debian-mirror-gitlab/lib/gitlab/regex.rb

501 lines
16 KiB
Ruby
Raw Normal View History

2018-12-13 13:39:08 +05:30
# frozen_string_literal: true
2014-09-02 18:07:02 +05:30
module Gitlab
module Regex
2020-06-23 00:09:42 +05:30
module Packages
2020-10-24 23:57:45 +05:30
CONAN_RECIPE_FILES = %w[conanfile.py conanmanifest.txt conan_sources.tgz conan_export.tgz].freeze
2020-06-23 00:09:42 +05:30
CONAN_PACKAGE_FILES = %w[conaninfo.txt conanmanifest.txt conan_package.tgz].freeze
2022-07-23 23:45:48 +05:30
PYPI_NORMALIZED_NAME_REGEX_STRING = '[-_.]+'
2021-06-08 01:23:25 +05:30
API_PATH_REGEX = %r{^/api/v\d+/(projects/[^/]+/|groups?/[^/]+/-/)?packages/[A-Za-z]+}.freeze
2020-06-23 00:09:42 +05:30
def conan_package_reference_regex
@conan_package_reference_regex ||= %r{\A[A-Za-z0-9]+\z}.freeze
end
def conan_revision_regex
@conan_revision_regex ||= %r{\A0\z}.freeze
end
2022-01-26 12:08:38 +05:30
def conan_recipe_user_channel_regex
%r{\A(_|#{conan_name_regex})\z}.freeze
end
2020-06-23 00:09:42 +05:30
def conan_recipe_component_regex
2022-01-26 12:08:38 +05:30
# https://docs.conan.io/en/latest/reference/conanfile/attributes.html#name
@conan_recipe_component_regex ||= %r{\A#{conan_name_regex}\z}.freeze
2020-06-23 00:09:42 +05:30
end
def composer_package_version_regex
2021-11-11 11:23:49 +05:30
# see https://github.com/composer/semver/blob/31f3ea725711245195f62e54ffa402d8ef2fdba9/src/VersionParser.php#L215
@composer_package_version_regex ||= %r{\Av?((\d++)(\.(?:\d++|[xX*]))?(\.(?:\d++|[xX*]))?(\.(?:\d++|[xX*]))?)?\z}.freeze
2020-06-23 00:09:42 +05:30
end
2021-01-29 00:20:46 +05:30
def composer_dev_version_regex
@composer_dev_version_regex ||= %r{(^dev-)|(-dev$)}.freeze
end
2020-06-23 00:09:42 +05:30
def package_name_regex
2021-01-08 16:13:35 +05:30
@package_name_regex ||=
%r{
\A\@?
(?> # atomic group to prevent backtracking
(([\w\-\.\+]*)\/)*([\w\-\.]+)
)
@?
(?> # atomic group to prevent backtracking
(([\w\-\.\+]*)\/)*([\w\-\.]*)
)
\z
}x.freeze
2020-06-23 00:09:42 +05:30
end
def maven_file_name_regex
@maven_file_name_regex ||= %r{\A[A-Za-z0-9\.\_\-\+]+\z}.freeze
end
def maven_path_regex
@maven_path_regex ||= %r{\A\@?(([\w\-\.]*)/)*([\w\-\.\+]*)\z}.freeze
end
def maven_app_name_regex
@maven_app_name_regex ||= /\A[\w\-\.]+\z/.freeze
end
2020-07-28 23:09:34 +05:30
def maven_version_regex
2021-12-07 22:27:20 +05:30
@maven_version_regex ||= /\A(?!.*\.\.)[\w+.-]+\z/.freeze
2020-07-28 23:09:34 +05:30
end
2020-06-23 00:09:42 +05:30
def maven_app_group_regex
maven_app_name_regex
end
2021-04-17 20:07:23 +05:30
def npm_package_name_regex
2022-08-27 11:52:29 +05:30
@npm_package_name_regex ||= %r{\A(?:@(#{Gitlab::PathRegex::NAMESPACE_FORMAT_REGEX})/)?[-+\.\_a-zA-Z0-9]+\z}o
2021-04-17 20:07:23 +05:30
end
2020-11-05 12:06:23 +05:30
def nuget_package_name_regex
@nuget_package_name_regex ||= %r{\A[-+\.\_a-zA-Z0-9]+\z}.freeze
end
2021-01-03 14:25:43 +05:30
def nuget_version_regex
@nuget_version_regex ||= /
2022-03-02 08:16:31 +05:30
\A#{_semver_major_regex}
\.#{_semver_minor_regex}
(\.#{_semver_patch_regex})?
