Assertion

Package assertions contains the implementations for all assertions which are referenced in goconvey's `convey` package (github.com/smartystreets/goconvey/convey) and gunit (github.com/smartystreets/gunit) for use with the So(...) method. They can also be used in traditional Go test functions and even in applications. https://smartystreets.com Many of the assertions lean heavily on work done by Aaron Jacobs in his excellent oglematchers library. (https://github.com/jacobsa/oglematchers) The ShouldResemble assertion leans heavily on work done by Daniel Jacques in his very helpful go-render library. (https://github.com/luci/go-render)

Package assertions contains the implementations for all assertions which are referenced in goconvey's `convey` package (github.com/smarty/goconvey/convey) and gunit (github.com/smarty/gunit) for use with the So(...) method. They can also be used in traditional Go test functions and even in applications. https://smarty.com Many of the assertions lean heavily on work done by Aaron Jacobs in his excellent oglematchers library. (https://github.com/jacobsa/oglematchers) The ShouldResemble assertion leans heavily on work done by Daniel Jacques in his very helpful go-render library. (https://github.com/luci/go-render)

** We are working on testify v2 and would love to hear what you'd like to see in it, have your say here: https://cutt.ly/testify ** Package testify is a set of packages that provide many tools for testifying that your code will behave as you intend. testify contains the following packages: The assert package provides a comprehensive set of assertion functions that tie in to the Go testing system. The mock package provides a system by which it is possible to mock your objects and verify calls are happening as expected. The suite package provides a basic structure for using structs as testing suites, and methods on those structs as tests. It includes setup/teardown functionality in the way of interfaces.

Package sessions provides cookie and filesystem sessions and infrastructure for custom session backends. The key features are: Let's start with an example that shows the sessions API in a nutshell: First we initialize a session store calling NewCookieStore() and passing a secret key used to authenticate the session. Inside the handler, we call store.Get() to retrieve an existing session or a new one. Then we set some session values in session.Values, which is a map[interface{}]interface{}. And finally we call session.Save() to save the session in the response. Note that in production code, we should check for errors when calling session.Save(r, w), and either display an error message or otherwise handle it. Save must be called before writing to the response, otherwise the session cookie will not be sent to the client. That's all you need to know for the basic usage. Let's take a look at other options, starting with flash messages. Flash messages are session values that last until read. The term appeared with Ruby On Rails a few years back. When we request a flash message, it is removed from the session. To add a flash, call session.AddFlash(), and to get all flashes, call session.Flashes(). Here is an example: Flash messages are useful to set information to be read after a redirection, like after form submissions. There may also be cases where you want to store a complex datatype within a session, such as a struct. Sessions are serialised using the encoding/gob package, so it is easy to register new datatypes for storage in sessions: As it's not possible to pass a raw type as a parameter to a function, gob.Register() relies on us passing it a value of the desired type. In the example above we've passed it a pointer to a struct and a pointer to a custom type representing a map[string]interface. (We could have passed non-pointer values if we wished.) This will then allow us to serialise/deserialise values of those types to and from our sessions. Note that because session values are stored in a map[string]interface{}, there's a need to type-assert data when retrieving it. We'll use the Person struct we registered above: By default, session cookies last for a month. This is probably too long for some cases, but it is easy to change this and other attributes during runtime. Sessions can be configured individually or the store can be configured and then all sessions saved using it will use that configuration. We access session.Options or store.Options to set a new configuration. The fields are basically a subset of http.Cookie fields. Let's change the maximum age of a session to one week: Sometimes we may want to change authentication and/or encryption keys without breaking existing sessions. The CookieStore supports key rotation, and to use it you just need to set multiple authentication and encryption keys, in pairs, to be tested in order: New sessions will be saved using the first pair. Old sessions can still be read because the first pair will fail, and the second will be tested. This makes it easy to "rotate" secret keys and still be able to validate existing sessions. Note: for all pairs the encryption key is optional; set it to nil or omit it and and encryption won't be used. Multiple sessions can be used in the same request, even with different session backends. When this happens, calling Save() on each session individually would be cumbersome, so we have a way to save all sessions at once: it's sessions.Save(). Here's an example: This is possible because when we call Get() from a session store, it adds the session to a common registry. Save() uses it to save all registered sessions.

Package testify is a set of packages that provide many tools for testifying that your code will behave as you intend. testify contains the following packages: The assert package provides a comprehensive set of assertion functions that tie in to the Go testing system. The http package contains tools to make it easier to test http activity using the Go testing system. The mock package provides a system by which it is possible to mock your objects and verify calls are happening as expected. The suite package provides a basic structure for using structs as testing suites, and methods on those structs as tests. It includes setup/teardown functionality in the way of interfaces.

Package testify is a set of packages that provide many tools for testifying that your code will behave as you intend. testify contains the following packages: The assert package provides a comprehensive set of assertion functions that tie in to the Go testing system. The http package contains tools to make it easier to test http activity using the Go testing system. The mock package provides a system by which it is possible to mock your objects and verify calls are happening as expected. The suite package provides a basic structure for using structs as testing suites, and methods on those structs as tests. It includes setup/teardown functionality in the way of interfaces.

Package redis is a client for the Redis database. The Redigo FAQ (https://github.com/gomodule/redigo/wiki/FAQ) contains more documentation about this package. The Conn interface is the primary interface for working with Redis. Applications create connections by calling the Dial, DialWithTimeout or NewConn functions. In the future, functions will be added for creating sharded and other types of connections. The application must call the connection Close method when the application is done with the connection. The Conn interface has a generic method for executing Redis commands: The Redis command reference (http://redis.io/commands) lists the available commands. An example of using the Redis APPEND command is: The Do method converts command arguments to bulk strings for transmission to the server as follows: Redis command reply types are represented using the following Go types: Use type assertions or the reply helper functions to convert from interface{} to the specific Go type for the command result. Connections support pipelining using the Send, Flush and Receive methods. Send writes the command to the connection's output buffer. Flush flushes the connection's output buffer to the server. Receive reads a single reply from the server. The following example shows a simple pipeline. The Do method combines the functionality of the Send, Flush and Receive methods. The Do method starts by writing the command and flushing the output buffer. Next, the Do method receives all pending replies including the reply for the command just sent by Do. If any of the received replies is an error, then Do returns the error. If there are no errors, then Do returns the last reply. If the command argument to the Do method is "", then the Do method will flush the output buffer and receive pending replies without sending a command. Use the Send and Do methods to implement pipelined transactions. Connections support one concurrent caller to the Receive method and one concurrent caller to the Send and Flush methods. No other concurrency is supported including concurrent calls to the Do and Close methods. For full concurrent access to Redis, use the thread-safe Pool to get, use and release a connection from within a goroutine. Connections returned from a Pool have the concurrency restrictions described in the previous paragraph. Use the Send, Flush and Receive methods to implement Pub/Sub subscribers. The PubSubConn type wraps a Conn with convenience methods for implementing subscribers. The Subscribe, PSubscribe, Unsubscribe and PUnsubscribe methods send and flush a subscription management command. The receive method converts a pushed message to convenient types for use in a type switch. The Bool, Int, Bytes, String, Strings and Values functions convert a reply to a value of a specific type. To allow convenient wrapping of calls to the connection Do and Receive methods, the functions take a second argument of type error. If the error is non-nil, then the helper function returns the error. If the error is nil, the function converts the reply to the specified type: The Scan function converts elements of a array reply to Go types: Connection methods return error replies from the server as type redis.Error. Call the connection Err() method to determine if the connection encountered non-recoverable error such as a network error or protocol parsing error. If Err() returns a non-nil value, then the connection is not usable and should be closed. This example implements ZPOP as described at http://redis.io/topics/transactions using WATCH/MULTI/EXEC and scripting.

Package saml contains a partial implementation of the SAML standard in golang. SAML is a standard for identity federation, i.e. either allowing a third party to authenticate your users or allowing third parties to rely on us to authenticate their users. In SAML parlance an Identity Provider (IDP) is a service that knows how to authenticate users. A Service Provider (SP) is a service that delegates authentication to an IDP. If you are building a service where users log in with someone else's credentials, then you are a Service Provider. This package supports implementing both service providers and identity providers. The core package contains the implementation of SAML. The package samlsp provides helper middleware suitable for use in Service Provider applications. The package samlidp provides a rudimentary IDP service that is useful for testing or as a starting point for other integrations. Version 0.4.0 introduces a few breaking changes to the _samlsp_ package in order to make the package more extensible, and to clean up the interfaces a bit. The default behavior remains the same, but you can now provide interface implementations of _RequestTracker_ (which tracks pending requests), _Session_ (which handles maintaining a session) and _OnError_ which handles reporting errors. Public fields of _samlsp.Middleware_ have changed, so some usages may require adjustment. See [issue 231](https://github.com/crewjam/saml/issues/231) for details. The option to provide an IDP metadata URL has been deprecated. Instead, we recommend that you use the `FetchMetadata()` function, or fetch the metadata yourself and use the new `ParseMetadata()` function, and pass the metadata in _samlsp.Options.IDPMetadata_. Similarly, the _HTTPClient_ field is now deprecated because it was only used for fetching metdata, which is no longer directly implemented. The fields that manage how cookies are set are deprecated as well. To customize how cookies are managed, provide custom implementation of _RequestTracker_ and/or _Session_, perhaps by extending the default implementations. The deprecated fields have not been removed from the Options structure, but will be in future. In particular we have deprecated the following fields in _samlsp.Options_: - `Logger` - This was used to emit errors while validating, which is an anti-pattern. - `IDPMetadataURL` - Instead use `FetchMetadata()` - `HTTPClient` - Instead pass httpClient to FetchMetadata - `CookieMaxAge` - Instead assign a custom CookieRequestTracker or CookieSessionProvider - `CookieName` - Instead assign a custom CookieRequestTracker or CookieSessionProvider - `CookieDomain` - Instead assign a custom CookieRequestTracker or CookieSessionProvider - `CookieDomain` - Instead assign a custom CookieRequestTracker or CookieSessionProvider Let us assume we have a simple web application to protect. We'll modify this application so it uses SAML to authenticate users. ```golang package main import ( ) ``` Each service provider must have an self-signed X.509 key pair established. You can generate your own with something like this: We will use `samlsp.Middleware` to wrap the endpoint we want to protect. Middleware provides both an `http.Handler` to serve the SAML specific URLs and a set of wrappers to require the user to be logged in. We also provide the URL where the service provider can fetch the metadata from the IDP at startup. In our case, we'll use [samltest.id](https://samltest.id/), an identity provider designed for testing. ```golang package main import ( ) ``` Next we'll have to register our service provider with the identity provider to establish trust from the service provider to the IDP. For [samltest.id](https://samltest.id/), you can do something like: Navigate to https://samltest.id/upload.php and upload the file you fetched. Now you should be able to authenticate. The flow should look like this: 1. You browse to `localhost:8000/hello` 1. The middleware redirects you to `https://samltest.id/idp/profile/SAML2/Redirect/SSO` 1. samltest.id prompts you for a username and password. 1. samltest.id returns you an HTML document which contains an HTML form setup to POST to `localhost:8000/saml/acs`. The form is automatically submitted if you have javascript enabled. 1. The local service validates the response, issues a session cookie, and redirects you to the original URL, `localhost:8000/hello`. 1. This time when `localhost:8000/hello` is requested there is a valid session and so the main content is served. Please see `example/idp/` for a substantially complete example of how to use the library and helpers to be an identity provider. The SAML standard is huge and complex with many dark corners and strange, unused features. This package implements the most commonly used subset of these features required to provide a single sign on experience. The package supports at least the subset of SAML known as [interoperable SAML](http://saml2int.org). This package supports the Web SSO profile. Message flows from the service provider to the IDP are supported using the HTTP Redirect binding and the HTTP POST binding. Message flows from the IDP to the service provider are supported via the HTTP POST binding. The package can produce signed SAML assertions, and can validate both signed and encrypted SAML assertions. It does not support signed or encrypted requests. The _RelayState_ parameter allows you to pass user state information across the authentication flow. The most common use for this is to allow a user to request a deep link into your site, be redirected through the SAML login flow, and upon successful completion, be directed to the originally requested link, rather than the root. Unfortunately, _RelayState_ is less useful than it could be. Firstly, it is not authenticated, so anything you supply must be signed to avoid XSS or CSRF. Secondly, it is limited to 80 bytes in length, which precludes signing. (See section 3.6.3.1 of SAMLProfiles.) The SAML specification is a collection of PDFs (sadly): - [SAMLCore](http://docs.oasis-open.org/security/saml/v2.0/saml-core-2.0-os.pdf) defines data types. - [SAMLBindings](http://docs.oasis-open.org/security/saml/v2.0/saml-bindings-2.0-os.pdf) defines the details of the HTTP requests in play. - [SAMLProfiles](http://docs.oasis-open.org/security/saml/v2.0/saml-profiles-2.0-os.pdf) describes data flows. - [SAMLConformance](http://docs.oasis-open.org/security/saml/v2.0/saml-conformance-2.0-os.pdf) includes a support matrix for various parts of the protocol. [SAMLtest](https://samltest.id/) is a testing ground for SAML service and identity providers. Please do not report security issues in the issue tracker. Rather, please contact me directly at ross@kndr.org ([PGP Key `78B6038B3B9DFB88`](https://keybase.io/crewjam)).

Package cli provides a thin CLI abstraction around the standard flag package. It is minimal, command-struct-oriented, and trades off "power" for flexibility and clarity at the caller level. It pretends that Go's single dash (-flag) support doesn't exist, and renders helps with --. Optional interface adherence can be asserted with a statement like

Package logr defines a general-purpose logging API and abstract interfaces to back that API. Packages in the Go ecosystem can depend on this package, while callers can implement logging with whatever backend is appropriate. Logging is done using a Logger instance. Logger is a concrete type with methods, which defers the actual logging to a LogSink interface. The main methods of Logger are Info() and Error(). Arguments to Info() and Error() are key/value pairs rather than printf-style formatted strings, emphasizing "structured logging". With Go's standard log package, we might write: With logr's structured logging, we'd write: Errors are much the same. Instead of: We'd write: Info() and Error() are very similar, but they are separate methods so that LogSink implementations can choose to do things like attach additional information (such as stack traces) on calls to Error(). Error() messages are always logged, regardless of the current verbosity. If there is no error instance available, passing nil is valid. Often we want to log information only when the application in "verbose mode". To write log lines that are more verbose, Logger has a V() method. The higher the V-level of a log line, the less critical it is considered. Log-lines with V-levels that are not enabled (as per the LogSink) will not be written. Level V(0) is the default, and logger.V(0).Info() has the same meaning as logger.Info(). Negative V-levels have the same meaning as V(0). Error messages do not have a verbosity level and are always logged. Where we might have written: We can write: Logger instances can have name strings so that all messages logged through that instance have additional context. For example, you might want to add a subsystem name: The WithName() method returns a new Logger, which can be passed to constructors or other functions for further use. Repeated use of WithName() will accumulate name "segments". These name segments will be joined in some way by the LogSink implementation. It is strongly recommended that name segments contain simple identifiers (letters, digits, and hyphen), and do not contain characters that could muddle the log output or confuse the joining operation (e.g. whitespace, commas, periods, slashes, brackets, quotes, etc). Logger instances can store any number of key/value pairs, which will be logged alongside all messages logged through that instance. For example, you might want to create a Logger instance per managed object: With the standard log package, we might write: With logr we'd write: Logger has very few hard rules, with the goal that LogSink implementations might have a lot of freedom to differentiate. There are, however, some things to consider. The log message consists of a constant message attached to the log line. This should generally be a simple description of what's occurring, and should never be a format string. Variable information can then be attached using named values. Keys are arbitrary strings, but should generally be constant values. Values may be any Go value, but how the value is formatted is determined by the LogSink implementation. Logger instances are meant to be passed around by value. Code that receives such a value can call its methods without having to check whether the instance is ready for use. The zero logger (= Logger{}) is identical to Discard() and discards all log entries. Code that receives a Logger by value can simply call it, the methods will never crash. For cases where passing a logger is optional, a pointer to Logger should be used. Keys are not strictly required to conform to any specification or regex, but it is recommended that they: These guidelines help ensure that log data is processed properly regardless of the log implementation. For example, log implementations will try to output JSON data or will store data for later database (e.g. SQL) queries. While users are generally free to use key names of their choice, it's generally best to avoid using the following keys, as they're frequently used by implementations: Implementations are encouraged to make use of these keys to represent the above concepts, when necessary (for example, in a pure-JSON output form, it would be necessary to represent at least message and timestamp as ordinary named values). Implementations may choose to give callers access to the underlying logging implementation. The recommended pattern for this is: Logger grants access to the sink to enable type assertions like this: Custom `With*` functions can be implemented by copying the complete Logger struct and replacing the sink in the copy: Don't use New to construct a new Logger with a LogSink retrieved from an existing Logger. Source code attribution might not work correctly and unexported fields in Logger get lost. Beware that the same LogSink instance may be shared by different logger instances. Calling functions that modify the LogSink will affect all of those.

Package httpexpect helps with end-to-end HTTP and REST API testing. See example directory: There are two common ways to test API with httpexpect: The second approach works only if the server is a Go module and its handler can be imported in tests. Concrete behaviour is determined by Client implementation passed to Config struct. If you're using http.Client, set its Transport field (http.RoundTriper) to one of the following: Note that http handler can be usually obtained from http framework you're using. E.g., echo framework provides either http.Handler or fasthttp.RequestHandler. You can also provide your own implementation of RequestFactory (creates http.Request), or Client (gets http.Request and returns http.Response). If you're starting server from tests, it's very handy to use net/http/httptest. Whenever values are checked for equality in httpexpect, they are converted to "canonical form": This is equivalent to subsequently json.Marshal() and json.Unmarshal() the value and currently is implemented so. When some check fails, failure is reported. If non-fatal failures are used (see Reporter interface), execution is continued and instance that was checked is marked as failed. If specific instance is marked as failed, all subsequent checks are ignored for this instance and for any child instances retrieved after failure. Example: If you want to be informed about every asserion made, successful or failed, you can use AssertionHandler interface. Default implementation of this interface ignores successful assertions and reports failed assertions using Formatter and Reporter objects. Custom AssertionHandler can handle all assertions (e.g. dump them in JSON format) and is free to use or not to use Formatter and Reporter in its sole discretion.

Package objx provides utilities for dealing with maps, slices, JSON and other data. Objx provides the `objx.Map` type, which is a `map[string]interface{}` that exposes a powerful `Get` method (among others) that allows you to easily and quickly get access to data within the map, without having to worry too much about type assertions, missing data, default values etc. Objx uses a predictable pattern to make access data from within `map[string]interface{}` easy. Call one of the `objx.` functions to create your `objx.Map` to get going: NOTE: Any methods or functions with the `Must` prefix will panic if something goes wrong, the rest will be optimistic and try to figure things out without panicking. Use `Get` to access the value you're interested in. You can use dot and array notation too: Once you have sought the `Value` you're interested in, you can use the `Is*` methods to determine its type. Or you can just assume the type, and use one of the strong type methods to extract the real value: If there's no value there (or if it's the wrong type) then a default value will be returned, or you can be explicit about the default value. If you're dealing with a slice of data as a value, Objx provides many useful methods for iterating, manipulating and selecting that data. You can find out more by exploring the index below. A simple example of how to use Objx: Since `objx.Map` is a `map[string]interface{}` you can treat it as such. For example, to `range` the data, do what you would expect:

Package wire implements the Decred wire protocol. For the complete details of the Decred protocol, see the official wiki entry at https://en.bitcoin.it/wiki/Protocol_specification. The following only serves as a quick overview to provide information on how to use the package. At a high level, this package provides support for marshalling and unmarshalling supported Decred messages to and from the wire. This package does not deal with the specifics of message handling such as what to do when a message is received. This provides the caller with a high level of flexibility. The Decred protocol consists of exchanging messages between peers. Each message is preceded by a header which identifies information about it such as which Decred network it is a part of, its type, how big it is, and a checksum to verify validity. All encoding and decoding of message headers is handled by this package. To accomplish this, there is a generic interface for Decred messages named Message which allows messages of any type to be read, written, or passed around through channels, functions, etc. In addition, concrete implementations of most of the currently supported Decred messages are provided. For these supported messages, all of the details of marshalling and unmarshalling to and from the wire using Decred encoding are handled so the caller doesn't have to concern themselves with the specifics. The following provides a quick summary of how the Decred messages are intended to interact with one another. As stated above, these interactions are not directly handled by this package. For more in-depth details about the appropriate interactions, see the official Decred protocol wiki entry at https://en.bitcoin.it/wiki/Protocol_specification. The initial handshake consists of two peers sending each other a version message (MsgVersion) followed by responding with a verack message (MsgVerAck). Both peers use the information in the version message (MsgVersion) to negotiate things such as protocol version and supported services with each other. Once the initial handshake is complete, the following chart indicates message interactions in no particular order. There are several common parameters that arise when using this package to read and write Decred messages. The following sections provide a quick overview of these parameters so the next sections can build on them. The protocol version should be negotiated with the remote peer at a higher level than this package via the version (MsgVersion) message exchange, however, this package provides the wire.ProtocolVersion constant which indicates the latest protocol version this package supports and is typically the value to use for all outbound connections before a potentially lower protocol version is negotiated. The Decred network is a magic number which is used to identify the start of a message and which Decred network the message applies to. This package provides the following constants: As discussed in the Decred message overview section, this package reads and writes Decred messages using a generic interface named Message. In order to determine the actual concrete type of the message, use a type switch or type assertion. An example of a type switch follows: In order to unmarshall Decred messages from the wire, use the ReadMessage function. It accepts any io.Reader, but typically this will be a net.Conn to a remote node running a Decred peer. Example syntax is: In order to marshall Decred messages to the wire, use the WriteMessage function. It accepts any io.Writer, but typically this will be a net.Conn to a remote node running a Decred peer. Example syntax to request addresses from a remote peer is: The errors returned by this package are either the raw errors provided by underlying calls to read/write from streams such as io.EOF, io.ErrUnexpectedEOF, and io.ErrShortWrite, or of type wire.MessageError. This allows the caller to differentiate between general IO errors and malformed messages through type assertions. This package includes spec changes outlined by the following BIPs:

Package anaconda provides structs and functions for accessing version 1.1 of the Twitter API. Successful API queries return native Go structs that can be used immediately, with no need for type assertions. If you already have the access token (and secret) for your user (Twitter provides this for your own account on the developer portal), creating the client is simple: Executing queries on an authenticated TwitterApi struct is simple. Certain endpoints allow separate optional parameter; if desired, these can be passed as the final parameter. Anaconda implements most of the endpoints defined in the Twitter API documentation: https://dev.twitter.com/docs/api/1.1. For clarity, in most cases, the function name is simply the name of the HTTP method and the endpoint (e.g., the endpoint `GET /friendships/incoming` is provided by the function `GetFriendshipsIncoming`). In a few cases, a shortened form has been chosen to make life easier (for example, retweeting is simply the function `Retweet`) More detailed information about the behavior of each particular endpoint can be found at the official Twitter API documentation.

