Dont format genfiles if disabled

[aseembits93]

PR Type

Bug fix, Enhancement

---

Description

• Skip formatting when formatter disabled

• Simplify diff generation for reviews

• Always compute optimization review

• Surface optimization review in LSP response

---

Diagram Walkthrough

flowchart LR
  Formatter["Formatting generated code"] -- "no formatter provided" --> NoFmt["Return normalized newlines"]
  AIReview["AI optimization review"] -- "simplified diff" --> ReviewReq["Service request payload"]
  Optimizer["Function optimizer"] -- "always run review" --> ReviewRes["optimization_review stored"]
  LSPResp["LSP optimization response"] -- "add field" --> Client["Editor client"]

Loading

File Walkthrough

	Relevant files
Enhancement	aiservice.py Simplify review diff and adjust logging                                    codeflash/api/aiservice.py Import Path only in TYPE_CHECKING. Remove root_dir parameter from review method. Simplify unified diff call without filenames. Update log message to "loading|Reviewing Optimization…". +4/-9      perform_optimization.py Add optimization review to LSP response                                    codeflash/lsp/features/perform_optimization.py Include optimization review (capitalized) in response payload. +1/-0      function_optimizer.py Always compute and store optimization review                          codeflash/optimization/function_optimizer.py Initialize optimization_review attribute. Always request optimization review regardless of PR flags. Store review in data and instance attribute. Limit PR creation based on review and flags. +10/-7
Bug fix	formatter.py Skip formatting when formatter is disabled                              codeflash/code_utils/formatter.py Detect missing formatter and short-circuit. Return code with normalized blank lines if disabled. +3/-0

---

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

Codeflash Bot seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[aseembits93]

buiilt from another branch

[github-actions]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 No relevant tests

🔒 No security concerns identified

⚡ Recommended focus areas for review Behavior Change Returning normalized newlines when formatter is disabled changes previous semantics; verify callers expect "disabled" to skip formatting and that downstream logic relying on formatter exit codes or formatting diffs still behaves correctly. formatter_name = formatter_cmds[0].lower() if formatter_cmds else "disabled" if formatter_name == "disabled": # nothing to do if no formatter provided return re.sub(r"\n{2,}", "\n\n", generated_test_source) with tempfile.TemporaryDirectory() as test_dir_str: Logging Change The log message prefix changed from '!lsp|Computing Optimization Review…' to 'loading|Reviewing Optimization…'; ensure any log parsers or UX elements depending on the old tag are updated. logger.info("loading|Reviewing Optimization…") Always Run Review Optimization review now runs unconditionally; confirm performance impact and that empty or failed responses are handled correctly where 'optimization_review' is consumed. # this will now run regardless of pr, staging review flags try: opt_review_response = self.aiservice_client.get_optimization_review( **data, calling_fn_details=function_references ) except Exception as e: logger.debug(f"optimization review response failed, investigate {e}") data["optimization_review"] = opt_review_response self.optimization_review = opt_review_response if raise_pr or staging_review:

[github-actions]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
Possible issue	Safely capitalize review value Guard against empty or non-string optimization_review to avoid AttributeError on .capitalize(). Provide a safe default for clients expecting a value. codeflash/lsp/features/perform_optimization.py [129-138] "explanation": best_optimization.explanation_v2, -"optimizationReview": function_optimizer.optimization_review.capitalize(), +"optimizationReview": (function_optimizer.optimization_review or "low").capitalize(), Suggestion importance[1-10]: 8 __ Why: Prevents a potential AttributeError when optimization_review is empty by providing a safe default before .capitalize(), directly improving robustness with minimal change.	Medium
	Restore filenames in diffs Preserve filenames in diffs to keep reviews meaningful. Pass stable file identifiers to unified_diff_strings so downstream consumers and humans can understand which file each hunk refers to. codeflash/api/aiservice.py [584-590] diff_str = "\n".join( [ - unified_diff_strings(code1=original_code[p], code2=new_code[p]) + unified_diff_strings( + code1=original_code[p], + code2=new_code[p], + fromfile=str(p), + tofile=str(p), + ) for p in original_code if not is_zero_diff(original_code[p], new_code[p]) ] ) Suggestion importance[1-10]: 7 __ Why: Correctly identifies that filenames were removed from unified_diff_strings, reducing diff clarity; the proposed change is accurate and aligns with the prior behavior, improving review usefulness without functional risk.	Medium
General	Avoid passing obsolete parameters The API signature dropped root_dir, but you still set it in data for PR/staging, risking accidental leakage to other calls. Only add root_dir when it's actually required downstream and ensure it's not passed to get_optimization_review. codeflash/optimization/function_optimizer.py [1507-1516] opt_review_response = "" # this will now run regardless of pr, staging review flags try: opt_review_response = self.aiservice_client.get_optimization_review( **data, calling_fn_details=function_references ) except Exception as e: logger.debug(f"optimization review response failed, investigate {e}") data["optimization_review"] = opt_review_response self.optimization_review = opt_review_response if raise_pr or staging_review: - data["root_dir"] = git_root_dir() + pr_data = {**data, "root_dir": git_root_dir()} Suggestion importance[1-10]: 6 __ Why: It rightly avoids polluting data with root_dir for the API that no longer accepts it, reducing risk of accidental leakage; however, the change is relatively minor and partly stylistic since the current code doesn't pass data again until PR creation.	Low