在Bazel中,给定构建目标,脚本(在Bazel外部运行)如何获取生成文件的路径?
场景:我正在使用Bazel进行构建,然后在完成后将结果复制到服务器。我只需要知道要复制什么文件。我可以对文件列表进行硬编码,但我不希望这样做。
一个简单的例子:这个Bazel脚本:
genrule(
name = "main",
srcs = ["main.in"],
outs = ["main.out"],
cmd = "cp $< $@",
)
如果您随后创建一个名为main.in
然后运行的文件,则bazel build :main
bazel报告:
INFO: Found 1 target...
Target //:main up-to-date:
bazel-genfiles/main.out
INFO: Elapsed time: 6.427s, Critical Path: 0.40s
所以有:bazel-genfiles/main.out
。但是,我可以使用哪种机器可读技术来获得该路径?(我可以解析的输出bazel build
,但不建议这样做。)
我发现最接近的是使用bazel query --output=xml :main
,它:main
以XML格式转储有关信息。输出包括以下行:
<rule-output name="//:main.out"/>
那太接近我想要的了。但是采用name
Bazel的标签格式;我不知道如何获得它。
我可以在该name
字段上进行某种字符串替换,以将其转换为bazel-genfiles/main.out
; 但即使那样也不可靠。如果my genrule
was include output_to_bindir = 1
,那么输出将bazel-bin/main.out
改为。
此外,并非所有规则<rule-output>
在XML输出中都有一个字段。例如,如果我的BUILD
文件包含以下代码以创建C库:
cc_library(
name = "mylib",
srcs = glob(["*.c"])
)
的输出bazel query --output=xml :mylib
不包含<rule-output>
或任何其他有用的信息:
<?xml version="1.1" encoding="UTF-8" standalone="no"?>
<query version="2">
<rule class="cc_library" location="/Users/mikemorearty/src/bazel/test1/BUILD:8:1" name="//:mylib">
<string name="name" value="mylib"/>
<list name="srcs">
<label value="//:foo.c"/>
</list>
<rule-input name="//:foo.c"/>
<rule-input name="//tools/defaults:crosstool"/>
<rule-input name="@bazel_tools//tools/cpp:stl"/>
</rule>
</query>
您可以通过bazel aquery
查询操作图来获取此信息。
这是一个稍微丰富一点的示例,其中有一个单一规则的两个输出文件:
$ ls
BUILD main.in WORKSPACE
$ cat WORKSPACE
$ cat BUILD
genrule(
name = "main",
srcs = ["main.in"],
outs = ["main.o1", "main.o2"],
cmd = "cp $< $(location main.o1); cp $< $(location main.o2)",
)
$ cat main.in
hello
使用bazel aquery //:main --output=textproto
查询与机器可读的输出动作图(原癌是analysis.ActionGraphContainer
):
$ bazel aquery //:main --output=textproto >aquery_result 2>/dev/null
$ cat aquery_result
artifacts {
id: "0"
exec_path: "main.in"
}
artifacts {
id: "1"
exec_path: "external/bazel_tools/tools/genrule/genrule-setup.sh"
}
artifacts {
id: "2"
exec_path: "bazel-out/k8-fastbuild/genfiles/main.o1"
}
artifacts {
id: "3"
exec_path: "bazel-out/k8-fastbuild/genfiles/main.o2"
}
actions {
target_id: "0"
action_key: "dd7fd759bbecce118a399c6ce7b0c4aa"
mnemonic: "Genrule"
configuration_id: "0"
arguments: "/bin/bash"
arguments: "-c"
arguments: "source external/bazel_tools/tools/genrule/genrule-setup.sh; cp main.in bazel-out/k8-fastbuild/genfiles/main.o1; cp main.in bazel-out/k8-fastbuild/genfiles/main.o2"
input_dep_set_ids: "0"
output_ids: "2"
output_ids: "3"
}
targets {
id: "0"
label: "//:main"
rule_class_id: "0"
}
dep_set_of_files {
id: "0"
direct_artifact_ids: "0"
direct_artifact_ids: "1"
}
configuration {
id: "0"
mnemonic: "k8-fastbuild"
platform_name: "k8"
}
rule_classes {
id: "0"
name: "genrule"
}
数据并非完全集中在一处,但请注意:
2
和3
对应于我们的两个期望的输出文件,以及列表中的那些工件作为路径上相对磁盘文件到工作区根的输出位置;artifacts
具有目标ID的条目与0
工件ID2
和相关联3
;和targets
ID"0"
为ID的条目与//:main
标签相关联。有了这种简单的结构,我们可以轻松地将脚本打包在一起,以列出与提供的标签相对应的所有输出文件。我找不到直接依赖Bazelanalysis.proto
对外部存储库的定义或其语言绑定的方法,因此您可以将以下脚本修补到bazelbuild/bazel
存储库本身:
工具/list_outputs/list_outputs.py
# Copyright 2019 The Bazel Authors. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
r"""Parse an `aquery` result to list outputs created for a target.
