Open
Conversation
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
Walkthrough重构了应用的资源定位与打包流程：新增 Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
诗篇
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
src/one_dragon_qt/app/installer.py (1)
28-30: 这里的图标设置已经和父类重复了。
src/one_dragon_qt/windows/app_window_base.py:27-29已经处理过app_icon。继续在子类保留这段，会让安装器窗口的初始化逻辑继续双轨维护，后面很容易出现父子类行为不一致。🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/one_dragon_qt/app/installer.py` around lines 28 - 30, The installer subclass duplicates icon setup already done in the parent (see app_window_base.py handling of app_icon); remove the duplicated block in src/one_dragon_qt/app/installer.py that calls os_utils.get_resource_path(...) and self.setWindowIcon(QIcon(...)) so the child no longer sets app_icon itself and instead inherits the parent's behavior (leave any other installer init intact).
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/one_dragon/utils/yolo_config_utils.py`:
- Around line 8-9: get_model_category_dir currently returns a resource
(read-only) path via get_resource_path which is then passed to OnnxModelLoader
as the download/write directory; change the API to separate read vs write paths:
keep a read helper (e.g. get_builtin_model_category_dir using get_resource_path)
for loading bundled models and add/keep a write helper (e.g.
get_user_model_category_dir using get_path_under_work_dir) for
downloads/updates, and update calls in yolo_model_card.py that pass
get_model_category_dir to OnnxModelLoader to use the user/write helper instead
while retaining the built-in/read helper where models are only read.
---
Nitpick comments:
In `@src/one_dragon_qt/app/installer.py`:
- Around line 28-30: The installer subclass duplicates icon setup already done
in the parent (see app_window_base.py handling of app_icon); remove the
duplicated block in src/one_dragon_qt/app/installer.py that calls
os_utils.get_resource_path(...) and self.setWindowIcon(QIcon(...)) so the child
no longer sets app_icon itself and instead inherits the parent's behavior (leave
any other installer init intact).
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: bc89d64a-7000-461c-a309-21ef58d9a7a2
📒 Files selected for processing (7)
deploy/OneDragon ScriptChainer.specsrc/one_dragon/utils/i18_utils.pysrc/one_dragon/utils/os_utils.pysrc/one_dragon/utils/yolo_config_utils.pysrc/one_dragon_qt/app/installer.pysrc/one_dragon_qt/view/like_interface.pysrc/one_dragon_qt/windows/app_window_base.py
There was a problem hiding this comment.
🧹 Nitpick comments (2)
.github/workflows/build-release.yml (2)
123-123: 建议在复制 Runner 前增加显式存在性校验。Line 123 当前逻辑可用，但在路径变化时报错上下文不够直观。建议在复制前用
Test-Path做前置校验并给出明确错误信息，方便排查构建失败。可选修改示例
- Copy-Item "dist/OneDragon ScriptChainer Runner.exe" -Destination "dist/OneDragon ScriptChainer/" -Force + $runnerExe = "dist/OneDragon ScriptChainer Runner.exe" + $editorDir = "dist/OneDragon ScriptChainer/" + if (-not (Test-Path $runnerExe)) { + throw "未找到 Runner 可执行文件：$runnerExe" + } + if (-not (Test-Path $editorDir)) { + throw "未找到 Editor 目录：$editorDir" + } + Copy-Item $runnerExe -Destination $editorDir -Force🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.github/workflows/build-release.yml at line 123, 在执行 Copy-Item "dist/OneDragon ScriptChainer Runner.exe" -Destination "dist/OneDragon ScriptChainer/" -Force 之前增加显式文件存在性校验：使用 Test-Path 检查 "dist/OneDragon ScriptChainer Runner.exe" 是否存在，若不存在则输出明确的错误信息（包含被查找的路径）并终止流程（例如用 Write-Error/throw 并返回非零退出码），这样在路径变更或构建产物缺失时能快速定位问题；保留原有 Copy-Item 调用用于实际复制。
