Skip to content

Fix parse error on quoted scalars beginning with ... or ---#598

Merged
biojppm merged 1 commit into
masterfrom
fix/597_quoted_ellipsis
May 7, 2026
Merged

Fix parse error on quoted scalars beginning with ... or ---#598
biojppm merged 1 commit into
masterfrom
fix/597_quoted_ellipsis

Conversation

@biojppm
Copy link
Copy Markdown
Owner

@biojppm biojppm commented May 7, 2026

Fixes #597

@codecov
Copy link
Copy Markdown

codecov Bot commented May 7, 2026

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 98.73%. Comparing base (b52185e) to head (da5cc7a).

Additional details and impacted files
@@           Coverage Diff           @@
##           master     #598   +/-   ##
=======================================
  Coverage   98.73%   98.73%           
=======================================
  Files          44       44           
  Lines       13799    13801    +2     
=======================================
+ Hits        13624    13626    +2     
  Misses        175      175           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@biojppm biojppm merged commit 5ab774e into master May 7, 2026
266 checks passed
@biojppm biojppm deleted the fix/597_quoted_ellipsis branch May 7, 2026 17:17
Hancock33 added a commit to Hancock33/batocera.piboy that referenced this pull request May 9, 2026
-------------------------------------------------------------------------------------
azahar.mk 929a51afc69dc8e0306fd884fd7b0ea883cf76c9 # Version: Commits on May 08, 2026
-------------------------------------------------------------------------------------
audio: Add option to simulate headphones plugged in (#2099),

-----------------------------------------------------------------------------------
cemu.mk f3fecba313f146d5cf822491f0b3786ad606fceb # Version: Commits on May 08, 2026
-----------------------------------------------------------------------------------
Migrate from SDL2 to SDL3 3.4.2 (#1847),

------------------------------------------------------------------------------------------
dolphin-emu.mk d19952cc113940391ebcf264ced7ddc56310ba75 # Version: Commits on May 08, 2026
------------------------------------------------------------------------------------------
Merge pull request #14634 from JosJuice/jitarm64-msr-dr-bit

JitArm64: Fix DR check in MSRUpdated,

------------------------------------------------------------------------------------------
duckstation.mk 015bfedb4b8116d5483367adfc3d78902081bd6f # Version: Commits on May 08, 2026
------------------------------------------------------------------------------------------
FullscreenUI: Align fade transition times to frame rate,

------------------------------------------------------------------------------------
rpcs3.mk c102d1ec0d486a615faca55807e16e06cdc32925 # Version: Commits on May 08, 2026
------------------------------------------------------------------------------------
Add missing hover effect in ‎GustavoGraziano's theme,

--------------------------------------------------------------------------------------
shadps4.mk 7132795b7a753c5a150dfd410f4fbe3f0568a960 # Version: Commits on May 08, 2026
--------------------------------------------------------------------------------------
Refactor memory size handling to better handle vdec titles (#4373)

* Refactor memory size constants and calculations

Updated memory size handling in video decoder functions.

* Refactor video decoder frame size calculations

Updated frame size computation to use worst-case dimensions and adjusted alignment values.

* Refactor alignment logic to use Common::AlignUp,

--------------------------------------------------------------------------------------
tsugaru.mk 59fbb9fe2a87c2d6566e369838ef64799168047b # Version: Commits on May 07, 2026
--------------------------------------------------------------------------------------
Resolved conflict.,

-------------------------------------------------
vice.mk r46094 # Version: Commits on May 07, 2026
-------------------------------------------------
null

-------------------------------------------------------------------------------------------
xenia-canary.mk 2c68a2fc4bbf360fb79c373664331c03ba70ac31 # Version: Commits on May 08, 2026
-------------------------------------------------------------------------------------------
[XAM] Fixed XamUserReadProfileSettingsEx setting buffer_size_ptr

Fixes Kinect Dashboard from crashing when selecting Switch Profiles.,

-----------------------------------------------------------------------------------------
xenia-edge.mk 17941be044321846d32887f96e74081e1d6817de # Version: Commits on May 08, 2026
-----------------------------------------------------------------------------------------
[Docs] Update build link,

