Commit Graph
1584 Commits
Author SHA1 Message Date
BFlat 72dfda7653 Merge pull request #1979 from Plastic-time/bugfix/r3.12-wasm-component-layout
A valid WebAssembly component showed an empty layout, because the inspector stopped once it had identified the component header. The outer section framing is now read and listed, embedded modules and components are left to the existing validator, and a cut-short section keeps the layout read so far.
2026-11-17 08:23:38 +07:00
BFlat c301f9c33c Merge pull request #1978 from Plastic-time/bugfix/r3.12-sqlite-type-boundaries
Follow-up to #1974. A quoted INTEGER followed by GENERATED was taken for a rowid alias although SQLite stores that key separately, and one followed by DEFERRABLE was not taken for one although it is. The whole declared type is now checked, with SQLite's own type boundaries, its trailing ALWAYS handling and its display of bracketed types. The cases in the tests were checked against SQLite 3.50.4.
2026-11-17 08:23:32 +07:00
RRFZ 07dfa1a52b Keep source map annotations before trailing comments 2026-11-16 20:12:13 -05:00
RRFZ 98b05760ce Keep the affinity of bracketed SQLite types with inner quotes 2026-11-16 20:01:33 -05:00
RRFZ b226a475f7 Accept GNU cpio CRC symlinks without a checksum (R3.12) 2026-11-16 19:59:06 -05:00
RRFZ 8c88747c09 Require complete block source map annotations (R3.12) 2026-11-16 19:59:05 -05:00
RRFZ 5d83a84b26 Refuse raw line breaks in quoted GraphQL strings (R3.12) 2026-11-16 19:59:04 -05:00
RRFZ 43bd04745f Show classic ECN flags in TCP packet reports (R3.12) 2026-11-16 19:59:03 -05:00
RRFZ 6f0a6f4677 Keep bogus HTML comments out of script inventories (R3.13) 2026-11-16 19:59:02 -05:00
RRFZ 7d60ac9f89 Report top-level WebAssembly component sections (R3.12) 2026-11-16 19:59:01 -05:00
RRFZ 73423d7b5e Claim R3.12: cpio-symlink-crc 2026-11-16 19:56:04 -05:00
RRFZ 48ae95d530 Claim R3.12: sourcemap-block-boundaries 2026-11-16 19:56:01 -05:00
RRFZ a159a12334 Claim R3.12: graphql-string-line-breaks 2026-11-16 19:55:58 -05:00
RRFZ d2a1c26fe9 Claim R3.12: tcp-ecn-flags 2026-11-16 19:55:55 -05:00
RRFZ 9f61b4b81f Claim R3.13: html-bogus-comments 2026-11-16 19:55:53 -05:00
RRFZ 282afb2433 Claim R3.12: wasm-component-layout 2026-11-16 19:55:50 -05:00
RRFZ 180660bb01 Use the full SQLite type to decide rowid aliases (R3.12) 2026-11-16 19:44:00 -05:00
RRFZ 64c2a1e5b8 Claim R3.12: use the full SQLite primary-key type 2026-11-16 19:40:03 -05:00
BFlat 9437e05641 Merge pull request #1977 from Plastic-time/bugfix/r3.13-html-comment-endings
Browsers also close a comment at `--!>`, `<!-->` and `<!--->`. The scanner looked only for `-->`, so it missed scripts and preloads after such a comment. It now resumes at the first ending a browser recognises, and still hides what is inside ordinary and unfinished comments.
2026-11-17 07:38:10 +07:00
BFlat 5b3067ca3e Merge pull request #1976 from Plastic-time/bugfix/r3.12-sourcemap-line-endings
The source map annotation reader split lines on LF only, so a sourceMappingURL comment after a lone CR, U+2028 or U+2029 was missed. It now scans backward over every JavaScript line terminator, counting CRLF once and keeping the 64-line window.
2026-11-17 07:38:04 +07:00
BFlat e78462bc77 Merge pull request #1975 from Plastic-time/bugfix/r3.12-graphql-block-strings
An escaped triple quote inside a GraphQL block string was taken as the end of the string, so braces and comment marks after it were read as document syntax and the operation and later definitions were lost. The escaped delimiter is now consumed as part of the string.
2026-11-17 07:37:59 +07:00
BFlat 61d947b0da Merge pull request #1974 from Plastic-time/bugfix/r3.12-sqlite-quoted-types