130-145: 建议统一为所有上传步骤开启缺失文件即失败。Line 145 的
Upload Dist已设置if-no-files-found: error，但Upload Editor/Runner（Line 125-138）未设置。建议统一开启，避免产物丢失时“上传成功但内容为空/缺失”。可选修改示例
- name: Upload Editor if: ${{ env.CREATE_RELEASE == 'false' }} uses: actions/upload-artifact@v4 with: name: Editor path: deploy/dist/OneDragon ScriptChainer/ + if-no-files-found: error - name: Upload Runner if: ${{ env.CREATE_RELEASE == 'false' }} uses: actions/upload-artifact@v4 with: name: Runner path: deploy/dist/OneDragon ScriptChainer Runner.exe + if-no-files-found: error🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.github/workflows/build-release.yml around lines 130 - 145, Add the missing if-no-files-found: error to all upload-artifact steps so uploads fail when artifacts are missing: update the "Upload Runner" step (and any other upload steps like "Upload Editor") to include with: if-no-files-found: error alongside their name/path entries, matching the existing "Upload Dist" behavior; ensure the steps using uses: actions/upload-artifact@v4 include that key to prevent silent empty uploads.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In @.github/workflows/build-release.yml:
- Line 123: 在执行 Copy-Item "dist/OneDragon ScriptChainer Runner.exe" -Destination
"dist/OneDragon ScriptChainer/" -Force 之前增加显式文件存在性校验：使用 Test-Path 检查
"dist/OneDragon ScriptChainer Runner.exe" 是否存在，若不存在则输出明确的错误信息（包含被查找的路径）并终止流程（例如用
Write-Error/throw 并返回非零退出码），这样在路径变更或构建产物缺失时能快速定位问题；保留原有 Copy-Item 调用用于实际复制。
- Around line 130-145: Add the missing if-no-files-found: error to all
upload-artifact steps so uploads fail when artifacts are missing: update the
"Upload Runner" step (and any other upload steps like "Upload Editor") to
include with: if-no-files-found: error alongside their name/path entries,
matching the existing "Upload Dist" behavior; ensure the steps using uses:
actions/upload-artifact@v4 include that key to prevent silent empty uploads.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: e5576cac-f38a-4e4d-a537-da9727f44e83
📒 Files selected for processing (1)
.github/workflows/build-release.yml
There was a problem hiding this comment.
🧹 Nitpick comments (2)
src/one_dragon/utils/os_utils.py (1)
34-46: 资源路径解析逻辑清晰，实现正确。函数正确处理了开发模式和 PyInstaller 打包模式下的资源查找优先级。有一个小问题：
文档字符串中使用了全角逗号
，，建议改为半角逗号,以保持一致性。📝 建议修复
def get_resource_path(*sub_paths: str) -> str: """获取资源文件路径。 - 优先查找工作目录下的路径，不存在时回退到 PyInstaller _MEIPASS。 + 优先查找工作目录下的路径,不存在时回退到 PyInstaller _MEIPASS。 """🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/one_dragon/utils/os_utils.py` around lines 34 - 46, The docstring in get_resource_path uses a full-width comma '，' — update the docstring to replace the full-width comma with a standard ASCII comma ',' so punctuation is consistent (e.g., change "优先查找工作目录下的路径，不存在时回退到 PyInstaller _MEIPASS。" to use a half-width comma).src/one_dragon/base/config/yaml_operator.py (1)