-----------------------------------------------------------------------------------
ymir.mk ec866d7506c09d6a4d855be182e58860fb6e83f8 # Version: Commits on May 08, 2026
-----------------------------------------------------------------------------------
Merge pull request #845 from mmkzer0/feat/headless-dbg

feat(dbg): bootstrap headless and dbg targets,

-------------------------------------------------------------------------------------
ikemen.mk 50eff9faa34bd07e095277500433f81d4f0e2af0 # Version: Commits on May 08, 2026
-------------------------------------------------------------------------------------
Merge pull request #3593 from Alceris/explodhitpause

fix: Fix explod ignorehitpause bugs,

-----------------------------------------------------------------------------------------
lightspark.mk 56071d2699abc8d57b1d0982748c6ef865fcff99 # Version: Commits on May 07, 2026
-----------------------------------------------------------------------------------------
fix several refcounting issues for AVM1 mouse event handlers,

---------------------------------------------------------------------------------------
thextech.mk 67686040dfe193e57700fbd56cea702c39e0c62d # Version: Commits on May 08, 2026
---------------------------------------------------------------------------------------
Translated using Weblate (Chinese (Simplified Han script))

Currently translated at 100.0% (393 of 393 strings)

Translated using Weblate (Chinese (Traditional Han script))

Currently translated at 100.0% (393 of 393 strings)

Co-authored-by: 迎春心情 <lovebodhi@foxmail.com>

Translate-URL: https://hosted.weblate.org/projects/thextech/engine-assets-specific/zh_Hans/

Translate-URL: https://hosted.weblate.org/projects/thextech/engine-assets-specific/zh_Hant/

Translation: TheXTech Engine/Engine Assets-specific,

-------------------------------------------------
ecm.mk v6.26.0 # Version: Commits on Apr 24, 2026
-------------------------------------------------
ECMGeneratePythonBindings.cmake - hide more compiler warnings

Add -Wno-shadow and -Wno-keyword-macro to the target_compile_options

to hide even more compile warnings from the generated code.

For gcc and clang compilers only.,

-------------------------------------------------------
rapidyaml.mk v0.12.1 # Version: Commits on May 07, 2026
-------------------------------------------------------
- Fix [#597](biojppm/rapidyaml#597): parse error when quoted scalars start with `...` or `---` ([PR#598](biojppm/rapidyaml#598)). This fixes a regression introduced in ([PR#587](biojppm/rapidyaml#587)), while trying to ensure a parse error when `...` or `---` occur in quoted scalars at a line begin.

------------------------------------------------------------------------------------
rtw88.mk 274f0d866576c4a8c51ad453a1a32f018c0b16c3 # Version: Commits on Mar 19, 2026
------------------------------------------------------------------------------------
wifi: rtw88: Enable receiving control frames in monitor mode

By default RTL8723D, RTL8703B, RTL8812A, RTL8821A, and RTL8814A are

configured to filter out all control frames except PS-Poll, even in

monitor mode.

Handle FIF_CONTROL in rtw_ops_configure_filter(). When it's set,

configure REG_RXFLTMAP1 to let all control frames through. When it's

unset, restore the original value. Because some drivers configure

REG_RXFLTMAP1 differently, keep track of its value in a new member of

struct rtw_hal.

Signed-off-by: Bitterblue Smith <rtl8821cerfe2@gmail.com>,

-------------------------------------------------------------------------------------------
jazz2-native.mk 460b97759a4b8809f059c55225d7f6757b3c3e82 # Version: Commits on May 08, 2026
-------------------------------------------------------------------------------------------
Fixed build,

-----------------------------------------------------
mergerfs.mk 2.42.0 # Version: Commits on May 08, 2026
-----------------------------------------------------
# mergerfs v2.42.0

## Donations / Sponsorship

If you find mergerfs useful please consider supporting its ongoing development.

https://github.com/trapexit/support

## New features

* `lup` (least used percentage) policy: selects the branch with the lowest used space percentage.

* `mount.mergerfs` now includes the `-n` flag to support mounting without updating /etc/mtab.

## Improvements

* Better lock management and behavior with open files. Reduction of contention.

* Custom filename dedup strategy improving directory reading performance

* Pre-calculate part of the file inode within readdir significantly reducing cost of overall calculation.

* Replace sched_yield with nanosleep for improved scheduling of contention situations.

* improved stat auto cache fingerprinting to reduce stale reads.

## Behavior changes

* credential handling reworked to be compatible with chroot and idmap. removed ability to disable `default_permissions` option as a result as mergerfs now requires the kernel to manage entitlements.

* broken mounts with ENOTCONN (common due to crashed FUSE instances) are automatically unmounted if possible on start instead of hard erroring. (mostly useful while developing).

* `fusermount3` will be used if available and other options missing.

* `remember-nodes` option is deprecated and the feature removed. There is now only `never-forget-nodes` behavior. Any non-zero value in `remember-nodes` will enable `never-forget-nodes`. It was removed because time based really didn't save much memory and complicated the code. Doing this also reduced the size of each node helping offset increased node count.

