stripe-ruby/lib/stripe/stripe_object.rb
Brandur 44766516d9 stripe-ruby V5 (#815)
* Convert library to use built-in `Net::HTTP`

Moves the library off of Faraday and over onto the standard library's
built-in `Net::HTTP` module. The upside of the transition is that we
break away from a few dependencies that have caused us a fair bit of
trouble in the past, the downside is that we need more of our own code
to do things (although surprisingly, not that much more).

The biggest new pieces are:

* `ConnectionManager`: A per-thread class that manages a connection to
  each Stripe infrastructure URL (like `api.stripe.com`,
  `connect.stripe.com`, etc.) so that we can reuse them between
  requests. It's also responsible for setting up and configuring new
  `Net::HTTP` connections, which is a little more heavyweight
  code-wise compared to other libraries. All of this could have lived in
  `StripeClient`, but I extracted it because that class has gotten so
  big.

* `MultipartEncoder`: A class that does multipart form encoding for file
  uploads. Unfortunately, Ruby doesn't bundle anything like this. I
  built this by referencing the Go implementation because the original
  RFC is not very detailed or well-written. I also made sure that it was
  behaving similarly to our other custom implementations like
  stripe-node, and that it can really upload a file outside the test
  suite.

There's some risk here in that it's easy to miss something across one of
these big transitions. I've tried to test out various error cases
through tests, but also by leaving scripts running as I terminate my
network connection and bring it back. That said, we'd certainly release
on a major version bump because some of the interface (like setting
`Stripe.default_client`) changes.

* Drop support for old versions of Ruby

Drops support for Ruby 2.1 (EOL March 31, 2017) and 2.2 (EOL March 31,
2018). They're removed from `.travis.yml` and the gemspec and RuboCop
configuration have also been updated to the new lower bound.

Most of the diff here are minor updates to styling as required by
RuboCop:

* String literals are frozen by default, so the `.freeze` we had
  everywhere is now considered redundant.

* We can now use Ruby 1.9 style hash syntax with string keys like `{
  "foo": "bar" }`.

* Converted a few heredocs over to use squiggly (leading whitespace
  removed) syntax.

As discussed in Slack, I didn't drop support for Ruby 2.3 (EOL March 31,
2019) as we still have quite a few users on it. As far as I know
dropping it doesn't get us access to any major syntax improvements or
anything, so it's probably not a big deal.

* Make `CardError`'s `code` parameter named instead of positional (#816)

Makes the `code` parameter on `CardError` named instead of positional.
This makes it more consistent with the rest of the constructor's
parameters and makes instantiating `CardError` from `StripeClient`
cleaner.

This is a minor breaking change so we're aiming to release it for the
next major version of stripe-ruby.

* Bump Rubocop to latest version (#818)

* Ruby minimum version increase followup (#819)

* Remove old deprecated methods (#820)

* Remove all alias for list methods (#823)

* Remove UsageRecord.create method (#826)

* Remove IssuerFraudRecord (#827)

* Add ErrorObject to StripeError exceptions (#811)

* Tweak retry logic to be a little more like stripe-node (#828)

Tweaks the retry logic to be a little more like stripe-node's. In
particular, we also retry under these conditions:

* If we receive a 500 on a non-`POST` request.
* If we receive a 503.

I made it slightly different from stripe-node which checks for a 500
with `>= 500`. I don't really like that -- if we want to retry specific
status codes we should be explicit about it.

We're actively re-examining ways on how to make it easier for clients to
figure out when to retry right now, but I figure V5 is a good time to
tweak this because the modifications change the method signature of
`should_retry?` slightly, and it's technically a public method.

* Fix inverted sign for 500 retries (#830)

I messed up in #828 by (1) accidentally flipping the comparison against
`:post` when checking whether to retry on 500, and (2) forgetting to
write new tests for the condition, which is how (1) got through.

This patch fixes both those problems.

* Remove a few more very old deprecated methods (#831)

I noticed that we had a couple of other deprecated methods on `Stripe`
and `StripeObject` that have been around for a long time. May as well
get rid of them too -- luckily they were using `Gem::Deprecate` so
they've been producing annoying deprecated warnings for quite a while
now.

* Remove extraneous slash at the end of the line

* Reset connections when connection-changing configuration changes (#829)

Adds a few basic features around connection and connection manager
management:

* `clear` on connection manager, which calls `finish` on each active
  connection and then disposes of it.