Package is provides a lightweight extension to the standard library's testing capabilities. Comments on the assertion lines are used to add a description. The following failing test: Will output: The following code shows a range of useful ways you can use the helper methods:

Package txscript implements the Decred transaction script language. This package provides data structures and functions to parse and execute decred transaction scripts. Decred transaction scripts are written in a stack-base, FORTH-like language. The Decred script language consists of a number of opcodes which fall into several categories such pushing and popping data to and from the stack, performing basic and bitwise arithmetic, conditional branching, comparing hashes, and checking cryptographic signatures. Scripts are processed from left to right and intentionally do not provide loops. The vast majority of Decred scripts at the time of this writing are of several standard forms which consist of a spender providing a public key and a signature which proves the spender owns the associated private key. This information is used to prove the the spender is authorized to perform the transaction. One benefit of using a scripting language is added flexibility in specifying what conditions must be met in order to spend decreds. Errors returned by this package are of type txscript.Error. This allows the caller to programmatically determine the specific error by examining the ErrorCode field of the type asserted txscript.Error while still providing rich error messages with contextual information. A convenience function named IsErrorCode is also provided to allow callers to easily check for a specific error code. See ErrorCode in the package documentation for a full list.

Package blockchain implements Decred block handling and chain selection rules. The Decred block handling and chain selection rules are an integral, and quite likely the most important, part of decred. Unfortunately, at the time of this writing, these rules are also largely undocumented and had to be ascertained from the bitcoind source code. At its core, Decred is a distributed consensus of which blocks are valid and which ones will comprise the main block chain (public ledger) that ultimately determines accepted transactions, so it is extremely important that fully validating nodes agree on all rules. At a high level, this package provides support for inserting new blocks into the block chain according to the aforementioned rules. It includes functionality such as rejecting duplicate blocks, ensuring blocks and transactions follow all rules, orphan handling, and best chain selection along with reorganization. Since this package does not deal with other Decred specifics such as network communication or wallets, it provides a notification system which gives the caller a high level of flexibility in how they want to react to certain events such as orphan blocks which need their parents requested and newly connected main chain blocks which might result in wallet updates. Before a block is allowed into the block chain, it must go through an intensive series of validation rules. The following list serves as a general outline of those rules to provide some intuition into what is going on under the hood, but is by no means exhaustive: Errors returned by this package are either the raw errors provided by underlying calls or of type blockchain.RuleError. This allows the caller to differentiate between unexpected errors, such as database errors, versus errors due to rule violations through type assertions. In addition, callers can programmatically determine the specific rule violation by examining the ErrorCode field of the type asserted blockchain.RuleError. This package includes spec changes outlined by the following BIPs:

Package dcrjson provides primitives for working with the Decred JSON-RPC API. When communicating via the JSON-RPC protocol, all of the commands need to be marshalled to and from the the wire in the appropriate format. This package provides data structures and primitives to ease this process. In addition, it also provides some additional features such as custom command registration, command categorization, and reflection-based help generation. This information is not necessary in order to use this package, but it does provide some intuition into what the marshalling and unmarshalling that is discussed below is doing under the hood. As defined by the JSON-RPC spec, there are effectively two forms of messages on the wire: Request Objects {"jsonrpc":"1.0","id":"SOMEID","method":"SOMEMETHOD","params":[SOMEPARAMS]} NOTE: Notifications are the same format except the id field is null. Response Objects {"result":SOMETHING,"error":null,"id":"SOMEID"} {"result":null,"error":{"code":SOMEINT,"message":SOMESTRING},"id":"SOMEID"} For requests, the params field can vary in what it contains depending on the method (a.k.a. command) being sent. Each parameter can be as simple as an int or a complex structure containing many nested fields. The id field is used to identify a request and will be included in the associated response. When working with asynchronous transports, such as websockets, spontaneous notifications are also possible. As indicated, they are the same as a request object, except they have the id field set to null. Therefore, servers will ignore requests with the id field set to null, while clients can choose to consume or ignore them. Unfortunately, the original Bitcoin JSON-RPC API (and hence anything compatible with it) doesn't always follow the spec and will sometimes return an error string in the result field with a null error for certain commands. However, for the most part, the error field will be set as described on failure. Based upon the discussion above, it should be easy to see how the types of this package map into the required parts of the protocol To simplify the marshalling of the requests and responses, the MarshalCmd and MarshalResponse functions are provided. They return the raw bytes ready to be sent across the wire. Unmarshalling a received Request object is a two step process: This approach is used since it provides the caller with access to the additional fields in the request that are not part of the command such as the ID. Unmarshalling a received Response object is also a two step process: As above, this approach is used since it provides the caller with access to the fields in the response such as the ID and Error. This package provides two approaches for creating a new command. This first, and preferred, method is to use one of the New<Foo>Cmd functions. This allows static compile-time checking to help ensure the parameters stay in sync with the struct definitions. The second approach is the NewCmd function which takes a method (command) name and variable arguments. The function includes full checking to ensure the parameters are accurate according to provided method, however these checks are, obviously, run-time which means any mistakes won't be found until the code is actually executed. However, it is quite useful for user-supplied commands that are intentionally dynamic. The command handling of this package is built around the concept of registered commands. This is true for the wide variety of commands already provided by the package, but it also means caller can easily provide custom commands with all of the same functionality as the built-in commands. Use the RegisterCmd function for this purpose. A list of all registered methods can be obtained with the RegisteredCmdMethods function. All registered commands are registered with flags that identify information such as whether the command applies to a chain server, wallet server, or is a notification along with the method name to use. These flags can be obtained with the MethodUsageFlags flags, and the method can be obtained with the CmdMethod function. To facilitate providing consistent help to users of the RPC server, this package exposes the GenerateHelp and function which uses reflection on registered commands or notifications, as well as the provided expected result types, to generate the final help text. In addition, the MethodUsageText function is provided to generate consistent one-line usage for registered commands and notifications using reflection. There are 2 distinct type of errors supported by this package: The first category of errors (type Error) typically indicates a programmer error and can be avoided by properly using the API. Errors of this type will be returned from the various functions available in this package. They identify issues such as unsupported field types, attempts to register malformed commands, and attempting to create a new command with an improper number of parameters. The specific reason for the error can be detected by type asserting it to a *dcrjson.Error and accessing the ErrorCode field. The second category of errors (type RPCError), on the other hand, are useful for returning errors to RPC clients. Consequently, they are used in the previously described Response type. This example demonstrates how to unmarshal a JSON-RPC response and then unmarshal the result field in the response to a concrete type.

Package rpcclient implements a websocket-enabled Decred JSON-RPC client. This client provides a robust and easy to use client for interfacing with a Decred RPC server that uses a mostly btcd/bitcoin core style Decred JSON-RPC API. This client has been tested with dcrd (https://github.com/decred/dcrd) and dcrwallet (https://github.com/decred/dcrwallet). In addition to the compatible standard HTTP POST JSON-RPC API, dcrd and dcrwallet provide a websocket interface that is more efficient than the standard HTTP POST method of accessing RPC. The section below discusses the differences between HTTP POST and websockets. By default, this client assumes the RPC server supports websockets and has TLS enabled. In practice, this currently means it assumes you are talking to dcrd or dcrwallet by default. However, configuration options are provided to fall back to HTTP POST and disable TLS to support talking with inferior bitcoin core style RPC servers. In HTTP POST-based JSON-RPC, every request creates a new HTTP connection, issues the call, waits for the response, and closes the connection. This adds quite a bit of overhead to every call and lacks flexibility for features such as notifications. In contrast, the websocket-based JSON-RPC interface provided by dcrd and dcrwallet only uses a single connection that remains open and allows asynchronous bi-directional communication. The websocket interface supports all of the same commands as HTTP POST, but they can be invoked without having to go through a connect/disconnect cycle for every call. In addition, the websocket interface provides other nice features such as the ability to register for asynchronous notifications of various events. The client provides both a synchronous (blocking) and asynchronous API. The synchronous (blocking) API is typically sufficient for most use cases. It works by issuing the RPC and blocking until the response is received. This allows straightforward code where you have the response as soon as the function returns. The asynchronous API works on the concept of futures. When you invoke the async version of a command, it will quickly return an instance of a type that promises to provide the result of the RPC at some future time. In the background, the RPC call is issued and the result is stored in the returned instance. Invoking the Receive method on the returned instance will either return the result immediately if it has already arrived, or block until it has. This is useful since it provides the caller with greater control over concurrency. The first important part of notifications is to realize that they will only work when connected via websockets. This should intuitively make sense because HTTP POST mode does not keep a connection open! All notifications provided by dcrd require registration to opt-in. For example, if you want to be notified when funds are received by a set of addresses, you register the addresses via the NotifyReceived (or NotifyReceivedAsync) function. Notifications are exposed by the client through the use of callback handlers which are setup via a NotificationHandlers instance that is specified by the caller when creating the client. It is important that these notification handlers complete quickly since they are intentionally in the main read loop and will block further reads until they complete. This provides the caller with the flexibility to decide what to do when notifications are coming in faster than they are being handled. In particular this means issuing a blocking RPC call from a callback handler will cause a deadlock as more server responses won't be read until the callback returns, but the callback would be waiting for a response. Thus, any additional RPCs must be issued an a completely decoupled manner. By default, when running in websockets mode, this client will automatically keep trying to reconnect to the RPC server should the connection be lost. There is a back-off in between each connection attempt until it reaches one try per minute. Once a connection is re-established, all previously registered notifications are automatically re-registered and any in-flight commands are re-issued. This means from the caller's perspective, the request simply takes longer to complete. The caller may invoke the Shutdown method on the client to force the client to cease reconnect attempts and return ErrClientShutdown for all outstanding commands. The automatic reconnection can be disabled by setting the DisableAutoReconnect flag to true in the connection config when creating the client. Minor RPC Server Differences and Chain/Wallet Separation Some of the commands are extensions specific to a particular RPC server. For example, the DebugLevel call is an extension only provided by dcrd (and dcrwallet passthrough). Therefore if you call one of these commands against an RPC server that doesn't provide them, you will get an unimplemented error from the server. An effort has been made to call out which commmands are extensions in their documentation. Also, it is important to realize that dcrd intentionally separates the wallet functionality into a separate process named dcrwallet. This means if you are connected to the dcrd RPC server directly, only the RPCs which are related to chain services will be available. Depending on your application, you might only need chain-related RPCs. In contrast, dcrwallet provides pass through treatment for chain-related RPCs, so it supports them in addition to wallet-related RPCs. There are 3 categories of errors that will be returned throughout this package: The first category of errors are typically one of ErrInvalidAuth, ErrInvalidEndpoint, ErrClientDisconnect, or ErrClientShutdown. NOTE: The ErrClientDisconnect will not be returned unless the DisableAutoReconnect flag is set since the client automatically handles reconnect by default as previously described. The second category of errors typically indicates a programmer error and as such the type can vary, but usually will be best handled by simply showing/logging it. The third category of errors, that is errors returned by the server, can be detected by type asserting the error in a *dcrjson.RPCError. For example, to detect if a command is unimplemented by the remote RPC server: The following full-blown client examples are in the examples directory:

Package dcrjson provides infrastructure for working with Decred JSON-RPC APIs. When communicating via the JSON-RPC protocol, all requests and responses must be marshalled to and from the wire in the appropriate format. This package provides infrastructure and primitives to ease this process. This information is not necessary in order to use this package, but it does provide some intuition into what the marshalling and unmarshalling that is discussed below is doing under the hood. As defined by the JSON-RPC spec, there are effectively two forms of messages on the wire: Request Objects {"jsonrpc":"1.0","id":"SOMEID","method":"SOMEMETHOD","params":[SOMEPARAMS]} NOTE: Notifications are the same format except the id field is null. Response Objects {"result":SOMETHING,"error":null,"id":"SOMEID"} {"result":null,"error":{"code":SOMEINT,"message":SOMESTRING},"id":"SOMEID"} For requests, the params field can vary in what it contains depending on the method (a.k.a. command) being sent. Each parameter can be as simple as an int or a complex structure containing many nested fields. The id field is used to identify a request and will be included in the associated response. When working with streamed RPC transports, such as websockets, spontaneous notifications are also possible. As indicated, they are the same as a request object, except they have the id field set to null. Therefore, servers will ignore requests with the id field set to null, while clients can choose to consume or ignore them. Unfortunately, the original Bitcoin JSON-RPC API (and hence anything compatible with it) doesn't always follow the spec and will sometimes return an error string in the result field with a null error for certain commands. However, for the most part, the error field will be set as described on failure. To simplify the marshalling of the requests and responses, the MarshalCmd and MarshalResponse functions are provided. They return the raw bytes ready to be sent across the wire. Unmarshalling a received Request object is a two step process: This approach is used since it provides the caller with access to the additional fields in the request that are not part of the command such as the ID. Unmarshalling a received Response object is also a two step process: As above, this approach is used since it provides the caller with access to the fields in the response such as the ID and Error. This package provides the NewCmd function which takes a method (command) name and variable arguments. The function includes full checking to ensure the parameters are accurate according to provided method, however these checks are, obviously, run-time which means any mistakes won't be found until the code is actually executed. However, it is quite useful for user-supplied commands that are intentionally dynamic. External packages can and should implement types implementing Command for use with MarshalCmd/ParseParams. The command handling of this package is built around the concept of registered commands. This is true for the wide variety of commands already provided by the package, but it also means caller can easily provide custom commands with all of the same functionality as the built-in commands. Use the RegisterCmd function for this purpose. A list of all registered methods can be obtained with the RegisteredCmdMethods function. All registered commands are registered with flags that identify information such as whether the command applies to a chain server, wallet server, or is a notification along with the method name to use. These flags can be obtained with the MethodUsageFlags flags, and the method can be obtained with the CmdMethod function. To facilitate providing consistent help to users of the RPC server, this package exposes the GenerateHelp and function which uses reflection on registered commands or notifications to generate the final help text. In addition, the MethodUsageText function is provided to generate consistent one-line usage for registered commands and notifications using reflection. There are 2 distinct type of errors supported by this package: The first category of errors (type Error) typically indicates a programmer error and can be avoided by properly using the API. Errors of this type will be returned from the various functions available in this package. They identify issues such as unsupported field types, attempts to register malformed commands, and attempting to create a new command with an improper number of parameters. The specific reason for the error can be detected by type asserting it to a *dcrjson.Error and accessing the ErrorCode field. The second category of errors (type RPCError), on the other hand, are useful for returning errors to RPC clients. Consequently, they are used in the previously described Response type. This example demonstrates how to unmarshal a JSON-RPC response and then unmarshal the result field in the response to a concrete type.

Package validator sets up several methods for simple validation as well as setting up an interface which when implemented allows for custom validation logic by the implementor. `Valid()` returns a boolean indicating if the value(s) passed into it are valid based on the different types. `Assert()` returns an error indicating `nil` if the values are valid. Otherwise `Assert()` returns an error indicating the index of the value(s) passed into it as to which was determined to be invalid.

Package txscript implements the Decred transaction script language. This package provides data structures and functions to parse and execute decred transaction scripts. Decred transaction scripts are written in a stack-base, FORTH-like language. The Decred script language consists of a number of opcodes which fall into several categories such pushing and popping data to and from the stack, performing basic and bitwise arithmetic, conditional branching, comparing hashes, and checking cryptographic signatures. Scripts are processed from left to right and intentionally do not provide loops. The vast majority of Decred scripts at the time of this writing are of several standard forms which consist of a spender providing a public key and a signature which proves the spender owns the associated private key. This information is used to prove the spender is authorized to perform the transaction. One benefit of using a scripting language is added flexibility in specifying what conditions must be met in order to spend decreds. Errors returned by this package are of type txscript.Error. This allows the caller to programmatically determine the specific error by examining the ErrorCode field of the type asserted txscript.Error while still providing rich error messages with contextual information. A convenience function named IsErrorCode is also provided to allow callers to easily check for a specific error code. See ErrorCode in the package documentation for a full list.

Package mempool provides a policy-enforced pool of unmined Decred transactions. A key responsibility of the Decred network is mining transactions – regular transactions and stake transactions – into blocks. In order to facilitate this, the mining process relies on having a readily-available source of transactions to include in a block that is being solved. At a high level, this package satisfies that requirement by providing an in-memory pool of fully validated transactions that can also optionally be further filtered based upon a configurable policy. The Policy configuration options has flags that control whether or not "standard" transactions and old votes are accepted into the mempool. In essence, a "standard" transaction is one that satisfies a fairly strict set of requirements that are largely intended to help provide fair use of the system to all users. It is important to note that what is considered to be a "standard" transaction changes over time as policy and consensus rules evolve. For some insight, at the time of this writing, an example of _some_ of the criteria that are required for a transaction to be considered standard are that it is of the most-recently supported version, finalized, does not exceed a specific size, and only consists of specific script forms. Since this package does not deal with other Decred specifics such as network communication and transaction relay, it returns a list of transactions that were accepted which gives the caller a high level of flexibility in how they want to proceed. Typically, this will involve things such as relaying the transactions to other peers on the network and notifying the mining process that new transactions are available. This package has intentionally been designed so it can be used as a standalone package for any projects needing the ability create an in-memory pool of Decred transactions that are not only valid by consensus rules, but also adhere to a configurable policy ## Feature Overview The following is a quick overview of the major features. It is not intended to be an exhaustive list. - Maintain a pool of fully validated transactions - Stake transaction support (ticket purchases, votes and revocations) - Orphan transaction support (transactions that spend from unknown outputs) - Configurable transaction acceptance policy - Additional metadata tracking for each transaction - Manual control of transaction removal Errors returned by this package are either the raw errors provided by underlying calls or of type mempool.RuleError. Since there are two classes of rules (mempool acceptance rules and blockchain (consensus) acceptance rules), the mempool.RuleError type contains a single Err field which will, in turn, either be a mempool.TxRuleError or a blockchain.RuleError. The first indicates a violation of mempool acceptance rules while the latter indicates a violation of consensus acceptance rules. This allows the caller to easily differentiate between unexpected errors, such as database errors, versus errors due to rule violations through type assertions. In addition, callers can programmatically determine the specific rule violation by type asserting the Err field to one of the aforementioned types and examining their underlying ErrorCode field.

Package testify is a set of packages that provide many tools for testifying that your code will behave as you intend. testify contains the following packages: The assert package provides a comprehensive set of assertion functions that tie in to the Go testing system. The http package contains tools to make it easier to test http activity using the Go testing system. The mock package provides a system by which it is possible to mock your objects and verify calls are happening as expected. The suite package provides a basic structure for using structs as testing suites, and methods on those structs as tests. It includes setup/teardown functionality in the way of interfaces.

Package standalone provides standalone functions useful for working with the Decred blockchain consensus rules. The primary goal of offering these functions via a separate module is to reduce the required dependencies to a minimum as compared to the blockchain module. It is ideal for applications such as lightweight clients that need to ensure basic security properties hold and calculate appropriate vote subsidies and block explorers. For example, some things an SPV wallet needs to prove are that the block headers all connect together, that they satisfy the proof of work requirements, and that a given transaction tree is valid for a given header. The provided functions fall into the following categories: Errors returned by this package are of type standalone.RuleError. This allows the caller to differentiate between errors further up the call stack through type assertions. In addition, callers can programmatically determine the specific rule violation by examining the ErrorCode field of the type asserted standalone.RuleError.

This is the official Go SDK for Oracle Cloud Infrastructure Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#installing for installation instructions. Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring for configuration instructions. The following example shows how to get started with the SDK. The example belows creates an identityClient struct with the default configuration. It then utilizes the identityClient to list availability domains and prints them out to stdout More examples can be found in the SDK Github repo: https://github.com/oracle/oci-go-sdk/tree/master/example Optional fields are represented with the `mandatory:"false"` tag on input structs. The SDK will omit all optional fields that are nil when making requests. In the case of enum-type fields, the SDK will omit fields whose value is an empty string. The SDK uses pointers for primitive types in many input structs. To aid in the construction of such structs, the SDK provides functions that return a pointer for a given value. For example: The SDK exposes functionality that allows the user to customize any http request before is sent to the service. You can do so by setting the `Interceptor` field in any of the `Client` structs. For example: The Interceptor closure gets called before the signing process, thus any changes done to the request will be properly signed and submitted to the service. The SDK exposes a stand-alone signer that can be used to signing custom requests. Related code can be found here: https://github.com/oracle/oci-go-sdk/blob/master/common/http_signer.go. The example below shows how to create a default signer. The signer also allows more granular control on the headers used for signing. For example: You can combine a custom signer with the exposed clients in the SDK. This allows you to add custom signed headers to the request. Following is an example: Bear in mind that some services have a white list of headers that it expects to be signed. Therefore, adding an arbitrary header can result in authentications errors. To see a runnable example, see https://github.com/oracle/oci-go-sdk/blob/master/example/example_identity_test.go For more information on the signing algorithm refer to: https://docs.cloud.oracle.com/Content/API/Concepts/signingrequests.htm Some operations accept or return polymorphic JSON objects. The SDK models such objects as interfaces. Further the SDK provides structs that implement such interfaces. Thus, for all operations that expect interfaces as input, pass the struct in the SDK that satisfies such interface. For example: In the case of a polymorphic response you can type assert the interface to the expected type. For example: An example of polymorphic JSON request handling can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_test.go#L63 When calling a list operation, the operation will retrieve a page of results. To retrieve more data, call the list operation again, passing in the value of the most recent response's OpcNextPage as the value of Page in the next list operation call. When there is no more data the OpcNextPage field will be nil. An example of pagination using this logic can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_pagination_test.go The SDK has a built-in logging mechanism used internally. The internal logging logic is used to record the raw http requests, responses and potential errors when (un)marshalling request and responses. Built-in logging in the SDK is controlled via the environment variable "OCI_GO_SDK_DEBUG" and its contents. The below are possible values for the "OCI_GO_SDK_DEBUG" variable 1. "info" or "i" enables all info logging messages 2. "debug" or "d" enables all debug and info logging messages 3. "verbose" or "v" or "1" enables all verbose, debug and info logging messages 4. "null" turns all logging messages off. If the value of the environment variable does not match any of the above then default logging level is "info". If the environment variable is not present then no logging messages are emitted. The default destination for logging is Stderr and if you want to output log to a file you can set via environment variable "OCI_GO_SDK_LOG_OUTPUT_MODE". The below are possible values 1. "file" or "f" enables all logging output saved to file 2. "combine" or "c" enables all logging output to both stderr and file You can also customize the log file location and name via "OCI_GO_SDK_LOG_FILE" environment variable, the value should be the path to a specific file If this environment variable is not present, the default location will be the project root path Sometimes you may need to wait until an attribute of a resource, such as an instance or a VCN, reaches a certain state. An example of this would be launching an instance and then waiting for the instance to become available, or waiting until a subnet in a VCN has been terminated. You might also want to retry the same operation again if there's network issue etc... This can be accomplished by using the RequestMetadata.RetryPolicy. You can find the examples here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_retry_test.go The GO SDK uses the net/http package to make calls to OCI services. If your environment requires you to use a proxy server for outgoing HTTP requests then you can set this up in the following ways: 1. Configuring environment variable as described here https://golang.org/pkg/net/http/#ProxyFromEnvironment 2. Modifying the underlying Transport struct for a service client In order to modify the underlying Transport struct in HttpClient, you can do something similar to (sample code for audit service client): The Object Storage service supports multipart uploads to make large object uploads easier by splitting the large object into parts. The Go SDK supports raw multipart upload operations for advanced use cases, as well as a higher level upload class that uses the multipart upload APIs. For links to the APIs used for multipart upload operations, see Managing Multipart Uploads (https://docs.cloud.oracle.com/iaas/Content/Object/Tasks/usingmultipartuploads.htm). Higher level multipart uploads are implemented using the UploadManager, which will: split a large object into parts for you, upload the parts in parallel, and then recombine and commit the parts as a single object in storage. This code sample shows how to use the UploadManager to automatically split an object into parts for upload to simplify interaction with the Object Storage service: https://github.com/oracle/oci-go-sdk/blob/master/example/example_objectstorage_test.go Some response fields are enum-typed. In the future, individual services may return values not covered by existing enums for that field. To address this possibility, every enum-type response field is a modeled as a type that supports any string. Thus if a service returns a value that is not recognized by your version of the SDK, then the response field will be set to this value. When individual services return a polymorphic JSON response not available as a concrete struct, the SDK will return an implementation that only satisfies the interface modeling the polymorphic JSON response. If you are using a version of the SDK released prior to the announcement of a new region, you may need to use a workaround to reach it, depending on whether the region is in the oraclecloud.com realm. A region is a localized geographic area. For more information on regions and how to identify them, see Regions and Availability Domains(https://docs.cloud.oracle.com/iaas/Content/General/Concepts/regions.htm). A realm is a set of regions that share entities. You can identify your realm by looking at the domain name at the end of the network address. For example, the realm for xyz.abc.123.oraclecloud.com is oraclecloud.com. oraclecloud.com Realm: For regions in the oraclecloud.com realm, even if common.Region does not contain the new region, the forward compatibility of the SDK can automatically handle it. You can pass new region names just as you would pass ones that are already defined. For more information on passing region names in the configuration, see Configuring (https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring). For details on common.Region, see (https://github.com/oracle/oci-go-sdk/blob/master/common/common.go). Other Realms: For regions in realms other than oraclecloud.com, you can use the following workarounds to reach new regions with earlier versions of the SDK. NOTE: Be sure to supply the appropriate endpoints for your region. You can overwrite the target host with client.Host: If you are authenticating via instance principals, you can set the authentication endpoint in an environment variable: Got a fix for a bug, or a new feature you'd like to contribute? The SDK is open source and accepting pull requests on GitHub https://github.com/oracle/oci-go-sdk Licensing information available at: https://github.com/oracle/oci-go-sdk/blob/master/LICENSE.txt To be notified when a new version of the Go SDK is released, subscribe to the following feed: https://github.com/oracle/oci-go-sdk/releases.atom Please refer to this link: https://github.com/oracle/oci-go-sdk#help

