See b/30806813 for more context. Copied from there: This appears to be happening if we get an EPP domain create command that is missing any contacts (but has a registrant; with no registrant we exercise a different codepath). In that case, JAXB leaves the contacts field on the Create null, and we try to pass it into Sets.union() as a result of Corey's refactoring in [] that changed contact loading to load the contacts and registrant all at once. The fix is just to apply nullToEmpty() first.
Note that it's always an error to try to create a domain without any non-registrant contacts, but that's supposed to happen later on in BaseDomainCreateFlow.verifyCreateIsAllowed() via validateRequiredContactsPresent(), which will produce a nice error message.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130309019
Adds the ability to search for entities (contacts and registrars) by name.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130305930
The ICANN operational profile says:
1.4.12. An entity object within an RDAP response MUST contain an events
member with the following events:
o An event of eventAction type registration.
o An event of eventAction type last changed. The event of eventAction type last changed MUST be omitted if the Contact Object (as defined in RFC5733) has not been updated since it was created.
o An event of eventAction type last update of RDAP database.
It has similar wording for domains and hosts. The registration and last changed events are already being generated. This CL adds a "last update of RDAP database" event for all three object types. It's very redundant, and I have warned ICANN that "last update" doesn't mean what they might expect in the context of an eventually consistent database, but there we are.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130279688
The instrumentation implementation adds the ability to track counters for dimensional data in the codebase. Instrumentation points will be added to different parts of the codebase in upcoming CLs. The ability to export the data to Stackdriver will also be added in an upcoming CL.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130195199
Support creating domain with gtech_tool, instead of creating temporary epp file
and run execute_epp manually. Also changes CreateContactCommand for consistency.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130127280
This CL adds the hooks necessary to implement TLD-specific flow info and update flow logic. Usage of the hooks follows in a separate CL.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130108832
I forgot to inject the CreateTldCommand/UpdateTldCommand commands, which now
need to be injected so that the list of available DnsWriter implementations is
instantiated. This CL also adds a new DI Set<String> with just the name of the
writer, so that the instantiated writer map (which may have many DI dependencies)
doesn't need to be pulled in.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130042215
According to Gustavo Lozano of ICANN: "In the case of the RDAP profile (gTLD space), the “port43” element is not expected to be used, because Whois/43 tcp will be deprecated in the future." So it sounds like we should not include the port43 element for the moment.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130017966
ECatcher seems to trigger on warnings, even if no exception is thrown (?). This warning doesn't really need to be a warning if it's WAI and not something we're actually going to bother to reach out to registrars about (in this case, 101domain apparently isn't declaring the launch extension, but we let them use it anyway).
If at some point we decide to be stricter about declaring extensions and want to go determine how often this is happening, making it INFO vs WARNING doesn't make that any harder.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=130016606
This CL has minimal[*] immediate impact on the targets, but it has two consequences moving forward:
- When [] moves to Java 8, compatible_with=appengine targets will remain on Java 7. (If this target is a non-java_* target, you can ignore this bullet.)
- compatible_with=appengine targets can depend only on other compatible_with=appengine targets. (Fortunately, most common libraries will be compatible_with=appengine: []
This CL is one of thousands in a migration on the critical path for Java 8, so please respond promptly, and, if you are the wrong reviewer, please reassign the CL.
[*] It does mean that your target will go back to being compiled to Java 7 bytecode after its recent switch to Java 8 ([]If your target compiles the same files as another non-compatible_with=appengine target, this CL may break one_definition_test targets.
More information:
[]
Tested:
[] build --nobuild
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129660908
Also had to add an EnumParameter class to support
List<T extends Enum<T>>, as these aren't natively supported by
JCommander (although single Enum parameters are.)
