repo stringlengths 5 53 | pr_number int32 1 321k | task_type stringclasses 2
values | issue_text stringlengths 0 81.2k | pr_title stringlengths 1 319 | pr_body stringlengths 0 105k | base_sha stringlengths 40 40 | head_sha stringlengths 40 40 | gold_diff stringlengths 0 202M | changed_files listlengths 0 100 | review_threads listlengths 0 100 | test_patch stringlengths 0 23.4M | merged bool 1
class |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
SerenityOS/serenity | 26,791 | comment_to_fix | Kernel: First batch of wait fixes and tests | Is this even reachable if setup_blocker() always returns false if there is no applicable waitee?
Even if it is, returning `false` seems wrong. Shouldn't the waiter be unblocked if there is no child so we can return the error? | aed345cae965706d0a44f94c19583ba917a25779 | 6e71a499351e304cd980b864ca31165f5a9b4503 | diff --git a/Kernel/Tasks/ThreadBlockers.cpp b/Kernel/Tasks/ThreadBlockers.cpp
index 080649cdfe1d10..dd363d17a6cf02 100644
--- a/Kernel/Tasks/ThreadBlockers.cpp
+++ b/Kernel/Tasks/ThreadBlockers.cpp
@@ -528,8 +528,8 @@ void Thread::WaitBlockerSet::try_unblock(Thread::WaitBlocker& blocker)
if (blocker.unblock(i... | [
"Kernel/Tasks/ThreadBlockers.cpp"
] | [
{
"comment": "Is this even reachable if setup_blocker() always returns false if there is no applicable waitee?\n\nEven if it is, returning `false` seems wrong. Shouldn't the waiter be unblocked if there is no child so we can return the error?",
"path": "Kernel/Tasks/ThreadBlockers.cpp",
"hunk": "@@ -730... | true | ||
SerenityOS/serenity | 26,791 | comment_to_fix | Kernel: First batch of wait fixes and tests | Do these lambdas need to be mutable? | aed345cae965706d0a44f94c19583ba917a25779 | 6e71a499351e304cd980b864ca31165f5a9b4503 | diff --git a/Tests/Kernel/Wait/TestWaitNoChild.cpp b/Tests/Kernel/Wait/TestWaitNoChild.cpp
new file mode 100644
index 00000000000000..55a53e4f847705
--- /dev/null
+++ b/Tests/Kernel/Wait/TestWaitNoChild.cpp
@@ -0,0 +1,45 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SPDX-License-Id... | [
"Tests/Kernel/Wait/TestWaitNoChild.cpp"
] | [
{
"comment": "Do these lambdas need to be mutable?",
"path": "Tests/Kernel/Wait/TestWaitNoChild.cpp",
"hunk": "@@ -0,0 +1,45 @@\n+/*\n+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>\n+ *\n+ * SPDX-License-Identifier: BSD-2-Clause\n+ */\n+\n+#include <LibTest/TestCase.h>\n+#include <... | true | ||
SerenityOS/serenity | 26,791 | comment_to_fix | Kernel: First batch of wait fixes and tests | (as discussed on discord)
I assume you need this because you don't want to use wait(WNOWAIT) here, because that's also what you're testing here, right?
Rather than a timeout, I would much prefer if you could maybe check if the process died (or rather is a zombie) in ProcFS or /sys/kernel/processes somehow. (/sys/kerne... | aed345cae965706d0a44f94c19583ba917a25779 | 6e71a499351e304cd980b864ca31165f5a9b4503 | diff --git a/Tests/Kernel/Wait/TestWaitDoesReap.cpp b/Tests/Kernel/Wait/TestWaitDoesReap.cpp
new file mode 100644
index 00000000000000..d026bcf50d39ec
--- /dev/null
+++ b/Tests/Kernel/Wait/TestWaitDoesReap.cpp
@@ -0,0 +1,58 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SPDX-License... | [
"Tests/Kernel/Wait/TestWaitDoesReap.cpp"
] | [
{
"comment": "(as discussed on discord)\nI assume you need this because you don't want to use wait(WNOWAIT) here, because that's also what you're testing here, right?\n\nRather than a timeout, I would much prefer if you could maybe check if the process died (or rather is a zombie) in ProcFS or /sys/kernel/proce... | true | ||
SerenityOS/serenity | 26,798 | issue_to_patch | LibGfx: Select largest ICO image before bpp tie-break | When choosing the best image from an ICO file, prefer the entry with the largest pixel area first, and only use bits-per-pixel as a tie-breaker. This had regressed in commit b10fe7c1365c917ec11e998efc4354f196331fb5.
This fixes ICO files like Discord's favicon, where 16x16, 32x32, 48x48, and 256x256 entries all adver... | bdf3d17e7521a7f09e48931837c1ea3e48afe73b | e29e4041ce971207dd1c926f465784079fabee5d | diff --git a/Userland/Libraries/LibGfx/ImageFormats/ICOLoader.cpp b/Userland/Libraries/LibGfx/ImageFormats/ICOLoader.cpp
index 29da8cafd9f86c..ad6fb738fb2452 100644
--- a/Userland/Libraries/LibGfx/ImageFormats/ICOLoader.cpp
+++ b/Userland/Libraries/LibGfx/ImageFormats/ICOLoader.cpp
@@ -6,6 +6,7 @@
#include <AK/ByteB... | [
"Tests/LibGfx/TestImageDecoder.cpp",
"Tests/LibGfx/test-inputs/ico/multiple-sizes-with-same-bpp.ico",
"Userland/Libraries/LibGfx/ImageFormats/ICOLoader.cpp"
] | [] | diff --git a/Tests/LibGfx/TestImageDecoder.cpp b/Tests/LibGfx/TestImageDecoder.cpp
index ed6768543a1d33..2b86a3eebde2cd 100644
--- a/Tests/LibGfx/TestImageDecoder.cpp
+++ b/Tests/LibGfx/TestImageDecoder.cpp
@@ -107,6 +107,15 @@ TEST_CASE(test_ico_malformed_frame)
}
}
+TEST_CASE(test_ico_selects_largest_image_wi... | true | |
SerenityOS/serenity | 26,797 | issue_to_patch | LibTest: Flush stdout before running tests | This avoids data to appears multiple time when a test forks. It also makes the "Running {} cases out of {}." line to always appear before all "Running test '{}'.".
Before:
(This specific test is not on master, but all you need to know is that if `fork`s once.)
<img width="596" height="408" alt="Screenshot From 2... | 1ad8304b1d0c1bafaed8a0e8e4182f2ee3249bab | 639b08f4a9aa91c3810f46df0929166fbfe58ccf | diff --git a/Userland/Libraries/LibTest/TestSuite.cpp b/Userland/Libraries/LibTest/TestSuite.cpp
index 6a41746ce3d7ec..c4ae2d8601c936 100644
--- a/Userland/Libraries/LibTest/TestSuite.cpp
+++ b/Userland/Libraries/LibTest/TestSuite.cpp
@@ -151,6 +151,10 @@ int TestSuite::main(ByteString const& suite_name, Span<StringVie... | [
"Userland/Libraries/LibTest/TestSuite.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,764 | issue_to_patch | Nix: Add fuse2fs as a dependency | This allows building the image on NixOS without sudo | 1ad8304b1d0c1bafaed8a0e8e4182f2ee3249bab | 833b2dc08760d77e2cb215e8b355b9ca03dc3293 | diff --git a/Toolchain/default.nix b/Toolchain/default.nix
index 8d776608ae391a..bdaef19c750061 100644
--- a/Toolchain/default.nix
+++ b/Toolchain/default.nix
@@ -46,8 +46,7 @@ mkShell.override { stdenv = gccStdenv; } {
pre-commit
]
++ lib.optionals stdenv.isLinux [
- fuse
- fuse-ext2
+ fuse2fs
... | [
"Toolchain/default.nix"
] | [
{
"comment": "Hm, this was previously added in 743f494a9c6115e7710cc59ab9a2105e2b15696d but then later replaced with `fuse-ext2` in fb897913bbb7e7ef25cbc2a01b4f7370737e6310 without a reason.\n\nWe only use `fuse-ext2` on Darwin and FreeBSD. Not sure why we can't use either `fuse-ext2` or `fuse2fs` on all system... | true | ||
SerenityOS/serenity | 26,764 | comment_to_fix | Nix: Add fuse2fs as a dependency | Hm, this was previously added in 743f494a9c6115e7710cc59ab9a2105e2b15696d but then later replaced with `fuse-ext2` in fb897913bbb7e7ef25cbc2a01b4f7370737e6310 without a reason.
We only use `fuse-ext2` on Darwin and FreeBSD. Not sure why we can't use either `fuse-ext2` or `fuse2fs` on all systems.
Either way, this is ... | 1ad8304b1d0c1bafaed8a0e8e4182f2ee3249bab | 833b2dc08760d77e2cb215e8b355b9ca03dc3293 | diff --git a/Toolchain/default.nix b/Toolchain/default.nix
index 8d776608ae391a..bdaef19c750061 100644
--- a/Toolchain/default.nix
+++ b/Toolchain/default.nix
@@ -46,8 +46,7 @@ mkShell.override { stdenv = gccStdenv; } {
pre-commit
]
++ lib.optionals stdenv.isLinux [
- fuse
- fuse-ext2
+ fuse2fs
... | [
"Toolchain/default.nix"
] | [
{
"comment": "Hm, this was previously added in 743f494a9c6115e7710cc59ab9a2105e2b15696d but then later replaced with `fuse-ext2` in fb897913bbb7e7ef25cbc2a01b4f7370737e6310 without a reason.\n\nWe only use `fuse-ext2` on Darwin and FreeBSD. Not sure why we can't use either `fuse-ext2` or `fuse2fs` on all system... | true | ||
SerenityOS/serenity | 26,792 | issue_to_patch | Kernel/aarch64: Allow userspace to use cache maintenance instructions | This allows userspace to execute the DC CVAU, DC CIVAC, DC CVAC, and IC IVAU data/instruction cache maintenance instructions. These instructions operate on virtual addresses and are affected by the permissions specified in the page tables, so it should be safe to allow userspace to use them.
Both FreeBSD and Linux a... | 50d8d3670f85e51724ef64b35374e74a744b9984 | 88c1cdd9616494cf34d99daed8c66dd5603e3897 | diff --git a/Kernel/Arch/aarch64/Registers.h b/Kernel/Arch/aarch64/Registers.h
index bd13d4f2ab8162..57c65f87d1f9dc 100644
--- a/Kernel/Arch/aarch64/Registers.h
+++ b/Kernel/Arch/aarch64/Registers.h
@@ -822,6 +822,9 @@ struct alignas(u64) SCTLR_EL1 {
// Allow EL0 to read CTR_EL0 (the cache type register).
... | [
"Kernel/Arch/aarch64/Registers.h"
] | [] | true | ||
SerenityOS/serenity | 26,778 | issue_to_patch | FileManager: Drag and drop of a directory in columns view copies contents
Without any keyboard shortcuts or options dragging and dropping a folder over another folder while in columns view causes the folder, and all of its contents to be copied into the new directory and not moved to the new directory. Files are not l... | FileManager: Make drag-drop move by default | Fixes #1423 | 15ea8e970bc30c7e754e5009fb1d057a8c8dcf18 | ac34136f954bc3bf0fde377e8e800b91a5cdb77d | diff --git a/Userland/Applications/FileManager/FileUtils.cpp b/Userland/Applications/FileManager/FileUtils.cpp
index 405f11ef88761b..865b9549831f4e 100644
--- a/Userland/Applications/FileManager/FileUtils.cpp
+++ b/Userland/Applications/FileManager/FileUtils.cpp
@@ -141,7 +141,7 @@ ErrorOr<bool> handle_drop(GUI::DropEv... | [
"Userland/Applications/FileManager/FileUtils.cpp"
] | [
{
"comment": "Please make atomic commits, so don't include such style changes next to actual code changes.\n\nAlso, this variable should be prefixed with `s_`.",
"path": "Userland/Applications/FileManager/FileUtils.cpp",
"hunk": "@@ -17,7 +17,7 @@\n \n namespace FileManager {\n \n-HashTable<NonnullRefPt... | true | |
SerenityOS/serenity | 26,778 | comment_to_fix | FileManager: Make drag-drop move by default | Please make atomic commits, so don't include such style changes next to actual code changes.
Also, this variable should be prefixed with `s_`. | 15ea8e970bc30c7e754e5009fb1d057a8c8dcf18 | ac34136f954bc3bf0fde377e8e800b91a5cdb77d | diff --git a/Userland/Applications/FileManager/FileUtils.cpp b/Userland/Applications/FileManager/FileUtils.cpp
index 405f11ef88761b..865b9549831f4e 100644
--- a/Userland/Applications/FileManager/FileUtils.cpp
+++ b/Userland/Applications/FileManager/FileUtils.cpp
@@ -141,7 +141,7 @@ ErrorOr<bool> handle_drop(GUI::DropEv... | [
"Userland/Applications/FileManager/FileUtils.cpp"
] | [
{
"comment": "Please make atomic commits, so don't include such style changes next to actual code changes.\n\nAlso, this variable should be prefixed with `s_`.",
"path": "Userland/Applications/FileManager/FileUtils.cpp",
"hunk": "@@ -17,7 +17,7 @@\n \n namespace FileManager {\n \n-HashTable<NonnullRefPt... | true | ||
SerenityOS/serenity | 26,778 | comment_to_fix | FileManager: Make drag-drop move by default | Again, please don't blindly apply clangd suggestions.