(\.\d*)?
#{_semver_prerelease_build_regex}\z
2021-01-03 14:25:43 +05:30
/x.freeze
end
2021-06-08 01:23:25 +05:30
def terraform_module_package_name_regex
@terraform_module_package_name_regex ||= %r{\A[-a-z0-9]+\/[-a-z0-9]+\z}.freeze
end
2020-11-24 15:15:51 +05:30
def pypi_version_regex
# See the official regex: https://github.com/pypa/packaging/blob/16.7/packaging/version.py#L159
@pypi_version_regex ||= %r{
\A(?:
v?
(?:([0-9]+)!)? (?# epoch)
([0-9]+(?:\.[0-9]+)*) (?# release segment)
([-_\.]?((a|b|c|rc|alpha|beta|pre|preview))[-_\.]?([0-9]+)?)? (?# pre-release)
((?:-([0-9]+))|(?:[-_\.]?(post|rev|r)[-_\.]?([0-9]+)?))? (?# post release)
([-_\.]?(dev)[-_\.]?([0-9]+)?)? (?# dev release)
(?:\+([a-z0-9]+(?:[-_\.][a-z0-9]+)*))? (?# local version)
)\z}xi.freeze
end
2021-01-03 14:25:43 +05:30
def debian_package_name_regex
# See official parser
# https://git.dpkg.org/cgit/dpkg/dpkg.git/tree/lib/dpkg/parsehelp.c?id=9e0c88ec09475f4d1addde9cdba1ad7849720356#n122
# @debian_package_name_regex ||= %r{\A[a-z0-9][-+\._a-z0-9]*\z}i.freeze
# But we prefer a more strict version from Lintian
# https://salsa.debian.org/lintian/lintian/-/blob/5080c0068ffc4a9ddee92022a91d0c2ff53e56d1/lib/Lintian/Util.pm#L116
@debian_package_name_regex ||= %r{\A[a-z0-9][-+\.a-z0-9]+\z}.freeze
end
def debian_version_regex
# See official parser: https://git.dpkg.org/cgit/dpkg/dpkg.git/tree/lib/dpkg/parsehelp.c?id=9e0c88ec09475f4d1addde9cdba1ad7849720356#n205
@debian_version_regex ||= %r{
\A(?:
2022-07-23 23:45:48 +05:30
(?:([0-9]{1,9}):)? (?# epoch)
([0-9][0-9a-z\.+~]*-?){1,15} (?# version-revision)
(?<!-)
2021-01-03 14:25:43 +05:30
)\z}xi.freeze
end
def debian_architecture_regex
# See official parser: https://git.dpkg.org/cgit/dpkg/dpkg.git/tree/lib/dpkg/arch.c?id=9e0c88ec09475f4d1addde9cdba1ad7849720356#n43
# But we limit to lower case
2022-08-27 11:52:29 +05:30
@debian_architecture_regex ||= %r{\A#{::Packages::Debian::ARCHITECTURE_REGEX}\z}o.freeze
2021-01-03 14:25:43 +05:30
end
def debian_distribution_regex
2022-08-27 11:52:29 +05:30
@debian_distribution_regex ||= %r{\A#{::Packages::Debian::DISTRIBUTION_REGEX}\z}io.freeze
2021-01-03 14:25:43 +05:30
end
def debian_component_regex
2022-08-27 11:52:29 +05:30
@debian_component_regex ||= %r{\A#{::Packages::Debian::COMPONENT_REGEX}\z}o.freeze
2021-01-03 14:25:43 +05:30
end
2021-06-08 01:23:25 +05:30
def helm_channel_regex
2021-11-18 22:05:49 +05:30
@helm_channel_regex ||= %r{\A([a-zA-Z0-9](\.|-|_)?){1,255}(?<!\.|-|_)\z}.freeze
2021-06-08 01:23:25 +05:30
end
def helm_package_regex
@helm_package_regex ||= %r{#{helm_channel_regex}}.freeze
end
def helm_version_regex
2021-09-04 01:27:46 +05:30
# identical to semver_regex, with optional preceding 'v'
@helm_version_regex ||= Regexp.new("\\Av?#{::Gitlab::Regex.unbounded_semver_regex.source}\\z", ::Gitlab::Regex.unbounded_semver_regex.options)
