#89 Redesign HTML parsing for securing PGP decrypted HTML

This commit is contained in:
the-djmaze 2022-02-01 11:46:12 +01:00
parent d9e81d6123
commit 6b26ddec7a
6 changed files with 227 additions and 195 deletions

View file

@ -1,11 +1,11 @@
import { ComposeType/*, FolderType*/ } from 'Common/EnumsUser';
import { EmailModel } from 'Model/Email';
import { encodeHtml } from 'Common/Html';
import { isArray } from 'Common/Utils';
import { createElement } from 'Common/Globals';
import { FolderUserStore } from 'Stores/User/Folder';
import { SettingsUserStore } from 'Stores/User/Settings';
import * as Local from 'Storage/Client';
import { plainToHtml } from 'Common/Html';
export const
@ -20,163 +20,6 @@ sortFolders = folders => {
}
},
/**
* @param {string} html
* @returns {string}
*/
htmlToPlain = (html) => {
let pos = 0,
limit = 800,
iP1 = 0,
iP2 = 0,
iP3 = 0,
text = '';
const
tpl = createElement('template'),
convertBlockquote = (blockquoteText) => {
blockquoteText = '> ' + blockquoteText.trim().replace(/\n/gm, '\n> ');
return blockquoteText.replace(/(^|\n)([> ]+)/gm, (...args) =>
args && 2 < args.length ? args[1] + args[2].replace(/[\s]/g, '').trim() + ' ' : ''
);
},
convertDivs = (...args) => {
let divText = 1 < args.length ? args[1].trim() : '';
if (divText.length) {
divText = '\n' + divText.replace(/<div[^>]*>([\s\S\r\n]*)<\/div>/gim, convertDivs).trim() + '\n';
}
return divText;
},
convertPre = (...args) =>
1 < args.length
? args[1]
.toString()
.replace(/[\n]/gm, '<br/>')
.replace(/[\r]/gm, '')
: '',
fixAttibuteValue = (...args) => (1 < args.length ? args[1] + encodeHtml(args[2]) : ''),
convertLinks = (...args) => (1 < args.length ? args[1].trim() : '');
tpl.innerHTML = html
.replace(/<p[^>]*><\/p>/gi, '')
.replace(/<pre[^>]*>([\s\S\r\n\t]*)<\/pre>/gim, convertPre)
.replace(/[\s]+/gm, ' ')
.replace(/((?:href|data)\s?=\s?)("[^"]+?"|'[^']+?')/gim, fixAttibuteValue)
.replace(/<br[^>]*>/gim, '\n')
.replace(/<\/h[\d]>/gi, '\n')
.replace(/<\/p>/gi, '\n\n')
.replace(/<ul[^>]*>/gim, '\n')
.replace(/<\/ul>/gi, '\n')
.replace(/<li[^>]*>/gim, ' * ')
.replace(/<\/li>/gi, '\n')
.replace(/<\/td>/gi, '\n')
.replace(/<\/tr>/gi, '\n')
.replace(/<hr[^>]*>/gim, '\n_______________________________\n\n')
.replace(/<div[^>]*>([\s\S\r\n]*)<\/div>/gim, convertDivs)
.replace(/<blockquote[^>]*>/gim, '\n__bq__start__\n')
.replace(/<\/blockquote>/gim, '\n__bq__end__\n')
.replace(/<a [^>]*>([\s\S\r\n]*?)<\/a>/gim, convertLinks)
.replace(/<\/div>/gi, '\n')
.replace(/&nbsp;/gi, ' ')
.replace(/&quot;/gi, '"')
.replace(/<[^>]*>/gm, '');
text = tpl.content.textContent;
if (text) {
text = text
.replace(/\n[ \t]+/gm, '\n')
.replace(/[\n]{3,}/gm, '\n\n')
.replace(/&gt;/gi, '>')
.replace(/&lt;/gi, '<')
.replace(/&amp;/gi, '&')
// wordwrap max line length 100
.match(/.{1,100}(\s|$)|\S+?(\s|$)/g).join('\n');
}
while (0 < --limit) {
iP1 = text.indexOf('__bq__start__', pos);
if (0 > iP1) {
break;
}
iP2 = text.indexOf('__bq__start__', iP1 + 5);
iP3 = text.indexOf('__bq__end__', iP1 + 5);
if ((-1 === iP2 || iP3 < iP2) && iP1 < iP3) {
text = text.slice(0, iP1) + convertBlockquote(text.slice(iP1 + 13, iP3)) + text.slice(iP3 + 11);
pos = 0;
} else if (-1 < iP2 && iP2 < iP3) {
pos = iP2 - 1;
} else {
pos = 0;
}
}
return text.replace(/__bq__start__|__bq__end__/gm, '').trim();
},
/**
* @param {string} plain
* @param {boolean} findEmailAndLinksInText = false
* @returns {string}
*/
plainToHtml = (plain) => {
plain = plain.toString().replace(/\r/g, '');
plain = plain.replace(/^>[> ]>+/gm, ([match]) => (match ? match.replace(/[ ]+/g, '') : match));
let bIn = false,
bDo = true,
bStart = true,
aNextText = [],
aText = plain.split('\n');
do {
bDo = false;
aNextText = [];
aText.forEach(sLine => {
bStart = '>' === sLine.slice(0, 1);
if (bStart && !bIn) {
bDo = true;
bIn = true;
aNextText.push('~~~blockquote~~~');
aNextText.push(sLine.slice(1));
} else if (!bStart && bIn) {
if (sLine) {
bIn = false;
aNextText.push('~~~/blockquote~~~');
aNextText.push(sLine);
} else {
aNextText.push(sLine);
}
} else if (bStart && bIn) {
aNextText.push(sLine.slice(1));
} else {
aNextText.push(sLine);
}
});
if (bIn) {
bIn = false;
aNextText.push('~~~/blockquote~~~');
}
aText = aNextText;
} while (bDo);
return aText.join('\n')
// .replace(/~~~\/blockquote~~~\n~~~blockquote~~~/g, '\n')
.replace(/&/g, '&amp;')
.replace(/>/g, '&gt;')
.replace(/</g, '&lt;')
.replace(/~~~blockquote~~~[\s]*/g, '<blockquote>')
.replace(/[\s]*~~~\/blockquote~~~/g, '</blockquote>')
.replace(/\n/g, '<br/>');
},
/**
* @param {Array=} aDisabled
* @param {Array=} aHeaderLines
@ -484,8 +327,3 @@ setLayoutResizer = (source, target, sClientSideKeyName, mode) =>
source.observer && source.observer.disconnect();
}
};
rl.Utils = {
htmlToPlain: htmlToPlain,
plainToHtml: plainToHtml
};