## Bug fixes

* fix race condition between runtime changing of branches and certain requests.

* `rmdir` no longer returns success if any rmdir call returns ENOTEMPTY.

* Invalid policy names are properly rejected.

* statfs calculation could theoretically have overflowed with impossibly large branch values.

* Fixed `readlink` buffer size management. The size does not include a nul terminating value.

* Minor calculation bugs in `copy_file_range` and `futimens`.

* `moveonenospc` no longer crashes when all destination branches are rejected. Returns ENOSPC.

* Better / more proper error calculations in a number of functions.

* rare ioctl call crash fixed

## Hardening

* many \the kernel should never do this bug **just** in case\ enhancements

## Full Changelog

trapexit/mergerfs@2.41.1...2.42.0-rc2

,

-------------------------------------------------------------------------------------------------
shadps4-qtlauncher.mk e2c0b57ecb3d7557c87dfa56515ff4bfd18c800d # Version: Commits on May 08, 2026
-------------------------------------------------------------------------------------------------
Hotkeys: add screenshot and settings hotkeys to GUI, change layout (#339)

* fix signals

* new UI

* add new hotkeys

* proper element names,

----------------------------------------------------------------------------------------------------
sdl2-gamecontrollerdb.mk f23f0c14ab3ca1926c5ea37082fb15f6ebfaac0a # Version: Commits on May 08, 2026
----------------------------------------------------------------------------------------------------
Update README.md,

----------------------------------------------------------------------------------------
retroarch.mk 39368171391c941ae3baed6d21b87fa3b1a66f36 # Version: Commits on May 08, 2026
----------------------------------------------------------------------------------------
win32: decouple Vulkan WGL context from OpenGL gating

The previous fix (commit 1e45eaf) widened the wnd_proc_wgl_* guard in

win32_common.c with HAVE_VULKAN to silence link errors when Vulkan was

enabled on Win32 without any OpenGL variant. That treated the symptom:

the real bug is that Vulkan does not need wgl_ctx.o at all - it has its

own w_vk_ctx.o with dedicated wnd_proc_vk_common / _dinput / _winraw.

Forcing WANT_WGL=1 from the Vulkan path conflated two independent

context drivers.

Untangle them:

- Makefile.common: under HAVE_VULKAN+Win32, add w_vk_ctx.o directly

  instead of setting WANT_WGL=1. The WANT_WGL block now only adds

  wgl_ctx.o, since w_vk_ctx.o is owned by the Vulkan block.

- Drop LIBS += -lcomctl32 from the WANT_WGL block: no code in the tree

  calls any comctl32 API. ui_win32.c includes <commctrl.h> but uses

  nothing from it, and its #pragma comment(lib, ...) is MSVC-only.

- griffin/griffin.c: drop HAVE_VULKAN from the wgl_ctx.c include guard

  to match the Makefile separation, and add the missing HAVE_OPENGL_CORE

  for consistency with the gl_common.c guard a few lines above.

- Revert the HAVE_VULKAN widening in win32_common.c - with wgl_ctx.o

  no longer compiled in Vulkan-only builds, the wnd_proc_wgl_* symbols

  are never referenced and the original OpenGL-only guard is correct.

Build matrix verified for Win32: OpenGL+Vulkan, OpenGL-only, Vulkan-only,

and neither. Vulkan-only now links cleanly without dragging in WGL

window procs that have no OpenGL context to create.,

-------------------------------------------------------------------------------------
libdof.mk c98a9e44841b56eb2089ea1dffd96c75005b2eff # Version: Commits on May 06, 2026
-------------------------------------------------------------------------------------
misc: update to support Visual Studio 2026,

---------------------------------------------------------------------------------------
vpinball.mk 5499f515df0da6275309eac665f00676ddac778c # Version: Commits on May 07, 2026
---------------------------------------------------------------------------------------
update,

----------------------------------------------------------------------------------------
doomretro.mk 64a84f1b2bd284d123554ed50c03724c30b41d58 # Version: Commits on May 07, 2026
----------------------------------------------------------------------------------------
Update releasenotes.md,

--------------------------------------------------------------------------------------
yquake2.mk debc11892ab6946bf2945fdb88222a5038730ef8 # Version: Commits on May 07, 2026
--------------------------------------------------------------------------------------
renders: replace unsigned char to byte,