9-21: YAML 缓存机制实现合理。基于文件 mtime 的缓存失效策略是正确的做法。需要注意：
如果文件在
os.path.exists()检查和os.path.getmtime()调用之间被删除，会抛出FileNotFoundError。当前设计依赖调用方（__read_from_file）在调用前检查文件存在性，但存在 TOCTOU（检查时间到使用时间）竞态条件。模块级缓存
cached_yaml_data是全局状态，在多线程环境下可能存在并发访问问题。如果应用是单线程的，这不是问题。💡 可选：增加防御性处理
def read_cache_or_load(file_path: str): cached = cached_yaml_data.get(file_path) - last_modify = os.path.getmtime(file_path) + try: + last_modify = os.path.getmtime(file_path) + except OSError: + return None if cached is not None and cached[0] == last_modify: return cached[1]🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/one_dragon/base/config/yaml_operator.py` around lines 9 - 21, read_cache_or_load currently reads os.path.getmtime and accesses module-level cached_yaml_data without defenses against the TOCTOU FileNotFoundError and concurrent access; wrap the getmtime/open/cache-read/write sequence in a try/except that catches FileNotFoundError (and optionally OSError) and returns None (or propagates a clear error) to avoid crashing if the file is removed between exists/check and use, and add a thread lock (e.g., cached_yaml_lock = threading.Lock()) to serialize accesses to cached_yaml_data so reads and writes around cached_yaml_data[file_path] and cached_yaml_data.get(file_path) are done under the lock; update references to cached_yaml_data and read_cache_or_load accordingly.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@src/one_dragon/base/config/yaml_operator.py`:
- Around line 9-21: read_cache_or_load currently reads os.path.getmtime and
accesses module-level cached_yaml_data without defenses against the TOCTOU
FileNotFoundError and concurrent access; wrap the getmtime/open/cache-read/write
sequence in a try/except that catches FileNotFoundError (and optionally OSError)
and returns None (or propagates a clear error) to avoid crashing if the file is
removed between exists/check and use, and add a thread lock (e.g.,
cached_yaml_lock = threading.Lock()) to serialize accesses to cached_yaml_data
so reads and writes around cached_yaml_data[file_path] and
cached_yaml_data.get(file_path) are done under the lock; update references to
cached_yaml_data and read_cache_or_load accordingly.
In `@src/one_dragon/utils/os_utils.py`:
- Around line 34-46: The docstring in get_resource_path uses a full-width comma
'，' — update the docstring to replace the full-width comma with a standard ASCII
comma ',' so punctuation is consistent (e.g., change "优先查找工作目录下的路径，不存在时回退到
PyInstaller _MEIPASS。" to use a half-width comma).
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 0c6d9a61-92b9-48f4-bb7c-945005a4a8ba
📒 Files selected for processing (6)
deploy/OneDragon ScriptChainer Runner.specdeploy/OneDragon ScriptChainer.specsrc/one_dragon/base/config/yaml_operator.pysrc/one_dragon/utils/i18_utils.pysrc/one_dragon/utils/os_utils.pysrc/script_chainer/win_exe/script_runner.py
🚧 Files skipped from review as they are similar to previous changes (1)
- src/one_dragon/utils/i18_utils.py
There was a problem hiding this comment.
