Commit Graph
1562 Commits
Author SHA1 Message Date
bflatastic 8f12dd262a Take a pasted JSON array as a plugin's arguments (R2.9) 2026-11-19 06:09:04 +07:00
bflatastic 6b2390c4b2 Claim pasted JSON plugin arguments (R2.9) 2026-11-19 06:08:41 +07:00
Dylan M. Taylor cf91402bdd Update craft-tar to storytold/craft-libs main (#2151) 2026-11-18 17:45:02 -05:00
BFlat 7f0c877f8c Merge pull request #2147 from mansur-akhmadi/bugfix/r2.9-stack-steps-literal-values
The build steps filled {editor}, {player} and {ext} with chained replace calls, so an editor
installed under a folder named like a later placeholder reached the builder with a different
path. The one-pass fill moves to platform::text::fill, which both the build steps and i18n::trf
use.
2026-11-19 04:58:24 +07:00
Mansur Akhmadi 4d6e834d11 Keep the steps test to paths only one-pass filling keeps; note the fill in platform's docs 2026-11-19 04:39:02 +07:00
Mansur Akhmadi 58a5989fdf Fill the engine build steps' placeholders in one pass (R2.9) 2026-11-19 04:31:11 +07:00
Mansur Akhmadi 93a8d30137 Claim the one-pass fill for the Unity build steps (R2.9) 2026-11-19 04:20:14 +07:00
BFlat d6a1f888aa Merge pull request #2142 from MacallanTheRoot/bugfix/r5.5-just-ask-builder-notice
Answers to questions now carry the replaced and replacing AI names as an optional
builder_replacement field (saved, used). The desktop fills them into the same catalog entry
Make and build starts use, so the notice after Just ask follows the UI language. The English
builder_notice is unchanged for CLI and MCP callers.
2026-11-19 04:11:01 +07:00
Efe Şahinbaş 479c287661 Localize saved-builder replacement notices after Just ask (R5.5) 2026-11-18 23:37:39 +03:00
Efe Şahinbaş 061b84b082 Claim R5.5 Just ask localization on current main 2026-11-18 23:37:31 +03:00
BFlat 8f693094b5 Merge pull request #2143 from mansur-akhmadi/bugfix/r2.9-trf-literal-values
i18n::trf filled {name} placeholders with one replace per argument, so a value that looked like
a later placeholder was filled in turn. It now fills them in one pass over the translated
template: inserted values stay literal, and unknown names and stray braces are kept as written.
The saved-builder replacement notice uses it instead of its own split.
2026-11-19 03:23:38 +07:00
BFlat ebfbd999be Merge pull request #2144 from MacallanTheRoot/bugfix/r3.11-bounded-managed-sources
The check compared .cs files against the sizes its folder walk recorded, then read each one
whole, so a file that grew in between could be read past the 64 MiB source budget. Each read is
now capped at what is left of the budget. When no file could be read, the error keeps the first
read errors and says when the folder walk was cut short.
2026-11-19 03:14:03 +07:00
Mansur Akhmadi 02ef4dffc2 Keep inserted values literal when filling UI placeholders (R2.9) 2026-11-19 02:40:56 +07:00
macallantheroot 9569dad045 Link the bounded managed-source draft (R3.11) 2026-11-18 22:32:28 +03:00
macallantheroot 60e1b0928f Keep the independent status updates conflict-free (R3.11) 2026-11-18 22:31:41 +03:00
macallantheroot 6a28746956 Bound managed reconstruction source reads to the remaining budget (R3.11) 2026-11-18 22:30:11 +03:00
macallantheroot 4bcc127887 Claim bounded managed source reads (R3.11) 2026-11-18 22:21:36 +03:00
Mansur Akhmadi ad06e5e3ee Claim literal placeholder values in i18n::trf (R2.9) 2026-11-19 02:20:05 +07:00
BFlat 5fee83713f Merge pull request #2140 from MacallanTheRoot/fix/r5.5-localize-builder-replacement-notice
The notice shown on Make and at build starts when a project's saved AI is off or gone now comes
from the UI catalogs. Turkish has a reviewed translation; the other languages keep the English
wording until a native speaker reviews theirs. Builder names are filled in literally, and the
CLI, MCP and headless notice are unchanged.
2026-11-19 01:55:15 +07:00
macallantheroot b277e09438 Localize saved-builder replacement notices in the desktop UI (R5.5) 2026-11-18 21:21:12 +03:00
macallantheroot fb40433a7e Claim saved-builder replacement notice localization (R5.5) 2026-11-18 21:21:12 +03:00
BFlat 5895463ed6 Merge pull request #2136 from mansur-akhmadi/bugfix/r3.11-reconstruction-check-cut-coverage
The check reports partial coverage, with no total, when a .cs file under the source folder may
have been left unread: the file or size budget stopped the reading, a listed file could not be
read, or the folder walk was cut short. Before, members declared only in an unread file read as
missing in a check marked complete.