Same for the other `const` change. | 15ea8e970bc30c7e754e5009fb1d057a8c8dcf18 | ac34136f954bc3bf0fde377e8e800b91a5cdb77d | diff --git a/Userland/Applications/FileManager/FileUtils.cpp b/Userland/Applications/FileManager/FileUtils.cpp
index 405f11ef88761b..865b9549831f4e 100644
--- a/Userland/Applications/FileManager/FileUtils.cpp
+++ b/Userland/Applications/FileManager/FileUtils.cpp
@@ -141,7 +141,7 @@ ErrorOr<bool> handle_drop(GUI::DropEv... | [
"Userland/Applications/FileManager/FileUtils.cpp"
] | [
{
"comment": "Again, please don't blindly apply clangd suggestions.\nSame for the other `const` change.",
"path": "Userland/Applications/FileManager/FileUtils.cpp",
"hunk": "@@ -69,7 +69,7 @@ ErrorOr<void> run_file_operation(FileOperation operation, Vector<ByteString> con\n VERIFY_NOT_REACHE... | true | ||
SerenityOS/serenity | 26,778 | comment_to_fix | FileManager: Make drag-drop move by default | Is lexical path resolution safe here? (non-lexical path resolution is usually a better choice) | 15ea8e970bc30c7e754e5009fb1d057a8c8dcf18 | ac34136f954bc3bf0fde377e8e800b91a5cdb77d | diff --git a/Userland/Applications/FileManager/FileUtils.cpp b/Userland/Applications/FileManager/FileUtils.cpp
index 405f11ef88761b..865b9549831f4e 100644
--- a/Userland/Applications/FileManager/FileUtils.cpp
+++ b/Userland/Applications/FileManager/FileUtils.cpp
@@ -141,7 +141,7 @@ ErrorOr<bool> handle_drop(GUI::DropEv... | [
"Userland/Applications/FileManager/FileUtils.cpp"
] | [
{
"comment": "Is lexical path resolution safe here? (non-lexical path resolution is usually a better choice)",
"path": "Userland/Applications/FileManager/FileUtils.cpp",
"hunk": "@@ -128,11 +129,11 @@ ErrorOr<bool> handle_drop(GUI::DropEvent const& event, ByteString const& destina\n return has_a... | true | ||
SerenityOS/serenity | 26,778 | comment_to_fix | FileManager: Make drag-drop move by default | Why did you choose Alt here? Ctrl is the usual keybdind for creating copies I think. At least on KDE and Windows. | 15ea8e970bc30c7e754e5009fb1d057a8c8dcf18 | ac34136f954bc3bf0fde377e8e800b91a5cdb77d | diff --git a/Userland/Applications/FileManager/FileUtils.cpp b/Userland/Applications/FileManager/FileUtils.cpp
index 405f11ef88761b..865b9549831f4e 100644
--- a/Userland/Applications/FileManager/FileUtils.cpp
+++ b/Userland/Applications/FileManager/FileUtils.cpp
@@ -141,7 +141,7 @@ ErrorOr<bool> handle_drop(GUI::DropEv... | [
"Userland/Applications/FileManager/FileUtils.cpp"
] | [
{
"comment": "Why did you choose Alt here? Ctrl is the usual keybdind for creating copies I think. At least on KDE and Windows.",
"path": "Userland/Applications/FileManager/FileUtils.cpp",
"hunk": "@@ -141,7 +141,7 @@ ErrorOr<bool> handle_drop(GUI::DropEvent const& event, ByteString const& destina\n ... | true | ||
SerenityOS/serenity | 26,783 | issue_to_patch | SSHServer: Pledge sigaction | I mainly test the SSHServer on Lagom (where pledges are ignored), so I didn't realize that sigaction was required on Serenity. This should have been done in 93cabc5a. | bde5e585afbc8682828cda872f857877a031805a | 3e6ff54e8a28597404d92f7acf973d41c86eec89 | diff --git a/Userland/Services/SSHServer/main.cpp b/Userland/Services/SSHServer/main.cpp
index f3d9ff89c666ed..27f1fd60e5a063 100644
--- a/Userland/Services/SSHServer/main.cpp
+++ b/Userland/Services/SSHServer/main.cpp
@@ -60,7 +60,7 @@ ErrorOr<void> accept_connection()
ErrorOr<int> serenity_main(Main::Arguments arg... | [
"Userland/Services/SSHServer/main.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,777 | issue_to_patch | Kernel/xHCI: Properly determine the transfer size for short transfers | Previously, we only set the remainder (number of untransmitted bytes) to the TRB transfer length of the Transfer Event.
(Note: this field is confusingly named. It's not actually the transfer length, but the number untransmitted bytes)
However, this field only tracks the number of untransmitted bytes for the specific ... | 32d409ee538c7edbe4144fae067ab670db3e5cc7 | 6e33c905160fc5b31f3919353ba1f631dfd545d0 | diff --git a/Kernel/Bus/USB/xHCI/xHCIController.cpp b/Kernel/Bus/USB/xHCI/xHCIController.cpp
index cfbb0455d3bcff..33d5169761e120 100644
--- a/Kernel/Bus/USB/xHCI/xHCIController.cpp
+++ b/Kernel/Bus/USB/xHCI/xHCIController.cpp
@@ -754,9 +754,9 @@ ErrorOr<void> xHCIController::enqueue_transfer(u8 slot, u8 endpoint, Pipe... | [
"Kernel/Bus/USB/xHCI/xHCIController.cpp",
"Kernel/Bus/USB/xHCI/xHCIController.h"
] | [] | true | ||
SerenityOS/serenity | 26,780 | issue_to_patch | Base: Add Landslide, Trombone emoji | π - U+1F6D8 Landslide
πͺ - U+1FA8A Trombone | 8fb7b20e014922959756f89159b00b01887a553b | dc581281c819b7361ae7ddb2227d056294b12ec5 | diff --git a/Base/home/anon/Documents/emoji-serenity.txt b/Base/home/anon/Documents/emoji-serenity.txt
index ae991de5d72fa7..11d13d83552883 100644
--- a/Base/home/anon/Documents/emoji-serenity.txt
+++ b/Base/home/anon/Documents/emoji-serenity.txt
@@ -290,3 +290,5 @@ Emoji 17
----------
π«ͺ - U+1FAEA DISTORTED FACE
π«... | [
"Base/home/anon/Documents/emoji-serenity.txt",
"Base/res/emoji/U+1F6D8.png",
"Base/res/emoji/U+1FA8A.png",
"Meta/emoji-file-list.txt"
] | [] | true | ||
SerenityOS/serenity | 26,768 | issue_to_patch | SSHServer: Use signals to detect exec's child death | This PR make this test more stable, and I didn't observe any hangs:
```bash
i=0; while ssh -p 2222 127.0.0.1 cat - < ~/progress_file_1MB.txt; do echo $i && ((i++)); done
```
There are still issues as this failure randomly appears (~1% of runs):
```
ssh_dispatch_run_fatal: Connection to [127.0.0.1 port 2222]: in... | e8e1a1ee7331312f5ae7d87d26f889f0c426cadf | 472f355b9cab7f4eac15194fc1be5e72b29d88fa | diff --git a/Userland/Libraries/LibCore/Socket.h b/Userland/Libraries/LibCore/Socket.h
index 172ec48bad4c8d..bc23be98d4bafa 100644
--- a/Userland/Libraries/LibCore/Socket.h
+++ b/Userland/Libraries/LibCore/Socket.h
@@ -196,6 +196,8 @@ class TCPSocket final : public Socket {
ErrorOr<void> set_blocking(bool enabled)... | [
"Userland/Libraries/LibCore/Socket.h",
"Userland/Libraries/LibSSH/Session.h",
"Userland/Services/SSHServer/SSHClient.cpp",
"Userland/Services/SSHServer/SSHClient.h",
"Userland/Services/SSHServer/main.cpp"
] | [
{
"comment": "nit (next time you're in the vicinity, if ever; doesn't have to be in this pr) \"will be close_d_\"",
"path": "Userland/Services/SSHServer/main.cpp",
"hunk": "",
"resolving_sha": "472f355b9cab7f4eac15194fc1be5e72b29d88fa",
"resolving_diff": "diff --git a/Userland/Services/SSHServer... | true | ||
SerenityOS/serenity | 26,768 | comment_to_fix | SSHServer: Use signals to detect exec's child death | nit (next time you're in the vicinity, if ever; doesn't have to be in this pr) "will be close_d_" | e8e1a1ee7331312f5ae7d87d26f889f0c426cadf | 472f355b9cab7f4eac15194fc1be5e72b29d88fa | diff --git a/Userland/Services/SSHServer/main.cpp b/Userland/Services/SSHServer/main.cpp
index 0759f94e014a42..f3d9ff89c666ed 100644
--- a/Userland/Services/SSHServer/main.cpp
+++ b/Userland/Services/SSHServer/main.cpp
@@ -27,20 +27,30 @@ OwnPtr<SSH::Server::TCPClient> g_client;
ErrorOr<void> accept_connection()
{
... | [
"Userland/Services/SSHServer/main.cpp"
] | [
{
"comment": "nit (next time you're in the vicinity, if ever; doesn't have to be in this pr) \"will be close_d_\"",
"path": "Userland/Services/SSHServer/main.cpp",
"hunk": "",
"resolving_sha": "472f355b9cab7f4eac15194fc1be5e72b29d88fa",
"resolving_diff": "diff --git a/Userland/Services/SSHServer... | true | ||
SerenityOS/serenity | 26,770 | issue_to_patch | SSHServer: Reap children when they die | Since 4c8188d6, we fork the main server on new connections, but I forgot to make the main server reap its children. This commit addresses the shortcoming.
---
With the patch applied on top of #26768 and #26769, the usual `while` loop seems to be a true infinite loop, on the one I started an hour ago it has done 6... | d7a7121d6a7b7feeeb3455fe37cff8ee20fab724 | 6609ed2bc821e88bd9fc85e0f01ab5b5099c097b | diff --git a/Userland/Services/SSHServer/main.cpp b/Userland/Services/SSHServer/main.cpp
index 2e1fe9ac143e71..0759f94e014a42 100644
--- a/Userland/Services/SSHServer/main.cpp
+++ b/Userland/Services/SSHServer/main.cpp
@@ -81,6 +81,21 @@ ErrorOr<int> serenity_main(Main::Arguments args)
Core::EventLoop loop;
+ ... | [
"Userland/Services/SSHServer/main.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,769 | issue_to_patch | LibSSH: Support encoding numbers with leading zeroes as mpint | As I failed to write a VERIFY that would hit this case, when the shared secret started with two null bytes, we were silently creating an invalid mpint which would make the ssh client abort the connection with this message:
ssh_dispatch_run_fatal: Connection to [Server IP]: incorrect signature
AFAICT, this patch mak... | 32d409ee538c7edbe4144fae067ab670db3e5cc7 | 1fd0d8337d3cb182d6202ede677bc304bb7ca506 | diff --git a/Userland/Libraries/LibSSH/DataTypes.cpp b/Userland/Libraries/LibSSH/DataTypes.cpp
index b1cf67633ad0a2..a0a5515f7c9d1e 100644
--- a/Userland/Libraries/LibSSH/DataTypes.cpp
+++ b/Userland/Libraries/LibSSH/DataTypes.cpp
@@ -29,12 +29,12 @@ ErrorOr<void> encode_mpint(AllocatingMemoryStream& stream, ReadonlyBy... | [
"Tests/LibSSH/TestSSHEncodeDecode.cpp",
"Userland/Libraries/LibSSH/DataTypes.cpp"
] | [
{
"comment": "nit: grammar. (better: \"Encoding this wrong makes openssh's client return\", \"Encoding this wrong results in openssh's client returning:\")",
"path": "Tests/LibSSH/TestSSHEncodeDecode.cpp",
"hunk": "@@ -83,11 +83,23 @@ TEST_CASE(mpint)\n return {};\n };\n \n+ TRY_OR_FAIL(t... | diff --git a/Tests/LibSSH/TestSSHEncodeDecode.cpp b/Tests/LibSSH/TestSSHEncodeDecode.cpp
index 7b2b57ffc563c2..c737900e4e49b0 100644
--- a/Tests/LibSSH/TestSSHEncodeDecode.cpp
+++ b/Tests/LibSSH/TestSSHEncodeDecode.cpp
@@ -83,11 +83,23 @@ TEST_CASE(mpint)
return {};
};
+ TRY_OR_FAIL(test(to_array<u8>... | true | |
SerenityOS/serenity | 26,769 | comment_to_fix | LibSSH: Support encoding numbers with leading zeroes as mpint | nit: grammar. (better: "Encoding this wrong makes openssh's client return", "Encoding this wrong results in openssh's client returning:") | 32d409ee538c7edbe4144fae067ab670db3e5cc7 | 1fd0d8337d3cb182d6202ede677bc304bb7ca506 | diff --git a/Tests/LibSSH/TestSSHEncodeDecode.cpp b/Tests/LibSSH/TestSSHEncodeDecode.cpp
index 7b2b57ffc563c2..c737900e4e49b0 100644
--- a/Tests/LibSSH/TestSSHEncodeDecode.cpp
+++ b/Tests/LibSSH/TestSSHEncodeDecode.cpp
@@ -83,11 +83,23 @@ TEST_CASE(mpint)
return {};
};
+ TRY_OR_FAIL(test(to_array<u8>... | [
"Tests/LibSSH/TestSSHEncodeDecode.cpp"
] | [
{
"comment": "nit: grammar. (better: \"Encoding this wrong makes openssh's client return\", \"Encoding this wrong results in openssh's client returning:\")",
"path": "Tests/LibSSH/TestSSHEncodeDecode.cpp",
"hunk": "@@ -83,11 +83,23 @@ TEST_CASE(mpint)\n return {};\n };\n \n+ TRY_OR_FAIL(t... | true | ||
SerenityOS/serenity | 26,706 | issue_to_patch | LibSSH+SSHServer: Add basic support for sftp | <img width="1027" height="833" alt="Screenshot From 2026-04-08 14-06-17" src="https://github.com/user-attachments/assets/8c93cf02-f2ed-4373-8726-25ce33ea40f5" />
---
There is quite a lot of boilerplate in the test code to generate the packets, but most of it could eventually land in a `Client` class in the future... | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 1a2dfeeb4a343a7c78243a7738a7c7314989a4b5 | diff --git a/Userland/Libraries/LibCore/System.cpp b/Userland/Libraries/LibCore/System.cpp
index 4820882cc452d6..0700934c6cb22f 100644
--- a/Userland/Libraries/LibCore/System.cpp
+++ b/Userland/Libraries/LibCore/System.cpp
@@ -674,6 +674,17 @@ ErrorOr<ssize_t> read(int fd, Bytes buffer)
return rc;
}
+ErrorOr<ss... | [
"Tests/LibSSH/CMakeLists.txt",
"Tests/LibSSH/SFTP/Shared.h",
"Tests/LibSSH/SFTP/TestSFTPPeer.cpp",
"Tests/LibSSH/SFTP/TestSFTPRead.cpp",
"Tests/LibSSH/SFTP/TestSFTPServerBasic.cpp",
"Tests/LibSSH/SFTP/TestSFTPStat.cpp",
"Userland/Libraries/LibCore/System.cpp",
"Userland/Libraries/LibCore/System.h",
... | [
{
"comment": "Could this be a flags enum?\n(AK_MAKE_BITWISE_ENUM_OPS I think it was) ",
"path": "Userland/Libraries/LibSSH/SFTP/Server.cpp",
"hunk": "@@ -144,4 +162,60 @@ ErrorOr<void> Server::send_file_attribute_message(u32 id, struct stat const& s)\n return {};\n }\n \n+// 6.3 Opening, Creating, a... | diff --git a/Tests/LibSSH/CMakeLists.txt b/Tests/LibSSH/CMakeLists.txt
index 66f98e7d7cd0fd..250c1cdcce1490 100644
--- a/Tests/LibSSH/CMakeLists.txt
+++ b/Tests/LibSSH/CMakeLists.txt
@@ -1,4 +1,8 @@
set(TEST_SOURCES
+ SFTP/TestSFTPPeer.cpp
+ SFTP/TestSFTPServerBasic.cpp
+ SFTP/TestSFTPStat.cpp
+ SFTP/TestS... | true | |
SerenityOS/serenity | 26,706 | comment_to_fix | LibSSH+SSHServer: Add basic support for sftp | Could this be a flags enum?