----------------------------------------------------------------------------------------------
libretro-azahar.mk 929a51afc69dc8e0306fd884fd7b0ea883cf76c9 # Version: Commits on May 08, 2026
----------------------------------------------------------------------------------------------
audio: Add option to simulate headphones plugged in (#2099),

-------------------------------------------------------------------------------------------------------
libretro-beetle-pce-fast.mk cb4d9d71b9ea63b6546f7d66399d506f74c242b8 # Version: Commits on May 08, 2026
-------------------------------------------------------------------------------------------------------
Fetch translations & Recreate libretro_core_options_intl.h,

--------------------------------------------------------------------------------------------------
libretro-beetle-psx.mk d69917827bff3c8dffda9b794fbf069ae266e349 # Version: Commits on May 08, 2026
--------------------------------------------------------------------------------------------------
Drop 'using namespace std' from shader.cpp

Twelve sites of unqualified std:: name use in shader.cpp:

- Two 'const string &' parameter types (line 96 get_stock_sampler,

  and lines 153, 199 inside the Shader constructor's resource loops)

- Eight 'string::npos' comparisons in the get_stock_sampler dispatch

  (lines 98-112)

- One 'string(\cache://spirv/\)' constructor call (line 125)

- One 'to_string(hash)' call (line 125)

All qualified explicitly as std::string, std::string::npos, and

std::to_string. The std::vector<SpecializationConstant> at line 261

was already qualified.

Removed 'using namespace std' at line 31. The other two 'using

namespace' directives ('using namespace spirv_cross' and 'using

namespace Util' at lines 32-33) stay — they pull in this codebase's

own namespaces and the equivalent Util:: cleanup is a separate

concern.

The string.find(...) calls in get_stock_sampler are member methods

on the parameter (a string reference), not std::find — unaffected.

Same with pipelines.find(hash) at the IntrusiveHashMap call site.

HAVE_VULKAN=1 library size unchanged at 16538360 bytes (purely a

parse-time name resolution change). Both HAVE_VULKAN=1 and plain

builds compile cleanly.

Cleanup progress: 8 of 10 files done. Remaining: renderer.cpp (~47

names), device.cpp (~73). Each will be its own commit.,

--------------------------------------------------------------------------------------------
libretro-dice.mk 9e348796528a590ee88a88c126955c0901a21210 # Version: Commits on May 08, 2026
--------------------------------------------------------------------------------------------
Fetch translations & Recreate libretro_core_options_intl.h,

---------------------------------------------------------------------------------------------------
libretro-dosbox-pure.mk 04a1b953118baed5fa068d769ca0863301e3404f # Version: Commits on May 08, 2026
---------------------------------------------------------------------------------------------------
Allow mounting of ZIP files via IMGMOUNT command

I.e. IMGMOUNT X: \file.zip\ -t zip,

---------------------------------------------------------------------------------------------
libretro-fbneo.mk 18bf8481410a18df20ac73e4bfdd07a5f2ec5e92 # Version: Commits on May 08, 2026
---------------------------------------------------------------------------------------------
(libretro) better fix for subsets & pgm2,

------------------------------------------------------------------------------------------------
libretro-gambatte.mk 897bc1386cff38860b149049f515f1fcd1e85621 # Version: Commits on May 08, 2026
------------------------------------------------------------------------------------------------
Fetch translations & Recreate libretro_core_options_intl.h,

--------------------------------------------------------------------------------------------------
libretro-gearcoleco.mk f336da73f64917a2889b183e7e5025485bcd0e79 # Version: Commits on May 07, 2026
--------------------------------------------------------------------------------------------------
Fix SGM memory mapping initialization,

--------------------------------------------------------------------------------------------
libretro-pcsx.mk 583f8462197f124ac2cdf5c1fd871f4f96a12fe1 # Version: Commits on May 07, 2026
--------------------------------------------------------------------------------------------
Merge branch 'master' (May 8 2026) into libretro,

-------------------------------------------------------------------------------------------
libretro-ps2.mk f1df451134822c70cdc611d06c0b6aa0e5d53be7 # Version: Commits on May 08, 2026
-------------------------------------------------------------------------------------------
MTGS: drop dead vsync-queue + MTVU coordination, force MTVU off

The libretro topology has no VsyncQueueSize-style frame queue and no

working MTVU support. The state and synchronization for both has been

sitting dead-but-paying-overhead-every-frame. Cut it.

What goes:

- VsyncQueueSize config field (always 0 in libretro; no value > 0

  buys parallelism here, the libretro thread can only drain the GS

  ring during retro_run regardless of queue depth).