2021-06-08 01:23:25 +05:30
end
2020-06-23 00:09:42 +05:30
def unbounded_semver_regex
# See the official regex: https://semver.org/#is-there-a-suggested-regular-expression-regex-to-check-a-semver-string
# The order of the alternatives in <prerelease> are intentionally
# reordered to be greedy. Without this change, the unbounded regex would
# only partially match "v0.0.0-20201230123456-abcdefabcdef".
@unbounded_semver_regex ||= /
2021-01-03 14:25:43 +05:30
#{_semver_major_minor_patch_regex}#{_semver_prerelease_build_regex}
/x.freeze
end
def semver_regex
2021-06-08 01:23:25 +05:30
@semver_regex ||= Regexp.new("\\A#{::Gitlab::Regex.unbounded_semver_regex.source}\\z", ::Gitlab::Regex.unbounded_semver_regex.options).freeze
2021-01-03 14:25:43 +05:30
end
# These partial semver regexes are intended for use in composing other
# regexes rather than being used alone.
def _semver_major_minor_patch_regex
@_semver_major_minor_patch_regex ||= /
2022-03-02 08:16:31 +05:30
#{_semver_major_regex}\.#{_semver_minor_regex}\.#{_semver_patch_regex}
/x.freeze
end
def _semver_major_regex
@_semver_major_regex ||= /
2020-06-23 00:09:42 +05:30
(?<major>0|[1-9]\d*)
2022-03-02 08:16:31 +05:30
/x.freeze
end
def _semver_minor_regex
@_semver_minor_regex ||= /
(?<minor>0|[1-9]\d*)
/x.freeze
end
def _semver_patch_regex
@_semver_patch_regex ||= /
(?<patch>0|[1-9]\d*)
2021-01-03 14:25:43 +05:30
/x.freeze
end
def _semver_prerelease_build_regex
@_semver_prerelease_build_regex ||= /
2020-06-23 00:09:42 +05:30
(?:-(?<prerelease>(?:\d*[a-zA-Z-][0-9a-zA-Z-]*|[1-9]\d*|0)(?:\.(?:\d*[a-zA-Z-][0-9a-zA-Z-]*|[1-9]\d*|0))*))?
(?:\+(?<build>[0-9a-zA-Z-]+(?:\.[0-9a-zA-Z-]+)*))?