Actionable comments posted: 4
🧹 Nitpick comments (2)
src/one_dragon/base/config/yaml_operator.py (1)
8-8: 全局 YAML 缓存缺少淘汰策略，长期运行可能持续占用内存。建议给
cached_yaml_data增加上限（LRU/FIFO 均可），避免路径数量增长导致常驻内存不可控。Also applies to: 10-20
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/one_dragon/base/config/yaml_operator.py` at line 8, cached_yaml_data is an unbounded global dict and needs a size limit to prevent unbounded memory growth; introduce a configurable max (e.g., YAML_CACHE_MAX) and use an eviction policy (LRU recommended) when inserting into cached_yaml_data — implement this by replacing the plain dict with an ordered mapping (collections.OrderedDict or a small LRU wrapper) or by adding logic in the function that populates cached_yaml_data (look for functions that set cached_yaml_data entries) to pop the least-recently-used/oldest key once len(cached_yaml_data) > YAML_CACHE_MAX; ensure access updates recency if using LRU and add a default max constant and a short comment explaining the policy.src/one_dragon/utils/os_utils.py (1)
142-151:get_monday_dt的文档与变量名建议同步到“周一”语义。当前返回说明与局部变量名仍是“星期天”，易误导维护者。
建议修复
def get_monday_dt(dt: str) -> str: @@ - :return: 星期天日期 yyyyMMdd 格式 + :return: 星期一日期 yyyyMMdd 格式 @@ - sunday_date = date + datetime.timedelta(days=-weekday) - return sunday_date.strftime("%Y%m%d") + monday_date = date + datetime.timedelta(days=-weekday) + return monday_date.strftime("%Y%m%d")🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/one_dragon/utils/os_utils.py` around lines 142 - 151, The get_monday_dt function's docstring and local variable names incorrectly reference "Sunday" while the logic returns Monday; rename the misleading symbol(s) and update the docstring to reflect "Monday" semantics (e.g., change sunday_date to monday_date and update the return description and parameter doc to say 星期一/周一), and ensure any inline comment about weekday mapping consistently states 0 = Monday; keep the existing computation (date + datetime.timedelta(days=-weekday)) which correctly yields Monday.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/one_dragon/base/config/yaml_config.py`:
- Line 42: The docstring line """是否只读取sample文件（即使.yml文件存在也只读sample）""" contains
full-width parentheses which trigger Ruff; update that string to use
ASCII/half-width parentheses like """是否只读取sample文件(即使.yml文件存在也只读sample)"""
(locate the exact docstring in yaml_config.py), save and run ruff/linters to
confirm the warning is resolved.
- Around line 82-84: 当前实现：在判断 self._copy_from_sample 为 True 后执行
shutil.copyfile(sample_yml_path, yml_path) 但仍返回
sample_yml_path，导致后续写入错误。修改该方法使其在执行复制后返回目标路径 yml_path（而非 sample_yml_path）；即保留
shutil.copyfile(sample_yml_path, yml_path) 调用不变，但在返回值处根据 self._copy_from_sample
或复制结果返回 yml_path，确保后续保存落在正确的 <module>.yml
文件（引用符号：self._copy_from_sample、sample_yml_path、yml_path）。
In `@src/one_dragon/base/config/yaml_operator.py`:
- Around line 10-20: The cache currently returns the same mutable object to
callers and doesn't enforce the YAML root to be a dict, causing shared-mutation
bugs and type-contract violations; in read_cache_or_load you should (1) validate
the result of yaml_utils.safe_load(file) is a dict (raise a clear TypeError or
convert to an empty dict if your contract requires) and (2) when returning
cached_yaml_data[file_path][1] return a defensive copy (e.g., a deep copy) so
callers do not share the same dict instance; update any callers that assume
YamlOperator.data is always a dict to rely on the validated/defensively-copied
value.
In `@src/one_dragon/utils/os_utils.py`:
- Around line 64-66: get_work_dir() currently returns os.getcwd() when
run_in_exe() is true which yields an unstable working directory; instead detect
the frozen/exe context and return the directory of the running executable/bundle
so relative paths remain stable. Modify get_work_dir()/run_in_exe() branch to
use the executable location (e.g., os.path.dirname(sys.executable) or the frozen
bundle path such as getattr(sys, "frozen", False) and sys._MEIPASS when
applicable) rather than os.getcwd(); ensure callers like
src/one_dragon/utils/cmd_utils.py that rely on get_work_dir() (e.g.,
subprocess.Popen default cwd) receive the executable directory.
---
Nitpick comments:
In `@src/one_dragon/base/config/yaml_operator.py`:
- Line 8: cached_yaml_data is an unbounded global dict and needs a size limit to
prevent unbounded memory growth; introduce a configurable max (e.g.,
YAML_CACHE_MAX) and use an eviction policy (LRU recommended) when inserting into
cached_yaml_data — implement this by replacing the plain dict with an ordered
mapping (collections.OrderedDict or a small LRU wrapper) or by adding logic in
the function that populates cached_yaml_data (look for functions that set
cached_yaml_data entries) to pop the least-recently-used/oldest key once
len(cached_yaml_data) > YAML_CACHE_MAX; ensure access updates recency if using
LRU and add a default max constant and a short comment explaining the policy.