Package gcs provides an API for building and using a Golomb-coded set filter. A Golomb-Coded Set (GCS) is a space-efficient probabilistic data structure that is used to test set membership with a tunable false positive rate while simultaneously preventing false negatives. In other words, items that are in the set will always match, but items that are not in the set will also sometimes match with the chosen false positive rate. This package currently implements two different versions for backwards compatibility. Version 1 is deprecated and therefore should no longer be used. Version 2 is the GCS variation that follows the specification details in DCP0005: https://github.com/decred/dcps/blob/master/dcp-0005/dcp-0005.mediawiki#golomb-coded-sets. Version 2 sets do not permit empty items (data of zero length) to be added and are parameterized by the following: * A parameter `B` that defines the remainder code bit size * A parameter `M` that defines the false positive rate as `1/M` * A key for the SipHash-2-4 function * The items to include in the set Errors returned by this package are of type gcs.Error. This allows the caller to programmatically determine the specific error by examining the ErrorCode field of the type asserted gcs.Error while still providing rich error messages with contextual information. A convenience function named IsErrorCode is also provided to allow callers to easily check for a specific error code. See ErrorCode in the package documentation for a full list. GCS is used as a mechanism for storing, transmitting, and committing to per-block filters. Consensus-validating full nodes commit to a single filter for every block and serve the filter to SPV clients that match against the filter locally to determine if the block is potentially relevant. The required parameters for Decred are defined by the blockcf2 package. For more details, see the the Block Filters section of DCP0005: https://github.com/decred/dcps/blob/master/dcp-0005/dcp-0005.mediawiki#block-filters

Package goworker is a Resque-compatible, Go-based background worker. It allows you to push jobs into a queue using an expressive language like Ruby while harnessing the efficiency and concurrency of Go to minimize job latency and cost. goworker workers can run alongside Ruby Resque clients so that you can keep all but your most resource-intensive jobs in Ruby. To create a worker, write a function matching the signature and register it using Here is a simple worker that prints its arguments: To create workers that share a database pool or other resources, use a closure to share variables. goworker worker functions receive the queue they are serving and a slice of interfaces. To use them as parameters to other functions, use Go type assertions to convert them into usable types. For testing, it is helpful to use the redis-cli program to insert jobs onto the Redis queue: will insert 100 jobs for the MyClass worker onto the myqueue queue. It is equivalent to: After building your workers, you will have an executable that you can run which will automatically poll a Redis server and call your workers as jobs arrive. There are several flags which control the operation of the goworker client. -queues="comma,delimited,queues" — This is the only required flag. The recommended practice is to separate your Resque workers from your goworkers with different queues. Otherwise, Resque worker classes that have no goworker analog will cause the goworker process to fail the jobs. Because of this, there is no default queue, nor is there a way to select all queues (à la Resque's * queue). Queues are processed in the order they are specififed. If you have multiple queues you can assign them weights. A queue with a weight of 2 will be checked twice as often as a queue with a weight of 1: -queues='high=2,low=1'. -interval=5.0 — Specifies the wait period between polling if no job was in the queue the last time one was requested. -concurrency=25 — Specifies the number of concurrently executing workers. This number can be as low as 1 or rather comfortably as high as 100,000, and should be tuned to your workflow and the availability of outside resources. -connections=2 — Specifies the maximum number of Redis connections that goworker will consume between the poller and all workers. There is not much performance gain over two and a slight penalty when using only one. This is configurable in case you need to keep connection counts low for cloud Redis providers who limit plans on maxclients. -uri=redis://localhost:6379/ — Specifies the URI of the Redis database from which goworker polls for jobs. Accepts URIs of the format redis://user:pass@host:port/db or unix:///path/to/redis.sock. The flag may also be set by the environment variable $($REDIS_PROVIDER) or $REDIS_URL. E.g. set $REDIS_PROVIDER to REDISTOGO_URL on Heroku to let the Redis To Go add-on configure the Redis database. -namespace=resque: — Specifies the namespace from which goworker retrieves jobs and stores stats on workers. -exit-on-complete=false — Exits goworker when there are no jobs left in the queue. This is helpful in conjunction with the time command to benchmark different configurations. -use-number=false — Uses json.Number when decoding numbers in the job payloads. This will avoid issues that occur when goworker and the json package decode large numbers as floats, which then get encoded in scientific notation, losing pecision. This will default to true soon. You can also configure your own flags for use within your workers. Be sure to set them before calling goworker.Main(). It is okay to call flags.Parse() before calling goworker.Main() if you need to do additional processing on your flags. To stop goworker, send a QUIT, TERM, or INT signal to the process. This will immediately stop job polling. There can be up to $CONCURRENCY jobs currently running, which will continue to run until they are finished. Like Resque, goworker makes no guarantees about the safety of jobs in the event of process shutdown. Workers must be both idempotent and tolerant to loss of the job in the event of failure. If the process is killed with a KILL or by a system failure, there may be one job that is currently in the poller's buffer that will be lost without any representation in either the queue or the worker variable. If you are running Goworker on a system like Heroku, which sends a TERM to signal a process that it needs to stop, ten seconds later sends a KILL to force the process to stop, your jobs must finish within 10 seconds or they may be lost. Jobs will be recoverable from the Redis database under as a JSON object with keys queue, run_at, and payload, but the process is manual. Additionally, there is no guarantee that the job in Redis under the worker key has not finished, if the process is killed before goworker can flush the update to Redis.

Package crypto11 enables access to cryptographic keys from PKCS#11 using Go crypto API. PKCS#11 tokens are accessed via Context objects. Each Context connects to one token. Context objects are created by calling Configure or ConfigureFromFile. In the latter case, the file should contain a JSON representation of a Config. There is support for generating DSA, RSA and ECDSA keys. These keys can be found later using FindKeyPair. All three key types implement the crypto.Signer interface and the RSA keys also implement crypto.Decrypter. RSA keys obtained through FindKeyPair will need a type assertion to be used for decryption. Assert either crypto.Decrypter or SignerDecrypter, as you prefer. Symmetric keys can also be generated. These are found later using FindKey. See the documentation for SecretKey for further information. Note that PKCS#11 session handles must not be used concurrently from multiple threads. Consumers of the Signer interface know nothing of this and expect to be able to sign from multiple threads without constraint. We address this as follows. 1. When a Context is created, a session is created and the user is logged in. This session remains open until the Context is closed, to ensure all object handles remain valid and to avoid repeatedly calling C_Login. 2. The Context also maintains a pool of read-write sessions. The pool expands dynamically as needed, but never beyond the maximum number of r/w sessions supported by the token (as reported by C_GetInfo). If other applications are using the token, a lower limit should be set in the Config. 3. Each operation transiently takes a session from the pool. They have exclusive use of the session, meeting PKCS#11's concurrency requirements. Sessions are returned to the pool afterwards and may be re-used. Behaviour of the pool can be tweaked via Config fields: - PoolWaitTimeout controls how long an operation can block waiting on a session from the pool. A zero value means there is no limit. Timeouts occur if the pool is fully used and additional operations are requested. - MaxSessions sets an upper bound on the number of sessions. If this value is zero, a default maximum is used (see DefaultMaxSessions). In every case the maximum supported sessions as reported by the token is obeyed. The PKCS1v15DecryptOptions SessionKeyLen field is not implemented and an error is returned if it is nonzero. The reason for this is that it is not possible for crypto11 to guarantee the constant-time behavior in the specification. See https://github.com/thalesignite/crypto11/issues/5 for further discussion. Symmetric crypto support via cipher.Block is very slow. You can use the BlockModeCloser API but you must call the Close() interface (not found in cipher.BlockMode). See https://github.com/ThalesIgnite/crypto11/issues/6 for further discussion.

Package quicktest provides a collection of Go helpers for writing tests. Quicktest helpers can be easily integrated inside regular Go tests, for instance: An assertion looks like this, where qt.Equals could be replaced by any available checker. If the assertion fails, the underlying Fatal method is called to describe the error and abort the test. If you don’t want to abort on failure, use Check instead, which calls Error instead of Fatal: For really short tests, the extra line for instantiating *qt.C can be avoided: The library provides some base checkers like Equals, DeepEquals, Matches, ErrorMatches, IsNil and others. More can be added by implementing the Checker interface. Below, we list the checkers implemented by the package in alphabetical order. All returns a Checker that uses the given checker to check elements of slice or array or the values of a map. It succeeds if all elements pass the check. On failure it prints the error from the first index that failed. For example: See also Any and Contains. Any returns a Checker that uses the given checker to check elements of a slice or array or the values from a map. It succeeds if any element passes the check. For example: See also All and Contains. CmpEquals checks equality of two arbitrary values according to the provided compare options. DeepEquals is more commonly used when no compare options are required. Example calls: CodecEquals returns a checker that checks for codec value equivalence. It expects two arguments: a byte slice or a string containing some codec-marshaled data, and a Go value. It uses unmarshal to unmarshal the data into an interface{} value. It marshals the Go value using marshal, then unmarshals the result into an interface{} value. It then checks that the two interface{} values are deep-equal to one another, using CmpEquals(opts) to perform the check. See JSONEquals for an example of this in use. Contains checks that a map, slice, array or string contains a value. It's the same as using Any(Equals), except that it has a special case for strings - if the first argument is a string, the second argument must also be a string and strings.Contains will be used. For example: ContentEquals is is like DeepEquals but any slices in the compared values will be sorted before being compared. For example: DeepEquals checks that two arbitrary values are deeply equal. The comparison is done using the github.com/google/go-cmp/cmp package. When comparing structs, by default no exported fields are allowed. If a more sophisticated comparison is required, use CmpEquals (see below). Example call: Equals checks that two values are equal, as compared with Go's == operator. For instance: Note that the following will fail: Use the IsNil checker below for this kind of nil check. ErrorAs checks that the error is or wraps a specific error type. If so, it assigns it to the provided pointer. This is analogous to calling errors.As. For instance: ErrorIs checks that the error is or wraps a specific error value. This is analogous to calling errors.Is. For instance: ErrorMatches checks that the provided value is an error whose message matches the provided regular expression. For instance: HasLen checks that the provided value has the given length. For instance: Implements checks that the provided value implements an interface. The interface is specified with a pointer to an interface variable. For instance: IsFalse checks that the provided value is false. The value must have a boolean underlying type. For instance: IsNil checks that the provided value is nil. For instance: As a special case, if the value is nil but implements the error interface, it is still considered to be non-nil. This means that IsNil will fail on an error value that happens to have an underlying nil value, because that's invariably a mistake. See https://golang.org/doc/faq#nil_error. So it's just fine to check an error like this: IsNotNil is a Checker checking that the provided value is not nil. IsNotNil is the equivalent of qt.Not(qt.IsNil) For instance: IsTrue checks that the provided value is true. The value must have a boolean underlying type. For instance: JSONEquals checks whether a byte slice or string is JSON-equivalent to a Go value. See CodecEquals for more information. It uses DeepEquals to do the comparison. If a more sophisticated comparison is required, use CodecEquals directly. For instance: Matches checks that a string or result of calling the String method (if the value implements fmt.Stringer) matches the provided regular expression. For instance: Not returns a Checker negating the given Checker. For instance: PanicMatches checks that the provided function panics with a message matching the provided regular expression. For instance: Satisfies checks that the provided value, when used as argument of the provided predicate function, causes the function to return true. The function must be of type func(T) bool, having got assignable to T. For instance: The testing.TB.Cleanup helper provides the ability to defer the execution of functions that will be run when the test completes. This is often useful for creating OS-level resources such as temporary directories (see c.Mkdir). When targeting Go versions that don't have Cleanup (< 1.14), the same can be achieved using c.Defer. In this case, to trigger the deferred behavior, calling c.Done is required. For instance, if you create a *C instance at the top level, you’ll have to add a defer to trigger the cleanups at the end of the test: However, if you use quicktest to create a subtest, Done will be called automatically at the end of that subtest. For example: The c.Patch, c.Setenv, c.Unsetenv and c.Mkdir helpers use t.Cleanup for cleaning up resources when available, and fall back to Defer otherwise.

