Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

integrate changes from cmark-gfm 0.29.0.gfm.12 and gfm.13 #61

Merged
merged 26 commits into from
Aug 1, 2023

Conversation

QuietMisdreavus
Copy link

Resolves rdar://112876969

This PR merges in changes from the following cmark-gfm releases:

Most importantly, this resolves GHSA-w4qg-3vf7-m9x5 (a/k/a CVE-2023-37463) by resolving some quadratic complexity in the table parser and HTML printer.

I had to make some changes on top of GitHub's patch to integrate their table-cell calculation with our data storage for row- and column-span. The only changes in the merge commit resolved merge conflicts introduced by the rowspan/colspan feature; the real changes are in the last commit in this branch.

I've tested this with the repo's own tests (make test), in Swift-Markdown, and in Swift-DocC, and all of the tests from these repos still pass.

waldyrious and others added 26 commits January 24, 2023 18:22
The code for the table extension used the term 'marker row',
but the spec calls it 'delimiter row'.

This change normalizes the terminology so that it's consistent.
```
../../../../ext/markly/autolink.c: In function ‘postprocess_text’:
../../../../ext/markly/autolink.c:364:31: warning: passing argument 1 of ‘validate_protocol’ discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
  364 |         if (validate_protocol("mailto:", data + start + offset + max_rewind, rewind, max_rewind)) {
      |                               ^~~~~~~~~
../../../../ext/markly/commonmark.c: In function ‘S_render_node’:
../../../../ext/markly/autolink.c:299:36: note: expected ‘char *’ but argument is of type ‘const char *’
  299 | static bool validate_protocol(char protocol[], uint8_t *data, size_t rewind, size_t max_rewind) {
      |                               ~~~~~^~~~~~~~~~
```
```
../../../../ext/markly/commonmark.c:405:18: warning: assignment discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
  405 |       emph_delim = "_";
      |                  ^
```
warning: assignment discards ‘const’ qualifier from pointer target type
warning: passing argument 1 of ‘validate_protocol’ discards ‘const’ qualifier from pointer target type
In addition, fix a bug where the length of the literal value was
calculated AFTER the actual literal string (null terminated) was
allocated.
Co-authored-by: Phill MV <phillmv@github.com>
Fix GHSL-2023-117, GHSL-2023-118, GHSL-2023-119
…on-0.29.0.12

Changelog and version bump for 0.29.0.12
Fix format specifier for printing a size_t
Expose CMARK_NODE_FOOTNOTE_DEFINITION literal value.
Normalize nomenclature: marker row vs. delimiter row
…on-0.29.0.13

Changelog and version bump for 0.29.0.13
@daniel-grumberg daniel-grumberg merged commit 3bc2f3e into gfm Aug 1, 2023
@daniel-grumberg daniel-grumberg deleted the QuietMisdreavus/gfm.13 branch August 1, 2023 13:07
daniel-grumberg pushed a commit to daniel-grumberg/swift-cmark that referenced this pull request Aug 1, 2023
* Normalize nomenclature: marker row vs. delimiter row

The code for the table extension used the term 'marker row',
but the spec calls it 'delimiter row'.

This change normalizes the terminology so that it's consistent.

* Update autolink.c

```
../../../../ext/markly/autolink.c: In function ‘postprocess_text’:
../../../../ext/markly/autolink.c:364:31: warning: passing argument 1 of ‘validate_protocol’ discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
  364 |         if (validate_protocol("mailto:", data + start + offset + max_rewind, rewind, max_rewind)) {
      |                               ^~~~~~~~~
../../../../ext/markly/commonmark.c: In function ‘S_render_node’:
../../../../ext/markly/autolink.c:299:36: note: expected ‘char *’ but argument is of type ‘const char *’
  299 | static bool validate_protocol(char protocol[], uint8_t *data, size_t rewind, size_t max_rewind) {
      |                               ~~~~~^~~~~~~~~~
```

* Update commonmark.c

```
../../../../ext/markly/commonmark.c:405:18: warning: assignment discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
  405 |       emph_delim = "_";
      |                  ^
```

* Fix GHSL-2023-119: prevent quadratic performance by not allowing very deeply nested footnote definitions.

* Fix GHSL-2023-117: store cell index on node so that it doesn't need to be recomputed during rendering.

* Fix GHSL-2023-118: limit number of autocompleted table cells to prevent DOS.

* Expose CMARK_NODE_FOOTNOTE_DEFINION literal value.

In addition, fix a bug where the length of the literal value was
calculated AFTER the actual literal string (null terminated) was
allocated.

* Update src/node.h

Co-authored-by: Phill MV <phillmv@github.com>

* Rename custom_int -> cell_index.

* Add newline

* Remove unnecessary scope.

* Create codeql.yml

* Changelog and version bump for 0.29.0.12

* Fix format specifier for printing a size_t

* Changelog and version bump for 0.29.0.13

* move cell index into node_cell_data

---------

Co-authored-by: Waldir Pimenta <waldyrious@gmail.com>
Co-authored-by: Samuel Williams <samuel.williams@oriontransfer.co.nz>
Co-authored-by: Kevin Backhouse <kevinbackhouse@github.com>
Co-authored-by: Phill MV <phillmv@github.com>
Co-authored-by: Bas Alberts <13686387+anticomputer@users.noreply.github.com>
Co-authored-by: Bas Alberts <anticomputer@github.com>
daniel-grumberg added a commit that referenced this pull request Aug 1, 2023
* Normalize nomenclature: marker row vs. delimiter row

The code for the table extension used the term 'marker row',
but the spec calls it 'delimiter row'.

This change normalizes the terminology so that it's consistent.

* Update autolink.c

```
../../../../ext/markly/autolink.c: In function ‘postprocess_text’:
../../../../ext/markly/autolink.c:364:31: warning: passing argument 1 of ‘validate_protocol’ discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
  364 |         if (validate_protocol("mailto:", data + start + offset + max_rewind, rewind, max_rewind)) {
      |                               ^~~~~~~~~
../../../../ext/markly/commonmark.c: In function ‘S_render_node’:
../../../../ext/markly/autolink.c:299:36: note: expected ‘char *’ but argument is of type ‘const char *’
  299 | static bool validate_protocol(char protocol[], uint8_t *data, size_t rewind, size_t max_rewind) {
      |                               ~~~~~^~~~~~~~~~
```

* Update commonmark.c

```
../../../../ext/markly/commonmark.c:405:18: warning: assignment discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
  405 |       emph_delim = "_";
      |                  ^
```

* Fix GHSL-2023-119: prevent quadratic performance by not allowing very deeply nested footnote definitions.

* Fix GHSL-2023-117: store cell index on node so that it doesn't need to be recomputed during rendering.

* Fix GHSL-2023-118: limit number of autocompleted table cells to prevent DOS.

* Expose CMARK_NODE_FOOTNOTE_DEFINION literal value.

In addition, fix a bug where the length of the literal value was
calculated AFTER the actual literal string (null terminated) was
allocated.

* Update src/node.h



* Rename custom_int -> cell_index.

* Add newline

* Remove unnecessary scope.

* Create codeql.yml

* Changelog and version bump for 0.29.0.12

* Fix format specifier for printing a size_t

* Changelog and version bump for 0.29.0.13

* move cell index into node_cell_data

---------

Co-authored-by: QuietMisdreavus <QuietMisdreavus@users.noreply.github.com>
Co-authored-by: Waldir Pimenta <waldyrious@gmail.com>
Co-authored-by: Samuel Williams <samuel.williams@oriontransfer.co.nz>
Co-authored-by: Kevin Backhouse <kevinbackhouse@github.com>
Co-authored-by: Phill MV <phillmv@github.com>
Co-authored-by: Bas Alberts <13686387+anticomputer@users.noreply.github.com>
Co-authored-by: Bas Alberts <anticomputer@github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

6 participants