feat(pack): modularizeImport support skip type import

[fireairforce]

Summary

modularizeImport transform support skip type import.

Test Plan

updated.

[gemini-code-assist]

Summary of Changes

Hello, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request enhances the modularizeImport transformation to properly support TypeScript files. By integrating a pre-processing step that removes type-only imports, the system avoids incorrect modularization of type declarations, ensuring better compatibility and correctness for projects utilizing TypeScript and this optimization. The changes are validated with an updated test case that now includes React syntax.

Highlights

• Modularize Imports for TypeScript: The modularizeImport transform now correctly handles TypeScript files by stripping type-only imports before applying modularization, preventing issues with type references.
• TypeScript File Detection: A new utility function is_typescript_like_file was introduced to identify TypeScript, TSX, MJS, and CJS files based on their extensions.
• Updated Test Case: An existing test case for modularize_imports was updated to use a .tsx file, include React components, and verify that type-only imports are correctly skipped, resolving previous module not found errors.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request enhances the modularizeImport transform to correctly handle type-only imports in TypeScript files. By running a TypeScript transform pass with verbatim_module_syntax: false before the main modularization logic, it effectively strips out type-only imports, preventing them from being processed as runtime dependencies. This resolves issues where modularizeImport would attempt to create invalid import paths for types. The changes are well-targeted, and the accompanying test case effectively demonstrates the fix. I've kept the suggestion to improve the file type detection logic for better robustness and readability.

[github-actions]

📊 Performance Benchmark Report (with-antd)

Utoopack Performance Report

Report ID: utoopack_performance_report_20260318_104553
Generated: 2026-03-18 10:45:53
Trace File: trace_antd.json (0.6GB, 1.74M spans)
Test Project: examples/with-antd

---

Executive Summary

Metric	Value	Assessment
Total Wall Time	14,977.4 ms	Baseline
Total Thread Work (de-duped)	51,633.1 ms	Non-overlapping busy time
Effective Parallelism	3.4x	thread_work / wall_time
Working Threads	5	Threads with actual spans
Thread Utilization	68.9%	🆗 Average
Total Spans	1,742,933	All B/E + X events
Meaningful Spans (>= 10us)	649,748	(37.3% of total)
Tracing Noise (< 10us)	1,093,185	(62.7% of total)

---

Build Phase Timeline

Shows when each build phase is active and how much CPU it consumes.
Self-Time is the time spent exclusively in that phase (excluding children).

Phase	Spans	Inclusive (ms)	Self-Time (ms)	Wall Range (ms)
Resolve	146,562	9,129.6	7,329.9	10,786.5
Parse	12,407	2,815.8	2,510.5	14,785.8
Analyze	378,501	37,250.7	28,386.3	14,227.9
Chunk	29,358	3,105.4	2,909.9	11,921.2
Codegen	69,253	5,135.3	3,849.0	12,017.0
Emit	75	76.7	38.2	11,958.7
Other	13,592	2,379.9	2,051.8	14,977.4

---

Workload Distribution by Diagnostic Tier

Category	Spans	Inclusive (ms)	% Work	Self-Time (ms)	% Self
P0: Scheduling & Resolution	534,830	48,048.3	93.1%	37,083.8	71.8%
P1: I/O & Heavy Tasks	3,317	204.6	0.4%	166.1	0.3%
P2: Architecture (Locks/Memory)	0	0.0	0.0%	0.0	0.0%
P3: Asset Pipeline	109,847	11,065.0	21.4%	9,278.1	18.0%
P4: Bridge/Interop	0	0.0	0.0%	0.0	0.0%
Other	1,754	575.4	1.1%	547.8	1.1%

---

Top 20 Tasks by Self-Time

Self-time is the exclusive duration: time spent in the task itself, not in sub-tasks.
This is the most accurate indicator of where CPU cycles are actually spent.