SQLite accepts a quoted column type such as `id "INTEGER" PRIMARY KEY`, which makes the column an alias for the rowid. The reader kept the quotes, so it showed the stored NULL in place of the id. A quoted type is now read with the identifier reader, in all four quoting forms.
2026-11-17 07:37:53 +07:00
BFlat b8d468bbb1 Merge pull request #1973 from Plastic-time/bugfix/r3.12-gnu-sparse-extensions
An old GNU sparse member can carry several 512-byte map blocks between its header and its stored data. The listing took the first map block for file data and stopped before the next member. The extension flags are now followed across whole blocks before the data offset is computed; a missing extension ends the listing with a warning.
2026-11-17 07:21:22 +07:00
RRFZ 2d6fab1bf6 Require the full SQLite type for a rowid alias 2026-11-16 19:16:24 -05:00
RRFZ 30172fd238 Keep scripts after an HTML comment closed by the browser (R3.13) 2026-11-16 19:00:03 -05:00
RRFZ 7fe4bcfe12 Find source map annotations after JavaScript line endings (R3.12) 2026-11-16 19:00:02 -05:00
RRFZ 4d39d020f7 Keep escaped triple quotes inside GraphQL block strings (R3.12) 2026-11-16 19:00:01 -05:00
RRFZ 388b6fafe8 Read rowids when SQLite quotes the INTEGER type (R3.12) 2026-11-16 19:00:00 -05:00
RRFZ 8ea2ec66a1 Keep listing TAR members after old GNU sparse files (R3.12) 2026-11-16 18:59:58 -05:00
NetterSilver 86c2dd0227 Reject '+' signs in hex of EVM metadata solc and ipfs values (#1971)
The EVM metadata reader parsed each two-character pair with from_str_radix, which accepts a leading '+', so a malformed solc value such as 0x+1+2+3 was reported as version 1.2.3 and '+f' decoded to a byte. Both characters must now be hex digits; malformed values stay as the raw string or are skipped.
2026-11-17 06:58:27 +07:00
RRFZ c6429efb81 Claim R3.13: html comment endings 2026-11-16 18:43:37 -05:00
RRFZ 2f37c8d290 Claim R3.12: sourcemap line endings 2026-11-16 18:43:35 -05:00
RRFZ cd50f30803 Claim R3.12: graphql block strings 2026-11-16 18:43:33 -05:00
RRFZ e7be207271 Claim R3.12: sqlite quoted types 2026-11-16 18:43:31 -05:00
RRFZ d3daccad84 Claim R3.12: gnu sparse extensions 2026-11-16 18:43:29 -05:00
Nathaniel Caldwell ad0a68a596 Find a Windows game's program in its folder on Linux and macOS (R2.4) (#1809)
A Windows game's folder copied to Linux or macOS has no file with an executable bit, so no program was found and Deep mode skipped the decompiler. A Windows .exe now counts as the program to study on every system, while runtime study starts only a program this computer can run, in both its window and console runs.
2026-11-17 02:52:44 +07:00
BFlat 9b8d6c2b17 Merge pull request #1966 from AlexJoaquimPereira/bugfix/r2.10-pin-estimate-project
A pin size estimate that finished after the person cancelled, or after they opened another project's Pin dialog, wrote its size into the open dialog, and a late estimate error could show as a toast. The estimate result and every project_pin_create error now name their project; the dialog takes only a result for the project it is open for, and a failed pin creation still tells the person.
2026-11-17 01:47:35 +07:00
AlexJoaquimPereira 80926a20e4 Keep a pin size estimate in its own project's dialog (R2.10)
A pin estimate that finished after the person cancelled or opened another project's dialog wrote its size into the open draft; an estimate error could also fall through to a toast. The estimate result and every project_pin_create error now name the project (estimate errors also say estimate), and the dialog fills only when the Pin dialog is open for the same normalized project. Estimate errors the open dialog rejects are swallowed; a failed pin creation still tells the person.