(AK_MAKE_BITWISE_ENUM_OPS I think it was) | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 1a2dfeeb4a343a7c78243a7738a7c7314989a4b5 | diff --git a/Userland/Libraries/LibSSH/SFTP/Server.cpp b/Userland/Libraries/LibSSH/SFTP/Server.cpp
new file mode 100644
index 00000000000000..08db2f0c6c0688
--- /dev/null
+++ b/Userland/Libraries/LibSSH/SFTP/Server.cpp
@@ -0,0 +1,327 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SP... | [
"Userland/Libraries/LibSSH/SFTP/Server.cpp"
] | [
{
"comment": "Could this be a flags enum?\n(AK_MAKE_BITWISE_ENUM_OPS I think it was) ",
"path": "Userland/Libraries/LibSSH/SFTP/Server.cpp",
"hunk": "@@ -144,4 +162,60 @@ ErrorOr<void> Server::send_file_attribute_message(u32 id, struct stat const& s)\n return {};\n }\n \n+// 6.3 Opening, Creating, a... | true | ||
SerenityOS/serenity | 26,706 | comment_to_fix | LibSSH+SSHServer: Add basic support for sftp | Similar with these | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 1a2dfeeb4a343a7c78243a7738a7c7314989a4b5 | diff --git a/Userland/Libraries/LibSSH/SFTP/Server.cpp b/Userland/Libraries/LibSSH/SFTP/Server.cpp
new file mode 100644
index 00000000000000..08db2f0c6c0688
--- /dev/null
+++ b/Userland/Libraries/LibSSH/SFTP/Server.cpp
@@ -0,0 +1,327 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SP... | [
"Userland/Libraries/LibSSH/SFTP/Server.cpp"
] | [
{
"comment": "Similar with these",
"path": "Userland/Libraries/LibSSH/SFTP/Server.cpp",
"hunk": "@@ -218,4 +220,108 @@ ErrorOr<void> Server::send_file_handle(u32 id, File const& file)\n return {};\n }\n \n+ErrorOr<Server::File*> Server::find_file(ReadonlyBytes handle)\n+{\n+ auto maybe_file = m_o... | true | ||
SerenityOS/serenity | 26,706 | comment_to_fix | LibSSH+SSHServer: Add basic support for sftp | it's a bit weird that this is a subclass that has a method with the same name and signature as the base class, but it doesn't override the base class and the base class version isn't virtualβ¦oh, is this just to make the methods public? I think for making things more visible, just doing
```
class PeerMock⦠{
public:
... | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 1a2dfeeb4a343a7c78243a7738a7c7314989a4b5 | diff --git a/Tests/LibSSH/SFTP/Shared.h b/Tests/LibSSH/SFTP/Shared.h
new file mode 100644
index 00000000000000..5f728cf038588c
--- /dev/null
+++ b/Tests/LibSSH/SFTP/Shared.h
@@ -0,0 +1,60 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#p... | [
"Tests/LibSSH/SFTP/Shared.h"
] | [
{
"comment": "it's a bit weird that this is a subclass that has a method with the same name and signature as the base class, but it doesn't override the base class and the base class version isn't virtualβ¦oh, is this just to make the methods public? I think for making things more visible, just doing\n\n```\ncla... | true | ||
SerenityOS/serenity | 26,706 | comment_to_fix | LibSSH+SSHServer: Add basic support for sftp | Putting `static` (for storage specifiers on globals, not for static methods, and not for statics within inline functions) in headers is almost never what you want in C++. `static` means each TU gets its own independent copy. Instead, you want to make these `inline` (C++ grew support for inline variables a while ago), t... | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 1a2dfeeb4a343a7c78243a7738a7c7314989a4b5 | diff --git a/Tests/LibSSH/SFTP/Shared.h b/Tests/LibSSH/SFTP/Shared.h
new file mode 100644
index 00000000000000..5f728cf038588c
--- /dev/null
+++ b/Tests/LibSSH/SFTP/Shared.h
@@ -0,0 +1,60 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#p... | [
"Tests/LibSSH/SFTP/Shared.h"
] | [
{
"comment": "Putting `static` (for storage specifiers on globals, not for static methods, and not for statics within inline functions) in headers is almost never what you want in C++. `static` means each TU gets its own independent copy. Instead, you want to make these `inline` (C++ grew support for inline var... | true | ||
SerenityOS/serenity | 26,706 | comment_to_fix | LibSSH+SSHServer: Add basic support for sftp | Are all types invalid, or are some types just not yet implemented? | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 1a2dfeeb4a343a7c78243a7738a7c7314989a4b5 | diff --git a/Userland/Libraries/LibSSH/SFTP/Server.cpp b/Userland/Libraries/LibSSH/SFTP/Server.cpp
new file mode 100644
index 00000000000000..08db2f0c6c0688
--- /dev/null
+++ b/Userland/Libraries/LibSSH/SFTP/Server.cpp
@@ -0,0 +1,327 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SP... | [
"Userland/Libraries/LibSSH/SFTP/Server.cpp"
] | [
{
"comment": "Are all types invalid, or are some types just not yet implemented?",
"path": "Userland/Libraries/LibSSH/SFTP/Server.cpp",
"hunk": "@@ -54,4 +55,15 @@ ErrorOr<void> Server::send_version_message()\n return {};\n }\n \n+ErrorOr<void> Server::handle_packet(FixedMemoryStream& stream)\n+{\n+... | true | ||
SerenityOS/serenity | 26,706 | comment_to_fix | LibSSH+SSHServer: Add basic support for sftp | Might want to get to this FIXME earlier rather than later as each new type with this FIXME makes it harder to do it :) | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 1a2dfeeb4a343a7c78243a7738a7c7314989a4b5 | diff --git a/Userland/Libraries/LibSSH/SFTP/Server.cpp b/Userland/Libraries/LibSSH/SFTP/Server.cpp
new file mode 100644
index 00000000000000..08db2f0c6c0688
--- /dev/null
+++ b/Userland/Libraries/LibSSH/SFTP/Server.cpp
@@ -0,0 +1,327 @@
+/*
+ * Copyright (c) 2026, Lucas Chollet <lucas.chollet@serenityos.org>
+ *
+ * SP... | [
"Userland/Libraries/LibSSH/SFTP/Server.cpp"
] | [
{
"comment": "Might want to get to this FIXME earlier rather than later as each new type with this FIXME makes it harder to do it :)",
"path": "Userland/Libraries/LibSSH/SFTP/Server.cpp",
"hunk": "@@ -218,4 +220,108 @@ ErrorOr<void> Server::send_file_handle(u32 id, File const& file)\n return {};\n }... | true | ||
SerenityOS/serenity | 26,779 | issue_to_patch | Meta: Install libcxx's __config_site to a known location | `libcxx` expects the `__config_site` file to be available in the default include folders. As far as I can tell, it is usually installed next to the `__config` file. Before this patch, the file was installed in its own folder, outside of clang's default includes paths.
As the `__config_site` file is dependent of CMak... | 32d409ee538c7edbe4144fae067ab670db3e5cc7 | f34049de967cd337bca459e9c36472fe6fc7b6c3 | diff --git a/Meta/build-root-filesystem.sh b/Meta/build-root-filesystem.sh
index b81453afea0dfd..c7688aa873ac94 100755
--- a/Meta/build-root-filesystem.sh
+++ b/Meta/build-root-filesystem.sh
@@ -39,9 +39,8 @@ SERENITY_ARCH="${SERENITY_ARCH:-x86_64}"
if [ "$SERENITY_TOOLCHAIN" = "Clang" ]; then
TOOLCHAIN_DIR="$SER... | [
"Meta/build-root-filesystem.sh"
] | [] | true | ||
SerenityOS/serenity | 26,772 | issue_to_patch | LibEDID: Accept yyyy-mm-dd date format in the PNP ID parser | The PNP ID file switched the date format from mm/dd/yyyy to yyyy-mm-dd.
Accept both formats, since the last version of https://uefi.org/uefi-pnp-export that was archived by archive.org still uses the old format.
We need to keep supporting that old version since uefi.org responds with an HTTP 403 error when trying t... | d7a7121d6a7b7feeeb3455fe37cff8ee20fab724 | 093d3f82c3e7428f3657d990651ad3e162c4bc39 | diff --git a/Meta/Lagom/Tools/CodeGenerators/LibEDID/GeneratePnpIDs.cpp b/Meta/Lagom/Tools/CodeGenerators/LibEDID/GeneratePnpIDs.cpp
index d821ba3af41837..4a91b756453115 100644
--- a/Meta/Lagom/Tools/CodeGenerators/LibEDID/GeneratePnpIDs.cpp
+++ b/Meta/Lagom/Tools/CodeGenerators/LibEDID/GeneratePnpIDs.cpp
@@ -22,28 +22... | [
"Meta/Lagom/Tools/CodeGenerators/LibEDID/GeneratePnpIDs.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,775 | issue_to_patch | Solitaire: Don't ignore the Play pile when solving | Hard to measure if this actually fixes the issue always, but it was weird to see it get stuck when I had a card there it refused to use. | 8823118b0cedb90c86d78ce2564dc0be24cbdc21 | f871b8de4cd2bd2c03760d69d2d27277f43bd3dc | diff --git a/Userland/Games/Solitaire/Game.cpp b/Userland/Games/Solitaire/Game.cpp
index 66ac1d827ed9db..48573d79a4f974 100644
--- a/Userland/Games/Solitaire/Game.cpp
+++ b/Userland/Games/Solitaire/Game.cpp
@@ -715,7 +715,7 @@ void Game::start_solving()
void Game::step_solve()
{
for (auto& stack : stacks()) {
- ... | [
"Userland/Games/Solitaire/Game.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,766 | issue_to_patch | Toolchain: Bump CMake to `v3.30.3` | This update fixes a build failure caused by a missing `<stdint.h>` include in `Utilities/cmcppdap/include/dap/network.h
Ref: https://github.com/Kitware/CMake/compare/v3.30.0...v3.30.3 | 26f8cccbbd9cc28ca53fd4bb9b82e6962e8f1910 | d7b1667bc2c6aa2a1ce2da4afc31bdc7faa9852e | diff --git a/Meta/CMake/cmake-version.cmake b/Meta/CMake/cmake-version.cmake
index d09c49c41e28fe..05a60ef3f19310 100644
--- a/Meta/CMake/cmake-version.cmake
+++ b/Meta/CMake/cmake-version.cmake
@@ -1,6 +1,6 @@
# Note: Update this alongside Toolchain/BuildCMake.sh
set(version_ok 0)
-if (CMAKE_VERSION VERSION_GREATER_... | [
"Meta/CMake/cmake-version.cmake",
"Ports/AvailablePorts.md",
"Ports/cmake/package.sh",
"Toolchain/BuildCMake.sh"
] | [] | true | ||
SerenityOS/serenity | 26,767 | issue_to_patch | SSHServer: Handle partial packets | With this PR, we are now able to run this command for quite some time:
```bash
i=0;while true; do ssh -p 2222 127.0.0.1 cat - < ~/progress_file_1MB.txt && echo $i && ((i++)); done
```
It used to crash in less than 10 attempts, but now I've seen it go above 200 hundreds loops. There are unfortunately still issues ... | 08792a0fd4c4d0b670d4cc9d9a128b1ffa8b15eb | 5e9c7dbccfb4469a285a601f7e438604257d7b83 | diff --git a/Userland/Libraries/LibSSH/Cipher.cpp b/Userland/Libraries/LibSSH/Cipher.cpp
index ad35ba9b8290b8..a1dbc78c16659f 100644
--- a/Userland/Libraries/LibSSH/Cipher.cpp
+++ b/Userland/Libraries/LibSSH/Cipher.cpp
@@ -40,6 +40,11 @@ ErrorOr<void> Cipher::encrypt(u32 packet_sequence_number, Bytes bytes)
return... | [
"Tests/LibSSH/TestSSHPeer.cpp",
"Userland/Libraries/LibSSH/Cipher.cpp",
"Userland/Libraries/LibSSH/Cipher.h",
"Userland/Libraries/LibSSH/Peer.cpp",
"Userland/Libraries/LibSSH/Peer.h",
"Userland/Services/SSHServer/SSHClient.cpp",
"Userland/Services/SSHServer/SSHClient.h",
"Userland/Services/SSHServer/T... | [] | diff --git a/Tests/LibSSH/TestSSHPeer.cpp b/Tests/LibSSH/TestSSHPeer.cpp
index e300c3bd35c59c..da6b59569c1edb 100644
--- a/Tests/LibSSH/TestSSHPeer.cpp
+++ b/Tests/LibSSH/TestSSHPeer.cpp
@@ -48,8 +48,16 @@ class PeerMock final : public SSH::Peer {
set_shared_secret(move(shared_secret));
}
+ PeerMock(... | true | |
SerenityOS/serenity | 26,765 | issue_to_patch | SSHServer: Support streaming more data to the child | This make the following command not crash every single time :sweat_smile::
```bash
ssh -p 2222 127.0.0.1 cat - < ~/progress_file_2MB.txt
``` | 26f8cccbbd9cc28ca53fd4bb9b82e6962e8f1910 | 53004933a2268344114fc4f022a9ba4acb531626 | diff --git a/Userland/Libraries/LibSSH/Session.cpp b/Userland/Libraries/LibSSH/Session.cpp
index 1c291d9fba8ab4..06163c94dd457c 100644
--- a/Userland/Libraries/LibSSH/Session.cpp
+++ b/Userland/Libraries/LibSSH/Session.cpp
@@ -8,6 +8,22 @@
namespace SSH {
+ErrorOr<ExecData> ExecData::create(Core::Process&& process... | [
"Userland/Libraries/LibSSH/Session.cpp",
"Userland/Libraries/LibSSH/Session.h",
"Userland/Services/SSHServer/SSHClient.cpp",
"Userland/Services/SSHServer/SSHClient.h"
] | [] | true | ||
SerenityOS/serenity | 26,763 | issue_to_patch | Kernel/MM: Flush data cache after mapping non-cacheable Regions | Without this PR, the (not yet upstream) VideoCore VII 3D driver will hang after drawing a couple of frames, due to the cache writing back stale data into MemoryType::NonCacheable regions.
This PR flushes the data cache immediately after mapping non-cacheable regions to ensure that the data caches don't contain any s... | f49ef7ac037df2cff32ae09c4894a594335d963a | c2edfa2812583d12f19e8d39a1239f99c2893fa6 | diff --git a/Kernel/Arch/Processor.h b/Kernel/Arch/Processor.h
index faa7ce4a54e97d..310dca1c6f235f 100644
--- a/Kernel/Arch/Processor.h
+++ b/Kernel/Arch/Processor.h
@@ -63,6 +63,8 @@ class ProcessorBase {
static void flush_tlb_local(VirtualAddress vaddr, size_t page_count);
static void flush_tlb(Memory::Pag... | [
"Kernel/Arch/Processor.h",
"Kernel/Arch/aarch64/Processor.cpp",
"Kernel/Arch/aarch64/Processor.h",
"Kernel/Arch/riscv64/Processor.cpp",
"Kernel/Arch/riscv64/Processor.h",
"Kernel/Arch/x86_64/Processor.cpp",
"Kernel/Arch/x86_64/Processor.h",
"Kernel/Memory/Region.cpp",
"Kernel/Memory/Region.h",
"Ke... | [
{
"comment": "I'm not sure if we should maybe instead panic here, same for RISC-V.",
"path": "Kernel/Arch/x86_64/Processor.cpp",
"hunk": "@@ -667,6 +671,9 @@ UNMAP_AFTER_INIT void ProcessorBase::initialize(u32 cpu)\n if (self->m_has_qemu_hvf_quirk.was_set())\n dmesgln(\"CPU[{}]: Applied corr... | true | ||
SerenityOS/serenity | 26,759 | issue_to_patch | SSHServer: Take functors as plain values in async_stream_std_data | Taking a universal reference may not always lead to the functor being owned by the coroutine frame. It can cause issues, if the functor is not available anymore when the coroutine resumes.