* A centralized cross-thread tracking system for connection managers in
  `StripeClient` and `clear_all_connection_managers` which clears all
  known connection managers across all threads in a thread-safe way.

The addition of these allow us to modify the implementation of some of
our configuration on `Stripe` so that it can reset all currently open
connections when its value changes.

This fixes a currently problem with the library whereby certain
configuration must be set before the first request or it remains fixed
on any open connections. For example, if `Stripe.proxy` is set after a
request is made from the library, it has no effect because the proxy
must have been set when the connection was originally being initialized.

The impetus for getting this out is that I noticed that we will need
this internally in a few places when we're upgrading to stripe-ruby V5.
Those spots used to be able to hack around the unavailability of this
feature by just accessing the Faraday connection directly and resetting
state on it, but in V5 `StripeClient#conn` is gone, and that's no longer
possible.

* Minor cleanup in `StripeClient` (#832)

I ended up having to relax the maximum method line length in a few
previous PRs, so I wanted to try one more cleanup pass in
`execute_request` to see if I could get it back at all.

The answer was "not by much" (without reducing clarity), but I found a
few places that could be tweaked. Unfortunately, ~50 lines is probably
the "right" length for this method in that you _could_ extract it
further, but you'd end up passing huge amounts of state all over the
place in method parameters, and it really wouldn't look that good.

* Do better bookkeeping when tracking state in `Thread.current` (#833)

This is largely just another cleanup patch, but does a couple main
things:

* Hoists the `last_response` value into thread state. This is a very
  minor nicety, but effectively makes `StripeClient` fully thread-safe,
  which seems like a minor nicety. Two calls to `#request` to the same
  `StripeObject` can now be executed on two different threads and their
  results won't interfere with each other.

* Moves state off one-off `Thread.current` keys and into a single one
  for the whole client which stores a new simple type of record called
  `ThreadContext`. Again, this doesn't change much, but adds some minor
  type safety and lets us document each field we expect to have in a
  thread's context.

