chore(deps): bump the actions group with 3 updates#137
Open
dependabot[bot] wants to merge 277 commits into
Open
chore(deps): bump the actions group with 3 updates#137dependabot[bot] wants to merge 277 commits into
dependabot[bot] wants to merge 277 commits into
Conversation
VSCode spawns the debug adapter with an arbitrary working directory (often /), which broke require's git-root search (.lisp/ was never found) and any relative load-file path. The launch request now accepts a `cwd` argument applied with os.Chdir before the program runs; the extension defaults it to the workspace folder (launch.json can override) and also spawns the LSP server at the workspace root for the same reason. Extension version bumped to 0.3.0. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
go test ./... skips every build-tagged package (debugadapter, the DAP/LSP command wiring), so the debugger tests never ran in CI. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The only DAP request paths without any coverage: Shift-F11 from inside a function body must stop at the caller's level, and pause must interrupt a running program at the next evaluated form. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Future goroutines inherited the spawning context's runtime.Thread, so their EVALs pushed/popped frames on the same live stack as the main goroutine (corrupting stack traces and step depths) and could pause a goroutine the client did not know about. - concurrent.NewFuture now detaches the debug thread from the future's context (runtime.DetachThread, no-op in release builds) - the DAP hook only reacts on the goroutine carrying the session's Thread; evaluations elsewhere (futures, Debug Console) never pause Known limitation documented in the extension README: code inside (future …) does not hit breakpoints and cannot be stepped. Full multi-thread debugging is future work. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two editor features: - lisp.languageServer.includeDirs: extra require search directories for the editor, sent as LSP initializationOptions and appended to the require cascade (the editor equivalent of the interpreter's -i). Without it, projects relying on -i showed unresolved requires in VSCode. - textDocument/definition (F12): document-local definitions jump within the file; definitions imported through require jump into the module's file, using the position information the import analysis already carried. Extension version bumped to 0.4.0. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Backlog with effort/risk/impact estimates and code entry points, ordered from surgical quick wins to invasive changes. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
$NAME placeholders — the READWithPreamble mechanism Go embedders use to pass data into lisp code — were unusable for scripts run from the CLI or the debugger: load-file goes through read-string, which handed the reader a nil placeholder map, and any $NAME in the file crashed with "invalid memory address or nil pointer dereference". - reader: a placeholder without provided values is now a proper parse error naming the placeholder instead of a panic - new -P/--preamble flag (repeatable): `-P '$NUMBER 1984'`. Values also come from leading `;; $NAME <expr>` lines in the file itself (defaults); flags win. A placeholder with no value anywhere reads as nil (the historical READWithPreamble behaviour) - runScript: with placeholders the file is evaluated directly with the same `;; $MODULE` wrapper load-file builds, so source rows — and therefore breakpoints — keep matching the on-disk file (verified end-to-end: BP on line 3 of a placeholder-bearing script); without placeholders the classic load-file path runs unchanged - DAP: --preamble applies to debugged scripts; the extension forwards a `preamble` array from launch.json as --preamble flags (v0.5.0) - LSP: analysed documents accept placeholders (treated as nil) instead of flagging every placeholder-bearing file as a parse error Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
test-preamble.lisp carries in-file placeholder defaults; the "LISP DEBUGGER: PREAMBLE TEST" launch configuration overrides them through the preamble attribute (forwarded as --preamble flags). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The single "Locals" scope enumerated the whole environment chain, so a paused frame showed the user's bindings drowned among ~150 core builtins. Scopes now mirror the env nesting, one per level, each listing only its own bindings: "Locals" (the frame's immediate env), "Closure"/"Closure N" (enclosing fn/let levels) and "Globals" (the root env, complete — user defs and builtins alike, useful to review what is loaded — marked expensive so clients keep it collapsed by default). env gains an Outer() accessor alongside LocalSymbols(). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
$NAME placeholders are read-time substitutions, invisible at run time, so the editor is the only place to surface them: hovering one shows its in-file default and origin line (or, when the file provides none, an explanation that the value arrives at run time via --preamble / launch.json / Go embedding and reads as nil when absent). F12 jumps to the `;; $NAME <expr>` line and completion offers in-file placeholders with their default. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds highlighting the old grammar missed: $NAME preamble placeholders
(including the leading `;; $NAME <expr>` default lines), ¬…¬ notation
strings with the ¬¬ escape, #{ set literals, the ^ metadata marker,
core macros (when, cond, and, or, ->, ->>, require, future, …), the
name being defined by def/defn/defmacro, and symbols in call position
as function names. Extension 0.5.1.