As an example:
```
template<typename Func>
f(Func&&) {}
void g() {
f([](){}); // The lambda is owned by g().
aut... | 9748ec51d0a4d03e74f14794d52a0a876b855ad1 | 078e1a55f94adcbdef12e7a7f98054452e62f44e | diff --git a/Userland/Services/SSHServer/SSHClient.cpp b/Userland/Services/SSHServer/SSHClient.cpp
index 515392d91d1ae5..de7fee58855ff0 100644
--- a/Userland/Services/SSHServer/SSHClient.cpp
+++ b/Userland/Services/SSHServer/SSHClient.cpp
@@ -675,7 +675,7 @@ ErrorOr<void> SSHClient::handle_channel_exec(NonnullRefPtr<Se... | [
"Userland/Services/SSHServer/SSHClient.cpp",
"Userland/Services/SSHServer/SSHClient.h"
] | [] | true | ||
SerenityOS/serenity | 26,760 | issue_to_patch | SSHServer: Stream stdin data to children | This makes command like `cat -` work with SSHServer. | f49ef7ac037df2cff32ae09c4894a594335d963a | 5985e981175e59a9f26b5c2641fd5bef91b42b40 | diff --git a/Userland/Libraries/LibSSH/Session.cpp b/Userland/Libraries/LibSSH/Session.cpp
index ec74650e88e0b4..1c291d9fba8ab4 100644
--- a/Userland/Libraries/LibSSH/Session.cpp
+++ b/Userland/Libraries/LibSSH/Session.cpp
@@ -8,13 +8,36 @@
namespace SSH {
-ErrorOr<Session> Session::create(u32 sender_channel_id, u... | [
"Userland/Libraries/LibSSH/Session.cpp",
"Userland/Libraries/LibSSH/Session.h",
"Userland/Services/SSHServer/SSHClient.cpp",
"Userland/Services/SSHServer/SSHClient.h"
] | [] | true | ||
SerenityOS/serenity | 26,715 | issue_to_patch | Terminal: Advertise 24-bit "true color" support | According to https://github.com/termstandard/colors?tab=readme-ov-file#checking-for-colorterm, a widespread method for advertising 24-bit color support is by defining the "COLORTERM" environment variable to either "truecolor" or "24bit".
The other suggested truecolor detection method of querying the terminal doesn't... | 3e0ef9397a1a9b3c32a0655543586ad43740dd96 | 97aa510ea400efd3e9a927c93c078e8fa1853b8a | diff --git a/Userland/Applications/Terminal/main.cpp b/Userland/Applications/Terminal/main.cpp
index e953b7df33a568..d092a3d0cbf858 100644
--- a/Userland/Applications/Terminal/main.cpp
+++ b/Userland/Applications/Terminal/main.cpp
@@ -148,7 +148,7 @@ static ErrorOr<void> run_command(StringView command, bool keep_open)
... | [
"Userland/Applications/Terminal/main.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,757 | issue_to_patch | SSHServer: Send exit-status messages | This is twofold: it makes the SSH client exit with a zero status when the subcommand succeeds, and it also ensures that the server properly closes the connection when the subcommand fails.
---
Before:
```bash
$ ssh -p 2222 127.0.0.1 echo "Hello"; echo $?
Hello
255
$ ssh -p 2222 127.0.0.1 cat IDontExist.txt ;... | 2e1860c98415ffda1efa8b9c765b7e026eafdaf0 | 31be3067dc429f87de1a36197a7fb6aa13e52ed2 | diff --git a/Userland/Services/SSHServer/SSHClient.cpp b/Userland/Services/SSHServer/SSHClient.cpp
index 622fa6a375afd3..5b13efce75abc8 100644
--- a/Userland/Services/SSHServer/SSHClient.cpp
+++ b/Userland/Services/SSHServer/SSHClient.cpp
@@ -661,8 +661,8 @@ Coroutine<void> SSHClient::async_wait_for_child(u32 sender_ch... | [
"Userland/Services/SSHServer/SSHClient.cpp",
"Userland/Services/SSHServer/SSHClient.h"
] | [] | true | ||
SerenityOS/serenity | 26,755 | issue_to_patch | SSHServer: Stream exec data back to the client | This PR makes the following command behave normally (results appear progressively and stderr is displayed):
`ssh -p 2222 anon@127.0.0.1 "echo 0; sleep 1; echo 1 1>&2; sleep 1; echo 2;"` | e00fd8ab4e36ad80e1d9734b892a508d8e086289 | 142ea10aaf813761cf1e8182fd26785bbfc37968 | diff --git a/Userland/Libraries/LibCore/EventLoop.h b/Userland/Libraries/LibCore/EventLoop.h
index bfaa9609c476ac..a871c832b498e8 100644
--- a/Userland/Libraries/LibCore/EventLoop.h
+++ b/Userland/Libraries/LibCore/EventLoop.h
@@ -73,7 +73,7 @@ class EventLoop {
void wake();
- void adopt_coroutine(Coroutine... | [
"Userland/Libraries/LibCore/EventLoop.h",
"Userland/Libraries/LibCore/File.h",
"Userland/Libraries/LibCore/Process.cpp",
"Userland/Libraries/LibCore/Process.h",
"Userland/Libraries/LibSSH/Session.h",
"Userland/Services/SSHServer/SSHClient.cpp",
"Userland/Services/SSHServer/SSHClient.h"
] | [
{
"comment": "Why are those prefixed with an underscore? If you're trying to use a reserved identifier, we usually add that underscore as a suffix.",
"path": "Userland/Libraries/LibSSH/Session.h",
"hunk": "@@ -11,16 +11,19 @@\n #include <AK/NonnullOwnPtr.h>\n #include <AK/Variant.h>\n #include <LibCore/... | true | ||
SerenityOS/serenity | 26,755 | comment_to_fix | SSHServer: Stream exec data back to the client | Why are those prefixed with an underscore? If you're trying to use a reserved identifier, we usually add that underscore as a suffix. | e00fd8ab4e36ad80e1d9734b892a508d8e086289 | 142ea10aaf813761cf1e8182fd26785bbfc37968 | diff --git a/Userland/Libraries/LibSSH/Session.h b/Userland/Libraries/LibSSH/Session.h
index bf67e3a4dbab3a..4858d4170378be 100644
--- a/Userland/Libraries/LibSSH/Session.h
+++ b/Userland/Libraries/LibSSH/Session.h
@@ -8,11 +8,24 @@
#include <AK/ByteBuffer.h>
#include <AK/Error.h>
+#include <AK/NonnullOwnPtr.h>
+#i... | [
"Userland/Libraries/LibSSH/Session.h"
] | [
{
"comment": "Why are those prefixed with an underscore? If you're trying to use a reserved identifier, we usually add that underscore as a suffix.",
"path": "Userland/Libraries/LibSSH/Session.h",
"hunk": "@@ -11,16 +11,19 @@\n #include <AK/NonnullOwnPtr.h>\n #include <AK/Variant.h>\n #include <LibCore/... | true | ||
SerenityOS/serenity | 26,754 | issue_to_patch | Toolchain+Ports: Update binutils to 2.46.0 and GCC to 16.1.0 | See commit descriptions for details.
Both the GCC update and binutils update commit include extra patches to fix upstream issues.
Additionally, I included some fixes for issues in our codebase found by this toolchain update. | 522d187b82eab1cbd1163e87de9b91a953b62f2b | 270145b67d68560405bb06255d292b60b3bc687e | diff --git a/AK/Function.h b/AK/Function.h
index 7ba7960750bf49..c97d3ef92a8923 100644
--- a/AK/Function.h
+++ b/AK/Function.h
@@ -52,6 +52,9 @@ namespace AK {
# pragma GCC diagnostic push
// FIXME: GCC does not like the union, thinking we access a zero-sized array out of bounds.
# pragma GCC diagnostic ignore... | [
"AK/Function.h",
"AK/StdLibExtraDetails.h",
"CMakeLists.txt",
"Meta/serenity.sh",
"Ports/AvailablePorts.md",
"Ports/binutils/package.sh",
"Ports/gcc/package.sh",
"Tests/Kernel/crash.cpp",
"Toolchain/BuildGNU.sh",
"Toolchain/Patches/binutils/0001-Add-support-for-SerenityOS.patch",
"Toolchain/Patc... | [] | diff --git a/Tests/Kernel/crash.cpp b/Tests/Kernel/crash.cpp
index b989e98c683828..b397297012f46f 100644
--- a/Tests/Kernel/crash.cpp
+++ b/Tests/Kernel/crash.cpp
@@ -131,14 +131,8 @@ int main(int argc, char** argv)
asm volatile("ud2" :);
#elif ARCH(RISCV64)
// Invalid instructions are not re... | true | |
SerenityOS/serenity | 26,750 | issue_to_patch | Toolchain: Add /usr/local Clang's default search path again | Of course, I found after the fact my note that was supposed to prevent me for forgetting about it when rebasing the patches.
---
This is the only part of the driver patch that was not upstreamed, so we
need to keep it locally to build ports with Clang.
It regressed in https://github.com/SerenityOS/serenity/c... | 311372abf1d884156b2334bd3e3e8ca7ad670a9f | 210941893cc3e839e0aa751ffb7bd76562c70589 | diff --git a/Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch b/Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch
new file mode 100644
index 00000000000000..ac04d8301c538b
--- /dev/null
+++ b/Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-se... | [
"Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch",
"Toolchain/Patches/llvm/0002-llvm-Add-support-for-building-LLVM-on-SerenityOS.patch",
"Toolchain/Patches/llvm/0003-tools-Support-building-shared-libLLVM-and-libClang-f.patch",
"Toolchain/Patches/llvm/0004-compiler-rt-Enable-pr... | [
{
"comment": "The commit message (of the commit adding this .patch file, not the commit in the patch file) is missing a \"to\" in front of \"Clang's\"",
"path": "Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch",
"hunk": "@@ -0,0 +1,36 @@\n+From 00000000000000000000000000... | true | ||
SerenityOS/serenity | 26,750 | comment_to_fix | Toolchain: Add /usr/local Clang's default search path again | The commit message (of the commit adding this .patch file, not the commit in the patch file) is missing a "to" in front of "Clang's" | 311372abf1d884156b2334bd3e3e8ca7ad670a9f | 210941893cc3e839e0aa751ffb7bd76562c70589 | diff --git a/Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch b/Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch
new file mode 100644
index 00000000000000..ac04d8301c538b
--- /dev/null
+++ b/Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-se... | [
"Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch"
] | [
{
"comment": "The commit message (of the commit adding this .patch file, not the commit in the patch file) is missing a \"to\" in front of \"Clang's\"",
"path": "Toolchain/Patches/llvm/0001-clang-Add-usr-local-to-the-default-search-path.patch",
"hunk": "@@ -0,0 +1,36 @@\n+From 00000000000000000000000000... | true | ||
SerenityOS/serenity | 26,753 | issue_to_patch | Ports: Correct the oniguruma version in AvailablePorts.md | This was missed in f327b4e6b7f. | f327b4e6b7fad78ad7ee460e32a006d0a0499e61 | 16d2e3dd771252f02ae8562182501526954bfb18 | diff --git a/Ports/AvailablePorts.md b/Ports/AvailablePorts.md
index 3c429bd3431380..e30f33aa583b15 100644
--- a/Ports/AvailablePorts.md
+++ b/Ports/AvailablePorts.md
@@ -248,7 +248,7 @@ This list is also available at [ports.serenityos.net](https://ports.serenityos.n
| [`nyancat`](nyancat/) |... | [
"Ports/AvailablePorts.md"
] | [] | true | ||
SerenityOS/serenity | 26,741 | issue_to_patch | Ports/oniguruma: Upgrade to 6.9.10 | This also fixes a build failure with LLVM/clang: the `alloca.h` include was guarded with `!defined(__GNUC__)`, which clang also defines for GCC compatibility, so the header was never included. Upstream fixed this in 6.9.10. | caea8bf48b9429e2bd56320e10c09fe683e068e4 | a571e9c239f1914eff958b6f511139b3295664c9 | diff --git a/Ports/oniguruma/package.sh b/Ports/oniguruma/package.sh
index 80ea8505fe5ab3..e892b6e9e8d5bd 100755
--- a/Ports/oniguruma/package.sh
+++ b/Ports/oniguruma/package.sh
@@ -1,9 +1,9 @@
#!/usr/bin/env -S bash ../.port_include.sh
port='oniguruma'
-version='6.9.9'
+version='6.9.10'
useconfigure='true'
use_fr... | [
"Ports/oniguruma/package.sh"
] | [] | true | ||
SerenityOS/serenity | 26,744 | issue_to_patch | Meta: Don't use virtio-net | Our virtio-net driver is unstable and causes kernel panics, see issue #26477. Instead, use the e1000 device, like we already do on x86. (`ethernet_device_type` defaults to "e1000") | 727c4a3d1a6748221b383207dac354e564b6150d | 9445416cff9d32971350a76b91e9044846e86a3a | diff --git a/Meta/run.py b/Meta/run.py
index e71be26d3de147..6f4ff070efda75 100755
--- a/Meta/run.py
+++ b/Meta/run.py
@@ -732,8 +732,6 @@ def set_up_network_hardware(config: Configuration):
provided_ethernet_device_type = environ.get("SERENITY_ETHERNET_DEVICE_TYPE")
if provided_ethernet_device_type is not No... | [
"Meta/run.py"
] | [] | true | ||
SerenityOS/serenity | 26,740 | issue_to_patch | Kernel+lsirq: Stop using `u8`s for interrupt numbers | This will be necessary for supporting the Pi 5 VideoCore V3D hardware, since it has interrupts 282 and 281.