* Add Invoice.list_upcoming_line_items method (#834)
2019-08-20 11:35:24 -07:00

565 lines
19 KiB
Ruby

# frozen_string_literal: true
module Stripe
class StripeObject
include Enumerable
@@permanent_attributes = Set.new([:id]) # rubocop:disable Style/ClassVars
# The default :id method is deprecated and isn't useful to us
undef :id if method_defined?(:id)
# Sets the given parameter name to one which is known to be an additive
# object.
#
# Additive objects are subobjects in the API that don't have the same
# semantics as most subobjects, which are fully replaced when they're set.
# This is best illustrated by example. The `source` parameter sent when
# updating a subscription is *not* additive; if we set it:
#
# source[object]=card&source[number]=123
#
# We expect the old `source` object to have been overwritten completely. If
# the previous source had an `address_state` key associated with it and we
# didn't send one this time, that value of `address_state` is gone.
#
# By contrast, additive objects are those that will have new data added to
# them while keeping any existing data in place. The only known case of its
# use is for `metadata`, but it could in theory be more general. As an
# example, say we have a `metadata` object that looks like this on the
# server side:
#
# metadata = { old: "old_value" }
#
# If we update the object with `metadata[new]=new_value`, the server side
# object now has *both* fields:
#
# metadata = { old: "old_value", new: "new_value" }
#
# This is okay in itself because usually users will want to treat it as
# additive:
#
# obj.metadata[:new] = "new_value"
# obj.save
#
# However, in other cases, they may want to replace the entire existing
# contents:
#
# obj.metadata = { new: "new_value" }
# obj.save
#
# This is where things get a little bit tricky because in order to clear
# any old keys that may have existed, we actually have to send an explicit
# empty string to the server. So the operation above would have to send
# this form to get the intended behavior:
#
# metadata[old]=&metadata[new]=new_value
#
# This method allows us to track which parameters are considered additive,
# and lets us behave correctly where appropriate when serializing
# parameters to be sent.
def self.additive_object_param(name)
@additive_params ||= Set.new
@additive_params << name
end
# Returns whether the given name is an additive object parameter. See
# `.additive_object_param` for details.
def self.additive_object_param?(name)
@additive_params ||= Set.new
@additive_params.include?(name)
end
def initialize(id = nil, opts = {})
id, @retrieve_params = Util.normalize_id(id)
@opts = Util.normalize_opts(opts)
@original_values = {}
@values = {}
# This really belongs in APIResource, but not putting it there allows us
# to have a unified inspect method
@unsaved_values = Set.new
@transient_values = Set.new
@values[:id] = id if id
end
def self.construct_from(values, opts = {})
values = Stripe::Util.symbolize_names(values)
# work around protected #initialize_from for now
new(values[:id]).send(:initialize_from, values, opts)
end
# Determines the equality of two Stripe objects. Stripe objects are
# considered to be equal if they have the same set of values and each one
# of those values is the same.
def ==(other)
other.is_a?(StripeObject) &&
@values == other.instance_variable_get(:@values)
end
# Hash equality. As with `#==`, we consider two equivalent Stripe objects
# equal.
def eql?(other)
# Defer to the implementation on `#==`.
self == other
end
# As with equality in `#==` and `#eql?`, we hash two Stripe objects to the
# same value if they're equivalent objects.
def hash
@values.hash
end
# Indicates whether or not the resource has been deleted on the server.
# Note that some, but not all, resources can indicate whether they have
# been deleted.
def deleted?
@values.fetch(:deleted, false)
end
def to_s(*_args)
JSON.pretty_generate(to_hash)
end
def inspect
id_string = respond_to?(:id) && !id.nil? ? " id=#{id}" : ""
"#<#{self.class}:0x#{object_id.to_s(16)}#{id_string}> JSON: " +
JSON.pretty_generate(@values)
end
# Mass assigns attributes on the model.
#
# This is a version of +update_attributes+ that takes some extra options
# for internal use.
#
# ==== Attributes
#
# * +values+ - Hash of values to use to update the current attributes of
# the object.
# * +opts+ - Options for +StripeObject+ like an API key that will be reused
# on subsequent API calls.
#
# ==== Options
#
# * +:dirty+ - Whether values should be initiated as "dirty" (unsaved) and
# which applies only to new StripeObjects being initiated under this
# StripeObject. Defaults to true.
def update_attributes(values, opts = {}, dirty: true)
values.each do |k, v|
add_accessors([k], values) unless metaclass.method_defined?(k.to_sym)
@values[k] = Util.convert_to_stripe_object(v, opts)
dirty_value!(@values[k]) if dirty
@unsaved_values.add(k)
end
end
def [](key)
@values[key.to_sym]
end
def []=(key, value)
send(:"#{key}=", value)
end
def keys
@values.keys
end
def values
@values.values
end
def to_json(*_opts)
# TODO: pass opts to JSON.generate?
JSON.generate(@values)
end
def as_json(*opts)
@values.as_json(*opts)
end
def to_hash
maybe_to_hash = lambda do |value|
return nil if value.nil?
value.respond_to?(:to_hash) ? value.to_hash : value
end
@values.each_with_object({}) do |(key, value), acc|
acc[key] = case value
when Array
value.map(&maybe_to_hash)
else
maybe_to_hash.call(value)
end
end
end
def each(&blk)
@values.each(&blk)