Self (ms)	Inclusive (ms)	Count	Avg Self (us)	P95 Self (ms)	Max Self (ms)	% Work	Task Name	Top Caller
18,406.9	20,938.3	235,635	78.1	0.1	51.4	35.6%	module	write all entrypoints to disk (1%)
5,086.4	6,843.9	40,401	125.9	0.2	267.3	9.9%	analyze ecmascript module	process module (73%)
3,708.8	4,065.7	68,519	54.1	0.1	18.5	7.2%	internal resolving	resolving (29%)
3,593.7	5,036.5	77,329	46.5	0.0	37.5	7.0%	resolving	module (34%)
3,285.5	7,825.9	81,009	40.6	0.1	31.8	6.4%	process module	module (16%)
2,103.5	2,193.1	9,015	233.3	0.6	47.2	4.1%	parse ecmascript	analyze ecmascript module (27%)
2,087.7	2,251.8	16,481	126.7	0.3	139.1	4.0%	chunking	write all entrypoints to disk (0%)
1,930.1	3,216.3	27,477	70.2	0.2	51.3	3.7%	code generation	chunking (7%)
1,582.5	1,582.5	39,694	39.9	0.1	17.9	3.1%	precompute code generation	code generation (29%)
1,466.2	1,760.0	10,821	135.5	0.0	337.6	2.8%	write all entrypoints to disk	None (0%)
1,417.2	1,417.2	18,046	78.5	0.3	158.0	2.7%	compute async module info	chunking (0%)
786.1	786.7	12,678	62.0	0.1	69.5	1.5%	compute async chunks	write all entrypoints to disk (0%)
496.0	514.6	1,514	327.6	0.9	22.6	1.0%	webpack loader	parse css (11%)
336.5	336.5	2,082	161.6	0.4	20.3	0.7%	generate source map	code generation (96%)
306.6	522.1	864	354.8	0.7	39.5	0.6%	parse css	module (6%)
109.1	109.1	1,357	80.4	0.0	26.0	0.2%	compute binding usage info	write all entrypoints to disk (0%)
100.2	100.2	2,509	39.9	0.1	4.5	0.2%	read file	parse ecmascript (85%)
64.3	64.3	2,018	31.9	0.0	10.0	0.1%	collect mergeable modules	compute merged modules (0%)
37.9	44.5	1,017	37.2	0.1	3.1	0.1%	async reference	write all entrypoints to disk (0%)
37.6	37.6	36	1045.0	5.5	12.5	0.1%	write file	apply effects (100%)

---

Critical Path Analysis

The longest sequential dependency chains that determine wall-clock time.
Focus on reducing the depth of these chains to improve parallelism.

Rank	Self-Time (ms)	Depth	Path
1	267.4	2	process module → analyze ecmascript module
2	67.7	2	code generation → generate source map
3	63.2	2	process module → analyze ecmascript module
4	57.6	2	process module → analyze ecmascript module
5	57.5	2	code generation → generate source map

---

Batching Candidates

High-volume tasks dominated by a single parent. If the parent can batch them,
it drastically reduces scheduler overhead.

Task Name	Count	Top Caller (Attribution)	Avg Self	P95 Self	Total Self
analyze ecmascript module	40,401	process module (73%)	125.9 us	0.20 ms	5,086.4 ms

---

Duration Distribution

Range	Count	Percentage
<10us	1,093,185	62.7%
10us-100us	613,773	35.2%
100us-1ms	27,641	1.6%
1ms-10ms	7,611	0.4%
10ms-100ms	715	0.0%
>100ms	8	0.0%

---

Action Items

1. [P0] Focus on tasks with the highest Self-Time — these are where CPU cycles are actually spent.
2. [P0] Use Batching Candidates to identify callers that should use try_join or reduce #[turbo_tasks::function] granularity.
3. [P1] Check Build Phase Timeline for phases with disproportionate wall range vs. self-time (= serialization).
4. [P1] Inspect P95 Self (ms) for heavy monolith tasks. Focus on long-tail outliers, not averages.
5. [P1] Review Critical Paths — reducing the longest chain depth directly improves wall-clock time.
6. [P2] If Thread Utilization < 60%, investigate scheduling gaps (lock contention or deep dependency chains).

Report generated by Utoopack Performance Analysis Agent