In `@src/one_dragon/utils/os_utils.py`:
- Around line 142-151: The get_monday_dt function's docstring and local variable
names incorrectly reference "Sunday" while the logic returns Monday; rename the
misleading symbol(s) and update the docstring to reflect "Monday" semantics
(e.g., change sunday_date to monday_date and update the return description and
parameter doc to say 星期一/周一), and ensure any inline comment about weekday
mapping consistently states 0 = Monday; keep the existing computation (date +
datetime.timedelta(days=-weekday)) which correctly yields Monday.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: dcf8c15f-4d0f-4761-b9b0-08602f6f84cb
📒 Files selected for processing (3)
src/one_dragon/base/config/yaml_config.pysrc/one_dragon/base/config/yaml_operator.pysrc/one_dragon/utils/os_utils.py
| """配置文件不存在时 是否从sample文件中读取""" | ||
|
|
||
| self._read_sample_only: bool = read_sample_only | ||
| """是否只读取sample文件（即使.yml文件存在也只读sample）""" |
There was a problem hiding this comment.
注释中的全角括号会持续触发 Ruff 告警，建议替换为半角。
这是低风险问题，但会增加静态检查噪音。
🧰 Tools
🪛 Ruff (0.15.9)
[warning] 42-42: String contains ambiguous （ (FULLWIDTH LEFT PARENTHESIS). Did you mean ( (LEFT PARENTHESIS)?
(RUF001)
[warning] 42-42: String contains ambiguous ） (FULLWIDTH RIGHT PARENTHESIS). Did you mean ) (RIGHT PARENTHESIS)?
(RUF001)
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/one_dragon/base/config/yaml_config.py` at line 42, The docstring line
"""是否只读取sample文件（即使.yml文件存在也只读sample）""" contains full-width parentheses which
trigger Ruff; update that string to use ASCII/half-width parentheses like
"""是否只读取sample文件(即使.yml文件存在也只读sample)""" (locate the exact docstring in
yaml_config.py), save and run ruff/linters to confirm the warning is resolved.
Comment on lines
+82
to
+84
| if self._copy_from_sample: | ||
| shutil.copyfile(sample_yml_path, yml_path) | ||
| return sample_yml_path |
There was a problem hiding this comment.
从 sample 复制后仍返回 sample 路径，会导致写入目标错误。
copy_from_sample=True 时已复制到 <module>.yml，但返回 sample_yml_path 会让后续保存仍落到 sample 文件。
建议修复
if self._sample and os.path.exists(sample_yml_path):
if self._copy_from_sample:
shutil.copyfile(sample_yml_path, yml_path)
- return sample_yml_path
+ return yml_path
+ return sample_yml_path📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| if self._copy_from_sample: | |
| shutil.copyfile(sample_yml_path, yml_path) | |
| return sample_yml_path | |
| if self._copy_from_sample: | |
| shutil.copyfile(sample_yml_path, yml_path) | |
| return yml_path | |
| return sample_yml_path |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/one_dragon/base/config/yaml_config.py` around lines 82 - 84, 当前实现：在判断
self._copy_from_sample 为 True 后执行 shutil.copyfile(sample_yml_path, yml_path)
但仍返回 sample_yml_path，导致后续写入错误。修改该方法使其在执行复制后返回目标路径 yml_path（而非
sample_yml_path）；即保留 shutil.copyfile(sample_yml_path, yml_path) 调用不变，但在返回值处根据
self._copy_from_sample 或复制结果返回 yml_path，确保后续保存落在正确的 <module>.yml
文件（引用符号：self._copy_from_sample、sample_yml_path、yml_path）。
Comment on lines
+10
to
+20
| def read_cache_or_load(file_path: str): | ||
| cached = cached_yaml_data.get(file_path) | ||
| last_modify = os.path.getmtime(file_path) | ||
| if cached is not None and cached[0] == last_modify: | ||
| return cached[1] | ||
|
|
||
| with open(file_path, 'r', encoding='utf-8') as file: | ||
| log.debug(f"加载yaml: {file_path}") | ||
| data = yaml_utils.safe_load(file) | ||
| cached_yaml_data[file_path] = (last_modify, data) | ||
| return data |
There was a problem hiding this comment.