Package standalone provides standalone functions useful for working with the Decred blockchain consensus rules. The primary goal of offering these functions via a separate module is to reduce the required dependencies to a minimum as compared to the blockchain module. It is ideal for applications such as lightweight clients that need to ensure basic security properties hold and calculate appropriate vote subsidies and block explorers. For example, some things an SPV wallet needs to prove are that the block headers all connect together, that they satisfy the proof of work requirements, and that a given transaction tree is valid for a given header. The provided functions fall into the following categories: The errors returned by this package are of type standalone.RuleError. This allows the caller to differentiate between errors further up the call stack through type assertions. In addition, callers can programmatically determine the specific rule violation by examining the ErrorCode field of the type asserted standalone.RuleError.

Package blockchain implements Decred block handling and chain selection rules. The Decred block handling and chain selection rules are an integral, and quite likely the most important, part of decred. At its core, Decred is a distributed consensus of which blocks are valid and which ones will comprise the main block chain (public ledger) that ultimately determines accepted transactions, so it is extremely important that fully validating nodes agree on all rules. At a high level, this package provides support for inserting new blocks into the block chain according to the aforementioned rules. It includes functionality such as rejecting duplicate blocks, ensuring blocks and transactions follow all rules, and best chain selection along with reorganization. Since this package does not deal with other Decred specifics such as network communication or wallets, it provides a notification system which gives the caller a high level of flexibility in how they want to react to certain events such as newly connected main chain blocks which might result in wallet updates. Before a block is allowed into the block chain, it must go through an intensive series of validation rules. The following list serves as a general outline of those rules to provide some intuition into what is going on under the hood, but is by no means exhaustive: Errors returned by this package are either the raw errors provided by underlying calls or of type blockchain.RuleError. This allows the caller to differentiate between unexpected errors, such as database errors, versus errors due to rule violations through type assertions. In addition, callers can programmatically determine the specific rule violation by examining the ErrorCode field of the type asserted blockchain.RuleError.

Package dcrjson provides primitives for working with the Decred JSON-RPC API. When communicating via the JSON-RPC protocol, all of the commands need to be marshalled to and from the the wire in the appropriate format. This package provides data structures and primitives to ease this process. In addition, it also provides some additional features such as custom command registration, command categorization, and reflection-based help generation. This information is not necessary in order to use this package, but it does provide some intuition into what the marshalling and unmarshalling that is discussed below is doing under the hood. As defined by the JSON-RPC spec, there are effectively two forms of messages on the wire: Request Objects {"jsonrpc":"1.0","id":"SOMEID","method":"SOMEMETHOD","params":[SOMEPARAMS]} NOTE: Notifications are the same format except the id field is null. Response Objects {"result":SOMETHING,"error":null,"id":"SOMEID"} {"result":null,"error":{"code":SOMEINT,"message":SOMESTRING},"id":"SOMEID"} For requests, the params field can vary in what it contains depending on the method (a.k.a. command) being sent. Each parameter can be as simple as an int or a complex structure containing many nested fields. The id field is used to identify a request and will be included in the associated response. When working with asynchronous transports, such as websockets, spontaneous notifications are also possible. As indicated, they are the same as a request object, except they have the id field set to null. Therefore, servers will ignore requests with the id field set to null, while clients can choose to consume or ignore them. Unfortunately, the original Bitcoin JSON-RPC API (and hence anything compatible with it) doesn't always follow the spec and will sometimes return an error string in the result field with a null error for certain commands. However, for the most part, the error field will be set as described on failure. Based upon the discussion above, it should be easy to see how the types of this package map into the required parts of the protocol To simplify the marshalling of the requests and responses, the MarshalCmd and MarshalResponse functions are provided. They return the raw bytes ready to be sent across the wire. Unmarshalling a received Request object is a two step process: This approach is used since it provides the caller with access to the additional fields in the request that are not part of the command such as the ID. Unmarshalling a received Response object is also a two step process: As above, this approach is used since it provides the caller with access to the fields in the response such as the ID and Error. This package provides two approaches for creating a new command. This first, and preferred, method is to use one of the New<Foo>Cmd functions. This allows static compile-time checking to help ensure the parameters stay in sync with the struct definitions. The second approach is the NewCmd function which takes a method (command) name and variable arguments. The function includes full checking to ensure the parameters are accurate according to provided method, however these checks are, obviously, run-time which means any mistakes won't be found until the code is actually executed. However, it is quite useful for user-supplied commands that are intentionally dynamic. The command handling of this package is built around the concept of registered commands. This is true for the wide variety of commands already provided by the package, but it also means caller can easily provide custom commands with all of the same functionality as the built-in commands. Use the RegisterCmd function for this purpose. A list of all registered methods can be obtained with the RegisteredCmdMethods function. All registered commands are registered with flags that identify information such as whether the command applies to a chain server, wallet server, or is a notification along with the method name to use. These flags can be obtained with the MethodUsageFlags flags, and the method can be obtained with the CmdMethod function. To facilitate providing consistent help to users of the RPC server, this package exposes the GenerateHelp and function which uses reflection on registered commands or notifications, as well as the provided expected result types, to generate the final help text. In addition, the MethodUsageText function is provided to generate consistent one-line usage for registered commands and notifications using reflection. There are 2 distinct type of errors supported by this package: The first category of errors (type Error) typically indicates a programmer error and can be avoided by properly using the API. Errors of this type will be returned from the various functions available in this package. They identify issues such as unsupported field types, attempts to register malformed commands, and attempting to create a new command with an improper number of parameters. The specific reason for the error can be detected by type asserting it to a *dcrjson.Error and accessing the ErrorCode field. The second category of errors (type RPCError), on the other hand, are useful for returning errors to RPC clients. Consequently, they are used in the previously described Response type. This example demonstrates how to unmarshal a JSON-RPC response and then unmarshal the result field in the response to a concrete type.

Package dcrjson provides infrastructure for working with Decred JSON-RPC APIs. When communicating via the JSON-RPC protocol, all requests and responses must be marshalled to and from the wire in the appropriate format. This package provides infrastructure and primitives to ease this process. This information is not necessary in order to use this package, but it does provide some intuition into what the marshalling and unmarshalling that is discussed below is doing under the hood. As defined by the JSON-RPC spec, there are effectively two forms of messages on the wire: Request Objects {"jsonrpc":"1.0","id":"SOMEID","method":"SOMEMETHOD","params":[SOMEPARAMS]} NOTE: Notifications are the same format except the id field is null. Response Objects {"result":SOMETHING,"error":null,"id":"SOMEID"} {"result":null,"error":{"code":SOMEINT,"message":SOMESTRING},"id":"SOMEID"} For requests, the params field can vary in what it contains depending on the method (a.k.a. command) being sent. Each parameter can be as simple as an int or a complex structure containing many nested fields. The id field is used to identify a request and will be included in the associated response. When working with streamed RPC transports, such as websockets, spontaneous notifications are also possible. As indicated, they are the same as a request object, except they have the id field set to null. Therefore, servers will ignore requests with the id field set to null, while clients can choose to consume or ignore them. Unfortunately, the original Bitcoin JSON-RPC API (and hence anything compatible with it) doesn't always follow the spec and will sometimes return an error string in the result field with a null error for certain commands. However, for the most part, the error field will be set as described on failure. To simplify the marshalling of the requests and responses, the MarshalCmd and MarshalResponse functions are provided. They return the raw bytes ready to be sent across the wire. Unmarshalling a received Request object is a two step process: This approach is used since it provides the caller with access to the additional fields in the request that are not part of the command such as the ID. Unmarshalling a received Response object is also a two step process: As above, this approach is used since it provides the caller with access to the fields in the response such as the ID and Error. This package provides the NewCmd function which takes a method (command) name and variable arguments. The function includes full checking to ensure the parameters are accurate according to provided method, however these checks are, obviously, run-time which means any mistakes won't be found until the code is actually executed. However, it is quite useful for user-supplied commands that are intentionally dynamic. External packages can and should implement types implementing Command for use with MarshalCmd/ParseParams. The command handling of this package is built around the concept of registered commands. This is true for the wide variety of commands already provided by the package, but it also means caller can easily provide custom commands with all of the same functionality as the built-in commands. Use the RegisterCmd function for this purpose. A list of all registered methods can be obtained with the RegisteredCmdMethods function. All registered commands are registered with flags that identify information such as whether the command applies to a chain server, wallet server, or is a notification along with the method name to use. These flags can be obtained with the MethodUsageFlags flags, and the method can be obtained with the CmdMethod function. To facilitate providing consistent help to users of the RPC server, this package exposes the GenerateHelp and function which uses reflection on registered commands or notifications to generate the final help text. In addition, the MethodUsageText function is provided to generate consistent one-line usage for registered commands and notifications using reflection. There are 2 distinct type of errors supported by this package: The first category of errors (type Error) typically indicates a programmer error and can be avoided by properly using the API. Errors of this type will be returned from the various functions available in this package. They identify issues such as unsupported field types, attempts to register malformed commands, and attempting to create a new command with an improper number of parameters. The specific reason for the error can be detected by type asserting it to a *dcrjson.Error and accessing the ErrorKind field. The second category of errors (type RPCError), on the other hand, are useful for returning errors to RPC clients. Consequently, they are used in the previously described Response type. This example demonstrates how to unmarshal a JSON-RPC response and then unmarshal the result field in the response to a concrete type.