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129464699
The XJC genrule in particular was dirtying /tmp which I believe may have
been causing problems for Donuts.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129451600
This is one of several CLs in a sequence for allowing per-TLD DNS
implementations.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129445641
The plumbing to support different DnsWriter implementations per TLD
using DI will come in a later CL.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129375856
Updates the RDAP code to reflect guidance and confirmation received from various mailing lists.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129254894
This is better than the previous way of using the canonical name of the class,
because the previous way did not allow for refactoring, and also required the
PremiumPricingEngine to live in the model package lest there be circular
dependencies, which does not seem ideal.
Note that, for reasons of backwards compatibility with existing persisted data,
the name of the static premium pricing engine has been set to its canonical
class name, but the class can now be refactored going forward so long as this
string remains unchanged, and any new pricing engine implementations can use
whatever string key they want (it doesn't have to be a canonical class name).
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129215185
These build rules allow Bazel to generate the .ear and .war files which
appcfg.sh (a tool that comes with the App Engine SDK) can then use to
perform a deployment.
Included in this CL are configurations for five separate production
environments: production, sandbox, alpha, crash, and local.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129163010
It turns out that if the script gets super huge, then Bazel will fail in
a mysterious way if you pass the script as a string to ctx.action.
This way of doing things is better anyway, because this makes it much
easier for the user to see the script that was actually generated for
troubleshooting purposes.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129114272
This doesn't work quite the same way as Fileset(). But it should be able
to serve as a decent replacement.
The important part of this design is that zips can depend on other zips.
Therefore definitions don't have to be monolithic. This will be
important when migrating the Domain Registry codebase, because our
Fileset() definitions are dispersed across many BUILD files. So we'll be
able to migrate to zip_file() with the least amount of intrusiveness.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129034885
In the case where currency has a zero scale (i.e. JPY),
Money.of(currency, 0) throws ArithmeticException because the second
argument is handled as a double, and JPY (for example) does not allow
decimal precision, even if the decimal is zero.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129023540
This change allows registries to customize the values returned by
RegistryConfig methods without needing to edit Domain Registry code in a
forked repository. This is accomplished by defining a custom
RegistryConfig implementation and specifying its name as a system
property in appengine-web.xml.
This change also open-sources the production configuration values that
Google has chosen to use for these methods. TestRegistryConfig was
hitherto used for production configuration in the open source world,
which is misleading and inappropriate, considering it tunes values such
as the number of commit log buckets to 1.
Another important benefit of this change is that it helps registry_tool
work out of the box in the open source world.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129022675
This will help [] be submitted without breaking the linter.
License headers are now added automatically where they were previously
added by hand. We're also now adding the license header to Soy and SQL
files.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=129017424
This is an internal-only feature that breaks the open source build.
CL created with:
dr-replace '(compatible_with.*)' '\1 # MOE:strip_line'
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=128852873
This CL fixes an issue preventing App Engine deploys in the open source
world, because an old version of Soy had its Maven jars compiled for Java
1.8. Upgrading to the latest Closure Rules release fixes this.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=128841077
This CL implements similar logic to deal with orphan glues as [] did
for ZonemanWriter. We are enforcing the policy that a glue record
should be deleted when authoritative NS record referring to it is
removed.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=128838082
goog.dom.createDom is about to support only the members of
goog.dom.TagName to improve the precision of its return type. This CL
prepares for that.
CL automatically created by:
replace_string --pcre
(\.createDom\(\s*)\'([a-z0-9]+)'
$1goog.dom.TagName.$2
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=128802686
This is one of several CLs in order to support per-TLD DnsWriter
implementations, modeled on the work done for PremiumPricingEngine.
Since DnsWriters will be set inside the Registry object, the DnsWriter
interface definition needs to be moved to models to create minimal
dependency on the rest of the registry codebase to avoid cyclic
dependency.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=128711643
When a domain refreshes, always delete all of its subordinate host
records and then add glue records for its in-bailiwick nameservers, if
the domain is in a publishable status. When a host refreshes, delete
its glue record (if any) and then refresh its superordinate domain. The
goal is to prevent A/AAAA records for hosts that are not used as
in-bailiwick nameservers from being published in the DNS.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=128354008