Semantic (LSP-computed) highlighting remains a roadmap item; this layer
is static TextMate config.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
<BINARY>PATH (LISPPATH for the default binary), OS path-list separated, extends the require search cascade. Placed after the git-root .lisp/ so a project's own modules always win over the shell environment — the conservative placement, and the per-binary name keeps it distinctive (GOPATH/PYTHONPATH style). -i/--include dirs still take precedence. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The server handles workspace/didChangeWatchedFiles by re-analysing
every open document (resolveRequires reads module files fresh from
disk), so editing a required module refreshes imported definitions and
require diagnostics in documents that use it — even when the module is
not open in an editor. The extension registers a **/*.{lisp,mal} file
watcher so the notification fires for those out-of-editor changes.
Extension 0.6.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
textDocument/signatureHelp shows the parameter list of the call form surrounding the cursor, highlighting the argument being typed. The enclosing call and active parameter are found by a bracket/string-aware text scan; parameters come from user-defined defn/defmacro, local or imported through require (qualified or :refer'd). Builtins carry no parameter metadata and yield no signature. Triggered on space and `(`. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signature help and hover now read parameter names from the interpreter environment: any lisp-defined library function or macro (defn/defmacro, e.g. reduce, partial, when, cond, ->, future) is a MalFunc whose Params vector carries the real names, so signature help and hover show a proper arglist without any annotation. The variadic `&` marker is dropped from parameter labels so trailing arguments map to the rest parameter. Pure Go builtins keep no parameter metadata (Go reflection exposes arity and types but not names), so they still yield no signature; a curated arglist map would be needed to cover them (noted in ROADMAP). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
(defn name "docstring" [params] body…) stores the docstring as
{:doc "…"} metadata on the function, retrievable with (doc name) (new
coreextended helper), (meta f), tooling. Implemented in the header lisp
macro (with-meta on the MalFunc); no interpreter core change. Fully
backwards compatible: a leading string is treated as a docstring only
when a parameter vector follows it, so (defn f [x] "literal") still
returns the string.
The LSP surfaces docstrings in hover, for document-local definitions
(definitionOf now parses the optional docstring so parameters are not
misread) and for functions/macros resolved from the interpreter
environment (via the {:doc …} MalFunc metadata).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
New leaf package docmeta is the single source of truth for what the interpreter cannot describe by itself: pure Go builtins (Go reflection exposes arity and types but not parameter names) and special forms (handled by EVAL's switch, never present in the environment). Each entry carries an arglist, a kind (function/special form), a group for the future reference-doc generator, and a one-line description. A consistency test asserts every Function entry resolves as a Go builtin in a loaded env and every SpecialForm entry is absent from it, so the table cannot silently drift from reality. The LSP consumes it: hover, signature help and completion now cover curated Go builtins and — for the first time — special forms (do, if, let, fn, quote, try/catch/finally, …), which resolve nowhere else. Lisp-defined functions and macros are intentionally not in the table: their MalFunc value already carries real parameter names and docstring metadata, read from the live environment. A representative common subset is filled in; the table extends incrementally, guarded by the consistency test. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
… table
Documentation for Go builtins now travels with the value in the
environment instead of a static table compiled into the LSP, so tooling
reflects exactly what an interpreter loads — including an embedder's own
builtins. types.Func gains Doc/Arglist fields, attached at registration
with the new call.Doc(env, name, arglist, doc); each namespace
documents its own builtins next to their registration (core, concurrent
so far). A new Go (doc name) builtin reads them (and the {:doc} metadata
of lisp-defined functions/macros), replacing the old lisp doc helper.
Crucially the documentation is kept OFF metadata: (meta +) stays nil, as
kanaka/mal requires — an earlier attempt to store it in Func.Meta broke
the canonical stepA_mal.mal assertion. This matches Clojure too, where a
function value's meta is nil and docs live on the var.
docmeta shrinks to just special forms (do, if, let, fn, quote,
try/catch/finally, …), which are handled by EVAL and never bound in an
environment, so a static table is the only place for them. Per-namespace
consistency tests assert every documented builtin is registered with nil
meta, and every documented special form is absent from the env.