See commit descriptions for details. | 459c2658b87eb2a88834fed93aa959c9cd82dbd9 | 03d43e87766f395d8e650044ede93ada9e21250f | diff --git a/Kernel/Arch/Interrupts.h b/Kernel/Arch/Interrupts.h
index 9aacfa7a424bef..1bc818099b58f5 100644
--- a/Kernel/Arch/Interrupts.h
+++ b/Kernel/Arch/Interrupts.h
@@ -9,6 +9,7 @@
#include <AK/Error.h>
#include <AK/Platform.h>
#include <AK/Types.h>
+#include <Kernel/Interrupts/Interrupts.h>
#if ARCH(X86_64... | [
"Kernel/Arch/Interrupts.h",
"Kernel/Arch/aarch64/IRQController.h",
"Kernel/Arch/aarch64/InterruptManagement.cpp",
"Kernel/Arch/aarch64/InterruptManagement.h",
"Kernel/Arch/aarch64/Interrupts.cpp",
"Kernel/Arch/aarch64/Interrupts/GICv2.cpp",
"Kernel/Arch/aarch64/Interrupts/GICv2.h",
"Kernel/Arch/aarch6... | [
{
"comment": "This could technically go into the for-header, I think\n\n`while` sadly does not allow compound statements,\nunless:\n```c++\nOptional<u8> maybe_interrupt_number;\nwhile ((maybe_interrupt_number = interrupt_controller->pending_interrupt()), maybe_interrupt_number.has_value()) {\n ...\n}\n```\nb... | true | ||
SerenityOS/serenity | 26,740 | comment_to_fix | Kernel+lsirq: Stop using `u8`s for interrupt numbers | Why not make find_redirection_entry_by_vector take InterruptNumber? | 459c2658b87eb2a88834fed93aa959c9cd82dbd9 | 03d43e87766f395d8e650044ede93ada9e21250f | diff --git a/Kernel/Arch/x86_64/Interrupts/IOAPIC.cpp b/Kernel/Arch/x86_64/Interrupts/IOAPIC.cpp
index bc944019d3b93c..471071bb7b6eb3 100644
--- a/Kernel/Arch/x86_64/Interrupts/IOAPIC.cpp
+++ b/Kernel/Arch/x86_64/Interrupts/IOAPIC.cpp
@@ -43,7 +43,7 @@ UNMAP_AFTER_INIT void IOAPIC::initialize()
{
}
-void IOAPIC::ma... | [
"Kernel/Arch/x86_64/Interrupts/IOAPIC.cpp"
] | [
{
"comment": "Why not make find_redirection_entry_by_vector take InterruptNumber?",
"path": "Kernel/Arch/x86_64/Interrupts/IOAPIC.cpp",
"hunk": "@@ -201,12 +201,12 @@ void IOAPIC::disable(GenericInterruptHandler const& handler)\n {\n InterruptDisabler disabler;\n VERIFY(!is_hard_disabled());\n- ... | true | ||
SerenityOS/serenity | 26,740 | comment_to_fix | Kernel+lsirq: Stop using `u8`s for interrupt numbers | (nit: this line should arguably be its own commit) | 459c2658b87eb2a88834fed93aa959c9cd82dbd9 | 03d43e87766f395d8e650044ede93ada9e21250f | diff --git a/Kernel/Arch/x86_64/Interrupts.cpp b/Kernel/Arch/x86_64/Interrupts.cpp
index 5270b114d33846..b69ea85fc20f9f 100644
--- a/Kernel/Arch/x86_64/Interrupts.cpp
+++ b/Kernel/Arch/x86_64/Interrupts.cpp
@@ -301,7 +301,7 @@ void handle_interrupt(TrapFrame* trap)
// reserved for when the PIC is disabled, so we c... | [
"Kernel/Arch/x86_64/Interrupts.cpp"
] | [
{
"comment": "(nit: this line should arguably be its own commit)",
"path": "Kernel/Arch/x86_64/Interrupts.cpp",
"hunk": "@@ -362,7 +362,7 @@ ErrorOr<u8> reserve_interrupt_handlers(u8 number_of_irqs)\n continue;\n \n found_range = true;\n- for (auto off = 1; off < number_of_irq... | true | ||
SerenityOS/serenity | 26,740 | comment_to_fix | Kernel+lsirq: Stop using `u8`s for interrupt numbers | (this should be in that commit too) | 459c2658b87eb2a88834fed93aa959c9cd82dbd9 | 03d43e87766f395d8e650044ede93ada9e21250f | diff --git a/Kernel/Arch/x86_64/Interrupts.cpp b/Kernel/Arch/x86_64/Interrupts.cpp
index 5270b114d33846..b69ea85fc20f9f 100644
--- a/Kernel/Arch/x86_64/Interrupts.cpp
+++ b/Kernel/Arch/x86_64/Interrupts.cpp
@@ -301,7 +301,7 @@ void handle_interrupt(TrapFrame* trap)
// reserved for when the PIC is disabled, so we c... | [
"Kernel/Arch/x86_64/Interrupts.cpp"
] | [
{
"comment": "(this should be in that commit too)",
"path": "Kernel/Arch/x86_64/Interrupts.cpp",
"hunk": "@@ -732,7 +732,7 @@ UNMAP_AFTER_INIT void initialize_interrupts()\n register_interrupt_handler(0xfe, interrupt_254_asm_entry);\n register_interrupt_handler(0xff, interrupt_255_asm_entry);\n ... | true | ||
SerenityOS/serenity | 26,740 | comment_to_fix | Kernel+lsirq: Stop using `u8`s for interrupt numbers | lhs has the result signed, rhs has it unsigned. intentional change? | 459c2658b87eb2a88834fed93aa959c9cd82dbd9 | 03d43e87766f395d8e650044ede93ada9e21250f | diff --git a/Kernel/Bus/PCI/Device.cpp b/Kernel/Bus/PCI/Device.cpp
index 7ebfac23d5a1cd..7a9e77046ec5d7 100644
--- a/Kernel/Bus/PCI/Device.cpp
+++ b/Kernel/Bus/PCI/Device.cpp
@@ -92,7 +92,7 @@ PCI::InterruptType Device::get_interrupt_type()
// is nothing left to do. The second parameter `msi` is used by the
// driver... | [
"Kernel/Bus/PCI/Device.cpp"
] | [
{
"comment": "lhs has the result signed, rhs has it unsigned. intentional change?",
"path": "Kernel/Bus/PCI/Device.cpp",
"hunk": "@@ -116,12 +116,11 @@ ErrorOr<InterruptType> Device::reserve_irqs(u8 number_of_irqs, bool msi)\n return m_interrupt_range.m_type;\n }\n \n-PhysicalAddress Device::msix_ta... | true | ||
SerenityOS/serenity | 26,752 | issue_to_patch | LibGfx/WebP: Remove an unused variable | Found by gcc 16's Wunused-but-set-variable.
No behavior change. | 311372abf1d884156b2334bd3e3e8ca7ad670a9f | ca8fa088d8bee0ca4b4d391ab8e2bd2dda08b7e2 | diff --git a/Userland/Libraries/LibGfx/ImageFormats/WebPLoaderLossy.cpp b/Userland/Libraries/LibGfx/ImageFormats/WebPLoaderLossy.cpp
index 0940407503647c..55bb4a59da61cc 100644
--- a/Userland/Libraries/LibGfx/ImageFormats/WebPLoaderLossy.cpp
+++ b/Userland/Libraries/LibGfx/ImageFormats/WebPLoaderLossy.cpp
@@ -1078,8 +1... | [
"Userland/Libraries/LibGfx/ImageFormats/WebPLoaderLossy.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,747 | issue_to_patch | Base: Add more emoji | Adds:
1FAEA fully-qualified # π«ͺ E17.0 distorted face
1F474 fully-qualified # π΄ E0.6 old man
1FACD fully-qualified # π« E17.0 orca | 727c4a3d1a6748221b383207dac354e564b6150d | 513c276b924f7393bc0330a616d4b7cc1f8a97be | diff --git a/Base/home/anon/Documents/emoji-serenity.txt b/Base/home/anon/Documents/emoji-serenity.txt
index 529aafc6171b27..ae991de5d72fa7 100644
--- a/Base/home/anon/Documents/emoji-serenity.txt
+++ b/Base/home/anon/Documents/emoji-serenity.txt
@@ -285,3 +285,8 @@ Emoji 16
πͺΎ - U+1FABE LEAFLESS TREE
πͺ - U+1FA8F SH... | [
"Base/home/anon/Documents/emoji-serenity.txt",
"Base/res/emoji/U+1F474.png",
"Base/res/emoji/U+1FACD.png",
"Base/res/emoji/U+1FAEA.png",
"Meta/emoji-file-list.txt"
] | [] | true | ||
SerenityOS/serenity | 26,746 | issue_to_patch | LibGfx+LibMedia: Very slightly nicer VP8 keyframe handling | No real behavior change. | 727c4a3d1a6748221b383207dac354e564b6150d | e33b06d7a9d8209fcede976d91b49588606c3895 | diff --git a/Userland/Libraries/LibGfx/ImageFormats/WebPLoader.cpp b/Userland/Libraries/LibGfx/ImageFormats/WebPLoader.cpp
index 5c319759941031..3aafff50ab9c17 100644
--- a/Userland/Libraries/LibGfx/ImageFormats/WebPLoader.cpp
+++ b/Userland/Libraries/LibGfx/ImageFormats/WebPLoader.cpp
@@ -448,7 +448,13 @@ static Error... | [
"Userland/Libraries/LibGfx/ImageFormats/WebPLoader.cpp",
"Userland/Libraries/LibGfx/ImageFormats/WebPLoaderLossy.cpp",
"Userland/Libraries/LibGfx/ImageFormats/WebPLoaderLossy.h",
"Userland/Libraries/LibMedia/Video/VP8/Decoder.cpp"
] | [] | true | ||
SerenityOS/serenity | 26,743 | issue_to_patch | Ports/git: Use the correct strip program | The host strip program will only work for the host architecture and throws an error during cross compilation. | 727c4a3d1a6748221b383207dac354e564b6150d | 6bb5cb18ee975a43aa152ad11e9a9b4fb6fdc28b | diff --git a/Ports/git/package.sh b/Ports/git/package.sh
index a749eb0139b788..e35b90ad0a42fd 100755
--- a/Ports/git/package.sh
+++ b/Ports/git/package.sh
@@ -18,7 +18,7 @@ build() {
run make \
"${makeopts[@]}" \
CURL_LDFLAGS='-lcurl -lssl -lcrypto -lz'
- run make strip
+ run make strip STR... | [
"Ports/git/package.sh"
] | [] | true | ||
Shopify/liquid | 2,094 | issue_to_patch | Let environment `self` shadow `SelfDrop` | Previously, bare `self` returned `SelfDrop` as soon as local scopes missed, which meant an environment or counter variable named `self` couldn't shadow it. This PR fixes that by checking environments before falling back to `SelfDrop`, while still suppressing strict-variable errors for the fallback path.
| 7b368dffb844c44a9466226d1a243b05aafc5be5 | 915d44fb90e1df795dcb89c8dab6e98a87698cfa | diff --git a/Gemfile b/Gemfile
index a6c448617..17463e74e 100644
--- a/Gemfile
+++ b/Gemfile
@@ -32,6 +32,7 @@ group :test do
end
group :spec do
- gem 'liquid-spec', github: 'Shopify/liquid-spec', branch: 'main'
+ # Using feature branch until https://github.com/Shopify/liquid-spec/pull/144 is merged
+ gem 'liqui... | [
"Gemfile",
"lib/liquid/context.rb"
] | [] | true | ||
Shopify/liquid | 2,091 | issue_to_patch | Fix `SelfDrop` equality | `Liquid::SelfDrop` currently behaves like a wrapper object for equality, even though every meaningful operation delegates back to the captured context. That means two separate `self` lookups from the same context compare as different values, leaking implementation detail into both Ruby callers and template behavior:
... | 742ac3dbf5432c6c2689ea00da604d7379f73799 | 771ceccf41822298cd5fce6d2273d0dfe5b38311 | diff --git a/lib/liquid/context.rb b/lib/liquid/context.rb
index 7902ebcff..26ff72852 100644
--- a/lib/liquid/context.rb
+++ b/lib/liquid/context.rb
@@ -208,7 +208,7 @@ def find_variable(key, raise_on_not_found: true)
# `self` resolves to a SelfDrop (enabling `self['var']` lookups),
# but only when it h... | [
"lib/liquid/context.rb",
"lib/liquid/self_drop.rb",
"test/integration/self_drop_context_test.rb"
] | [] | diff --git a/test/integration/self_drop_context_test.rb b/test/integration/self_drop_context_test.rb
index 338719e81..f57fdc393 100644
--- a/test/integration/self_drop_context_test.rb
+++ b/test/integration/self_drop_context_test.rb
@@ -77,6 +77,28 @@ def test_self_drop_context_setter_is_undefined
assert_template_... | true | |
Shopify/liquid | 2,027 | issue_to_patch | Make new tests serializable to liquid-spec | (No anonymous classes)
| 22e979a6fae0038c703a95c7421f159f71934a38 | bad29caaae0e374366d28eadf3b45fb07ae4ec01 | diff --git a/.github/workflows/liquid.yml b/.github/workflows/liquid.yml
index e496f55f8..bd0df8e5b 100644
--- a/.github/workflows/liquid.yml
+++ b/.github/workflows/liquid.yml
@@ -22,17 +22,17 @@ jobs:
rubyopt: "--enable-frozen-string-literal",
}
- { ruby: 3.4, allowed-failure: f... | [
".github/workflows/liquid.yml",
"test/integration/hash_rendering_test.rb",
"test/integration/standard_filter_test.rb",
"test/test_helper.rb"
] | [] | diff --git a/test/integration/hash_rendering_test.rb b/test/integration/hash_rendering_test.rb
index 47e6288b3..0c1b33e48 100644
--- a/test/integration/hash_rendering_test.rb
+++ b/test/integration/hash_rendering_test.rb
@@ -88,17 +88,11 @@ def test_render_hash_with_hash_key
end
def test_rendering_hash_with_cus... | true | |
Shopify/liquid | 2,082 | issue_to_patch | Prevent `SelfDrop` context mutation across render boundaries | This PR fixes `SelfDrop` scoping so a `self` assigned in one template preserves its original scope when passed to a rendered template.
```liquid
# template1.liquid
{% assign var = 42 %}
{% assign s = self %}
{% render 'template2', other_self: s %}
# template2.liquid
{% assign var = 43 %}
{{ other_self.var }... | 1954a2655cf4d427b6c9169354832638740f2db5 | 3f5a7faeb246d5d0b242292976893dedc421c58c | diff --git a/lib/liquid/self_drop.rb b/lib/liquid/self_drop.rb
index 357814653..4bfff0e45 100644
--- a/lib/liquid/self_drop.rb
+++ b/lib/liquid/self_drop.rb
@@ -16,23 +16,25 @@ module Liquid
# then the local value takes precedence over the `self` object.
# @liquid_access global
class SelfDrop < Drop
- def... | [
"lib/liquid/self_drop.rb",
"lib/liquid/template.rb",
"test/integration/self_drop_context_test.rb"
] | [
{
"comment": "Rename to `test/integration/self_drop_context_test.rb`",
"path": "test/integration/self_drop_context_test.rb",
"hunk": "",
"resolving_sha": "3f5a7faeb246d5d0b242292976893dedc421c58c",
"resolving_diff": "diff --git a/test/integration/self_drop_context_test.rb b/test/integration/self... | diff --git a/test/integration/self_drop_context_test.rb b/test/integration/self_drop_context_test.rb
new file mode 100644
index 000000000..338719e81
--- /dev/null
+++ b/test/integration/self_drop_context_test.rb
@@ -0,0 +1,98 @@
+# frozen_string_literal: true
+
+require 'test_helper'
+
+class SelfDropContextTest < Mini... | true | |
Shopify/liquid | 2,082 | comment_to_fix | Prevent `SelfDrop` context mutation across render boundaries | Rename to `test/integration/self_drop_context_test.rb` | 1954a2655cf4d427b6c9169354832638740f2db5 | 3f5a7faeb246d5d0b242292976893dedc421c58c | diff --git a/test/integration/self_drop_context_test.rb b/test/integration/self_drop_context_test.rb
new file mode 100644
index 000000000..338719e81
--- /dev/null
+++ b/test/integration/self_drop_context_test.rb
@@ -0,0 +1,98 @@
+# frozen_string_literal: true
+
+require 'test_helper'
+
+class SelfDropContextTest < Mini... | [
"test/integration/self_drop_context_test.rb"
] | [
{
"comment": "Rename to `test/integration/self_drop_context_test.rb`",
"path": "test/integration/self_drop_context_test.rb",
"hunk": "",
"resolving_sha": "3f5a7faeb246d5d0b242292976893dedc421c58c",
"resolving_diff": "diff --git a/test/integration/self_drop_context_test.rb b/test/integration/self... | true | ||
Shopify/liquid | 2,082 | comment_to_fix | Prevent `SelfDrop` context mutation across render boundaries | ```suggestion
undef context
```
this should be done so upstream implementations know the value is not contextualizable | 1954a2655cf4d427b6c9169354832638740f2db5 | 3f5a7faeb246d5d0b242292976893dedc421c58c | diff --git a/lib/liquid/self_drop.rb b/lib/liquid/self_drop.rb
index 357814653..4bfff0e45 100644
--- a/lib/liquid/self_drop.rb
+++ b/lib/liquid/self_drop.rb
@@ -16,23 +16,25 @@ module Liquid
# then the local value takes precedence over the `self` object.
# @liquid_access global
class SelfDrop < Drop
- def... | [
"lib/liquid/self_drop.rb"
] | [
{
"comment": "```suggestion\n undef context\n```\n\nthis should be done so upstream implementations know the value is not contextualizable",
"path": "lib/liquid/self_drop.rb",
"hunk": "@@ -16,21 +16,23 @@ module Liquid\n # then the local value takes precedence over the `self` object.\n # @liqui... | true | ||
Shopify/liquid | 1,422 | issue_to_patch | Add replace_last and remove_last filters | This PR adds two filters `replace_last` and `remove_last`. They do exactly what you think. The opposites of `replace_first` and `remove_first`.