缓存返回共享可变对象且未约束 YAML 根类型，存在串扰与运行时风险。
当前命中缓存会把同一个 dict 引用分发给多个实例；同时未校验 safe_load 的返回类型，遇到 list/scalar 配置时会破坏 YamlOperator.data 的字典契约。
可参考的修正 diff
import os
+import copy
import yaml
@@
-cached_yaml_data: dict[str, tuple[float, dict]] = {}
+cached_yaml_data: dict[str, tuple[float, dict]] = {}
@@
-def read_cache_or_load(file_path: str):
+def read_cache_or_load(file_path: str) -> dict:
cached = cached_yaml_data.get(file_path)
last_modify = os.path.getmtime(file_path)
if cached is not None and cached[0] == last_modify:
- return cached[1]
+ return copy.deepcopy(cached[1])
@@
with open(file_path, 'r', encoding='utf-8') as file:
log.debug(f"加载yaml: {file_path}")
- data = yaml_utils.safe_load(file)
+ raw = yaml_utils.safe_load(file)
+ data = raw if isinstance(raw, dict) else {}
cached_yaml_data[file_path] = (last_modify, data)
- return data
+ return copy.deepcopy(data)Also applies to: 50-50
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/one_dragon/base/config/yaml_operator.py` around lines 10 - 20, The cache
currently returns the same mutable object to callers and doesn't enforce the
YAML root to be a dict, causing shared-mutation bugs and type-contract
violations; in read_cache_or_load you should (1) validate the result of
yaml_utils.safe_load(file) is a dict (raise a clear TypeError or convert to an
empty dict if your contract requires) and (2) when returning
cached_yaml_data[file_path][1] return a defensive copy (e.g., a deep copy) so
callers do not share the same dict instance; update any callers that assume
YamlOperator.data is always a dict to rely on the validated/defensively-copied
value.
Comment on lines
64
to
66
| if run_in_exe(): | ||
| return os.path.dirname(sys.executable) | ||
| return os.getcwd() | ||
| dir_path: str = os.path.abspath(__file__) |
There was a problem hiding this comment.
exe 模式下使用 os.getcwd() 会引入不稳定工作目录。
这会影响依赖 get_work_dir() 的调用方（例如 src/one_dragon/utils/cmd_utils.py 中 subprocess.Popen 的默认 cwd），导致从不同启动位置运行时行为不一致，出现相对路径失效风险。
建议修复
def get_work_dir() -> str:
@@
if run_in_exe():
- return os.getcwd()
+ return os.path.dirname(sys.executable)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/one_dragon/utils/os_utils.py` around lines 64 - 66, get_work_dir()
currently returns os.getcwd() when run_in_exe() is true which yields an unstable
working directory; instead detect the frozen/exe context and return the
directory of the running executable/bundle so relative paths remain stable.
Modify get_work_dir()/run_in_exe() branch to use the executable location (e.g.,
os.path.dirname(sys.executable) or the frozen bundle path such as getattr(sys,
"frozen", False) and sys._MEIPASS when applicable) rather than os.getcwd();
ensure callers like src/one_dragon/utils/cmd_utils.py that rely on
get_work_dir() (e.g., subprocess.Popen default cwd) receive the executable
directory.
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.
Summary by CodeRabbit
发布说明
优化改进
修复
文档