feat: multi-language orchestration loop with per-language config discovery

.planning/STATE.md

@@ -0,0 +1,22 @@
+# Project State
+
+## Current Position
+- **Phase:** 08 - Sequential Multi-Language Orchestration
+- **Plan:** 02 (Complete)
+- **Status:** Complete
+
+## Progress
+- Plan 08-01: Complete (apply_language_config + orchestration loop)
+- Plan 08-02: Complete (error isolation + config normalization + summary logging)
+
+## Decisions
+- Multi-language orchestration uses sequential passes with deep-copied args
+- find_all_config_files walks up from CWD collecting per-language configs
+- apply_language_config mirrors process_pyproject_config for the multi-config path
+- normalize_toml_config is the shared helper for config normalization (path resolution, defaults, key conversion)
+- Per-language error isolation: try/except in loop with status tracking dict
+- Summary logging: comma-separated "lang: status" pairs via logger.info
+
+## Last Session
+- **Stopped at:** Completed 08-02-PLAN.md
+- **Timestamp:** 2026-03-18T04:40:02Z

.planning/config.json

@@ -0,0 +1,7 @@
+{
+  "executor_model": "opus",
+  "commit_docs": true,
+  "parallelization": false,
+  "branching_strategy": "none",
+  "verifier_enabled": false
+}

.planning/phases/08-sequential-multi-language-orchestration/08-02-PLAN.md

@@ -0,0 +1,59 @@
+---
+phase: 08-sequential-multi-language-orchestration
+plan: 02
+type: implementation
+autonomous: true
+wave: 1
+depends_on: [08-01]
+---
+
+# Plan 08-02: Error Isolation and Config Normalization for Multi-Language Orchestration
+
+## Objective
+
+Harden the multi-language orchestration loop with per-language error isolation and ensure `find_all_config_files` normalizes config values consistently with `parse_config_file`.
+
+## Context
+
+- @codeflash/main.py — multi-language orchestration loop from 08-01
+- @codeflash/cli_cmds/cli.py — apply_language_config from 08-01
+- @codeflash/code_utils/config_parser.py — find_all_config_files and parse_config_file
+- @tests/test_multi_language_orchestration.py — existing tests from 08-01
+
+## Tasks
+
+### Task 1: Normalize config values in find_all_config_files
+type="auto"
+
+`find_all_config_files` reads raw toml data but doesn't normalize it the way `parse_config_file` does (path resolution, hyphen-to-underscore key conversion, defaults for missing keys). Add a helper that normalizes the raw config dict so `apply_language_config` receives consistent data.
+
+**Done criteria:**
+- Config values from `find_all_config_files` have paths resolved relative to config file parent
+- Hyphenated keys are converted to underscored keys
+- Default values are applied for missing keys (formatter_cmds=[], disable_telemetry=False, etc.)
+- Tests verify normalization
+
+### Task 2: Add per-language error isolation in main.py orchestration loop
+type="auto"
+
+Wrap each language pass in a try/except so one language failure doesn't prevent other languages from being optimized. Log the error and continue.
+
+**Done criteria:**
+- Exception in one language pass logs the error and continues to next language
+- Test verifies that if optimizer.run_with_args raises for one language, the other language still runs
+- Summary logging at end of loop reports which languages succeeded/failed
+
+### Task 3: Add summary logging for multi-language orchestration results
+type="auto"
+
+After the orchestration loop completes, log a summary showing which languages were processed and their status (success/failure/skipped).
+
+**Done criteria:**
+- Summary log message after loop shows per-language status
+- Test verifies summary includes correct language names and statuses
+
+## Verification
+
+- All existing tests in test_multi_language_orchestration.py still pass
+- New tests cover normalization, error isolation, and summary logging
+- `uv run prek` passes

.planning/phases/08-sequential-multi-language-orchestration/08-02-SUMMARY.md

@@ -0,0 +1,95 @@
+---
+phase: 08-sequential-multi-language-orchestration
+plan: 02
+subsystem: cli
+tags: [multi-language, config-normalization, error-isolation, orchestration]
+
+requires:
+  - phase: 08-01
+    provides: apply_language_config, multi-language orchestration loop, find_all_config_files
+provides:
+  - normalize_toml_config helper for consistent config normalization
+  - Per-language error isolation in orchestration loop
+  - Orchestration summary logging with per-language status
+affects: [config-parser, main-entry-point, multi-language-support]
+
+tech-stack:
+  added: []
+  patterns: [shared-normalization-helper, error-isolation-loop, status-tracking-dict]
+
+key-files:
+  created: []
+  modified:
+    - codeflash/code_utils/config_parser.py
+    - codeflash/main.py
+    - tests/test_multi_language_orchestration.py
+
+key-decisions:
+  - "Extract normalize_toml_config as shared helper used by both find_all_config_files and parse_config_file"
+  - "Track per-language status as dict[str, str] with values success/failed/skipped"
+  - "Log orchestration summary after loop completes with all statuses"
+
+patterns-established:
+  - "Config normalization: always use normalize_toml_config for toml-based configs"
+  - "Error isolation: wrap per-language passes in try/except, track status, continue on failure"
+
+requirements-completed: []
+
+duration: 3min
+completed: 2026-03-18
+---
+
+# Phase 08 Plan 02: Error Isolation and Config Normalization Summary
+
+**Shared config normalization via normalize_toml_config, per-language error isolation with status tracking, and orchestration summary logging**
+
+## Performance
+
+- **Duration:** 3 min
+- **Started:** 2026-03-18T04:36:44Z
+- **Completed:** 2026-03-18T04:40:02Z
+- **Tasks:** 3
+- **Files modified:** 3
+
+## Accomplishments
+- Extracted `normalize_toml_config` helper that resolves paths, applies defaults, and converts hyphenated keys -- used by both `find_all_config_files` and `parse_config_file` to eliminate duplication
+- Added per-language error isolation so one language failure does not prevent other languages from being optimized
+- Added orchestration summary logging showing per-language status (success/failed/skipped) after the loop completes
+- 13 new tests covering normalization, error isolation, skipped status, and summary logging format
+
+## Task Commits
+
+Each task was committed atomically:
+
+1. **Task 1: Normalize config values in find_all_config_files** - `97e21aab` (feat)
+2. **Task 2: Per-language error isolation in orchestration loop** - `a3014ec0` (feat)
+3. **Task 3: Summary logging tests for orchestration results** - `dcf366e2` (test)
+
+## Files Created/Modified
+- `codeflash/code_utils/config_parser.py` - Added normalize_toml_config helper, used in find_all_config_files and parse_config_file
+- `codeflash/main.py` - Added error isolation try/except, status tracking dict, _log_orchestration_summary helper
+- `tests/test_multi_language_orchestration.py` - 13 new tests (6 normalization, 3 error isolation, 4 summary logging)
+
+## Decisions Made
+- Extracted normalization into a standalone function rather than keeping it duplicated between parse_config_file and find_all_config_files
+- Used a simple dict[str, str] for status tracking rather than a more complex result type
+- Summary logging uses logger.info with comma-separated "lang: status" pairs
+
+## Deviations from Plan
+
+None - plan executed exactly as written.
+
+## Issues Encountered
+None
+
+## User Setup Required
+None - no external service configuration required.
+
+## Next Phase Readiness
+- Multi-language orchestration is now robust against per-language failures
+- Config normalization is consistent across single-config and multi-config paths
+- Ready for further multi-language pipeline enhancements
+
+---
+*Phase: 08-sequential-multi-language-orchestration*
+*Completed: 2026-03-18*