The LSP reads builtin arglists/docs from the environment (Func fields or
MalFunc params + docstring) and special forms from docmeta, for hover,
signature help and completion.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fills in call.Doc metadata for the Go builtins that were still undocumented, each next to its registration: core (collections, base64/JSON, bytes, time, errors, set, macro?, assert, spew, …), concurrent (future-*), system (getenv/setenv/unsetenv) and require (require, resolve-require). Only internal deserialization constructors (new-*) and the embedder-provided eval are intentionally left undocumented. A system docs consistency test is added. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
scan sliced list.Val[2:] (and [1:]) in the def/defn/defmacro/fn/let/ catch/require cases without checking the length, so a half-typed form the reader returns as a short list — a bare (fn), (let), (def), … which happen constantly while editing — panicked with "slice bounds out of range [2:1]", crashing the language server (VSCode then stops restarting it after five crashes). All suffix slicing in scan now goes through a bounds-safe tail() helper. Regression test covers a range of incomplete and nested-incomplete forms. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add docstrings to every function/macro in header-coreextended.lisp so the
LSP surfaces them on hover and completion:
- plain fns via (defn name "doc" [params] body)
- macros and stateful closures via (with-meta … {:doc "…"})
Fix an inc docstring error (it is the integer successor, not predecessor)
and normalise the remaining (fn (args) …) forms to [args] so the LSP
renders clean arglists.
This makes coreextended depend on defn (from HeaderBasic); production load
order already satisfies that, so drop the redundant nscoreextended.Load
calls from the test helpers that loaded coreextended before HeaderBasic
(docstring_test, docmeta_test, require_test).
Extend debugadapter/test.lisp to exercise a docstring on sqr and hover on
a coreextended builtin (inc).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
core-test-base.mal and core-test_test.mal were leftovers from the standalone `maltest` interpreter (cmd/maltest no longer exists). They are not embedded, globbed or required anywhere, and core-test_test.mal even called `test.suite` instead of the real `test-suite` macro. Update the README to point at `lisp --test DIR`, which absorbed the maltest behaviour and runs every *_test.mal in the directory. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- test.yml: bump checkout/setup-go v3 → v5; add a `quality` job running gofmt check and `go vet` in both the default and lispdebug builds. - codeql.yml: CodeQL security scanning for Go (push/PR to main + weekly). - lint.yml + .golangci.yml: golangci-lint (v2, lispdebug tag), kept non-blocking (continue-on-error) until the tree is clean. - dependabot.yml: weekly updates for gomod, github-actions and the VSCode extension's npm deps, grouped to limit PR noise. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Configure golangci-lint to respect the interpreter's inherited kanaka/mal conventions rather than churn the code: - .golangci.yml: disable ST1001 (deliberate dot-import of `types`), ST1003 (`Foo_Q` predicate naming, part of the exported API), the ST1000/ST1020/ST1021 doc-comment format checks (section-header comment style), and the opt-in QF* quickfix suggestions. Fix the genuine findings so the tree is clean: - errcheck: handle the deferred/immediate Close() results on listeners, connections and pipes (command, repl, tests). - staticcheck ST1019: drop the redundant qualified `types` import in the files that also dot-import it (concurrent, system, mal_test, placeholder_test), qualifying nothing new. - staticcheck ST1016: unify *Position receiver names on `cursor`. - govet: reflect.Ptr → reflect.Pointer. - unused: remove the dead closeOnce/exited fields (and now-unused sync import) from the debug adapter. lint.yml: drop continue-on-error and pin golangci-lint to v2.12.2, so a future release with new linters surfaces as a Dependabot PR instead of breaking main. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
