Pypi: Rework to use Json::find_versions

This reworks the new `Pypi` JSON API implementation to use
`Json::find_versions` in `Pypi::find_versions`, borrowing some of the
approach from the `Crate` strategy.

Besides that, this pares down the fields in the
`::generate_input_values` return hash to only `:url`, as we're not
using a generated regex to match version information in this setup.

This adds a `provided_content` parameter to `::find_versions` as part
of this process and I will expand the `Pypi` tests to increase
coverage (like the `Crates` tests) in a later PR. 75% of `Pypi` checks
are failing at the moment (with some returning inaccurate version
information), so the current priority is getting this fix merged in
the short-term.
This commit is contained in:
Sam Ford 2024-12-07 10:56:46 -05:00
parent d49e01b82b
commit 935eb89eca
No known key found for this signature in database
GPG Key ID: 7AF5CBEE1DD6F76D
2 changed files with 36 additions and 39 deletions

View File

@ -1,26 +1,31 @@
# typed: strict # typed: strict
# frozen_string_literal: true # frozen_string_literal: true
require "json"
require "utils/curl"
module Homebrew module Homebrew
module Livecheck module Livecheck
module Strategy module Strategy
# The {Pypi} strategy identifies versions of software at pypi.org by # The {Pypi} strategy identifies the newest version of a PyPI package by
# using the JSON API endpoint. # checking the JSON API endpoint for the project and using the
# `info.version` field from the response.
# #
# PyPI URLs have a standard format: # PyPI URLs have a standard format:
# `https://files.pythonhosted.org/packages/<hex>/<hex>/<long_hex>/example-1.2.3.tar.gz`
# #
# * `https://files.pythonhosted.org/packages/<hex>/<hex>/<long_hex>/example-1.2.3.tar.gz` # Upstream documentation for the PyPI JSON API can be found at:
# # https://docs.pypi.org/api/json/#get-a-project
# This method uses the `info.version` field in the JSON response to
# determine the latest stable version.
# #
# @api public # @api public
class Pypi class Pypi
NICE_NAME = "PyPI" NICE_NAME = "PyPI"
# The default `strategy` block used to extract version information when
# a `strategy` block isn't provided.
DEFAULT_BLOCK = T.let(proc do |json|
json.dig("info", "version").presence
end.freeze, T.proc.params(
arg0: T::Hash[String, T.untyped],
).returns(T.nilable(String)))
# The `Regexp` used to extract the package name and suffix (e.g. file # The `Regexp` used to extract the package name and suffix (e.g. file
# extension) from the URL basename. # extension) from the URL basename.
FILENAME_REGEX = / FILENAME_REGEX = /
@ -46,8 +51,8 @@ module Homebrew
URL_MATCH_REGEX.match?(url) URL_MATCH_REGEX.match?(url)
end end
# Extracts the package name from the provided URL and generates the # Extracts the package name from the provided URL and uses it to
# PyPI JSON API endpoint. # generate the PyPI JSON API URL for the project.
# #
# @param url [String] the URL used to generate values # @param url [String] the URL used to generate values
# @return [Hash] # @return [Hash]
@ -58,48 +63,41 @@ module Homebrew
match = File.basename(url).match(FILENAME_REGEX) match = File.basename(url).match(FILENAME_REGEX)
return values if match.blank? return values if match.blank?
package_name = T.must(match[:package_name]).gsub(/[_-]/, "-") values[:url] = "https://pypi.org/pypi/#{T.must(match[:package_name]).gsub(/%20|_/, "-")}/json"
values[:url] = "https://pypi.org/project/#{package_name}/#files"
values[:regex] = %r{href=.*?/packages.*?/#{package_name}[._-]v?(\d+(?:\.\d+)*(?:[._-]post\d+)?)\.t}i
values values
end end
# Fetches the latest version of the package from the PyPI JSON API. # Generates a PyPI JSON API URL for the project and identifies new
# versions using {Json#find_versions} with a block.
# #
# @param url [String] the URL of the content to check # @param url [String] the URL of the content to check
# @param regex [Regexp] a regex used for matching versions in content (optional) # @param regex [Regexp] a regex used for matching versions in content
# @param provided_content [String, nil] content to check instead of
# fetching
# @return [Hash] # @return [Hash]
sig { sig {
params( params(
url: String, url: String,
regex: T.nilable(Regexp), regex: T.nilable(Regexp),
_unused: T.untyped, provided_content: T.nilable(String),
_block: T.nilable(Proc), unused: T.untyped,
block: T.nilable(Proc),
).returns(T::Hash[Symbol, T.untyped]) ).returns(T::Hash[Symbol, T.untyped])
} }
def self.find_versions(url:, regex: nil, **_unused, &_block) def self.find_versions(url:, regex: nil, provided_content: nil, **unused, &block)
match_data = { matches: {}, regex:, url: } match_data = { matches: {}, regex:, url: }
generated = generate_input_values(url) generated = generate_input_values(url)
return match_data if generated.blank? return match_data if generated.blank?
match_data[:url] = generated[:url] Json.find_versions(
url: generated[:url],
# Parse JSON and get the latest version regex:,
begin provided_content:,
response = Utils::Curl.curl_output(generated[:url]) **unused,
data = JSON.parse(response.stdout, symbolize_names: true) &block || DEFAULT_BLOCK
latest_version = data.dig(:info, :version) )
rescue => e
puts "Error fetching version from PyPI: #{e.message}"
return {}
end
# Return the version if found
return {} if latest_version.blank?
{ matches: { latest_version => Version.new(latest_version) } }
end end
end end
end end

View File

@ -10,8 +10,7 @@ RSpec.describe Homebrew::Livecheck::Strategy::Pypi do
let(:generated) do let(:generated) do
{ {
url: "https://pypi.org/project/example-package/#files", url: "https://pypi.org/pypi/example-package/json",
regex: %r{href=.*?/packages.*?/example-package[._-]v?(\d+(?:\.\d+)*(?:[._-]post\d+)?)\.t}i,
} }
end end