2026-11-19 01:18:35 +07:00
Mansur Akhmadi 7bd2834d28 Mark the managed reconstruction check partial when it leaves source files unread (R3.11) 2026-11-19 00:46:30 +07:00
Mansur Akhmadi f25f0de04b Claim the managed reconstruction check's cut coverage (R3.11) 2026-11-19 00:15:26 +07:00
FredApps 4932cca8c5 Use merged DEX bounds and completeness repairs (#2130) 2026-11-19 00:06:11 +07:00
BFlat 3b0a7ec1d4 Merge pull request #2133 from AlexJoaquimPereira/bugfix/r3.9-game-files-request-ids
Picking game A, then B, then A again while the first look at A was still running could show the older result, because the screen matched an answer to the folder's path. Each look and each archive listing is now matched to the job that started it, so a repeated pick ignores the older answer, and a job that fails settles only its own request: one a newer pick replaced is logged and says nothing. A folder whose name isn't valid text used to be sent under a changed name and left the screen loading forever; picking one now says its name can't be read and asks for another folder. A failed save of a picture or asset also says so.
2026-11-18 23:47:48 +07:00
FredApps 7e3843633a Keep the legacy MUTF-8 helper available only to parity tests (#2132) 2026-11-18 23:07:15 +07:00
AlexJoaquimPereira 64a388d06c Reject non-text game folders and settle only the failed job (R3.9)
Review follow-up: folder_picked now refuses a folder whose name isn't text before dispatching either tool, with a message, and sends the exact UTF-8 path. Game files keeps one map of in-flight jobs with their kind and whether the screen still waits for them, instead of a looks list and a listing id. A worker failure is always logged; a current folder scan or listing clears only its own spinner and says so, a failed triage holds neither, and a superseded job says nothing. Export save errors say so without stopping the archive listing, even when the archive path matches a reopened listing.

The same-folder regression now starts both looks through folder_picked and feeds results through CraftApp dispatch; new regressions cover the non-text refusal, a failed triage keeping the scan running, a superseded failure staying quiet, and an export error not clearing loading.
2026-11-18 21:26:03 +05:30
AlexJoaquimPereira 54f42f8230 Keep a Game files look with the job that asked for it (R3.9)
Picking game A, B and A again while the first A look ran could still apply the older A result, because the screen compared the shown folder's path. Worse, a folder whose name has bytes that aren't text was sent as a lossy display string, so every completion failed the same path check and the screen loaded forever. act now returns its job id; GameFiles records the folder looks and the archive listing still expected, and tool_done applies only those. A repeated pick ignores the older answer, and a non-text folder's current answer finishes with its error.

Regressions: a stale same-folder look is ignored and the newer one fills the list; a non-text folder's current error clears looking and is shown. The three existing wrong-folder/archive guards still pass.
2026-11-18 21:26:03 +05:30
AlexJoaquimPereira 169092b3cc Claim R3.9: correlate each Game files look with its request
Picking game A, B, then A again while the first A look runs can still apply the older A result, because the screen compares the shown folder's path. Record the pending change before fixing it.
2026-11-18 21:26:03 +05:30
Fury f9fbd77d1b Name the folders a walk couldn't list in every tool that walks one (R3.11) (#2106)
A folder walk says which folders it couldn't list and why, but five tools that walk a folder (two for Apple bundles, the app source import and comparison, the managed source check and the JavaScript source restore) read only "cut short". A folder that couldn't be opened was reported as a limit, with neither its name nor the reason, and some results still said they were complete.

- Each of those tools now names the folders and their reasons, and marks its result partial when a walk was cut or a folder couldn't be listed.
- When nothing is found and a folder couldn't be listed, the error says so ("no .cs files under …; 1 folder(s) couldn't be listed…") with what to do, since what was looked for may be in it.
- Two inputs with the same folder name are told apart in the notes ("dist (input 2)").
2026-11-18 22:53:40 +07:00
Dylan M. Taylor 7023bf8ccc Merge pull request #2131 from storytold/fix/dex-mutf8-dead-code
Gate the modified-UTF-8 helper to test builds
2026-11-18 10:16:15 -05:00
Dylan M. Taylor 2098b6560a Gate the modified-UTF-8 helper to test builds 2026-11-18 10:15:58 -05:00
Dylan M. Taylor 8a192033ed Merge pull request #2032 from storytold/refactor/shared-dex-reader
Use shared DEX reader and Dalvik opcode decoder
2026-11-18 09:46:54 -05:00
Justin John 5e09595d55 Let Settings → AI choose Codex's model (#2129)
Codex now shows the model it uses ("Codex default" until one is chosen) with Change…, like Claude Code (#2109). The choices are its default (no model is passed, so its own settings decide) and the models `codex debug models` lists, in its order, leaving out the ones it hides; over SSH the remote Codex is asked. `ai model Codex` lists and sets them the same way. Until now the model could only be set by adding `cli_model` to config.toml by hand.