actions/setup-go@v5 and golangci/golangci-lint-action@v7 still target Node 20, which the runners now force onto Node 24 with a warning. Bump to setup-go@v6 and golangci-lint-action@v9 (both node24); the latter keeps the `version` input, so the pinned golangci-lint v2.12.2 is unchanged. checkout@v5 already runs on node24. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Turn the mal-in-mal interpreter into a showcase of what jig/lisp adds over kanaka/mal: - rewrite every top-level definition as `defn` with a docstring (the multi-line ones use ¬…¬ strings, since " strings are single-line); these surface on LSP hover and via (doc name). - extend the hosted `core_ns` with jig/lisp's extra *function* builtins (range, merge, get-in, assoc-in, update, reduce, take, uuid, sets, …) so programs interpreted by this mini-mal can use them; macros and special forms such as future/context/try stay out, as they can't be bound with (eval sym). - teach the interpreted language a small `defn` macro of its own. - refresh the banner/prompt to jig/lisp. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Pr_data now prints reader-macro expansions with their sugar instead of the underlying call: (quote x)->'x, (quasiquote x)->`x, (unquote x)->~x, (splice-unquote x)->~@x, (deref x)->@x, and (with-meta form meta)-> ^meta form. Width-aware wrapping still applies (a long '[...] wraps with the form aligned under the prefix), and the forms round-trip: the reader re-expands the sugar to the same value. This makes .state files (which serialize via Pr_data) read naturally. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
feat(printer): reader-macro sugar in Pr_data
The integrity audit log was one JSON line to stderr — fine for server log ingestion, noisy for interactive use. Now, when stderr is a terminal, print a human-readable block using the term package: a green '✓ integrity verified' with ref/commit/signer each on its own line, or a red '✗ integrity check failed' with the reason. When stderr is redirected (pipe/file) the structured JSON line is kept, so machine consumers are unaffected. Honors NO_COLOR / CLICOLOR_FORCE. - lib/term: exported StderrStyle (Go-facing colorizer, stderr TTY + NO_COLOR/CLICOLOR_FORCE aware) and StderrIsTerminal; colorDecision refactored to take an fd. - command: reportIntegrity/integrityBlock; a terminal failure prints the red block and returns ErrIntegrityReported so main exits non-zero without a second 'Error:' line. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
feat(integrity): human-readable colored audit output on a terminal
Adds a regexp namespace, loaded by default, with first-class compiled patterns and Clojure-style semantics: - re-pattern compiles a reusable «regex …» value; the other functions accept a compiled regex or a raw pattern string (compiled on use). - re-matches? / re-find? return booleans (anchored whole-string / unanchored substring). - re-matches / re-find return Clojure-style match data: nil, the match string when there are no groups, or a vector [whole g1 g2 …] with an unmatched optional group as nil. Patterns are Go RE2 (linear-time; no backreferences or lookaround) and are written as raw ¬…¬ strings — jig/lisp's equivalent of Clojure's #"…" literal, since a normal "…" string rejects \d. re-matches anchors with \A(?:…)\z so it means the whole string regardless of (?m). re-replace/re-split/re-seq are left for a later iteration. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
feat(regexp): regular expressions library (Go RE2)
state-save becomes variadic — (state-save name value & [message]) — mirroring state-load. The message is the git commit message for the state commit; it defaults to "state: name" as before, so existing calls are unchanged. Nothing keys on the message (the .state/-only descent check is path-based), so a custom message is safe. Also make the command integrity tests' runGit helper hermetic (disable commit.gpgsign/tag.gpgsign) so they don't fail on a machine whose global git config enables signing. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
feat(integrity): optional commit message for state-save
Wrap the context-deadline tests (TestContextTimeoutFiresOnTime, TestContextNoTimeout, the two future variants, TestTimeoutOnTryCatch) in synctest.Test so the time package uses a fake clock: deadlines fire deterministically (no wall-clock jitter) and instantly (no real wait). This lets TestTimeoutOnTryCatch drop back to a tight 500ms deadline — the value that was flaky under the real clock and which an earlier commit had widened to 2s to paper over the jitter. Verified 30x under full CPU load, all green, with and without -tags debugger. Also drops the redundant concurrent.Load setup in the future tests (newEnv already loads it). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
test: run EVAL timeout tests under testing/synctest