- s_QueuedFrameCount + s_VsyncSignalListener + s_sem_Vsync. The

  fetch_add/check at PostVsyncStart was 'fetch_add(1) < 0' i.e. always

  false, then a Wait() on a semaphore that was already Posted by the

  preceding WaitGS. All atomic ops + sem ops per frame, all pure

  overhead.

- s_mtx_RingBufferBusy2. Only purpose was MTGS<->MTVU coordination

  (held during MainLoop iteration, busy-waited by WaitGS's MTVU

  branch). With MTVU dead, no contention, no need.

- isMTVU parameter on MTGS::WaitGS. Single use was the now-removed

  MTVU branch; all six call sites passed false anyway except two in

  Gif_Path that passed isMTVU() (always false since vuThread is

  forced off).

What stays:

- The WaitGS() at PostVsyncStart. This *is* the libretro frame-pacing

  mechanism: cpu_thread blocks until the libretro thread (which is

  also the MTGS thread) processes one VSYNC. The dead apparatus

  around it was redundant; the WaitGS itself is load-bearing.

- The GS_RINGTYPE_MTVU_GSPACKET case in MainLoop. With MTVU off, no

  one enqueues this packet type, so it's never executed - but kept

  defensively in the switch with a note for any future MTVU revival.

vuThread is also forced to false at libretro setup. Reasoning is in

the comment at the call site, but briefly: the GameDB only ever

explicitly disables MTVU (~60 entries with comments like 'Fixes

corrupted textures', zero entries enable it); the libretro topology

has no spare thread for VU work; MTVU breaks determinism for

netplay/replay/rewind. Locking it off here prevents stray .ini

overrides or future GameDB entries from silently turning it back on.

This is Stage 1 of dismantling MTVU support. Stage 2 (mechanical

removal of the ~40 'if (THREAD_VU1)' guards in Memory.cpp et al.,

now all 'if (false)') and Stage 3 (rip out vu1Thread, MTVU.cpp/h,

mtvuInterrupts plumbing) are deferred so this change can be

validated in isolation. The recompiler-internal THREAD_VU1 sites

in microVU_*.inl are deferred to last.,

----------------------------------------------------------------------------------------------
libretro-stella.mk 36ac4b1084e7a8dedf02c6d1bbe11ca9ae646f92 # Version: Commits on May 08, 2026
----------------------------------------------------------------------------------------------
Cleaned up NOLINT usage throughout codebase:

 - use specific warning instead of disabling all warnings

 - removed NOLINT where no longer needed

 - fixed code in several places to eliminate NOLINT,

---------------------------------------------------------------------------------------------------
libretro-swanstation.mk f11dfea0a7aaafcb75a705a6f89ec8ed5c4c5e73 # Version: Commits on May 08, 2026
---------------------------------------------------------------------------------------------------
system: honour AllowBootingWithoutSBIFile in the libretro core

CheckForSBIFile detects libcrypt-protected discs that are missing

their SBI sidecar. When the user has opted in via the

AllowBootingWithoutSBIFile=true setting the function was supposed to

boot the game anyway with a warning, but the libretro core has no

synchronous yes/no UI: the implementation routed through

HostInterface::ConfirmMessage, which in this fork is a stub that

unconditionally returns false. The result was that users who

explicitly set the override still got booting denied, and the only

way to actually boot a libcrypt game without SBI was unreachable.

Cut out the ConfirmMessage indirection. When the override is set,

post a 30-second OSD warning explaining that the game will likely

not run correctly and to consult the README about adding an SBI

file, and return true to proceed with boot. The off-by-default

branch is unchanged: report the error and refuse the boot, which

is what users who haven't opted in want to see.

Net effect:

  - Without the override: identical behaviour (refuse + error).

  - With the override: now actually boots, with a long, informative

    OSD warning. Previously: silently failed via the ConfirmMessage

    stub returning false.,

-----------------------------------------------------------------------------------------------
libretro-uae4arm.mk 276979efa4f862d1f84afeff5a2e794de4744024 # Version: Commits on May 08, 2026
-----------------------------------------------------------------------------------------------
Merge pull request #78 Fix ambiguous 'numbers' symbol conflict with std::numbers (GCC 16),

---------------------------------------------------------------------------------------------
libretro-vba-m.mk ab89a5b0bc4af0a76f4e13373fda612c0cece646 # Version: Commits on May 08, 2026
---------------------------------------------------------------------------------------------
translations: transifex pull

Signed-off-by: Rafael Kitover <rkitover@gmail.com>,
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Parsing error since latest release for strings starting with "... {any character}"

2 participants