end
# Sets all keys within the StripeObject as unsaved so that they will be
# included with an update when #serialize_params is called. This method is
# also recursive, so any StripeObjects contained as values or which are
# values in a tenant array are also marked as dirty.
def dirty!
@unsaved_values = Set.new(@values.keys)
@values.each_value do |v|
dirty_value!(v)
end
end
# Implements custom encoding for Ruby's Marshal. The data produced by this
# method should be comprehendable by #marshal_load.
#
# This allows us to remove certain features that cannot or should not be
# serialized.
def marshal_dump
# The StripeClient instance in @opts is not serializable and is not
# really a property of the StripeObject, so we exclude it when
# dumping
opts = @opts.clone
opts.delete(:client)
[@values, opts]
end
# Implements custom decoding for Ruby's Marshal. Consumes data that's
# produced by #marshal_dump.
def marshal_load(data)
values, opts = data
initialize(values[:id])
initialize_from(values, opts)
end
def serialize_params(options = {})
update_hash = {}
@values.each do |k, v|
# There are a few reasons that we may want to add in a parameter for
# update:
#
# 1. The `force` option has been set.
# 2. We know that it was modified.
# 3. Its value is a StripeObject. A StripeObject may contain modified
# values within in that its parent StripeObject doesn't know about.
#
unsaved = @unsaved_values.include?(k)
next unless options[:force] || unsaved || v.is_a?(StripeObject)
update_hash[k.to_sym] = serialize_params_value(
@values[k], @original_values[k], unsaved, options[:force], key: k
)
end
# a `nil` that makes it out of `#serialize_params_value` signals an empty
# value that we shouldn't appear in the serialized form of the object
update_hash.reject! { |_, v| v.nil? }
update_hash
end
# A protected field is one that doesn't get an accessor assigned to it
# (i.e. `obj.public = ...`) and one which is not allowed to be updated via
# the class level `Model.update(id, { ... })`.
def self.protected_fields
[]
end
protected def metaclass
class << self; self; end
end
protected def remove_accessors(keys)
# not available in the #instance_eval below
protected_fields = self.class.protected_fields
metaclass.instance_eval do
keys.each do |k|
next if protected_fields.include?(k)
next if @@permanent_attributes.include?(k)
# Remove methods for the accessor's reader and writer.
[k, :"#{k}=", :"#{k}?"].each do |method_name|
next unless method_defined?(method_name)
begin
remove_method(method_name)
rescue NameError
# In some cases there can be a method that's detected with
# `method_defined?`, but which cannot be removed with
# `remove_method`, even though it's on the same class. The only
# case so far that we've noticed this is when a class is
# reopened for monkey patching:
#
# https://github.com/stripe/stripe-ruby/issues/749
#
# Here we swallow that error and issue a warning so at least
# the program doesn't crash.
warn("WARNING: Unable to remove method `#{method_name}`; " \
"if custom, please consider renaming to a name that doesn't " \
"collide with an API property name.")
end
end
end
end
end
protected def add_accessors(keys, values)
# not available in the #instance_eval below
protected_fields = self.class.protected_fields
metaclass.instance_eval do
keys.each do |k|
next if protected_fields.include?(k)
next if @@permanent_attributes.include?(k)
if k == :method
# Object#method is a built-in Ruby method that accepts a symbol
# and returns the corresponding Method object. Because the API may
# also use `method` as a field name, we check the arity of *args
# to decide whether to act as a getter or call the parent method.
define_method(k) { |*args| args.empty? ? @values[k] : super(*args) }
else
define_method(k) { @values[k] }
end
define_method(:"#{k}=") do |v|
if v == ""
raise ArgumentError, "You cannot set #{k} to an empty string. " \
"We interpret empty strings as nil in requests. " \
"You may set (object).#{k} = nil to delete the property."
end
@values[k] = Util.convert_to_stripe_object(v, @opts)
dirty_value!(@values[k])
@unsaved_values.add(k)
end
if [FalseClass, TrueClass].include?(values[k].class)
define_method(:"#{k}?") { @values[k] }
end
end
end
end
# Disabling the cop because it's confused by the fact that the methods are
# protected, but we do define `#respond_to_missing?` just below. Hopefully
# this is fixed in more recent Rubocop versions.
# rubocop:disable Style/MissingRespondToMissing
protected def method_missing(name, *args)
# TODO: only allow setting in updateable classes.
if name.to_s.end_with?("=")
attr = name.to_s[0...-1].to_sym
# Pull out the assigned value. This is only used in the case of a
# boolean value to add a question mark accessor (i.e. `foo?`) for
# convenience.
val = args.first
# the second argument is only required when adding boolean accessors
add_accessors([attr], attr => val)
begin
mth = method(name)
rescue NameError
raise NoMethodError,
"Cannot set #{attr} on this object. HINT: you can't set: " \
"#{@@permanent_attributes.to_a.join(', ')}"
end
return mth.call(args[0])
elsif @values.key?(name)
return @values[name]
end
begin
super
rescue NoMethodError => e
# If we notice the accessed name if our set of transient values we can
# give the user a slightly more helpful error message. If not, just
# raise right away.
raise unless @transient_values.include?(name)
raise NoMethodError,
e.message + ". HINT: The '#{name}' attribute was set in the " \
