feat: fix target system live property - #42
Merged
Conversation
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…n plans Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…S.md Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…eld version guard
OsProcessor.querySourceRecord/queryTargetRecord now return the companion-table
row with data decompressed (same shape OsScanner yields), so addLiveField reads
the published revision's version correctly for draft-over-published entries.
addLiveField now uses readPositiveIntegerVersion (root then data fallback),
cache check is !== undefined instead of truthiness, and non-integer versions
produce live: null with a warning instead of live: { version: undefined }.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…o IDynamoDbClient updateAttribute does a conditional SET on a single attribute path, returning "written" or "condition-failed" (ConditionalCheckFailedException is caught, never retried). ScanOptions.limit stops the generator after N yielded items; sortKeyEquals adds a server-side FilterExpression. MockDynamoDbClient updated to support all three. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Appends line + newline, creating parent directory. Used by ChangeReport to write fix-live audit trail entries as they happen. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Pure decision logic: given a group of L, P, and REV# records for one PK, decides which records need their data.live field filled, cleared, or fixed. 24 test cases cover all skip conditions (no-L, invalid version, status contradictions, missing/mismatched REV#) and change reasons (missing-live, empty-live, wrong-version, stale-live). OS table skips REV# checks. Abstractions define the contract for runners (DDB/OS), ChangeReport (JSONL audit trail), and FixLiveState (dry-run gate). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Appends one JSON line per change/skip event under .transfer/<runId>/fix-live-report.jsonl. Undefined before values are serialized as null. Test container and MockChangeReport for runner tests. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
FixLiveState reads/writes .transfer/state/fix-live/<project>__<system>.json. recordDryRun and recordLiveRun preserve each other's data. FixLiveFeature registers LiveFieldReconciler, ChangeReport, and FixLiveState. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…strap wiring BaseLiveFieldRunner: scan L rows per segment → queryAll full PK group → prepareGroup → reconciler.decide → conditional UpdateItem (or dry-run report). DdbLiveFieldRunner writes ["data","live"] only, filters internal models. Helpers: createEmptyStats (zero-initialized reason counters), runConcurrently (bounded worker pool), cmsEntryGuards (isCmsEntryRow, isInternalModel). FixLiveFeature now registered in bootstrap.ts. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Three tests against a real DynamoDB (dynalite): dry-run leaves the table unchanged, live-run writes data.live via conditional UpdateItem (verifies empty-string preservation via convertEmptyValues and _md unchanged), condition-failed via MdBumpingClient proxy reports changed-during-run. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…ompresses OS subclass: acceptsRow always passes (modelId is inside the blob), prepareGroup decompresses all rows and filters internal models after, buildWrite recompresses the full data blob with only live changed. Decompress failure → skipped group with decompress-failed reason. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two tests against dynalite with real gzip: dry-run reports empty-live on L and leaves the table unchanged, live-run rewrites the L blob with only live changed and verifies P blob is untouched and idempotent re-run finds nothing. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…entations Every prompt method returns null on cancel (never exits). ClackPrompts, ClackUI, ClackSpinner wrap @clack/prompts; commands never import the library directly. StubPrompts/StubUI for tests with scripted answers. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
One Cli/Command token with many implementations; CommandRegistry collects them lazily via resolveAll on first access. menu() filters hidden commands, get() matches on the first token of the yargs command string. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
… dispatch Move commands/run/ to commands/transfer/; wrap all 5 commands as Command implementations (transfer visible, init/init-project/process-segment/ update-skills hidden). New src/cli.ts registers all commands via the CommandRegistry, adds a $0 [folder] default that preserves yarn transfer <folder> and yarn transfer --config --preset, and falls back to a clack select menu when invoked with no arguments. Cancel exits 130. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
StepOutcome<T> discriminated union (ok/cancelled/refused) for guided flow control. selectProject discovers projects and prompts or validates --project. selectSystem prompts source/target with DDB+region+OS hints. confirmSystem shows a summary note and asks for confirmation (--yes skips the prompt). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
guardV6 scans L rows to find a CMS entry and checks for root data object (v6) vs flat fields (v5); refuses v5 tables before the confirm. selectMode offers dry-run/live menu (live disabled without prior dry-run), confirms with last-run summary. format helpers for count grouping and timestamps. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Full flow: selectProject → loadConfig → bootstrap → selectSystem → guardV6 → confirmSystem → selectMode → runTable per DDB/OS → recordDryRun/recordLiveRun → summarise. Resolves DdbLiveFieldRunner or OsLiveFieldRunner directly. Registered in cliContainer as fix-live (visible in menu after transfer). Exit 0/1/130. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Commands guide: add menu section, fix-live section with flags/report/state. Troubleshooting: add published-not-showing-as-live entry. Hard-won decisions: 7 new entries for fix-live reconciler and CLI patterns. AGENTS.md: update runtime flow for menu, mark fix-live as implemented, add open work items. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…in TransferWizard TransferWizard now receives Prompts + UI via constructor injection. All select/input/confirm calls use the abstraction; cancel returns null instead of throwing ExitPromptError. Console output replaced with UI methods. process.exit(1) replaced with thrown errors. Tests use StubPrompts/StubUI with scripted answers. @inquirer/core and @inquirer/prompts removed from dependencies. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
![github-code-quality[bot]](https://avatars.githubusercontent.com/u/9919?s=60&v=4){kind=small}
FixLiveCommand: 8 tests covering cancel at each step, --live without dry run, --yes --dry-run flow, --table=os without OS, live with state. TransferCommand: 4 tests covering --config+--preset, wizard null, wizard result, name/hidden properties. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…, statements 81) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
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.
No description provided.