} = reset(c, entity, output)); | |
} | |
else | |
{ | |
state = FOUNDHEX; | |
entity += c; | |
} | |
} | |
break; | |
case GOTENTITY: | |
{ | |
if((isHex && withinHexRange(entity)) || (!isHex && withinDecRange(entity))) | |
{ | |
output += getCharacter(entity, isHex); | |
} | |
else | |
{ | |
output += entity; | |
} | |
({output, entity, state} = reset(c, '', output)); | |
} | |
break; | |
case FOUNDDEC: | |
{ | |
if (c === ';') | |
{ | |
state = GOTENTITY; | |
} | |
else if(digits.indexOf(c) === -1) | |
{ | |
({output, entity, state} = reset(c, entity, output)); | |
} | |
else | |
{ | |
state = FOUNDDEC; | |
entity += c; | |
} | |
} | |
break; | |
} | |
} | |
return output.slice(0, -1); | |
} | |
function reset(c, entity, output) | |
{ | |
let state = 0; | |
output += entity; | |
entity = ''; | |
if(c === '&') | |
{ | |
state = FOUNDAMP; | |
entity = c | |
} | |
else | |
{ | |
state = NORMAL; | |
output += c; | |
} | |
return {output: output, entity: entity, state: state}; | |
} | |
function withinHexRange(entity) | |
{ | |
let value = Number.parseInt(entity.slice(3), 16); | |
return value <= 0x10ffff; | |
} | |
function withinDecRange(entity) | |
{ | |
let value = Number.parseInt(entity.slice(2), 10); | |
return value <= 0x10ffff; | |
} | |
function getCharacter(entity, isHex) | |
{ | |
let value = 0; | |
if(isHex) | |
{ | |
value = Number.parseInt(entity.slice(3), 16); | |
} | |
else | |
{ | |
value = Number.parseInt(entity.slice(2), 10); | |
} | |
let cha = String.fromCodePoint(value); | |
return cha; | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment