[dreamwidth/dw-free] b023c1: Fix location of commentmanage.js in views/dev/tests

[github]

Branch: refs/heads/master Home: https://github.com/dreamwidth/dw-free Commit: b023c12a04e72a0ab4f81d47d704a9ae2d95b629 https://github.com/dreamwidth/dw-free/commit/b023c12a04e72a0ab4f81d47d704a9ae2d95b629 Author: Kareila kareila@dreamwidth.org Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M views/dev/tests/commentmanage.js

Log Message:

---

Fix location of commentmanage.js in views/dev/tests

Commit: 99195720b441df24c3d5da55a4750cd1c06ff1db https://github.com/dreamwidth/dw-free/commit/99195720b441df24c3d5da55a4750cd1c06ff1db Author: Kareila kareila@dreamwidth.org Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M views/dev/tests.tt

Log Message:

---

Don't include sinon-ie.js in dev/tests.tt

This breaks the test suite by overriding the built-in Date() object.

"If you don't require fake timers to work in IE, don't include this file."

Commit: 3cbe433e8196a30c12603c44908f6279cbcb711f https://github.com/dreamwidth/dw-free/commit/3cbe433e8196a30c12603c44908f6279cbcb711f Author: Kareila kareila@dreamwidth.org Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M views/dev/tests-all.tt

Log Message:

---

Don't include sample.js in dev/tests-all.tt

It's an example file, not a real test.

Commit: 3e855fc52d47ca37e2cd15cd092d20cd28843419 https://github.com/dreamwidth/dw-free/commit/3e855fc52d47ca37e2cd15cd092d20cd28843419 Author: Kareila kareila@dreamwidth.org Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M views/dev/tests/commentmanage.js

Log Message:

---

Fix timing issues in delcomment tests

Through trial and error, I proved that the delcomment tests were failing in a non-deterministic way due to timing issues between the comment visibility being changed and being observed. If I removed all the stop calls or changed their arguments, all the visibility tests failed consistently.

I eventually discovered that if I commented out the stop calls for the elements whose visibility was unchanged, the tests passed consistently.

Compare: https://github.com/dreamwidth/dw-free/compare/9be41d342503...3e855fc52d47

[dreamwidth/dw-free] 357c95: Make site-schemed comment page links nicer (#2801)

[github]

Branch: refs/heads/master Home: https://github.com/dreamwidth/dw-free Commit: 357c95d2cbfe9dec36cfb3068b38250665fddb48 https://github.com/dreamwidth/dw-free/commit/357c95d2cbfe9dec36cfb3068b38250665fddb48 Author: rshatch ruth.s.hatch@gmail.com Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M htdocs/scss/skins/entry-styles.scss M htdocs/scss/skins/celerity.scss M htdocs/scss/skins/gradation/gradation-base.scss M htdocs/scss/skins/lynx.scss M styles/siteviews/layout.s2 M views/components/pagination.tt

Log Message:

---

Make site-schemed comment page links nicer (#2801)

• Make site-schemed comment page links nicer

• Linting

• Update htdocs/scss/skins/_entry-styles.scss

Co-authored-by: Nick Fagerlund nick@hashicorp.com

• Update htdocs/scss/skins/_entry-styles.scss

Co-authored-by: Nick Fagerlund nick@hashicorp.com

• Make arrow styling match TT pagination

• Fix TT pagination component

• Make disabled pagination arrows more clearly noninteractive

Co-authored-by: Ruth momijizukamori@gmail.com Co-authored-by: Nick Fagerlund nick@hashicorp.com

[dreamwidth/dw-free] 429ead: New cleaner test: don't add bogus `` tags whe...

[github]

Branch: refs/heads/master Home: https://github.com/dreamwidth/dw-free Commit: 429ead8638d2f8f9819be8fe54e96700d8a87cf2 https://github.com/dreamwidth/dw-free/commit/429ead8638d2f8f9819be8fe54e96700d8a87cf2 Author: Nick Fagerlund nick.fagerlund@gmail.com Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M t/cleaner-event.t

Log Message:

---

New cleaner test: don't add bogus </br> tags when closing a paragraph

This test currently fails.

Commit: 13e222c36a7fb8f77b3e89bfd4e935e86c580d75 https://github.com/dreamwidth/dw-free/commit/13e222c36a7fb8f77b3e89bfd4e935e86c580d75 Author: Nick Fagerlund nick.fagerlund@gmail.com Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M cgi-bin/LJ/CleanHTML.pm

Log Message:

---

Cleaner: exclude empty tags from per-tag closer

I think a more correct approach is to not add empty tags to the tag stack in the first place! But the more parsimonious approach is to just do the same thing in both the end-of-document closer and the end-of-tag closer.

Commit: aa8f6cc1dc8e4b8fac189573d80d5f8fe892fd12 https://github.com/dreamwidth/dw-free/commit/aa8f6cc1dc8e4b8fac189573d80d5f8fe892fd12 Author: Nick Fagerlund nick.fagerlund@gmail.com Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M cgi-bin/LJ/CleanHTML.pm

Log Message:

---

Cleaner: Add HTML5 source and track tags to list of empty tags

Compare: https://github.com/dreamwidth/dw-free/compare/357c95d2cbfe...aa8f6cc1dc8e

[dreamwidth/dw-free] c8e8ae: [#2841] Properly record anon comment IPs in the te...

[github]

Branch: refs/heads/master Home: https://github.com/dreamwidth/dw-free Commit: c8e8ae70f324c67b7794439f3706e8a7500740c4 https://github.com/dreamwidth/dw-free/commit/c8e8ae70f324c67b7794439f3706e8a7500740c4 Author: Nick Fagerlund nick.fagerlund@gmail.com Date: 2020-09-07 (Mon, 07 Sep 2020)

Changed paths: M cgi-bin/LJ/Talk.pm

Log Message:

---

[#2841] Properly record anon comment IPs in the temp table for spam

This is another instance of #2738 -- edit_comment and enter_comment used to mutate their $comment argument by adding a talkid key with the jtalkid, and I missed a few things referenced that value via the mutated hashref rather than one of their other handles to the jtalkid. In this case, it was preventing anonymous comment spam reports by not properly writing the IP address of new anon comments to the tempanonips table.

This time I hunted real hard for anything in the code that might access any hashref key called talkid. As a result, this commit also:

• Fixes one more instance of That Thing (re-screening when editing).
• Just deletes the final remaining instance (an unused hook).
• Yanks a similar "mutate the $comment argument" behavior in enter_imported_comment -- nothing ever references that hashref argument again before it goes out of scope.

All the remaining hashrefs with talkid members seem to be either s2 comment objects or the parent element of one of prepare_and_validate_comment's comment hashrefs.