/x.freeze
end
2021-01-03 14:25:43 +05:30
def prefixed_semver_regex
# identical to semver_regex, except starting with 'v'
@prefixed_semver_regex ||= Regexp.new("\\Av#{::Gitlab::Regex.unbounded_semver_regex.source}\\z", ::Gitlab::Regex.unbounded_semver_regex.options)
2020-06-23 00:09:42 +05:30
end
def go_package_regex
# A Go package name looks like a URL but is not; it:
# - Must not have a scheme, such as http:// or https://
# - Must not have a port number, such as :8080 or :8443
2021-10-27 15:23:28 +05:30
@go_package_regex ||= %r{
2020-06-23 00:09:42 +05:30
\b (?# word boundary)
(?<domain>
[0-9a-z](?:(?:-|[0-9a-z]){0,61}[0-9a-z])? (?# first domain)
(?:\.[0-9a-z](?:(?:-|[0-9a-z]){0,61}[0-9a-z])?)* (?# inner domains)
\.[a-z]{2,} (?# top-level domain)
)
2021-10-27 15:23:28 +05:30
(?<path>/(?:
[-/$_.+!*'(),0-9a-z] (?# plain URL character)
2020-06-23 00:09:42 +05:30
| %[0-9a-f]{2})* (?# URL encoded character)
)? (?# path)
\b (?# word boundary)
2021-10-27 15:23:28 +05:30
}ix.freeze
2020-06-23 00:09:42 +05:30
end
2020-11-24 15:15:51 +05:30
def generic_package_version_regex
2021-04-29 21:17:54 +05:30
maven_version_regex
2020-11-24 15:15:51 +05:30
end
2021-01-03 14:25:43 +05:30
def generic_package_name_regex
maven_file_name_regex
end
def generic_package_file_name_regex
generic_package_name_regex
end
2022-01-26 12:08:38 +05:30
2022-05-03 16:02:30 +05:30
def sha256_regex
@sha256_regex ||= /\A[0-9a-f]{64}\z/i.freeze
end
2022-01-26 12:08:38 +05:30
private
def conan_name_regex
@conan_name_regex ||= %r{[a-zA-Z0-9_][a-zA-Z0-9_\+\.-]{1,49}}.freeze
end
2020-06-23 00:09:42 +05:30
end
2014-09-02 18:07:02 +05:30
extend self
2020-06-23 00:09:42 +05:30
extend Packages
2014-09-02 18:07:02 +05:30
def project_name_regex
2020-03-13 15:44:24 +05:30
# The character range \p{Alnum} overlaps with \u{00A9}-\u{1f9ff}
# hence the Ruby warning.
# https://gitlab.com/gitlab-org/gitlab/merge_requests/23165#not-easy-fixable
2021-11-18 22:05:49 +05:30
@project_name_regex ||= /\A[\p{Alnum}\u{00A9}-\u{1f9ff}_][\p{Alnum}\p{Pd}\u{002B}\u{00A9}-\u{1f9ff}_\. ]*\z/.freeze
2014-09-02 18:07:02 +05:30
end
2015-04-26 12:48:37 +05:30
def project_name_regex_message
2021-11-18 22:05:49 +05:30
"can contain only letters, digits, emojis, '_', '.', '+', dashes, or spaces. " \
"It must start with a letter, digit, emoji, or '_'."
2014-09-02 18:07:02 +05:30
end
2022-05-07 20:08:51 +05:30
# Project path must conform to this regex. See https://gitlab.com/gitlab-org/gitlab/-/issues/27483
def oci_repository_path_regex
@oci_repository_path_regex ||= %r{\A[a-zA-Z0-9]+([._-][a-zA-Z0-9]+)*\z}.freeze
end
def oci_repository_path_regex_message
'must not start or end with a special character and must not contain consecutive special characters.'
end
2020-03-28 13:19:24 +05:30
def group_name_regex
2020-11-24 15:15:51 +05:30
@group_name_regex ||= /\A#{group_name_regex_chars}\z/.freeze
end
def group_name_regex_chars
@group_name_regex_chars ||= /[\p{Alnum}\u{00A9}-\u{1f9ff}_][\p{Alnum}\p{Pd}\u{00A9}-\u{1f9ff}_()\. ]*/.freeze
2020-03-28 13:19:24 +05:30
end
def group_name_regex_message
2020-04-08 14:13:33 +05:30
"can contain only letters, digits, emojis, '_', '.', dash, space, parenthesis. " \
"It must start with letter, digit, emoji or '_'."
