When populating `StripeObject`s, we add accessors to them so that people
can access fields like `obj.currency`.
This was probably only meant to apply to API resources, but through
what might have been an accident of history, we've also traditionally
unmarshaled any hash that comes back from the API as a `StripeObject`,
including `metadata` fields. This allows some convenience because users
can access values like `obj.metadata.my_field`, but is also obviously a
minefield for potential problems.
In issue #969, what's essentially happening is that because there's a
metadata field named `class`, we've overwritten the object's normal
`class` method with our own custom one that accesses the metadata value.
Amazingly, the object can still marshal/unmarshal mostly properly, but
fails on this line as we try to access `obj.class` and that turns out to
be a metadata value instead of a class:
``` ruby
when StripeObject
obj.class.construct_from(
...
```
Here I solve the problem by banning accessors added with the name
`class`. This has a slight risk of backward incompatibility in that
users that previously had metadata named "class" will now have to use
square bracket accessors instead like `obj.metadata[:class]`, but
honestly, I just can't see anything good in allowing "class" to be used
as an accessor.
An alternative solution might be to alias `class` in `StripeObject` and
then make sure we always use that in places like `initialize_from` and
`deep_copy`.
The best long term solution would be to stop add accessors to metadata
objects. This just seems like a bad idea given that there are still
myriads of Ruby built-ins that could potentially be overwritten. This is
definitely a considerably-sized breaking change though, so we'd have to
do it on a major.
* 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)
* Bump Rubocop to 0.57.2
* Style/StderrPuts: Use warn instead of .puts
* Style/ExpandPathArguments: Use expand_path('../test_helper', __dir__) instead of expand_path('../../test_helper', __FILE__)
* Style/Encoding: Unnecessary utf-8 encoding comment
* Style/StringLiterals: Prefer double-quoted strings
* Style/AccessModifierDeclarations
* Style/FormatStringToken: Prefer annotated tokens
* Naming/UncommunicativeMethodParamName
* Metrics/LineLength: set maximum line length to 100 characters
* Style/IfUnlessModifier: Favor modifier if usage when having a single-line body
* Style/ClassVars
* Metrics/LineLength: set maximum line length to 80 characters (default)
* Style/AccessModifierDeclarations: EnforcedStyle: inline
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 like in
issue #749.
We'll still try to discourage this sort of use, but here we swallow the
error and issue a warning so at least the program doesn't crash.
Fixes#749.
Overrides `#eql?` (hash equality) and `#hash` so that Stripe objects can
be used more easily as Hash keys and that certain other frameworks that
rely on these methods will have an easier time (e.g. RSpec's `change`,
see #687).
I think this might be a little controversial if we weren't already
overriding the `#==` implementation, but because we are, I think it
makes sense to extent it to these two methods as well.
Adds the magic `frozen_string_literal: true` comment to every file and
enables a Rubocop rule to make sure that it's always going to be there
going forward as well.
See here for more background [1], but the basic idea is that unlike many
other languages, static strings in code are mutable by default. This has
since been acknowledged as not a particularly good idea, and the
intention is to rectify the mistake when Ruby 3 comes out, where all
string literals will be frozen. The `frozen_string_literal` magic
comment was introduced in Ruby 2.3 as a way of easing the transition,
and allows libraries and projects to freeze their literals in advance.
I don't think this is breaking in any way: it's possible that users
might've been pulling out one of are literals somehow and mutating it,
but that would probably not have been useful for anything and would
certainly not be recommended, so I'm quite comfortable pushing this
change through as a minor version.
As discussed in #641.
[1] https://stackoverflow.com/a/37799399
The test suite is currently throwing a bunch of warnings from some
recent changes I made -- although we initialize `@additive_params` when
setting one with `self.additive_object_param`, we don't when we check
one with `self.additive_object_param?`. This often isn't a problem
because every API resource sets `metadata`, but it is from the test
suite and probably for vanilla `StripeObject`s too.
So we have a bit of a problem right now when it comes to replacing a
`StripeObject` that's embedded in an API resource.