Regressions: a late estimate and a late estimate error for another project, an estimate after the dialog closed, and a failed creation that still toasts. Engine tests check the estimate result and both error details, including an open-project failure.
2026-11-16 23:56:16 +05:30
AlexJoaquimPereira 3cd1c9017f Claim R2.10: keep a late pin estimate in its own project's dialog
A pin size estimate that finishes after the person cancels or opens another project's dialog is applied to the open dialog. Record the pending change before fixing it.
2026-11-16 23:56:16 +05:30
BFlat d42418bdbe Merge pull request #1965 from AlexJoaquimPereira/bugfix/r2.9-build-file-read-cap
build_file_read read the whole file into memory before returning its 256 KiB page, so a multi-gigabyte generated file under the project could freeze or kill the app. It now reads through the platform's capped read with a 16 MiB limit: a larger file is refused with a plain error, and files at or under the limit page as before.
2026-11-17 01:07:18 +07:00
AlexJoaquimPereira 21affde516 Refuse a project file over 16 MB in the builder's read tool (R2.9)
build_file_read used std::fs::read, so a builder could point it at a multi-GB generated file under app/ or target/ and allocate the whole file before the 256 KiB page was produced; MAX_READ only capped the returned page.

It now reads through platform::fs::read_file_capped with a 16 MiB cap. A file over the cap returns OverLimit with a plain message and a remediation; files at or under the cap page as before. The helper also rejects non-regular files before reading.

Regression: a 16 MiB + 1 text file is refused and a file exactly at 16 MiB still returns its first page; the over-cap assertion failed before the fix, where the tool read it and returned a page. A follow-up roadmap line records build_file_edit, which still reads the whole file.
2026-11-16 23:02:53 +05:30
AlexJoaquimPereira f1fc90e72a Claim R2.9: bound a builder's project file read
build_file_read reads the whole file before paging, so a large generated file can freeze or kill the app. Record the pending change before fixing it; the parallel build_file_edit read stays as a follow-up.
2026-11-16 22:45:13 +05:30
BFlat 59f2e3f49f Merge pull request #1964 from Fadlalla-ai/bugfix/r5.7-pin-own-program
A build folder shared by several projects holds every project's programs, and a same-named one there was taken for this project's: Play started it and a new pin copied it. A program outside the project now counts only when Cargo's dep-info file beside it lists sources inside the project, and the newest of the project's own builds is started. A pin made before programs were recorded looks only inside its own copy.
2026-11-17 00:13:01 +07:00
BFlat 44293b47f0 Merge pull request #1962 from AlexJoaquimPereira/bugfix/r2.9-project-evidence-switch
A session that opened one project and then another kept the first project's evidence and open questions in its ledger, so the second project's report and evidence tools mixed both. Opening a different project now clears the ledger, its order, the open questions and the unloaded count before loading the new project's store; reopening the same project still refreshes it in place. Records are saved to their project as they are made, so nothing is lost by clearing.
2026-11-17 00:04:33 +07:00
BFlat 88b2490892 Merge pull request #1961 from Fadlalla-ai/bugfix/r5.7-engine-path-folder
A path set in Settings → Engines counted as found whenever it existed, so a JDK's folder set for the Java runtime was listed as installed and then run as a program ("permission denied"). A folder is now looked in the way an install folder is, newest version first with digit runs compared as numbers; a folder without the program isn't found, and the message names the folder.
2026-11-16 23:55:43 +07:00
Fadlalla-ai 00cf099a38 Take a program from a shared Cargo target directory only when it was built from the project, and start an older pin from its own copy 2026-11-16 20:47:03 +04:00
Fadlalla-ai b4f97d834e Claim: a pin or Play takes a same-named program from a shared CARGO_TARGET_DIR 2026-11-16 20:37:27 +04:00
AlexJoaquimPereira 12bfb1fafd Start a newly opened project from its own evidence (R2.9)
A session that opened project A and then B kept A's ledger and open questions: open_project loaded B's records into the same ledger. The desktop app has one engine for its lifetime, and each rebuild opens its project's investigation, so B's report and evidence tools mixed both projects.

Opening a different project now clears the session ledger, its order, open questions and unloaded count before loading the new project's store. Reopening the same project still refreshes it in place, and a failed unknowns read leaves the current project untouched. unloaded_evidence is recomputed on every open.

Regression: two projects with stored evidence and an open question; switching shows only the new project's records, reopening the same project keeps its ledger, and reopening the first restores its records. A follow-up roadmap line records the open targets and annotations that still survive a switch.
2026-11-16 22:04:36 +05:30
AlexJoaquimPereira 408ae82f63 Claim R2.9: start a newly opened project from its own evidence
A session that opens a second project currently keeps the first project's evidence and open questions in the ledger. Record the pending change before fixing it.
2026-11-16 22:04:36 +05:30
Fadlalla-ai a1c1c8abb9 Look for an engine's program in a folder set as its path, instead of running the folder 2026-11-16 20:18:11 +04:00