2020-03-28 13:19:24 +05:30
end
2017-08-17 22:00:37 +05:30
##
2017-09-10 17:25:29 +05:30
# Docker Distribution Registry repository / tag name rules
#
# See https://github.com/docker/distribution/blob/master/reference/regexp.go.
2017-08-17 22:00:37 +05:30
#
def container_repository_name_regex
2020-11-05 12:06:23 +05:30
@container_repository_regex ||= %r{\A[a-z0-9]+(([._/]|__|-*)[a-z0-9])*\z}
2017-08-17 22:00:37 +05:30
end
2017-09-10 17:25:29 +05:30
##
# We do not use regexp anchors here because these are not allowed when
# used as a routing constraint.
#
def container_registry_tag_regex
2021-06-08 01:23:25 +05:30
@container_registry_tag_regex ||= /\w[\w.-]{0,127}/
2017-09-10 17:25:29 +05:30
end
def environment_name_regex_chars
2018-03-17 18:26:18 +05:30
'a-zA-Z0-9_/\\$\\{\\}\\. \\-'
2017-09-10 17:25:29 +05:30
end
2018-03-27 19:54:05 +05:30
def environment_name_regex_chars_without_slash
'a-zA-Z0-9_\\$\\{\\}\\. -'
end
def environment_name_regex
2018-03-27 19:54:05 +05:30
@environment_name_regex ||= /\A[#{environment_name_regex_chars_without_slash}]([#{environment_name_regex_chars}]*[#{environment_name_regex_chars_without_slash}])?\z/.freeze
end
def environment_name_regex_message
2018-03-27 19:54:05 +05:30
"can contain only letters, digits, '-', '_', '/', '$', '{', '}', '.', and spaces, but it cannot start or end with '/'"
end
2017-08-17 22:00:37 +05:30
2019-10-12 21:52:04 +05:30
def environment_scope_regex_chars
"#{environment_name_regex_chars}\\*"
end
def environment_scope_regex
@environment_scope_regex ||= /\A[#{environment_scope_regex_chars}]+\z/.freeze
end
def environment_scope_regex_message
"can contain only letters, digits, '-', '_', '/', '$', '{', '}', '.', '*' and spaces"
end
2020-10-24 23:57:45 +05:30
# https://gitlab.com/gitlab-org/cluster-integration/gitlab-agent/-/blob/master/doc/identity_and_auth.md#agent-identity-and-name
def cluster_agent_name_regex
/\A[a-z0-9]([-a-z0-9]*[a-z0-9])?\z/
end
def cluster_agent_name_regex_message
%q{can contain only lowercase letters, digits, and '-', but cannot start or end with '-'}
end
2017-08-17 22:00:37 +05:30
def kubernetes_namespace_regex
/\A[a-z0-9]([-a-z0-9]*[a-z0-9])?\z/
end
def kubernetes_namespace_regex_message
2018-03-17 18:26:18 +05:30
"can contain only lowercase letters, digits, and '-'. " \
"Must start with a letter, and cannot end with '-'"
2017-08-17 22:00:37 +05:30
end
2020-05-24 23:13:21 +05:30
# Pod name adheres to DNS Subdomain Names(RFC 1123) naming convention
# https://kubernetes.io/docs/concepts/overview/working-with-objects/names/#dns-subdomain-names
def kubernetes_dns_subdomain_regex
/\A[a-z0-9]([a-z0-9\-\.]*[a-z0-9])?\z/
end
2017-08-17 22:00:37 +05:30
def environment_slug_regex
@environment_slug_regex ||= /\A[a-z]([a-z0-9-]*[a-z0-9])?\z/.freeze
end
def environment_slug_regex_message
"can contain only lowercase letters, digits, and '-'. " \
"Must start with a letter, and cannot end with '-'"
end
2018-03-17 18:26:18 +05:30
2021-01-03 14:25:43 +05:30
# The section start, e.g. section_start:12345678:NAME
def logs_section_prefix_regex
/section_((?:start)|(?:end)):(\d+):([a-zA-Z0-9_.-]+)/
end
# The optional section options, e.g. [collapsed=true]
def logs_section_options_regex
/(\[(?:\w+=\w+)(?:, ?(?:\w+=\w+))*\])?/
end
# The region end, always: \r\e\[0K
def logs_section_suffix_regex
/\r\033\[0K/
end
2018-03-17 18:26:18 +05:30
def build_trace_section_regex
2021-01-03 14:25:43 +05:30
@build_trace_section_regexp ||= %r{
#{logs_section_prefix_regex}
#{logs_section_options_regex}
#{logs_section_suffix_regex}
}x.freeze
2018-03-17 18:26:18 +05:30
end
2018-11-18 11:00:15 +05:30
def markdown_code_or_html_blocks
@markdown_code_or_html_blocks ||= %r{
(?<code>
# Code blocks:
# ```
# Anything, including `>>>` blocks which are ignored by this filter
# ```
^```
.+?