Use this binary in conjunction with `bazel aquery` to determine the
paths on disk to output files of a target.
Example usage: first, query the action graph for the target that you
want to analyze:
bazel aquery //path/to:target --output=textproto >/tmp/aquery_result
Then, from the Bazel repository:
bazel run //tools/list_outputs -- \
--aquery_result /tmp/aquery_result \
--label //path/to:target \
;
This will print a list of zero or more output files emitted by the given
target, like:
bazel-out/k8-fastbuild/foo.genfile
bazel-out/k8-fastbuild/bar.genfile
If the provided label does not appear in the output graph, an error will
be raised.
"""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import sys
from absl import app
from absl import flags
from google.protobuf import text_format
from src.main.protobuf import analysis_pb2
flags.DEFINE_string(
"aquery_result",
None,
"Path to file containing result of `bazel aquery ... --output=textproto`.",
)
flags.DEFINE_string(
"label",
None,
"Label whose outputs to print.",
)
def die(message):
sys.stderr.write("fatal: %s\n" % (message,))
sys.exit(1)
def main(unused_argv):
if flags.FLAGS.aquery_result is None:
raise app.UsageError("Missing `--aquery_result` argument.")
if flags.FLAGS.label is None:
raise app.UsageError("Missing `--label` argument.")
if flags.FLAGS.aquery_result == "-":
aquery_result = sys.stdin.read()
else:
with open(flags.FLAGS.aquery_result) as infile:
aquery_result = infile.read()
label = flags.FLAGS.label
action_graph_container = analysis_pb2.ActionGraphContainer()
text_format.Merge(aquery_result, action_graph_container)
matching_targets = [
t for t in action_graph_container.targets
if t.label == label
]
if len(matching_targets) != 1:
die(
"expected exactly one target with label %r; found: %s"
% (label, sorted(t.label for t in matching_targets))
)
target = matching_targets[0]
all_artifact_ids = frozenset(
artifact_id
for action in action_graph_container.actions
if action.target_id == target.id
for artifact_id in action.output_ids
)
for artifact in action_graph_container.artifacts:
if artifact.id in all_artifact_ids:
print(artifact.exec_path)
if __name__ == "__main__":
app.run(main)
工具/列表输出/建立
# Copyright 2019 The Bazel Authors. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
package(default_visibility = ["//visibility:public"])
licenses(["notice"]) # Apache 2.0
filegroup(
name = "srcs",
srcs = glob(["**"]),
)
py_binary(
name = "list_outputs",
srcs = ["list_outputs.py"],
srcs_version = "PY2AND3",
deps = [
"//third_party/py/abseil",
"//src/main/protobuf:analysis_py_proto",
],
)
作为Git补丁,为了您的方便:https : //gist.github.com/wchargin/5e6a43a203d6c95454aae2886c5b54e4
请注意,该代码尚未经过审查或验证是否正确;我主要以它为例。如果它对您有用,那么也许这个周末我可以为此编写一些测试,并针对Bazel进行PR。
本文收集自互联网,转载请注明来源。
如有侵权,请联系 [email protected] 删除。
我来说两句