…ve-all) Adds to the system library: - (chdir path) -> nil; os.Chdir, process-global working directory. - (cwd) -> the working directory as an absolute path. - (mkdtemp & [pfx]) -> a fresh unique temp directory (os.MkdirTemp). - (remove-all path) -> nil; os.RemoveAll, recursive, no error if absent. These let in-process tests drive state-save/state-load against a temporary git repo (mkdtemp + git-init + chdir) instead of the real one, verified end to end: state-save creates the first commit on an unborn branch and state-load reads it back, both without --integrity. The Go func is remove_all (snake_case) so the lisp name is remove-all; the chdir test restores cwd (registered after t.TempDir so it runs first) to avoid polluting other tests in the package. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
feat(system): directory/process primitives (chdir, cwd, mkdtemp, remove-all)
slurp, spit, load-file and load-file-once move from core to the system library. core becomes a pure base with no ambient filesystem access: it keeps eval/read-program/read-string (evaluate code held as data) but can no longer read a file to run it. File access is an explicit capability — +system for raw I/O, +require for structured module loading (require reads modules itself in Go, so it needs neither slurp nor system). - Move slurp/slurp_source/spit + the VerifySource integrity hook from lib/core to lib/system; move the load-file header (header-load-file.lisp) and load-file-once from nscore.LoadInput to nssystem.Load. read-program and eval stay in core. - command's setupIntegrity now installs system.VerifySource. - Not user-facing: the lisp binary loads system by default, so scripts and CLI script execution (which uses load-file) are unaffected. Breaking for embedders loading only core — they must add nssystem.Load (after core + its input layer). Documented in the CHANGELOG. - Updated the ~15 test harnesses that build core-only envs and use slurp/load-file to also load system/nssystem; docgen blurbs and LANGUAGE.md regenerated. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
fullEnv REPLed the load-file header but didn't register slurp-source (which moved to system), so load-file failed and the DAP server hung waiting for a stopped event that never came — a 10m timeout under -tags debugger ./... on CI. Add system.Load(ns). Verified the full tagged suite passes with no hang. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
refactor: move file I/O from core to system (capability model)
Complete the regexp library with the remaining Clojure-style builtins:
- re-seq: vector of every non-overlapping match, each shaped like
re-find (match string, or [whole g1 g2 …] with groups)
- re-replace / re-replace-first: rewrite matches with $1 / ${name}
group references ($$ for a literal $); replace-first leaves the
string unchanged when there is no match
- re-split: split around matches into a vector, with an optional
integer limit (Go Split semantics: trailing empties kept)
Docs (README, CHANGELOG, LANGUAGE.md) and tests updated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
feat(regexp): re-seq, re-replace/-first, re-split
Calendar/duration arithmetic and ordering over the epoch-millisecond
ints that time-ms/time-format already use:
- (time-add ms {:years .. :months .. :days .. :hours .. :minutes ..
:seconds .. :milliseconds ..}) — years/months/days go through
time.AddDate (calendar-aware, overflow normalised like Go: Jan 31 +
1 month -> March 2); the rest add a fixed duration. Missing keys are
0, negatives shift backwards. Unknown key or non-integer value errors.
The deltas map is read, never mutated.
- (time-before? t1 t2) / (time-after? t1 t2) — strict ordering.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
feat(core): time-add, time-before?, time-after?
Study for replacing the U+029E-prefixed-string keyword representation with a dedicated Go type, including confirmed correctness bugs, design options and benchmarks, a migration plan targeting 0.4, and the spec for seqable hash-maps and sets. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…ring
Hash-maps seq as sorted [key value] entry vectors and sets as their
sorted elements, via a single choke point in types.GetSlice (plus
ConvertFrom for vec and a seq case for Clojure parity: seq of an empty
map or set is nil). This makes seq, first, rest, map, filter, reduce,
apply, cons, concat, vec, into and lazy-seqs work over both, enabling
(into {} (map (fn [[k v]] ...) m)).
The order is sorted rather than left unspecified because first and
rest each seq the collection independently: with Go's per-call
randomised map iteration, a first/rest traversal would drop or repeat
entries. nth rejects maps and sets explicitly (they seq but are not
indexed; Clojure errors too).
Binding forms gain sequential destructuring, which the entry idiom
requires and the language lacked: vector patterns in fn params and let
bindings bind positionally and recursively, missing elements bind nil,
& binds the remainder as a list. Top-level fn arity stays strict;
loop/recur bindings remain symbol-only.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
feat(core): seqable hash-maps and sets, sequential destructuring
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Bumps the actions group with 3 updates: [actions/checkout](https://github.com/actions/checkout), [github/codeql-action](https://github.com/github/codeql-action) and [actions/setup-go](https://github.com/actions/setup-go). Updates `actions/checkout` from 5 to 7 - [Release notes](https://github.com/actions/checkout/releases) - [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md) - [Commits](actions/checkout@v5...v7) Updates `github/codeql-action` from 3 to 4 - [Release notes](https://github.com/github/codeql-action/releases) - [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md) - [Commits](github/codeql-action@v3...v4) Updates `actions/setup-go` from 6 to 7 - [Release notes](https://github.com/actions/setup-go/releases) - [Commits](actions/setup-go@v6...v7) --- updated-dependencies: - dependency-name: actions/checkout dependency-version: '7' dependency-type: direct:production update-type: version-update:semver-major dependency-group: actions - dependency-name: github/codeql-action dependency-version: '4' dependency-type: direct:production update-type: version-update:semver-major dependency-group: actions - dependency-name: actions/setup-go dependency-version: '7' dependency-type: direct:production update-type: version-update:semver-major dependency-group: actions ... Signed-off-by: dependabot[bot] <support@github.com>
dependabot
Bot
added
dependencies
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Bumps the actions group with 3 updates: actions/checkout, github/codeql-action and actions/setup-go.
Updates
actions/checkoutfrom 5 to 7Release notes
Sourced from actions/checkout's releases.
... (truncated)
Changelog
Sourced from actions/checkout's changelog.
... (truncated)
Commits
3d3c42eprep v7.0.1 release (#2531)2880268escape values passed to --unset (#2530)12cd223trim only ascii whitespace for branch (#2521)62661c4skip running unsafe pr check if input is default (#2518)e8d4307Bump the minor-actions-dependencies group with 2 updates (#2499)631c942eslint 9 (#2474)4f1f4aeBump actions/upload-artifact from 4 to 7 (#2476)ba09753Bump actions/checkout from 6 to 7 (#2488)b9e0990Bump docker/login-action from 3.3.0 to 4.2.0 (#2479)e8cb398Bump docker/build-push-action from 6.5.0 to 7.2.0 (#2478)Updates
github/codeql-actionfrom 3 to 4Release notes
Sourced from github/codeql-action's releases.
... (truncated)
Changelog
Sourced from github/codeql-action's changelog.
... (truncated)
Commits
da21ad6RemoveNewRemoteFileAddressesFF and default to its behaviourcf46341Merge pull request #4007 from github/mbg/use-registry-proxy-for-repo-auth7db34aeRefactor looking up proxy env vars intogetRegistryProxyConfig.8125f87Remove unusedgetApiFetch7c4a258Merge pull request #4021 from github/mergeback/v4.37.1-to-main-7188fc360297913Rebuild1226301Update changelog and version after v4.37.17188fc3Merge pull request #4020 from github/update-v4.37.1-9e7c07009c8b5f69Update changelog for v4.37.19e7c070Merge pull request #4014 from github/mbg/explicit-remote-prefixUpdates
actions/setup-gofrom 6 to 7Release notes
Sourced from actions/setup-go's releases.
... (truncated)
Commits
b7ad1dachore(deps): bump@actions/cacheto 6.2.0 (#771)0778a10Migrate to ESM and upgrade dependencies (#763)Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting
@dependabot rebase.Dependabot commands and options
You can trigger Dependabot actions by commenting on this PR:
@dependabot rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore <dependency name> major versionwill close this group update PR and stop Dependabot creating any more for the specific dependency's major version (unless you unignore this specific dependency's major version or upgrade to it yourself)@dependabot ignore <dependency name> minor versionwill close this group update PR and stop Dependabot creating any more for the specific dependency's minor version (unless you unignore this specific dependency's minor version or upgrade to it yourself)@dependabot ignore <dependency name>will close this group update PR and stop Dependabot creating any more for the specific dependency (unless you unignore this specific dependency or upgrade to it yourself)@dependabot unignore <dependency name>will remove all of the ignore conditions of the specified dependency@dependabot unignore <dependency name> <ignore condition>will remove the ignore condition of the specified dependency and ignore conditions