\n```\ *$
)
|
(?<html>
# HTML block:
# <tag>
# Anything, including `>>>` blocks which are ignored by this filter
# </tag>
^<[^>]+?>\ *\n
.+?
\n<\/[^>]+?>\ *$
)
}mx
end
2019-10-12 21:52:04 +05:30
# Based on Jira's project key format
# https://confluence.atlassian.com/adminjiraserver073/changing-the-project-key-format-861253229.html
def jira_issue_key_regex
2020-05-24 23:13:21 +05:30
@jira_issue_key_regex ||= /[A-Z][A-Z_0-9]+-\d+/
2019-10-12 21:52:04 +05:30
end
2022-07-29 17:44:30 +05:30
def jira_issue_key_project_key_extraction_regex
@jira_issue_key_project_key_extraction_regex ||= /-\d+/
end
2018-11-18 11:00:15 +05:30
def jira_transition_id_regex
@jira_transition_id_regex ||= /\d+/
end
2019-12-04 20:38:33 +05:30
def breakline_regex
@breakline_regex ||= /\r\n|\r|\n/
end
2019-12-21 20:55:43 +05:30
2019-12-26 22:10:19 +05:30
# https://docs.aws.amazon.com/general/latest/gr/acct-identifiers.html
def aws_account_id_regex
/\A\d{12}\z/
end
def aws_account_id_message
'must be a 12-digit number'
end
2019-12-21 20:55:43 +05:30
# https://docs.aws.amazon.com/general/latest/gr/aws-arns-and-namespaces.html
def aws_arn_regex
/\Aarn:\S+\z/
end
def aws_arn_regex_message
2019-12-26 22:10:19 +05:30
'must be a valid Amazon Resource Name'
end
def utc_date_regex
@utc_date_regex ||= /\A[0-9]{4}-[0-9]{2}-[0-9]{2}\z/.freeze
2019-12-21 20:55:43 +05:30
end
2020-04-08 14:13:33 +05:30
2020-07-28 23:09:34 +05:30
def merge_request_draft
2021-11-18 22:05:49 +05:30
/\A(?i)(\[draft\]|\(draft\)|draft:)/
2020-07-28 23:09:34 +05:30
end
2020-04-08 14:13:33 +05:30
def issue
2022-01-26 12:08:38 +05:30
@issue ||= /(?<issue>\d+)(?<format>\+)?(?=\W|\z)/
end
def merge_request
@merge_request ||= /(?<merge_request>\d+)(?<format>\+)?/
2020-04-08 14:13:33 +05:30
end
2020-06-23 00:09:42 +05:30
def base64_regex
2021-10-27 15:23:28 +05:30
@base64_regex ||= %r{(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{2}==|[A-Za-z0-9+/]{3}=)?}.freeze
2020-06-23 00:09:42 +05:30
end
2020-11-24 15:15:51 +05:30
def feature_flag_regex
/\A[a-z]([-_a-z0-9]*[a-z0-9])?\z/
end
def feature_flag_regex_message
"can contain only lowercase letters, digits, '_' and '-'. " \
"Must start with a letter, and cannot end with '-' or '_'"
end
2022-05-07 20:08:51 +05:30
def saved_reply_name_regex
@saved_reply_name_regex ||= /\A[a-z]([a-z0-9\-_]*[a-z0-9])?\z/.freeze
end
def saved_reply_name_regex_message
"can contain only lowercase letters, digits, '_' and '-'. " \
"Must start with a letter, and cannot end with '-' or '_'"
end
2022-07-23 23:45:48 +05:30
# One or more `part`s, separated by separator
def sep_by_1(separator, part)
%r(#{part} (#{separator} #{part})*)x
end
2022-08-13 15:12:31 +05:30
def x509_subject_key_identifier_regex
@x509_subject_key_identifier_regex ||= /\A(?:\h{2}:)*\h{2}\z/.freeze
end
2014-09-02 18:07:02 +05:30
end
end
2022-01-26 12:08:38 +05:30
Gitlab::Regex.prepend_mod