Most of the time when someone does this, they want to _replace_ an
object embedded in another object. Take setting a source on a
subscription for example:
``` ruby
subscription.source = {
object: 'card',
number: 123,
}
subscription.save
```
In the case above, the serialized parameters should come out as:
```
source[object]=card&source[number]=123
```
That should apply even if the previous source had something else set on
it which we're not going to set this time -- say an optional parameter
like `source[address_state]`. Those should not be present at all in the
final serialized parameters.
(Another example is setting a `payout_schedule` as seen in #631 which is
PR is intended to address.)
There is an exception to this rule in the form of metadata though.
Metadata is a bit of a strange case in that the API will treat it as
additive, so if we send `metadata[foo]`, that will set the `foo` key,
but it won't overwrite any other keys that were already present.
This is a problem because when a user fully sets `metadata` to a new
object in Ruby, what they're probably trying to do is _replace_ it
rather than add to it. For example:
``` ruby
subscription.metadata
=> { old: 'bar' }
subscription.metadata = {
new: 'baz'
}
subscription.save
```
To accomplish what the user is probably trying to do, we actually need
to send `metadata[old]=&metadata[new]=baz` so that we empty the value of
`old` while simultaneously setting `new` to `baz`.
In summary, metadata behaves different from other embedded objects in a
fairly fundamental way, and because the code is currently only set up to
handle the metadata case, it's not behaving correctly when other types
of objects are being set. A lot of the time emptying values like we do
for `metadata` is benign, but as we've seen in #631, sometimes it's not.
In this patch, I modify serialization to only empty out object values
when we see that parameter is `metadata`.
I'm really not crazy about the implementation here _at all_, but I'm
having trouble thinking of a better way to do it. One possibility is to
introduce a new class annotation like `empty_embedded_object :metadata`,
but that will have to go everywhere and might be error-prone in case
someone forgets it on a new resource type. If anyone has a suggestion
for an alternative (or can let me know if I'm missing something), I'd
love to hear it.
This PR is an alternate to #631.
Backtracks a little bit #586 by bringing back custom `StripeObject`
encoding and decoding methods for Ruby's `Marshal`. These work by just
persisting values and some opts, and skipping everything else. It's
mostly the same as what we had before, but implemented a little more
cleanly so that we don't actually need to invoke `Marshal` anywhere
ourselves.
In #586 we still managed to remove all the uses of `Marshal` in our own
codebase to make the linter happy. Even though we wouldn't recommend the
use of `Marshal`, this code at least enables it for anyone using a Rails
cache or similar mechanism.
Addresses #90.
We were previously using a bit of a hack to get a free deep copy
implementation through Ruby's marshaling framework. Lint call this out
as a security problem though, and rightfully so: when combined with
unsanitized user input, unmarshaling can result in very serious security
breaches involving arbitrary code execution.
This patch removes all uses of marshal/unmarshal in favor of
implementing a deep copy method for `StripeObject`. I also reworked some
of the constants around what keys are available for `opts`. I'm still
not completely happy with the results, but I think it's going to need a
slightly larger refactor in order to get somewhere truly good.
There is what could be a breaking change for people doing non-standard
stuff with the library: the opts that we copy with an object are now
whitelisted, so if they were being used to pass around extraneous data,
that might not work as expected anymore. But because this is a contract
that we never committed to, I don't think I'd bump the major version for
change.
Removes Rubocop TODO around guard clauses and fixes the outstanding
offenses.
This is starting to get into territory that feels of more dubious value
to me, but at least it did get me writing a couple more tests, so let's
see how it goes by keeping this on.
I wanted to see what fixing Rubocop TODOs was like, so I tried to
eliminate all the easy ones. Most of these were pretty easy, and the
changes required are relatively minimal.
Some of the stuff left is harder. Pretty much everything under
`Metrics/*` is going to be a pretty big yak shave. A few of the others
are just going to need a little more work (e.g. `Style/ClassVars` and
`Style/GuardClause`). Going to stop here for now.
Tweaks the serialization behavior so that when a resource is explicitly
set to a resource's field and that resource is subsequently saved, then
if it looks like the set resource was persisted we extract its ID and
send it up to the API.
By slight extension we also throw an `ArgumentError` if it looks like
that set resource was _not_ persisted because if the user set it
explicitly then it was probably not their intention to have it silently
ignored by the library in the event of a problem.
