我的工作有run
方法,如下所示:
def perform(share, document)
@history = document.scheduled_posts.find_by(provider: share[:provider])
job = ProviderJob.new(document, share)
begin
job.run
@history.update_columns(response: 'posted', status: 'complete')
rescue StandardError => e
@history.update_columns(response: e.message, status: 'error')
raise Errors::FailedJob, e.message
rescue FbGraph2::Exception::Unauthorized, Twitter::Error::Unauthorized, Mailchimp::UserUnknownError, LinkedIn::OAuthError, Errors::MissingAuth => e
@history.update_columns(response: e.message, status: 'unauthorised')
raise Errors::FailedJob, e.message
end
end
即使引发Errors::MissingAuth
,StandardError
块也会捕获它,因为它继承了它。如何确保正确的块捕获指定的异常?
答案 0 :(得分:3)
这些救援块按顺序执行。你应该先把更具体的一些。将StandardError
作为最后一个移动。
答案 1 :(得分:2)
救援区按顺序运行。因为Errors :: MissingAuth继承自StandardError,所以StandardError块将始终首先触发。您应该更改救援区块的优先级,例如:
def perform(share, document)
@history = document.scheduled_posts.find_by(provider: share[:provider])
job = ProviderJob.new(document, share)
begin
job.run
@history.update_columns(response: 'posted', status: 'complete')
rescue FbGraph2::Exception::Unauthorized, Twitter::Error::Unauthorized, Mailchimp::UserUnknownError, LinkedIn::OAuthError, Errors::MissingAuth => e
@history.update_columns(response: e.message, status: 'unauthorised')
raise Errors::FailedJob, e.message
rescue StandardError => e
@history.update_columns(response: e.message, status: 'error')
raise Errors::FailedJob, e.message
end
end
答案 2 :(得分:2)
如果其他答案有效,我认为这是一个更好的方法。我没有意识到这一点,并开始输入另一个答案,所以我还是要包括它。
我假设此处的所有错误都继承自StandardError。在这种情况下,您可以使用单个救援,并根据引发的错误类配置行为:
rescue StandardError => e
status = [
FbGraph2::Exception::Unauthorized, Twitter::Error::Unauthorized,
Mailchimp::UserUnknownError, LinkedIn::OAuthError, Errors::MissingAuth
].include?(e.class) ? 'unauthorized' : 'error'
@history.update_columns(response: e.message, status: status)
raise Errors::FailedJob, e.message
end