我有一个要解析的列表(但是我正在寻找一种通用的方法来解析任何这样的列表):
dev-libs/icu-63.1-r1 alpha amd64 arm arm64 ia64 ppc ppc64 x86 hppa s390
dev-libs/icu-layoutex-63.1 alpha amd64 ia64 ppc ppc64 x86 hppa sparc
dev-lang/perl-5.28-r1 s390
virtual/ruby_gems-0.3_pre24 amd64 x86
这似乎有时会下降,因为它尝试解析体系结构列表,例如以alpha
开头直到行尾,但是我真的想忽略软件包版本之后的所有内容,但在之后保留空格的可能性一个版本。
我的代码如下:(仅打印调试内容)
for line in args.list:
print(line)
package_category = re.search(r'((?<==)\w+-\w+|\w+-\w+|\w+)', line).group(0)
print(package_category)
package_name = re.search(r'(?<=/)[a-z]+.[a-z]+', line).group(0)
print(package_name)
package_version = re.search(r'(?<=-)\d+.\d-*\w*\s?', line).group(0)
我希望这样做:
package_category变量应包含如下类别:
dev-libs
dev-lang
virtual
package_name应该包含一个软件包名称,例如:
icu
icu-layoutex
perl
ruby_gems
package_version:
63.1-r1
63.1
0.3_pre24
其余的应该忽略不计
目前,我突然以某种方式进入了建筑列表:
dev-libs/icu-63.1-r1
dev-libs
icu
alpha
alpha
Traceback (most recent call last):
File "./repomator.py", line 47, in <module>
package_name = re.search(r'(?<=/)[a-z]+.[a-z]+', line).group(0)
AttributeError: 'NoneType' object has no attribute 'group'
答案 0 :(得分:1)
是您想要的吗?
npm install
说明:
package-lock.json
代码:
(?P<category>\w+(?:-\w+)?)/(?P<name>[a-z]+(?:[-_][a-z]+)?)-(?P<version>\S+)
输出:
(?<category> # named group category
\w+ # 1 or more word character
(?:-\w+)? # optional, a dash then 1 or more word character
) # end group
/ # a slash
(?<name> # named group name
[a-z]+ # 1 or more alpha
(?:[-_][a-z]+)? # optional, dash or underscore and 1 or more alpha
) # end group
- # a dash
(?<version> # named group version
\S+ # 1 or more non space character
) # end group