self-hosted/joplin
1
0
Fork 0
mirror of https://github.com/laurent22/joplin.git synced 2025-10-31 00:07:48 +02:00
Code Issues Releases Activity

Desktop: Security: Fix HTML parsing bug (#10876)

Browse Source
This commit is contained in:
Henry Heino
2024-08-15 11:37:58 -07:00
committed by GitHub
parent a6cf0a3a81
commit b220413a9b
5 changed files with 39 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
packages/app-cli/tests/md_to_html/sanitize_20.html Normal file
View File

@@ -0,0 +1,2 @@
<div class="jop-noMdConv">
&lt;.a

2
packages/app-cli/tests/md_to_html/sanitize_20.md Normal file
View File

@@ -0,0 +1,2 @@
<div>
<.a<iframe src="http://example.com/" >

2
packages/fork-htmlparser2/README.md
View File

@@ -54,6 +54,8 @@ index 44b4371..bcd7cc2 100644
if (this._cbs.onend)
```
To fix an HTML parsing issue (tags were allowed to start with non-alphanumeric characters), [this upstream commit](https://github.com/fb55/htmlparser2/commit/bc010de9df09f2d730a69734e05e5175ea8bd2d7) has also been applied.
* * *
# htmlparser2

21
packages/fork-htmlparser2/src/Tokenizer.ts
View File

@@ -32,6 +32,7 @@ const enum State {
//comments
BeforeComment,
InComment,
InSpecialComment,
AfterComment1,
AfterComment2,
@@ -87,6 +88,10 @@ function whitespace(c: string): boolean {
return c === " " || c === "\n" || c === "\t" || c === "\f" || c === "\r";
}
function isASCIIAlpha(c: string): boolean {
return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z');
}
interface Callbacks {
onattribdata(value: string): void; //TODO implement the new event
onattribend(): void;
@@ -282,6 +287,8 @@ export default class Tokenizer {
} else if (c === "?") {
this._state = State.InProcessingInstruction;
this._sectionStart = this._index + 1;
} else if (!isASCIIAlpha(c)) {
this._state = State.Text;
} else {
this._state =
!this._xmlMode && (c === "s" || c === "S")
@@ -309,6 +316,9 @@ export default class Tokenizer {
this._state = State.Text;
this._index--;
}
} else if (!isASCIIAlpha(c)) {
this._state = State.InSpecialComment;
this._sectionStart = this._index;
} else {
this._state = State.InClosingTagName;
this._sectionStart = this._index;
@@ -454,6 +464,15 @@ export default class Tokenizer {
_stateInComment(c: string) {
if (c === "-") this._state = State.AfterComment1;
}
_stateInSpecialComment(c: string) {
if (c === ">") {
this._cbs.oncomment(
this._buffer.substring(this._sectionStart, this._index)
);
this._state = State.Text;
this._sectionStart = this._index + 1;
}
}
_stateAfterComment1(c: string) {
if (c === "-") {
this._state = State.AfterComment2;
@@ -702,6 +721,8 @@ export default class Tokenizer {
this._stateInAttributeName(c);
} else if (this._state === State.InComment) {
this._stateInComment(c);
} else if (this._state === State.InSpecialComment) {
this._stateInSpecialComment(c);
} else if (this._state === State.BeforeAttributeName) {
this._stateBeforeAttributeName(c);
} else if (this._state === State.InTagName) {

12
packages/fork-htmlparser2/src/__fixtures__/Events/34-not-alpha-tags.json Normal file
View File

@@ -0,0 +1,12 @@
{
"name": "tag names are not ASCII alpha",
"options": {
"parser": {}
},
"html": "<12>text</12>",
"expected": [
{ "event": "text", "data": ["<12>text"] },
{ "event": "comment", "data": ["12"] },
{ "event": "commentend", "data": [] }
]
}