"past, however. It was then wiped when refreshing the object " \
"with the result returned by Stripe's API, probably as a " \
"result of a save(). The attributes currently available on " \
"this object are: #{@values.keys.join(', ')}"
end
end
# rubocop:enable Style/MissingRespondToMissing
protected def respond_to_missing?(symbol, include_private = false)
@values && @values.key?(symbol) || super
end
# Re-initializes the object based on a hash of values (usually one that's
# come back from an API call). Adds or removes value accessors as necessary
# and updates the state of internal data.
#
# Protected on purpose! Please do not expose.
#
# ==== Options
#
# * +:values:+ Hash used to update accessors and values.
# * +:opts:+ Options for StripeObject like an API key.
# * +:partial:+ Indicates that the re-initialization should not attempt to
# remove accessors.
protected def initialize_from(values, opts, partial = false)
@opts = Util.normalize_opts(opts)
# the `#send` is here so that we can keep this method private
@original_values = self.class.send(:deep_copy, values)
removed = partial ? Set.new : Set.new(@values.keys - values.keys)
added = Set.new(values.keys - @values.keys)
# Wipe old state before setting new. This is useful for e.g. updating a
# customer, where there is no persistent card parameter. Mark those
# values which don't persist as transient
remove_accessors(removed)
add_accessors(added, values)
removed.each do |k|
@values.delete(k)
@transient_values.add(k)
@unsaved_values.delete(k)
end
update_attributes(values, opts, dirty: false)
values.each_key do |k|
@transient_values.delete(k)
@unsaved_values.delete(k)
end
self
end
protected def serialize_params_value(value, original, unsaved, force,
key: nil)
if value.nil?
""
# The logic here is that essentially any object embedded in another
# object that had a `type` is actually an API resource of a different
# type that's been included in the response. These other resources must
# be updated from their proper endpoints, and therefore they are not
# included when serializing even if they've been modified.
#
# There are _some_ known exceptions though.
#
# For example, if the value is unsaved (meaning the user has set it), and
# it looks like the API resource is persisted with an ID, then we include
# the object so that parameters are serialized with a reference to its
# ID.
#
# Another example is that on save API calls it's sometimes desirable to
# update a customer's default source by setting a new card (or other)
# object with `#source=` and then saving the customer. The
# `#save_with_parent` flag to override the default behavior allows us to
# handle these exceptions.
#
# We throw an error if a property was set explicitly but we can't do
# anything with it because the integration is probably not working as the
# user intended it to.
elsif value.is_a?(APIResource) && !value.save_with_parent
if !unsaved
nil
elsif value.respond_to?(:id) && !value.id.nil?
value
else
raise ArgumentError, "Cannot save property `#{key}` containing " \
"an API resource. It doesn't appear to be persisted and is " \
"not marked as `save_with_parent`."
end
elsif value.is_a?(Array)
update = value.map { |v| serialize_params_value(v, nil, true, force) }
# This prevents an array that's unchanged from being resent.
update if update != serialize_params_value(original, nil, true, force)
# Handle a Hash for now, but in the long run we should be able to
# eliminate all places where hashes are stored as values internally by
# making sure any time one is set, we convert it to a StripeObject. This
# will simplify our model by making data within an object more
# consistent.
#
# For now, you can still run into a hash if someone appends one to an
# existing array being held by a StripeObject. This could happen for
# example by appending a new hash onto `additional_owners` for an
# account.
elsif value.is_a?(Hash)
Util.convert_to_stripe_object(value, @opts).serialize_params
elsif value.is_a?(StripeObject)
update = value.serialize_params(force: force)
# If the entire object was replaced and this is an additive object,
# then we need blank each field of the old object that held a value
# because otherwise the update to the keys of the object will be
# additive instead of a full replacement. The new serialized values
# will override any of these empty values.
if original && unsaved && key && self.class.additive_object_param?(key)
update = empty_values(original).merge(update)
end
update
else
value
end
end
# Produces a deep copy of the given object including support for arrays,
# hashes, and StripeObjects.
private_class_method def self.deep_copy(obj)
case obj
when Array
obj.map { |e| deep_copy(e) }
when Hash
obj.each_with_object({}) do |(k, v), copy|
copy[k] = deep_copy(v)
copy
end
when StripeObject
obj.class.construct_from(
deep_copy(obj.instance_variable_get(:@values)),
obj.instance_variable_get(:@opts).select do |k, _v|
Util::OPTS_COPYABLE.include?(k)
end
)
else
obj
end
end
private def dirty_value!(value)
case value
when Array
value.map { |v| dirty_value!(v) }
when StripeObject
value.dirty!
end
end
# Returns a hash of empty values for all the values that are in the given
# StripeObject.
private def empty_values(obj)
values = case obj
when Hash then obj
when StripeObject then obj.instance_variable_get(:@values)
else
raise ArgumentError,
"#empty_values got unexpected object type: " \
"#{obj.class.name}"
end
values.each_with_object({}) do |(k, _), update|
update[k] = ""
end
end
end
end