Related issue: https://github.com/Shopify/liquid/issues/778 | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/History.md b/History.md
index 935151bc0..5d0c092ba 100644
--- a/History.md
+++ b/History.md
@@ -1,10 +1,12 @@
# Liquid Change Log
-## 5.1.1 (unreleased)
+## 5.2.0 (unreleased)
-### Fixes
+### Features
+* Add `remove_last`, and `replace_last` filters (#1422) [Anders Hagbard]
-* Fix some internal erro... | [
"History.md",
"lib/liquid/standardfilters.rb",
"test/integration/standard_filter_test.rb"
] | [
{
"comment": "It appears ambiguous which `' a'` is being removed. For `replace_last`, please try `a a b a` as returning `a a b` will unambiguously prove it wouldn't have been `a b a`. The same goes for the numeric version. Try `1 2 1 3 1` which should return `1 2 1 3`, not `2 1 3 1` or `1 2 3 1`.\r\n\r\nAlso pl... | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index e7b34c164..694d24d77 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -539,19 +539,31 @@ def test_first_last
end
def test_replace
- assert_equal('2 2 2... | true | |
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | It appears ambiguous which `' a'` is being removed. For `replace_last`, please try `a a b a` as returning `a a b` will unambiguously prove it wouldn't have been `a b a`. The same goes for the numeric version. Try `1 2 1 3 1` which should return `1 2 1 3`, not `2 1 3 1` or `1 2 3 1`.
Also please update the cases for ... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index e7b34c164..694d24d77 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -539,19 +539,31 @@ def test_first_last
end
def test_replace
- assert_equal('2 2 2... | [
"test/integration/standard_filter_test.rb"
] | [
{
"comment": "It appears ambiguous which `' a'` is being removed. For `replace_last`, please try `a a b a` as returning `a a b` will unambiguously prove it wouldn't have been `a b a`. The same goes for the numeric version. Try `1 2 1 3 1` which should return `1 2 1 3`, not `2 1 3 1` or `1 2 3 1`.\r\n\r\nAlso pl... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | It appears the first test case for each filter is replacing a string with a numeric input. Shouldn't it be a string replaced with another string? The second test case for each filter covers numeric input for both the candidate and the replacement, so it's not necessary to cover numeric input in the first one.
This a... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index e7b34c164..694d24d77 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -539,19 +539,31 @@ def test_first_last
end
def test_replace
- assert_equal('2 2 2... | [
"test/integration/standard_filter_test.rb"
] | [
{
"comment": "It appears the first test case for each filter is replacing a string with a numeric input. Shouldn't it be a string replaced with another string? The second test case for each filter covers numeric input for both the candidate and the replacement, so it's not necessary to cover numeric input in th... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | **Tip:** If your parent commit is local (not pushed yet), simply use `git commit --amend` to overwrite the parent commit with simple corrections like this. That way you don't create another commit just for fixing mistakes π
Even if it's pushed to the PR recently, as long as I can be reasonably sure no one is pulli... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/lib/liquid/standardfilters.rb b/lib/liquid/standardfilters.rb
index 98e38f5f4..f361c02bf 100644
--- a/lib/liquid/standardfilters.rb
+++ b/lib/liquid/standardfilters.rb
@@ -308,14 +308,34 @@ def replace_first(input, string, replacement = '')
input.to_s.sub(string.to_s, replacement.to_s)
end
+ ... | [
"lib/liquid/standardfilters.rb"
] | [
{
"comment": "**Tip:** If your parent commit is local (not pushed yet), simply use `git commit --amend` to overwrite the parent commit with simple corrections like this. That way you don't create another commit just for fixing mistakes π \r\n\r\nEven if it's pushed to the PR recently, as long as I can be reaso... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | Great to see the other changes. π Over here I think my suggestion was to change `2` to `'2'` in the removed line, rather than to remove it. π First one proves string replaced with string. Second one proves numeric replaced with numeric.
In fact, to make it crystal clear, it should be more like this, with `assert_e... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index e7b34c164..694d24d77 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -539,19 +539,31 @@ def test_first_last
end
def test_replace
- assert_equal('2 2 2... | [
"test/integration/standard_filter_test.rb"
] | [
{
"comment": "Great to see the other changes. π Over here I think my suggestion was to change `2` to `'2'` in the removed line, rather than to remove it. π First one proves string replaced with string. Second one proves numeric replaced with numeric.\r\n\r\nIn fact, to make it crystal clear, it should be more... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | And for this (better organize `assert_template_result` as a copy of `assert_equal` below it β easier to maintain):
```ruby
assert_equal(' ', @filters.remove("a a a a", 'a'))
assert_equal(' ', @filters.remove("1 1 1 1", 1))
assert_equal('b a a', @filters.remove_first("a b a a", 'a '))
assert_equ... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index e7b34c164..694d24d77 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -539,19 +539,31 @@ def test_first_last
end
def test_replace
- assert_equal('2 2 2... | [
"test/integration/standard_filter_test.rb"
] | [
{
"comment": "And for this (better organize `assert_template_result` as a copy of `assert_equal` below it β easier to maintain):\r\n```ruby\r\n assert_equal(' ', @filters.remove(\"a a a a\", 'a'))\r\n assert_equal(' ', @filters.remove(\"1 1 1 1\", 1))\r\n assert_equal('b a a', @filters.remove_first... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | This seems quite inefficient. Couldn't this be done more efficiently with [String#rindex](https://ruby-doc.org/core-3.0.2/String.html#method-i-rindex) to search for the substring to remove?
It also looks like `remove_last` could delegate to `replace_last` with an empty replacement string. Seems like that simplific... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/lib/liquid/standardfilters.rb b/lib/liquid/standardfilters.rb
index 98e38f5f4..f361c02bf 100644
--- a/lib/liquid/standardfilters.rb
+++ b/lib/liquid/standardfilters.rb
@@ -308,14 +308,34 @@ def replace_first(input, string, replacement = '')
input.to_s.sub(string.to_s, replacement.to_s)
end
+ ... | [
"lib/liquid/standardfilters.rb"
] | [
{
"comment": "This seems quite inefficient. Couldn't this be done more efficiently with [String#rindex](https://ruby-doc.org/core-3.0.2/String.html#method-i-rindex) to search for the substring to remove?\r\n\r\nIt also looks like `remove_last` could delegate to `replace_last` with an empty replacement string. ... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | The replacement string default is kind of strange, considering that `remove_last` should actually be used in that case. Perhaps we should deprecate using the default argument with `replace_first` and replace that usage with `remove_first` in https://github.com/shopify/theme-check in which case it might not even make s... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/lib/liquid/standardfilters.rb b/lib/liquid/standardfilters.rb
index 98e38f5f4..f361c02bf 100644
--- a/lib/liquid/standardfilters.rb
+++ b/lib/liquid/standardfilters.rb
@@ -308,14 +308,34 @@ def replace_first(input, string, replacement = '')
input.to_s.sub(string.to_s, replacement.to_s)
end
+ ... | [
"lib/liquid/standardfilters.rb"
] | [
{
"comment": "The replacement string default is kind of strange, considering that `remove_last` should actually be used in that case. Perhaps we should deprecate using the default argument with `replace_first` and replace that usage with `remove_first` in https://github.com/shopify/theme-check in which case it... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | > Updated to using Regex. It seems like a better option.
Why? Compiling a regex on each call will certainly be slower.
The following benchmark shows it being about an order of magnitude slower for a simple string and replacement
```ruby
require 'benchmark/ips'
input = "hello world"
string = "o"
replacem... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/lib/liquid/standardfilters.rb b/lib/liquid/standardfilters.rb
index 98e38f5f4..f361c02bf 100644
--- a/lib/liquid/standardfilters.rb
+++ b/lib/liquid/standardfilters.rb
@@ -308,14 +308,34 @@ def replace_first(input, string, replacement = '')
input.to_s.sub(string.to_s, replacement.to_s)
end
+ ... | [
"lib/liquid/standardfilters.rb"
] | [
{
"comment": "> Updated to using Regex. It seems like a better option.\r\n\r\nWhy? Compiling a regex on each call will certainly be slower.\r\n\r\nThe following benchmark shows it being about an order of magnitude slower for a simple string and replacement\r\n\r\n```ruby\r\nrequire 'benchmark/ips'\r\n\r\ninput... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | The replace filters (replace, replace_first and replace_last) already coerce the input to a string, so we don't need to do it again on these remove filters
```suggestion
replace(input, string, '')
``` | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/lib/liquid/standardfilters.rb b/lib/liquid/standardfilters.rb
index 98e38f5f4..f361c02bf 100644
--- a/lib/liquid/standardfilters.rb
+++ b/lib/liquid/standardfilters.rb
@@ -308,14 +308,34 @@ def replace_first(input, string, replacement = '')
input.to_s.sub(string.to_s, replacement.to_s)
end
+ ... | [
"lib/liquid/standardfilters.rb"
] | [
{
"comment": "The replace filters (replace, replace_first and replace_last) already coerce the input to a string, so we don't need to do it again on these remove filters\r\n\r\n```suggestion\r\n replace(input, string, '')\r\n```",
"path": "lib/liquid/standardfilters.rb",
"hunk": "@@ -296,14 +296,24... | true | ||
Shopify/liquid | 1,422 | comment_to_fix | Add replace_last and remove_last filters | This is multiplying the number of times the filter is called with the same number of arguments when it takes less than 3 arguments, making the test significantly slower. Now the test is taking more than a second. I opened https://github.com/Shopify/liquid/pull/1476 to help speed up this test, generalize it and also h... | 0d5e01ae9825f8ae237fbfcf88e53ab05799ab32 | 63ae4cc2cc09518f17c67ce3f96c3cc5b6b11942 | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index e7b34c164..694d24d77 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -539,19 +539,31 @@ def test_first_last
end
def test_replace
- assert_equal('2 2 2... | [
"test/integration/standard_filter_test.rb"
] | [
{
"comment": "This is multiplying the number of times the filter is called with the same number of arguments when it takes less than 3 arguments, making the test significantly slower. Now the test is taking more than a second. I opened https://github.com/Shopify/liquid/pull/1476 to help speed up this test, ge... | true | ||
Shopify/liquid | 2,050 | issue_to_patch | Add squish filter | Hello,
This PR provides a clean solution to https://github.com/Shopify/liquid/issues/2040
This PR answers a common use case we're facing with Shopify themes. As themes now become more complex, it is often needed to compose classes or inline styles, for instance using capture tags:
```liquid
{% capture styles ... | d897899f6654c476e58e884bc8e24924600e5801 | d589c516974058d84ce0e17ab51cd37ae91bfa21 | diff --git a/lib/liquid/standardfilters.rb b/lib/liquid/standardfilters.rb
index 24e135d91..a91462f43 100644
--- a/lib/liquid/standardfilters.rb
+++ b/lib/liquid/standardfilters.rb
@@ -293,6 +293,19 @@ def split(input, pattern)
input.split(pattern)
end
+ # @liquid_public_docs
+ # @liquid_type filter... | [
"lib/liquid/standardfilters.rb",
"test/integration/standard_filter_test.rb"
] | [] | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index ef5f15493..b69bdf7af 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -164,6 +164,13 @@ def test_split
assert_equal(['A', 'Z'], @filters.split('A1Z', 1))
... | true | |
Shopify/liquid | 2,079 | issue_to_patch | Update liquid-spec adapters for `missing_features` | Updates our liquid-spec adapters for [Shopify/liquid-spec#134](https://github.com/Shopify/liquid-spec/pull/134), which inverted feature gating from adapter allowlists to explicit `missing_features` denylists.
A few follow-up fixes were needed once the suite started running more specs:
- `rake spec` now runs each adap... | 6d81b1b68cf576cc9643ada149c7aa06c25304b8 | 9ff501f9193a7a4f0ba5becbf38c725e205c8021 | diff --git a/.ruby-version b/.ruby-version
index 47b322c97..4d54daddb 100644
--- a/.ruby-version
+++ b/.ruby-version
@@ -1,1 +1,1 @@
-3.4.1
+4.0.2
diff --git a/Rakefile b/Rakefile
index 650923bf8..e14d49aa6 100755
--- a/Rakefile
+++ b/Rakefile
@@ -151,6 +151,8 @@ end
desc('run liquid-spec suite across all adapters')... | [
".ruby-version",
"Rakefile",
"spec/ruby_liquid.rb",
"spec/ruby_liquid_lax.rb",
"spec/ruby_liquid_with_active_support.rb",
"spec/ruby_liquid_yjit.rb",
"spec/support/liquid_spec_adapter_helper.rb"
] | [] | diff --git a/spec/ruby_liquid.rb b/spec/ruby_liquid.rb
index 628fbbfe3..c78049776 100644
--- a/spec/ruby_liquid.rb
+++ b/spec/ruby_liquid.rb
@@ -6,14 +6,24 @@
$LOAD_PATH.unshift(File.expand_path('../lib', __dir__))
require 'liquid'
+require_relative 'support/liquid_spec_adapter_helper'
LiquidSpec.configure do |c... | true | |
Shopify/liquid | 2,077 | issue_to_patch | Remove "strict2" from bare bracket error message | ## Summary
- Removes the `strict2 mode` qualifier from the bare bracket access error message in `lib/liquid/parser.rb` so the error reads generically regardless of which parser surfaced it.
## Test plan
- [ ] Existing tests (e.g. `test_assign_rejects_brackets_in_variable_name_in_strict2`, `test_tablerow_with_bracketed... | 95ce7e7fa1d4ae80cfd650c4e066d85508190905 | dfddd8f390038402fbf5ddf133eaf81375a7ab1a | diff --git a/lib/liquid/parser.rb b/lib/liquid/parser.rb
index c2e2c5d28..7a2136052 100644
--- a/lib/liquid/parser.rb
+++ b/lib/liquid/parser.rb
@@ -55,7 +55,7 @@ def expression
str << variable_lookups
when :open_square
if @reject_bare_brackets
- raise SyntaxError, "Bare bracket access... | [
"lib/liquid/parser.rb"
] | [] | true | ||
Shopify/liquid | 2,067 | issue_to_patch | Add strict2_parse to increment and decrement tags | > [!NOTE]
> Stacked on [Shopify/liquid#2065](https://github.com/Shopify/liquid/pull/2065) (assign/capture strict2_parse)
## Summary
Adds `strict2_parse` to `Increment` and `Decrement` tags. Both tags previously accepted any string as a variable name via `markup.strip` β no Parser validation was performed in any mode.... | 197d755e0ccf5ed216636f95f45f5725857d8e75 | c99036046eba682c9ce50afe37ba639279df84f0 | diff --git a/lib/liquid/tags/decrement.rb b/lib/liquid/tags/decrement.rb
index 8dc620740..70661716d 100644
--- a/lib/liquid/tags/decrement.rb
+++ b/lib/liquid/tags/decrement.rb
@@ -23,13 +23,29 @@ module Liquid
# {% decrement variable_name %}
# @liquid_syntax_keyword variable_name The name of the variable being... | [
"lib/liquid/tags/decrement.rb",
"lib/liquid/tags/increment.rb",
"test/integration/tags/increment_tag_test.rb",
"test/integration/tags/render_tag_test.rb"
] | [
{
"comment": "i thought of merging the code between increment and decrement since they share so much, but didn't see it with other similar tags like `case` and `if` so i just kept it verbose",
"path": "lib/liquid/tags/increment.rb",
"hunk": "@@ -23,13 +23,32 @@ module Liquid\n # {% increment variabl... | diff --git a/test/integration/tags/increment_tag_test.rb b/test/integration/tags/increment_tag_test.rb
index 1ad94254e..bb7245ed5 100644
--- a/test/integration/tags/increment_tag_test.rb
+++ b/test/integration/tags/increment_tag_test.rb
@@ -27,4 +27,50 @@ def test_dec
'{%decrement starboard %}',
)
end
+
... | true | |
Shopify/liquid | 2,065 | issue_to_patch | Add strict2_parse to assign and capture tags | ## Summary
Adds `strict2_parse` to the `assign` and `capture` tags so invalid variable names are rejected in strict2 mode.
Both tags previously used only regex (`VariableSignature`) to validate variable names across all parse modes. This allowed invalid identifiers like `(a(b(c)`, `[x.y]`, and `a.b` to be silently ac... | d0c5444db14178efc15b7ec3dc538ab2683da859 | 0d5c15a03e847b139ab8a696486d6e657390a424 | diff --git a/lib/liquid/tags/assign.rb b/lib/liquid/tags/assign.rb
index 4743ab0b8..901db40f9 100644
--- a/lib/liquid/tags/assign.rb
+++ b/lib/liquid/tags/assign.rb
@@ -18,6 +18,8 @@ module Liquid
# @liquid_syntax_keyword variable_name The name of the variable being created.
# @liquid_syntax_keyword value The val... | [
"lib/liquid/tags/assign.rb",
"lib/liquid/tags/capture.rb",
"lib/liquid/tags/include.rb",
"test/integration/assign_test.rb",
"test/integration/capture_test.rb",
"test/integration/tags/cycle_tag_test.rb",
"test/integration/tags/include_tag_test.rb"
] | [
{
"comment": "I don't think you need the include here because `Tag` already has it.",
"path": "lib/liquid/tags/assign.rb",
"hunk": "@@ -18,6 +18,8 @@ module Liquid\n # @liquid_syntax_keyword variable_name The name of the variable being created.\n # @liquid_syntax_keyword value The value you want to ... | diff --git a/test/integration/assign_test.rb b/test/integration/assign_test.rb
index fdb6c99ca..2388b357d 100644
--- a/test/integration/assign_test.rb
+++ b/test/integration/assign_test.rb
@@ -97,6 +97,46 @@ def test_assign_score_of_hash
assert_equal(12, assign_score_of('int' => 123, 'str' => 'abcd'))
end
+ ... | true | |
Shopify/liquid | 2,064 | issue_to_patch | Add for_loop? to Include tag for AST-based keyword detection | ## Summary
Stores `@is_for_loop` during parsing in the `Include` tag, matching the existing pattern in the `Render` tag. This allows consumers (like the liquid-rewriter) to determine the `with`/`for` keyword from the parsed AST instead of re-parsing raw markup with regex.
**Changes:**
- Added `FOR = 'for'` constant
-... | 532b4390634b111b0cae447347d7a6afd04b8d75 | 346166b600f08856250534faa3693742332b0349 | diff --git a/lib/liquid/tags/include.rb b/lib/liquid/tags/include.rb
index 969482d49..f0161c6e9 100644
--- a/lib/liquid/tags/include.rb
+++ b/lib/liquid/tags/include.rb
@@ -20,7 +20,8 @@ module Liquid
class Include < Tag
prepend Tag::Disableable
- SYNTAX = /(#{QuotedFragment}+)(\s+(?:with|for)\s+(#{QuotedF... | [
"lib/liquid/tags/include.rb",
"test/integration/tags/include_tag_test.rb"
] | [
{
"comment": "Looks good but let's remove that comment. IDK why claude loves these so much. I struggle with em too. ",
"path": "test/integration/tags/include_tag_test.rb",
"hunk": "@@ -439,4 +439,50 @@ def test_include_attribute_with_invalid_expression\n assert_match(/Unexpected character =/, erro... | diff --git a/test/integration/tags/include_tag_test.rb b/test/integration/tags/include_tag_test.rb
index 44c0dcda7..bd4085f21 100644
--- a/test/integration/tags/include_tag_test.rb
+++ b/test/integration/tags/include_tag_test.rb
@@ -439,4 +439,49 @@ def test_include_attribute_with_invalid_expression
assert_match... | true | |
Shopify/liquid | 2,064 | comment_to_fix | Add for_loop? to Include tag for AST-based keyword detection | nit: I kind of like how `render` uses `FOR` inside of `SYNTAX`: https://github.com/Shopify/liquid/blob/main/lib/liquid/tags/render.rb#L29-L30 | 532b4390634b111b0cae447347d7a6afd04b8d75 | 346166b600f08856250534faa3693742332b0349 | diff --git a/lib/liquid/tags/include.rb b/lib/liquid/tags/include.rb
index 969482d49..f0161c6e9 100644
--- a/lib/liquid/tags/include.rb
+++ b/lib/liquid/tags/include.rb
@@ -20,7 +20,8 @@ module Liquid
class Include < Tag
prepend Tag::Disableable
- SYNTAX = /(#{QuotedFragment}+)(\s+(?:with|for)\s+(#{QuotedF... | [
"lib/liquid/tags/include.rb"
] | [
{
"comment": "nit: I kind of like how `render` uses `FOR` inside of `SYNTAX`: https://github.com/Shopify/liquid/blob/main/lib/liquid/tags/render.rb#L29-L30",
"path": "lib/liquid/tags/include.rb",
"hunk": "@@ -20,8 +20,9 @@ module Liquid\n class Include < Tag\n prepend Tag::Disableable\n \n- SYN... | true | ||
Shopify/liquid | 2,060 | issue_to_patch | Reject bare-bracket syntax in strict2 and introduce `self` keyword | ## Summary
Adds bare-bracket rejection to the strict2 parser and introduces a `self` keyword for explicit variable lookups.
**Parser changes:**
- `Parser#expression` now raises a `SyntaxError` when it encounters bare-bracket access (`['product']`) in strict2/rigid mode
- `ParseContext#new_parser` passes `reject_bare_... | dd37353cca46f81c7708847708086ab91eaf382d | 532b4390634b111b0cae447347d7a6afd04b8d75 | diff --git a/lib/liquid.rb b/lib/liquid.rb
index 4d0a71a64..dce089772 100644
--- a/lib/liquid.rb
+++ b/lib/liquid.rb
@@ -65,6 +65,7 @@ module Liquid
require 'liquid/parser'
require 'liquid/i18n'
require 'liquid/drop'
+require 'liquid/self_drop'
require 'liquid/tablerowloop_drop'
require 'liquid/forloop_drop'
requ... | [
"lib/liquid.rb",
"lib/liquid/context.rb",
"lib/liquid/expression.rb",
"lib/liquid/parse_context.rb",
"lib/liquid/parser.rb",
"lib/liquid/self_drop.rb",
"lib/liquid/variable.rb",
"test/integration/context_test.rb",
"test/integration/variable_test.rb"
] | [
{
"comment": "<img width=\"499\" height=\"238\" alt=\"Image\" src=\"https://github.com/user-attachments/assets/d510613a-cc3e-40cf-8894-49829b73d624\" />",
"path": "lib/liquid/self_drop.rb",
"hunk": "@@ -0,0 +1,38 @@\n+# frozen_string_literal: true\n+\n+module Liquid\n+ # @liquid_public_docs\n+ # @liqu... | diff --git a/test/integration/context_test.rb b/test/integration/context_test.rb
index d230734f6..8fa83ddb5 100644
--- a/test/integration/context_test.rb
+++ b/test/integration/context_test.rb
@@ -296,8 +296,8 @@ def test_access_hashes_with_hash_notation
end
def test_access_variable_with_hash_notation
- asse... | true | |
Shopify/liquid | 2,060 | comment_to_fix | Reject bare-bracket syntax in strict2 and introduce `self` keyword | <img width="499" height="238" alt="Image" src="https://github.com/user-attachments/assets/d510613a-cc3e-40cf-8894-49829b73d624" /> | dd37353cca46f81c7708847708086ab91eaf382d | 532b4390634b111b0cae447347d7a6afd04b8d75 | diff --git a/lib/liquid/self_drop.rb b/lib/liquid/self_drop.rb
new file mode 100644
index 000000000..357814653
--- /dev/null
+++ b/lib/liquid/self_drop.rb
@@ -0,0 +1,38 @@
+# frozen_string_literal: true
+
+module Liquid
+ # @liquid_public_docs
+ # @liquid_type object
+ # @liquid_name self
+ # @liquid_summary
+ # ... | [
"lib/liquid/self_drop.rb"
] | [
{
"comment": "<img width=\"499\" height=\"238\" alt=\"Image\" src=\"https://github.com/user-attachments/assets/d510613a-cc3e-40cf-8894-49829b73d624\" />",
"path": "lib/liquid/self_drop.rb",
"hunk": "@@ -0,0 +1,38 @@\n+# frozen_string_literal: true\n+\n+module Liquid\n+ # @liquid_public_docs\n+ # @liqu... | true | ||
Shopify/liquid | 2,060 | comment_to_fix | Reject bare-bracket syntax in strict2 and introduce `self` keyword | Would love to see some additional tests for the new logic. | dd37353cca46f81c7708847708086ab91eaf382d | 532b4390634b111b0cae447347d7a6afd04b8d75 | diff --git a/lib/liquid/context.rb b/lib/liquid/context.rb
index 433b6d003..7902ebcff 100644
--- a/lib/liquid/context.rb
+++ b/lib/liquid/context.rb
@@ -187,6 +187,15 @@ def key?(key)
find_variable(key, raise_on_not_found: false) != nil
end
+ # Checks whether a variable is defined in any scope, includi... | [
"lib/liquid/context.rb"
] | [
{
"comment": "Would love to see some additional tests for the new logic.",
"path": "lib/liquid/context.rb",
"hunk": "",
"resolving_sha": "532b4390634b111b0cae447347d7a6afd04b8d75",
"resolving_diff": "diff --git a/lib/liquid/context.rb b/lib/liquid/context.rb\nindex 433b6d003..7902ebcff 100644\n-... | true | ||
Shopify/liquid | 2,066 | issue_to_patch | Skip slow test raising many exceptions on non-CRuby | fd68d076ddc9d9c65cabbccc6dd5a1b4b9ab0af3 | 9f4d7e78b880f2e72d01cbe610ea5b99d7dd6db8 | [
"test/integration/standard_filter_test.rb"
] | [] | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index fc9873e7b..94097ae1f 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -1181,6 +1181,8 @@ def test_where_array_of_only_unindexable_values
end
def test_all_... | true | |||
Shopify/liquid | 2,038 | issue_to_patch | Add TruffleRuby in CI | * Adapt slice filter to use the correct Array/String index range on TruffleRuby
* Add TruffleRuby in CI
Passing CI run: https://github.com/eregon/liquid/actions/runs/23550121288/job/68561453232 | dd37353cca46f81c7708847708086ab91eaf382d | 96aa47d13f19b5117385d31f12136fd9d409b39a | diff --git a/.github/workflows/liquid.yml b/.github/workflows/liquid.yml
index 3a3687d75..4389d2a44 100644
--- a/.github/workflows/liquid.yml
+++ b/.github/workflows/liquid.yml
@@ -22,6 +22,7 @@ jobs:
}
- { ruby: 4.0, allowed-failure: false, rubyopt: "--yjit" }
- { ruby: 4.0, allowed-... | [
".github/workflows/liquid.yml",
"lib/liquid/standardfilters.rb",
"test/integration/security_test.rb"
] | [] | diff --git a/test/integration/security_test.rb b/test/integration/security_test.rb
index f84058940..3a35d47cc 100644
--- a/test/integration/security_test.rb
+++ b/test/integration/security_test.rb
@@ -44,32 +44,39 @@ def test_no_instance_eval_later_in_chain
end
def test_does_not_permanently_add_filters_to_symbo... | true | |
Shopify/liquid | 2,062 | issue_to_patch | Update ruby/setup-ruby to v1.295.0 for ubuntu-24.04 support | ## Summary
- Updates `ruby/setup-ruby` from v1.273.0 to v1.295.0 in the CI workflow
## Why
GitHub Actions recently rolled `ubuntu-latest` from 22.04 to 24.04. The pinned `ruby/setup-ruby@v1.273.0` does not have prebuilt `ruby-head` binaries for `ubuntu-24.04`, causing CI to fail with:
> Error: Unavailable version h... | 59d8d0d22db3b4865deecde0ab2683bea25a1e9f | e80f775f898e8a6e928d9e78caa951d3c438bd67 | diff --git a/.github/workflows/liquid.yml b/.github/workflows/liquid.yml
index c018a4954..3a3687d75 100644
--- a/.github/workflows/liquid.yml
+++ b/.github/workflows/liquid.yml
@@ -35,7 +35,7 @@ jobs:
name: Test Ruby ${{ matrix.entry.ruby }} ${{ matrix.entry.rubyopt }} --${{ matrix.entry.allowed-failure && 'allowe... | [
".github/workflows/liquid.yml"
] | [] | true | ||
Shopify/liquid | 2,058 | issue_to_patch | Add cumulative resource score tracking across partial renders | ## Situation
`Template#render` unconditionally calls `resource_limits.reset` before rendering, which zeros `render_score` and `assign_score`. Since both `{% render %}` and `{% include %}` go through `Template#render` for each partial, the per-template scores reset on every partial invocation β making `render_score_l... | 5fa36267aa1f97ab1f87971ef948bc6730c42e58 | a911767c75c024299adf551d56b79eb4065f98e4 | diff --git a/lib/liquid/resource_limits.rb b/lib/liquid/resource_limits.rb
index 70fac24be..ee0c66cbb 100644
--- a/lib/liquid/resource_limits.rb
+++ b/lib/liquid/resource_limits.rb
@@ -2,24 +2,39 @@
module Liquid
class ResourceLimits
- attr_accessor :render_length_limit, :render_score_limit, :assign_score_limi... | [
"lib/liquid/resource_limits.rb",
"lib/liquid/version.rb",
"test/integration/template_test.rb",
"test/unit/resource_limits_unit_test.rb"
] | [
{
"comment": "is this necessary? why? ",
"path": "lib/liquid/block_body.rb",
"hunk": "@@ -232,6 +232,7 @@ def render_to_output_buffer(context, output)\n break if context.interrupt? # might have happened in a for-block\n end\n idx += 1\n+ break if context.resource_limits.... | diff --git a/test/integration/template_test.rb b/test/integration/template_test.rb
index c01c7161c..1ba4bafd7 100644
--- a/test/integration/template_test.rb
+++ b/test/integration/template_test.rb
@@ -179,6 +179,86 @@ def test_render_length_uses_number_of_bytes_not_characters
assert_equal("γγγ", t.render)
end
... | true | |
Shopify/liquid | 2,054 | issue_to_patch | Fix rubocop offenses in test file | Follow up to https://github.com/Shopify/liquid/pull/2050
Ran `bundle exec rubocop -a` | 3e76244cd28901cf2124cfa5974c58205d84bbef | a72b604680288bfbe82ccb782351e666bb8922e1 | [
"test/integration/standard_filter_test.rb"
] | [] | diff --git a/test/integration/standard_filter_test.rb b/test/integration/standard_filter_test.rb
index b69bdf7af..fc9873e7b 100644
--- a/test/integration/standard_filter_test.rb
+++ b/test/integration/standard_filter_test.rb
@@ -165,7 +165,7 @@ def test_split
end
def test_squish_filter
- assert_equal("foo ba... | true | ||
Shopify/liquid | 2,001 | issue_to_patch | Introduce Inline Snippets tag | This PR implements [RFC#1916](https://github.com/Shopify/liquid/discussions/1916) and introduces support for the new inline snippets `{% snippet %}` tag, which allows us to define reusable template components directly within Liquid files
Traditionally, snippets are stored as separate `.liquid` files in a `snippets/`... | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid.rb b/lib/liquid.rb
index 4d0a71a64..09cafe947 100644
--- a/lib/liquid.rb
+++ b/lib/liquid.rb
@@ -67,6 +67,7 @@ module Liquid
require 'liquid/drop'
require 'liquid/tablerowloop_drop'
require 'liquid/forloop_drop'
+require 'liquid/snippet_drop'
require 'liquid/extensions'
require 'liquid/erro... | [
"lib/liquid.rb",
"lib/liquid/locales/en.yml",
"lib/liquid/snippet_drop.rb",
"lib/liquid/tags.rb",
"lib/liquid/tags/render.rb",
"lib/liquid/tags/snippet.rb",
"test/integration/tags/render_tag_test.rb",
"test/integration/tags/snippet_test.rb"
] | [
{
"comment": "I believe we may remove this line now. Am I missing something?\r\n```suggestion\r\n```",
"path": "lib/liquid/tags/snippet.rb",
"hunk": "@@ -0,0 +1,39 @@\n+# frozen_string_literal: true\n+\n+module Liquid\n+ # @liquid_type tag\n+ # @liquid_category theme\n+ # @liquid_name snippet\n+ # @... | diff --git a/test/integration/tags/render_tag_test.rb b/test/integration/tags/render_tag_test.rb
index d6453fb51..6fe19abb4 100644
--- a/test/integration/tags/render_tag_test.rb
+++ b/test/integration/tags/render_tag_test.rb
@@ -101,10 +101,6 @@ def test_sub_contexts_count_towards_the_same_recursion_limit
end
e... | true | |
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | I believe we may remove this line now. Am I missing something?
```suggestion
``` | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/snippet.rb b/lib/liquid/tags/snippet.rb
new file mode 100644
index 000000000..87e2a0ded
--- /dev/null
+++ b/lib/liquid/tags/snippet.rb
@@ -0,0 +1,45 @@
+# frozen_string_literal: true
+
+module Liquid
+ # @liquid_public_docs
+ # @liquid_type tag
+ # @liquid_category variable
+ # @liquid_... | [
"lib/liquid/tags/snippet.rb"
] | [
{
"comment": "I believe we may remove this line now. Am I missing something?\r\n```suggestion\r\n```",
"path": "lib/liquid/tags/snippet.rb",
"hunk": "@@ -0,0 +1,39 @@\n+# frozen_string_literal: true\n+\n+module Liquid\n+ # @liquid_type tag\n+ # @liquid_category theme\n+ # @liquid_name snippet\n+ # @... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | I think we actually could use **var** (just to be consistent with the capture tag, I'd prefer **variable**, but being consistent here sounds like a better idea) :)
```suggestion
snippet: "Syntax Error in 'snippet' - Valid syntax: snippet [var]"
``` | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/locales/en.yml b/lib/liquid/locales/en.yml
index b99d490c8..7a9fb71e4 100644
--- a/lib/liquid/locales/en.yml
+++ b/lib/liquid/locales/en.yml
@@ -5,6 +5,7 @@
block_tag_unexpected_args: "Syntax Error in '%{tag}' - Valid syntax: {% %{tag} %}{% end%{tag} %}"
assign: "Syntax Error in 'a... | [
"lib/liquid/locales/en.yml"
] | [
{
"comment": "I think we actually could use **var** (just to be consistent with the capture tag, I'd prefer **variable**, but being consistent here sounds like a better idea) :)\r\n\r\n```suggestion\r\n snippet: \"Syntax Error in 'snippet' - Valid syntax: snippet [var]\"\r\n```",
"path": "lib/liquid/lo... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | I believe we could use `variable` to be consistent with the `capture` tag (also thinking it created a `SnippetDrop`)
```suggestion
# @liquid_category variable
``` | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/snippet.rb b/lib/liquid/tags/snippet.rb
new file mode 100644
index 000000000..87e2a0ded
--- /dev/null
+++ b/lib/liquid/tags/snippet.rb
@@ -0,0 +1,45 @@
+# frozen_string_literal: true
+
+module Liquid
+ # @liquid_public_docs
+ # @liquid_type tag
+ # @liquid_category variable
+ # @liquid_... | [
"lib/liquid/tags/snippet.rb"
] | [
{
"comment": "I believe we could use `variable` to be consistent with the `capture` tag (also thinking it created a `SnippetDrop`)\r\n\r\n```suggestion\r\n # @liquid_category variable\r\n```",
"path": "lib/liquid/tags/snippet.rb",
"hunk": "@@ -0,0 +1,39 @@\n+# frozen_string_literal: true\n+\n+module Li... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | I believe we could have something like a [`with_snippet`](https://github.com/Shopify/liquid/blob/1c1e7119065b0fa915149b9e2bdc303490a0dfff/lib/liquid/tags/capture.rb#L35) method here. This way, we make sure the assignments for inline snippets will respect the [resource limits](https://github.com/Shopify/liquid/blob/1c1e... | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/snippet.rb b/lib/liquid/tags/snippet.rb
new file mode 100644
index 000000000..87e2a0ded
--- /dev/null
+++ b/lib/liquid/tags/snippet.rb
@@ -0,0 +1,45 @@
+# frozen_string_literal: true
+
+module Liquid
+ # @liquid_public_docs
+ # @liquid_type tag
+ # @liquid_category variable
+ # @liquid_... | [
"lib/liquid/tags/snippet.rb"
] | [
{
"comment": "I believe we could have something like a [`with_snippet`](https://github.com/Shopify/liquid/blob/1c1e7119065b0fa915149b9e2bdc303490a0dfff/lib/liquid/tags/capture.rb#L35) method here. This way, we make sure the assignments for inline snippets will respect the [resource limits](https://github.com/Sh... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | Great stuff with this error -- really appreciate the care π | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/render.rb b/lib/liquid/tags/render.rb
index 4d29e420e..dec725ed4 100644
--- a/lib/liquid/tags/render.rb
+++ b/lib/liquid/tags/render.rb
@@ -27,7 +27,7 @@ module Liquid
# @liquid_syntax_keyword filename The name of the snippet to render, without the `.liquid` extension.
class Render < ... | [
"lib/liquid/tags/render.rb"
] | [
{
"comment": "Great stuff with this error -- really appreciate the care π ",
"path": "lib/liquid/tags/render.rb",
"hunk": "@@ -47,28 +47,48 @@ def render_to_output_buffer(context, output)\n end\n \n def render_tag(context, output)\n- # The expression should be a String literal, which parse... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | π @charlespwd @dejmedus, I believe that we actually want to fail when there's no string neither variable name, so I believe we actually want something like this, correct?
```suggestion
return p.consume(:string) if p.look(:string)
return p.consume(:id) if p.look(:id)
p.consume(:end_of_string)
... | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/render.rb b/lib/liquid/tags/render.rb
index 4d29e420e..dec725ed4 100644
--- a/lib/liquid/tags/render.rb
+++ b/lib/liquid/tags/render.rb
@@ -27,7 +27,7 @@ module Liquid
# @liquid_syntax_keyword filename The name of the snippet to render, without the `.liquid` extension.
class Render < ... | [
"lib/liquid/tags/render.rb"
] | [
{
"comment": "π @charlespwd @dejmedus, I believe that we actually want to fail when there's no string neither variable name, so I believe we actually want something like this, correct?\r\n\r\n```suggestion\r\n return p.consume(:string) if p.look(:string)\r\n return p.consume(:id) if p.look(:id)\r\n\r... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | I believe we might be a bit more strict in this regex to prevent the scenario below:
```ruby
require 'liquid'
source = <<~EOS
{% assign world = ', world!' %}
{% snippet hello %}
Hello{{ world }}
{% endsnippet %}
{% render hello, "..." %}
EOS
template = Liquid::Template.parse(source, erro... | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/render.rb b/lib/liquid/tags/render.rb
index 4d29e420e..dec725ed4 100644
--- a/lib/liquid/tags/render.rb
+++ b/lib/liquid/tags/render.rb
@@ -27,7 +27,7 @@ module Liquid
# @liquid_syntax_keyword filename The name of the snippet to render, without the `.liquid` extension.
class Render < ... | [
"lib/liquid/tags/render.rb"
] | [
{
"comment": "I believe we might be a bit more strict in this regex to prevent the scenario below:\r\n\r\n```ruby\r\nrequire 'liquid'\r\n\r\nsource = <<~EOS\r\n {% assign world = ', world!' %}\r\n\r\n {% snippet hello %}\r\n Hello{{ world }}\r\n {% endsnippet %}\r\n\r\n {% render hello, \"...\" %}\r\nEOS... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | Personally, I'd inline this constant for clarity. | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid.rb b/lib/liquid.rb
index 4d0a71a64..09cafe947 100644
--- a/lib/liquid.rb
+++ b/lib/liquid.rb
@@ -67,6 +67,7 @@ module Liquid
require 'liquid/drop'
require 'liquid/tablerowloop_drop'
require 'liquid/forloop_drop'
+require 'liquid/snippet_drop'
require 'liquid/extensions'
require 'liquid/erro... | [
"lib/liquid.rb"
] | [
{
"comment": "Personally, I'd inline this constant for clarity.",
"path": "lib/liquid.rb",
"hunk": "@@ -40,6 +40,7 @@ module Liquid\n QuotedString = /\"[^\"]*\"|'[^']*'/\n QuotedFragment = /#{QuotedString}|(?:[^\\s,\\|'\"]|#{QuotedString})+/o\n TagAttributes ... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | I think we have to be thinking that we might expand this to `...product`. If we do, it would probably mean that we'd implement this differently.
We need some kind of system for "spreading" props of something. Even if right now we're only spreading the props of the global context. | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/render.rb b/lib/liquid/tags/render.rb
index 4d29e420e..dec725ed4 100644
--- a/lib/liquid/tags/render.rb
+++ b/lib/liquid/tags/render.rb
@@ -27,7 +27,7 @@ module Liquid
# @liquid_syntax_keyword filename The name of the snippet to render, without the `.liquid` extension.
class Render < ... | [
"lib/liquid/tags/render.rb"
] | [
{
"comment": "I think we have to be thinking that we might expand this to `...product`. If we do, it would probably mean that we'd implement this differently. \n\nWe need some kind of system for \"spreading\" props of something. Even if right now we're only spreading the props of the global context. ",
"pat... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | I think we could remove this line:
```suggestion
```
(I'm almost sure it doesn't impact YARD pipeline, but I think it would be nice to remove to be 100% sure and also be cohesive with the other tags) π | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/snippet.rb b/lib/liquid/tags/snippet.rb
new file mode 100644
index 000000000..87e2a0ded
--- /dev/null
+++ b/lib/liquid/tags/snippet.rb
@@ -0,0 +1,45 @@
+# frozen_string_literal: true
+
+module Liquid
+ # @liquid_public_docs
+ # @liquid_type tag
+ # @liquid_category variable
+ # @liquid_... | [
"lib/liquid/tags/snippet.rb"
] | [
{
"comment": "I think we could remove this line:\n```suggestion\n```\n\n(I'm almost sure it doesn't impact YARD pipeline, but I think it would be nice to remove to be 100% sure and also be cohesive with the other tags) π ",
"path": "lib/liquid/tags/snippet.rb",
"hunk": "@@ -0,0 +1,45 @@\n+# frozen_stri... | true | ||
Shopify/liquid | 2,001 | comment_to_fix | Introduce Inline Snippets tag | May we we remove the new `p.read` method and call `p.consume` instead? | c357f91e0c677327180c8e745b6e1bf867d65ab2 | c4a6987a307be14a6714375b64c5878bdb246608 | diff --git a/lib/liquid/tags/render.rb b/lib/liquid/tags/render.rb
index 4d29e420e..dec725ed4 100644
--- a/lib/liquid/tags/render.rb
+++ b/lib/liquid/tags/render.rb
@@ -27,7 +27,7 @@ module Liquid
# @liquid_syntax_keyword filename The name of the snippet to render, without the `.liquid` extension.
class Render < ... | [
"lib/liquid/tags/render.rb"
] | [
{
"comment": "May we we remove the new `p.read` method and call `p.consume` instead? ",
"path": "lib/liquid/tags/render.rb",
"hunk": "@@ -101,14 +103,18 @@ def rigid_parse(markup)\n key = p.consume\n p.consume(:colon)\n @attributes[key] = safe_parse_expression(p)\n- p.cons... | true |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.