Build the bundled flatc only when its inputs changed, and build it in parallel
维护者通常 1 天内回复
已经有一个关联 PR 被合并了。
- #1998 来自 @oberstet —— 已合并
评估
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 新手友好度
- 22/100
- Issue 类型
- 重构
- 描述清晰度
- 基本清楚
- 活跃度
- 活跃
- 技术栈
- cmake, python
- 领域
- build-system
调研方向
Start in hatch_build.py, at _build_flatc(), which currently deletes deps/flatbuffers/build, reconfigures with cmake, and builds flatc on one core. Read the just recipes that invoke the build, and the linked zlmdb issue for the shared design. Done means a stamp stored beside the packaged binary lets an unchanged build be skipped, the build runs with parallel jobs, AUTOBAHN_FORCE_FLATC_BUILD=1 forces a rebuild, and just clean clears the stamp. The issue asks for one implementation synced to zlmdb, so the design needs maintainer agreement first.
由索引模型根据 Issue 内容生成。
描述
Same problem and same fix as crossbario/zlmdb#TODO. The two repositories share
hatch_build.py _build_flatc().
On every build, including every editable install, _build_flatc() deletes
deps/flatbuffers/build, configures with cmake again, and builds flatc on one core
(cmake --build without --parallel). Local just recipes that install first therefore
rebuild the same flatc from the same source again and again.
Proposal, as detailed in the zlmdb issue:
- a sha256 stamp over what
flatcis built from: thedeps/flatbufferscommit (and
whether it has local changes), the configure arguments, the target platform and
toolchain, and a version for the build logic. Store it next to the packaged binary; - skip the build when the binary exists and the stamp matches. Otherwise build
incrementally, keeping the build directory unless the configure arguments or the
toolchain changed; - build in parallel (
--parallel <os.cpu_count()>, unlessCMAKE_BUILD_PARALLEL_LEVEL
is set); - force a rebuild with
AUTOBAHN_FORCE_FLATC_BUILD=1, and injust clean.
The toolchain in the stamp also keeps a host-built flatc out of wheels built for
another architecture or libc in a container with the repository mounted.
Both repositories should use the same implementation, so it is written once and then
synced.
Note: This issue was drafted with AI assistance (Claude Code).
- 主要语言
- Python
- 星标
- 2.5k
- 派生
- 767
- 平均合并
- 23 小时 55 分钟
- 30 天内合并 PR
- 25
环境准备
- 没有 Dockerfile 或 Docker Compose 文件
- 有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
crossbario/autobahn-python 的其他 Issue
-
CI-CD enhancement
难度 2/5 1-3 小时 新手友好度 78/100
crossbario/autobahn-python#1945 · 1 条评论 ·
维护者通常 1 天内回复
-
bug CI-CD
难度 2/5 1-3 小时 新手友好度 76/100
crossbario/autobahn-python#1932 ·
维护者通常 1 天内回复
-
CI-CD enhancement
难度 2/5 1-3 小时 新手友好度 72/100
crossbario/autobahn-python#1829 · 1 条评论 ·
维护者通常 1 天内回复
-
bug CI-CD
难度 4/5 3-5 天 新手友好度 52/100
crossbario/autobahn-python#2013 ·
维护者通常 1 天内回复
-
难度 5/5 一周以上 新手友好度 25/100
crossbario/autobahn-python#1963 · 2 条评论 ·
维护者通常 1 天内回复
查看 crossbario/autobahn-python 的全部 Issue
相似的 Issue
-
难度 2/5 1-3 小时 新手友好度 66/100
-
[Bug]: Queen's own Ollama model loses num_ctx when a running session switches to it可能已有人在做 @anasamin26 今天认领。 未关闭bug enhancement
难度 2/5 1-3 小时 新手友好度 78/100
维护者通常 1 天内回复
-
llm translation
难度 2/5 1-3 小时 新手友好度 78/100
BerriAI/litellm#47884 · 1 条评论 ·
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 78/100
维护者通常 1 天内回复
-
bug 🐛
难度 2/5 1-3 小时 新手友好度 72/100
netalertx/NetAlertX#1856 · 1 个 reaction ·
维护者通常 1 天内回复