2026-11-18 21:35:57 +07:00
BFlat 9c9f5a10d2 Merge pull request #2128 from Plastic-time/bugfix/r3.11-pseudo-write-footprints
Follow-up to #2124. The pseudo-C could still show an argument's load in the call line after an instruction that overwrote the bytes it read, for writes the decompiler measured wrongly: 64-bit ARM pair and exclusive stores (both registers count, and non-temporal pairs are stores too), x86 bit writes through a register index, `movdir64b`, and a store that overwrites part of a Windows stack argument. Those are measured now. A register is treated as the stack frame only while the function really keeps its stack there, so a stack pointer loaded from elsewhere is no longer assumed clear of global data. 64-bit ARM loads also show the size they read (`*(uint8_t*)…`), and a pair store shows as two typed writes.

Measured on two real programs before and after (every function, default limits): an x86-64 `grep` folds 3,021 register arguments where it folded 3,023, the two lost being loads from a global with a store through a data pointer in between; an AArch64 `ls` folds 1,813 either way.
2026-11-18 21:27:16 +07:00
Dylan M. Taylor 2ab3f0ede8 Merge remote-tracking branch 'origin/main' into pr-2032 2026-11-18 09:14:42 -05:00
RRFZ 79c9aaa7f3 Retain scalar conversions and writable pair addresses (R3.11) 2026-11-18 08:53:21 -05:00
RRFZ e2df6bc410 Preserve pair writeback and split outgoing arguments (R3.11) 2026-11-18 08:49:22 -05:00
RRFZ f9f0866214 Keep pair destinations separate from register results and type scalar reads (R3.11) 2026-11-18 08:31:05 -05:00
BFlat b935cb33bf Merge pull request #2123 from Plastic-time/bugfix/r3.12-js-app-file-localhost-dot-folders
Follow-up to #2113. A window loading `file://localhost/ui/index.html` was looked up as `localhost/ui/index.html`, which could match the wrong page; `localhost` and an empty host both mean this computer now. And a page reference ending in a `.` or `..` part (`./boot.js/.`) names a folder, so it is no longer matched to the file before it.
2026-11-18 20:26:30 +07:00
RRFZ bdd284d030 Preserve A64 memory effects and truncated direct stores (R3.11) 2026-11-18 08:23:09 -05:00
RRFZ d1421498e6 Preserve argument loads across overlapping memory writes (R3.11) 2026-11-18 07:57:51 -05:00
RRFZ a39f0bf2f8 Cover terminal parent segments in both HTML reference types (R3.12) 2026-11-18 07:48:05 -05:00
BFlat de20759323 Merge pull request #2125 from Plastic-time/bugfix/r5.7-result-lists-cap-before-index
Follow-up to #2116. A result with thousands of lists at its top level was still gone through in full, though its first 64 already decide what is listed. The top level is read first and the work stops at the 64th list; nested lists are only looked for when there is room left. What is listed doesn't change.
2026-11-18 19:47:46 +07:00
RRFZ 4731a13375 Claim the pseudo-C memory-write follow-up (R3.11) 2026-11-18 07:45:18 -05:00
BFlat 194442dae5 Merge pull request #2124 from Plastic-time/bugfix/r3.11-pseudo-memory-write-overlap
Follow-up to #2119, which stopped a load being shown after a store that may have changed the memory it read, but still let it through in several cases: on 64-bit ARM, where the size of an access wasn't known and a 16-byte pair store looked clear of a load 8 bytes on; when a stack address had an index, or the two addresses used different segments; after an instruction the decompiler doesn't model, such as a byte store on 32-bit ARM or an x86 string operation; and for a value loaded from a fixed address. A Win64 argument passed on the stack is also no longer shown with the value stored before a later write to its slot. Where it can't tell, the decompiler now keeps the load as its own statement.
2026-11-18 19:38:01 +07:00
RRFZ a640dc31b8 Resolve local file URL authorities from the application root (R3.12) 2026-11-18 07:15:44 -05:00
RRFZ 55a9d46b24 Stop listing root arrays at the summary cap (R5.7) 2026-11-18 07:13:19 -05:00
RRFZ 4040a90d92 Claim the root list cap follow-up (R5.7) 2026-11-18 07:08:35 -05:00