Previously, the value of whatever accessor was missing was left out of
the call to build it. This had the effect of skipping over the
lazily-built predicate method when the missing accessor is for a
boolean.
Of course, I realize this all is a bit contrived as most of the time
folks aren't assigning these values manually to stripe objects. However,
in my testing it caught me by surprised that the behavior is asymmetric
depending on how and when values are assigned.
As such I believe this is a less surprising implementation.
As described in #481, adding a protected field like `legal_entity` as
part of an update API operation can cause some issues like a custom
encoding scheme not being considered and special handling around empty
values being ignored.
As a an easy fix for this, let's disallow access to protected fields in
the same way that we disallow them from being set directly on an
instance of a given model.
Helps address (but is not a complete fix for) #481.
Introduce a `#save_with_parent` flag that allows the default behavior of
never saving API resources nested under a parent to be overridden, a
feature that we so far only know to need for updating a source under a
customer.
I'm not sure exactly what changed here (did we change the `$VERBOSE`
setting?), but I'm not seeing a whole lot of warnings when running the
test suites locally and in CI. For example:
```
Started
........................................./home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
............../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
......../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
.../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
........./home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
...
..../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
....../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
..../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
......./home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
........./home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
........../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
................./home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
.../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
..../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
....../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
..........
........./home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
....../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
......../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
......../home/travis/build/stripe/stripe-ruby/lib/stripe/api_operations/list.rb:6: warning: instance variable @opts not initialized
............./home/travis/build/stripe/stripe-ruby/lib/stripe/stripe_object.rb:35: warning: instance variable @values not initialized
./home/travis/build/stripe/stripe-ruby/lib/stripe/stripe_object.rb:35: warning: instance variable @values not initialized
...................../home/travis/build/stripe/stripe-ruby/lib/stripe/transfer.rb:8: warning: instance variable @api_key not initialized
..............
..
Finished in 0.785621037 seconds.
```
Most of these are due to unused or uninitialized variables. This patch
fixes all warnings by fixing offending code.
Prior to my last major serialization refactor, there was a check in the
code that would remove any subobjects from serialization that were of
their own proper resource type (for example, if a charge contained a
customer, that customer would be removed).
What I didn't realize at the time is that the old serialization code had
a bug/quirk that would allow *certain types* of subobjects that were API
resources to make it through unscathed.
In short, the behavior requirement here is *directly* contradictory.
There was a test in place that would make sure that `customer` was
removed from this hash:
``` ruby
{
:id => 'ch_id',
:object => 'charge',
:customer => {
:object => 'customer',
:id => 'customer_id'
}
}
```
But, as reported in #406, we expect, and indeed need, for `source` (a
card) to make it through to the API in this hash:
``` ruby
{
:id => 'cus_id',
:object => 'customer',
:source => {
:object => 'card',
:id => 'card_id'
}
}
```
My proposal here is to just remove the check on serializing API
resources. The normal code that only sends up keys/hashes that have
changed is still in place, so in the first example, `customer` still
isn't sent unless the user has directly manipulated a field on that
subobject. I propose that in those cases we allow the API itself to
reject the request rather than try to cut it off at the client level.
Unfortunately, there is some possibility that removing those API
resources is important for some reason, but of course there's no
documentation on it beyond the after-the-fact post-justification that I
wrote during my last refactor. I can't think of any reason that it would
be too destructive, but there is some level of risk.
This pull does two major things:
1. Refactors `serialize_params` to be more concise and readable while
still complying to our existing test suite. Unfortunately over time
this method has become a ball of mud that's very difficult to reason
about, as recently evidenced by #384.
2. Moves `serialize_params` from class method to instance method (while
still keeping for old class method for backwards compatibility). This
is to give it a more sane interface.
Hashes are converted to StripeObject when used as params of save.
They need to be converted to hash on serialize.
Signed-off-by: François de Metz <francois@stormz.me>
Signed-off-by: Cyril Duez <cyril@stormz.me>
This is kind of a weird one because it'll only cause a failure when
serializing a subobject or hash of a `StripeObject`, but it's good
practice to initialize instance variables anyway.
Fixes#360.