Package rpcclient implements a websocket-enabled Decred JSON-RPC client. This client provides a robust and easy to use client for interfacing with a Decred RPC server that uses a mostly btcd/bitcoin core style Decred JSON-RPC API. This client has been tested with dcrd (https://github.com/decred/dcrd) and dcrwallet (https://github.com/decred/dcrwallet). In addition to the compatible standard HTTP POST JSON-RPC API, dcrd and dcrwallet provide a websocket interface that is more efficient than the standard HTTP POST method of accessing RPC. The section below discusses the differences between HTTP POST and websockets. By default, this client assumes the RPC server supports websockets and has TLS enabled. In practice, this currently means it assumes you are talking to dcrd or dcrwallet by default. However, configuration options are provided to fall back to HTTP POST and disable TLS to support talking with inferior bitcoin core style RPC servers. In HTTP POST-based JSON-RPC, every request creates a new HTTP connection, issues the call, waits for the response, and closes the connection. This adds quite a bit of overhead to every call and lacks flexibility for features such as notifications. In contrast, the websocket-based JSON-RPC interface provided by dcrd and dcrwallet only uses a single connection that remains open and allows asynchronous bi-directional communication. The websocket interface supports all of the same commands as HTTP POST, but they can be invoked without having to go through a connect/disconnect cycle for every call. In addition, the websocket interface provides other nice features such as the ability to register for asynchronous notifications of various events. The client provides both a synchronous (blocking) and asynchronous API. The synchronous (blocking) API is typically sufficient for most use cases. It works by issuing the RPC and blocking until the response is received. This allows straightforward code where you have the response as soon as the function returns. The asynchronous API works on the concept of futures. When you invoke the async version of a command, it will quickly return an instance of a type that promises to provide the result of the RPC at some future time. In the background, the RPC call is issued and the result is stored in the returned instance. Invoking the Receive method on the returned instance will either return the result immediately if it has already arrived, or block until it has. This is useful since it provides the caller with greater control over concurrency. The first important part of notifications is to realize that they will only work when connected via websockets. This should intuitively make sense because HTTP POST mode does not keep a connection open! All notifications provided by dcrd require registration to opt-in. For example, if you want to be notified when funds are received by a set of addresses, you register the addresses via the NotifyReceived (or NotifyReceivedAsync) function. Notifications are exposed by the client through the use of callback handlers which are setup via a NotificationHandlers instance that is specified by the caller when creating the client. It is important that these notification handlers complete quickly since they are intentionally in the main read loop and will block further reads until they complete. This provides the caller with the flexibility to decide what to do when notifications are coming in faster than they are being handled. In particular this means issuing a blocking RPC call from a callback handler will cause a deadlock as more server responses won't be read until the callback returns, but the callback would be waiting for a response. Thus, any additional RPCs must be issued an a completely decoupled manner. By default, when running in websockets mode, this client will automatically keep trying to reconnect to the RPC server should the connection be lost. There is a back-off in between each connection attempt until it reaches one try per minute. Once a connection is re-established, all previously registered notifications are automatically re-registered and any in-flight commands are re-issued. This means from the caller's perspective, the request simply takes longer to complete. The caller may invoke the Shutdown method on the client to force the client to cease reconnect attempts and return ErrClientShutdown for all outstanding commands. The automatic reconnection can be disabled by setting the DisableAutoReconnect flag to true in the connection config when creating the client. Minor RPC Server Differences and Chain/Wallet Separation Some of the commands are extensions specific to a particular RPC server. For example, the DebugLevel call is an extension only provided by dcrd (and dcrwallet passthrough). Therefore if you call one of these commands against an RPC server that doesn't provide them, you will get an unimplemented error from the server. An effort has been made to call out which commmands are extensions in their documentation. Also, it is important to realize that dcrd intentionally separates the wallet functionality into a separate process named dcrwallet. This means if you are connected to the dcrd RPC server directly, only the RPCs which are related to chain services will be available. Depending on your application, you might only need chain-related RPCs. In contrast, dcrwallet provides pass through treatment for chain-related RPCs, so it supports them in addition to wallet-related RPCs. There are 3 categories of errors that will be returned throughout this package: The first category of errors are typically one of ErrInvalidAuth, ErrInvalidEndpoint, ErrClientDisconnect, or ErrClientShutdown. NOTE: The ErrClientDisconnect will not be returned unless the DisableAutoReconnect flag is set since the client automatically handles reconnect by default as previously described. The second category of errors typically indicates a programmer error and as such the type can vary, but usually will be best handled by simply showing/logging it. The third category of errors, that is errors returned by the server, can be detected by type asserting the error in a *dcrjson.RPCError. For example, to detect if a command is unimplemented by the remote RPC server: The following full-blown client examples are in the examples directory:

Package assert provide assertions for testing

Package gcs provides an API for building and using a Golomb-coded set filter. A Golomb-Coded Set (GCS) is a space-efficient probabilistic data structure that is used to test set membership with a tunable false positive rate while simultaneously preventing false negatives. In other words, items that are in the set will always match, but items that are not in the set will also sometimes match with the chosen false positive rate. This package currently implements two different versions for backwards compatibility. Version 1 is deprecated and therefore should no longer be used. Version 2 is the GCS variation that follows the specification details in DCP0005: https://github.com/decred/dcps/blob/master/dcp-0005/dcp-0005.mediawiki#golomb-coded-sets. Version 2 sets do not permit empty items (data of zero length) to be added and are parameterized by the following: * A parameter `B` that defines the remainder code bit size * A parameter `M` that defines the false positive rate as `1/M` * A key for the SipHash-2-4 function * The items to include in the set Errors returned by this package are of type gcs.Error. This allows the caller to programmatically determine the specific error by examining the ErrorKind field of the type asserted gcs.Error while still providing rich error messages with contextual information. See ErrorKind in the package documentation for a full list. GCS is used as a mechanism for storing, transmitting, and committing to per-block filters. Consensus-validating full nodes commit to a single filter for every block and serve the filter to SPV clients that match against the filter locally to determine if the block is potentially relevant. The required parameters for Decred are defined by the blockcf2 package. For more details, see the Block Filters section of DCP0005: https://github.com/decred/dcps/blob/master/dcp-0005/dcp-0005.mediawiki#block-filters

Package rpcclient implements a websocket-enabled Decred JSON-RPC client. This client provides a robust and easy to use client for interfacing with a Decred RPC server that uses a mostly btcd/bitcoin core style Decred JSON-RPC API. This client has been tested with dcrd (https://github.com/decred/dcrd) and dcrwallet (https://github.com/decred/dcrwallet). In addition to the compatible standard HTTP POST JSON-RPC API, dcrd and dcrwallet provide a websocket interface that is more efficient than the standard HTTP POST method of accessing RPC. The section below discusses the differences between HTTP POST and websockets. By default, this client assumes the RPC server supports websockets and has TLS enabled. In practice, this currently means it assumes you are talking to dcrd or dcrwallet by default. However, configuration options are provided to fall back to HTTP POST and disable TLS to support talking with inferior bitcoin core style RPC servers. In HTTP POST-based JSON-RPC, every request creates a new HTTP connection, issues the call, waits for the response, and closes the connection. This adds quite a bit of overhead to every call and lacks flexibility for features such as notifications. In contrast, the websocket-based JSON-RPC interface provided by dcrd and dcrwallet only uses a single connection that remains open and allows asynchronous bi-directional communication. The websocket interface supports all of the same commands as HTTP POST, but they can be invoked without having to go through a connect/disconnect cycle for every call. In addition, the websocket interface provides other nice features such as the ability to register for asynchronous notifications of various events. The client provides both a synchronous (blocking) and asynchronous API. The synchronous (blocking) API is typically sufficient for most use cases. It works by issuing the RPC and blocking until the response is received. This allows straightforward code where you have the response as soon as the function returns. The asynchronous API works on the concept of futures. When you invoke the async version of a command, it will quickly return an instance of a type that promises to provide the result of the RPC at some future time. In the background, the RPC call is issued and the result is stored in the returned instance. Invoking the Receive method on the returned instance will either return the result immediately if it has already arrived, or block until it has. This is useful since it provides the caller with greater control over concurrency. The first important part of notifications is to realize that they will only work when connected via websockets. This should intuitively make sense because HTTP POST mode does not keep a connection open! All notifications provided by dcrd require registration to opt-in. For example, if you want to be notified when funds are received by a set of addresses, you register the addresses via the NotifyReceived (or NotifyReceivedAsync) function. Notifications are exposed by the client through the use of callback handlers which are setup via a NotificationHandlers instance that is specified by the caller when creating the client. It is important that these notification handlers complete quickly since they are intentionally in the main read loop and will block further reads until they complete. This provides the caller with the flexibility to decide what to do when notifications are coming in faster than they are being handled. In particular this means issuing a blocking RPC call from a callback handler will cause a deadlock as more server responses won't be read until the callback returns, but the callback would be waiting for a response. Thus, any additional RPCs must be issued an a completely decoupled manner. By default, when running in websockets mode, this client will automatically keep trying to reconnect to the RPC server should the connection be lost. There is a back-off in between each connection attempt until it reaches one try per minute. Once a connection is re-established, all previously registered notifications are automatically re-registered and any in-flight commands are re-issued. This means from the caller's perspective, the request simply takes longer to complete. The caller may invoke the Shutdown method on the client to force the client to cease reconnect attempts and return ErrClientShutdown for all outstanding commands. The automatic reconnection can be disabled by setting the DisableAutoReconnect flag to true in the connection config when creating the client. Minor RPC Server Differences and Chain/Wallet Separation Some of the commands are extensions specific to a particular RPC server. For example, the DebugLevel call is an extension only provided by dcrd (and dcrwallet passthrough). Therefore if you call one of these commands against an RPC server that doesn't provide them, you will get an unimplemented error from the server. An effort has been made to call out which commands are extensions in their documentation. Also, it is important to realize that dcrd intentionally separates the wallet functionality into a separate process named dcrwallet. This means if you are connected to the dcrd RPC server directly, only the RPCs which are related to chain services will be available. Depending on your application, you might only need chain-related RPCs. In contrast, dcrwallet provides pass through treatment for chain-related RPCs, so it supports them in addition to wallet-related RPCs. There are 3 categories of errors that will be returned throughout this package: The first category of errors are typically one of ErrInvalidAuth, ErrInvalidEndpoint, ErrClientDisconnect, or ErrClientShutdown. NOTE: The ErrClientDisconnect will not be returned unless the DisableAutoReconnect flag is set since the client automatically handles reconnect by default as previously described. The second category of errors typically indicates a programmer error and as such the type can vary, but usually will be best handled by simply showing/logging it. The third category of errors, that is errors returned by the server, can be detected by type asserting the error in a *dcrjson.RPCError. For example, to detect if a command is unimplemented by the remote RPC server: The following full-blown client examples are in the examples directory:

This is the official Go SDK for Oracle Cloud Infrastructure Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#installing for installation instructions. Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring for configuration instructions. The following example shows how to get started with the SDK. The example belows creates an identityClient struct with the default configuration. It then utilizes the identityClient to list availability domains and prints them out to stdout More examples can be found in the SDK Github repo: https://github.com/oracle/oci-go-sdk/tree/master/example Optional fields are represented with the `mandatory:"false"` tag on input structs. The SDK will omit all optional fields that are nil when making requests. In the case of enum-type fields, the SDK will omit fields whose value is an empty string. The SDK uses pointers for primitive types in many input structs. To aid in the construction of such structs, the SDK provides functions that return a pointer for a given value. For example: The SDK exposes functionality that allows the user to customize any http request before is sent to the service. You can do so by setting the `Interceptor` field in any of the `Client` structs. For example: The Interceptor closure gets called before the signing process, thus any changes done to the request will be properly signed and submitted to the service. The SDK exposes a stand-alone signer that can be used to signing custom requests. Related code can be found here: https://github.com/oracle/oci-go-sdk/blob/master/common/http_signer.go. The example below shows how to create a default signer. The signer also allows more granular control on the headers used for signing. For example: You can combine a custom signer with the exposed clients in the SDK. This allows you to add custom signed headers to the request. Following is an example: Bear in mind that some services have a white list of headers that it expects to be signed. Therefore, adding an arbitrary header can result in authentications errors. To see a runnable example, see https://github.com/oracle/oci-go-sdk/blob/master/example/example_identity_test.go For more information on the signing algorithm refer to: https://docs.cloud.oracle.com/Content/API/Concepts/signingrequests.htm Some operations accept or return polymorphic JSON objects. The SDK models such objects as interfaces. Further the SDK provides structs that implement such interfaces. Thus, for all operations that expect interfaces as input, pass the struct in the SDK that satisfies such interface. For example: In the case of a polymorphic response you can type assert the interface to the expected type. For example: An example of polymorphic JSON request handling can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_test.go#L63 When calling a list operation, the operation will retrieve a page of results. To retrieve more data, call the list operation again, passing in the value of the most recent response's OpcNextPage as the value of Page in the next list operation call. When there is no more data the OpcNextPage field will be nil. An example of pagination using this logic can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_pagination_test.go The SDK has a built-in logging mechanism used internally. The internal logging logic is used to record the raw http requests, responses and potential errors when (un)marshalling request and responses. Built-in logging in the SDK is controlled via the environment variable "OCI_GO_SDK_DEBUG" and its contents. The below are possible values for the "OCI_GO_SDK_DEBUG" variable 1. "info" or "i" enables all info logging messages 2. "debug" or "d" enables all debug and info logging messages 3. "verbose" or "v" or "1" enables all verbose, debug and info logging messages 4. "null" turns all logging messages off. If the value of the environment variable does not match any of the above then default logging level is "info". If the environment variable is not present then no logging messages are emitted. The default destination for logging is Stderr and if you want to output log to a file you can set via environment variable "OCI_GO_SDK_LOG_OUTPUT_MODE". The below are possible values 1. "file" or "f" enables all logging output saved to file 2. "combine" or "c" enables all logging output to both stderr and file You can also customize the log file location and name via "OCI_GO_SDK_LOG_FILE" environment variable, the value should be the path to a specific file If this environment variable is not present, the default location will be the project root path Sometimes you may need to wait until an attribute of a resource, such as an instance or a VCN, reaches a certain state. An example of this would be launching an instance and then waiting for the instance to become available, or waiting until a subnet in a VCN has been terminated. You might also want to retry the same operation again if there's network issue etc... This can be accomplished by using the RequestMetadata.RetryPolicy(request level configuration), alternatively, global(all services) or client level RetryPolicy configration is also possible. You can find the examples here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_retry_test.go If you are trying to make a PUT/POST API call with binary request body, please make sure the binary request body is resettable, which means the request body should inherit Seeker interface. The Retry behavior Precedence (Highest to lowest) is defined as below:- The OCI Go SDK defines a default retry policy that retries on the errors suitable for retries (see https://docs.oracle.com/en-us/iaas/Content/API/References/apierrors.htm), for a recommended period of time (up to 7 attempts spread out over at most approximately 1.5 minutes). The default retry policy is defined by : Default Retry-able Errors Below is the list of default retry-able errors for which retry attempts should be made. The following errors should be retried (with backoff). HTTP Code Customer-facing Error Code Apart from the above errors, retries should also be attempted in the following Client Side errors : 1. HTTP Connection timeout 2. Request Connection Errors 3. Request Exceptions 4. Other timeouts (like Read Timeout) The above errors can be avoided through retrying and hence, are classified as the default retry-able errors. Additionally, retries should also be made for Circuit Breaker exceptions (Exceptions raised by Circuit Breaker in an open state) Default Termination Strategy The termination strategy defines when SDKs should stop attempting to retry. In other words, it's the deadline for retries. The OCI SDKs should stop retrying the operation after 7 retry attempts. This means the SDKs will have retried for ~98 seconds or ~1.5 minutes have elapsed due to total delays. SDKs will make a total of 8 attempts. (1 initial request + 7 retries) Default Delay Strategy Default Delay Strategy - The delay strategy defines the amount of time to wait between each of the retry attempts. The default delay strategy chosen for the SDK – Exponential backoff with jitter, using: 1. The base time to use in retry calculations will be 1 second 2. An exponent of 2. When calculating the next retry time, the SDK will raise this to the power of the number of attempts 3. A maximum wait time between calls of 30 seconds (Capped) 4. Added jitter value between 0-1000 milliseconds to spread out the requests Configure and use default retry policy You can set this retry policy for a single request: or for all requests made by a client: or for all requests made by all clients: or setting default retry via environment varaible, which is a global switch for all services: Some services enable retry for operations by default, this can be overridden using any alternatives mentioned above. To know which service operations have retries enabled by default, look at the operation's description in the SDK - it will say whether that it has retries enabled by default Some resources may have to be replicated across regions and are only eventually consistent. That means the request to create, update, or delete the resource succeeded, but the resource is not available everywhere immediately. Creating, updating, or deleting any resource in the Identity service is affected by eventual consistency, and doing so may cause other operations in other services to fail until the Identity resource has been replicated. For example, the request to CreateTag in the Identity service in the home region succeeds, but immediately using that created tag in another region in a request to LaunchInstance in the Compute service may fail. If you are creating, updating, or deleting resources in the Identity service, we recommend using an eventually consistent retry policy for any service you access. The default retry policy already deals with eventual consistency. Example: This retry policy will use a different strategy if an eventually consistent change was made in the recent past (called the "eventually consistent window", currently defined to be 4 minutes after the eventually consistent change). This special retry policy for eventual consistency will: 1. make up to 9 attempts (including the initial attempt); if an attempt is successful, no more attempts will be made 2. retry at most until (a) approximately the end of the eventually consistent window or (b) the end of the default retry period of about 1.5 minutes, whichever is farther in the future; if an attempt is successful, no more attempts will be made, and the OCI Go SDK will not wait any longer 3. retry on the error codes 400-RelatedResourceNotAuthorizedOrNotFound, 404-NotAuthorizedOrNotFound, and 409-NotAuthorizedOrResourceAlreadyExists, for which the default retry policy does not retry, in addition to the errors the default retry policy retries on (see https://docs.oracle.com/en-us/iaas/Content/API/References/apierrors.htm) If there were no eventually consistent actions within the recent past, then this special retry strategy is not used. If you want a retry policy that does not handle eventual consistency in a special way, for example because you retry on all error responses, you can use DefaultRetryPolicyWithoutEventualConsistency or NewRetryPolicyWithOptions with the common.ReplaceWithValuesFromRetryPolicy(common.DefaultRetryPolicyWithoutEventualConsistency()) option: The NewRetryPolicy function also creates a retry policy without eventual consistency. Circuit Breaker can prevent an application repeatedly trying to execute an operation that is likely to fail, allowing it to continue without waiting for the fault to be rectified or wasting CPU cycles, of course, it also enables an application to detect whether the fault has been resolved. If the problem appears to have been rectified, the application can attempt to invoke the operation. Go SDK intergrates sony/gobreaker solution, wraps in a circuit breaker object, which monitors for failures. Once the failures reach a certain threshold, the circuit breaker trips, and all further calls to the circuit breaker return with an error, this also saves the service from being overwhelmed with network calls in case of an outage. Circuit Breaker Configuration Definitions 1. Failure Rate Threshold - The state of the CircuitBreaker changes from CLOSED to OPEN when the failure rate is equal or greater than a configurable threshold. For example when more than 50% of the recorded calls have failed. 2. Reset Timeout - The timeout after which an open circuit breaker will attempt a request if a request is made 3. Failure Exceptions - The list of Exceptions that will be regarded as failures for the circuit. 4. Minimum number of calls/ Volume threshold - Configures the minimum number of calls which are required (per sliding window period) before the CircuitBreaker can calculate the error rate. 1. Failure Rate Threshold - 80% - This means when 80% of the requests calculated for a time window of 120 seconds have failed then the circuit will transition from closed to open. 2. Minimum number of calls/ Volume threshold - A value of 10, for the above defined time window of 120 seconds. 3. Reset Timeout - 30 seconds to wait before setting the breaker to halfOpen state, and trying the action again. 4. Failure Exceptions - The failures for the circuit will only be recorded for the retryable/transient exceptions. This means only the following exceptions will be regarded as failure for the circuit. HTTP Code Customer-facing Error Code Apart from the above, the following client side exceptions will also be treated as a failure for the circuit : 1. HTTP Connection timeout 2. Request Connection Errors 3. Request Exceptions 4. Other timeouts (like Read Timeout) Go SDK enable circuit breaker with default configuration for most of the service clients, if you don't want to enable the solution, can disable the functionality before your application running Go SDK also supports customize Circuit Breaker with specified configurations. You can find the examples here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_circuitbreaker_test.go To know which service clients have circuit breakers enabled, look at the service client's description in the SDK - it will say whether that it has circuit breakers enabled by default The GO SDK uses the net/http package to make calls to OCI services. If your environment requires you to use a proxy server for outgoing HTTP requests then you can set this up in the following ways: 1. Configuring environment variable as described here https://golang.org/pkg/net/http/#ProxyFromEnvironment 2. Modifying the underlying Transport struct for a service client In order to modify the underlying Transport struct in HttpClient, you can do something similar to (sample code for audit service client): The Object Storage service supports multipart uploads to make large object uploads easier by splitting the large object into parts. The Go SDK supports raw multipart upload operations for advanced use cases, as well as a higher level upload class that uses the multipart upload APIs. For links to the APIs used for multipart upload operations, see Managing Multipart Uploads (https://docs.cloud.oracle.com/iaas/Content/Object/Tasks/usingmultipartuploads.htm). Higher level multipart uploads are implemented using the UploadManager, which will: split a large object into parts for you, upload the parts in parallel, and then recombine and commit the parts as a single object in storage. This code sample shows how to use the UploadManager to automatically split an object into parts for upload to simplify interaction with the Object Storage service: https://github.com/oracle/oci-go-sdk/blob/master/example/example_objectstorage_test.go Some response fields are enum-typed. In the future, individual services may return values not covered by existing enums for that field. To address this possibility, every enum-type response field is a modeled as a type that supports any string. Thus if a service returns a value that is not recognized by your version of the SDK, then the response field will be set to this value. When individual services return a polymorphic JSON response not available as a concrete struct, the SDK will return an implementation that only satisfies the interface modeling the polymorphic JSON response. If you are using a version of the SDK released prior to the announcement of a new region, you may need to use a workaround to reach it, depending on whether the region is in the oraclecloud.com realm. A region is a localized geographic area. For more information on regions and how to identify them, see Regions and Availability Domains(https://docs.cloud.oracle.com/iaas/Content/General/Concepts/regions.htm). A realm is a set of regions that share entities. You can identify your realm by looking at the domain name at the end of the network address. For example, the realm for xyz.abc.123.oraclecloud.com is oraclecloud.com. oraclecloud.com Realm: For regions in the oraclecloud.com realm, even if common.Region does not contain the new region, the forward compatibility of the SDK can automatically handle it. You can pass new region names just as you would pass ones that are already defined. For more information on passing region names in the configuration, see Configuring (https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring). For details on common.Region, see (https://github.com/oracle/oci-go-sdk/blob/master/common/common.go). Other Realms: For regions in realms other than oraclecloud.com, you can use the following workarounds to reach new regions with earlier versions of the SDK. NOTE: Be sure to supply the appropriate endpoints for your region. You can overwrite the target host with client.Host: If you are authenticating via instance principals, you can set the authentication endpoint in an environment variable: Got a fix for a bug, or a new feature you'd like to contribute? The SDK is open source and accepting pull requests on GitHub https://github.com/oracle/oci-go-sdk Licensing information available at: https://github.com/oracle/oci-go-sdk/blob/master/LICENSE.txt To be notified when a new version of the Go SDK is released, subscribe to the following feed: https://github.com/oracle/oci-go-sdk/releases.atom Please refer to this link: https://github.com/oracle/oci-go-sdk#help

This is the official Go SDK for Oracle Cloud Infrastructure Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#installing for installation instructions. Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring for configuration instructions. The following example shows how to get started with the SDK. The example belows creates an identityClient struct with the default configuration. It then utilizes the identityClient to list availability domains and prints them out to stdout More examples can be found in the SDK Github repo: https://github.com/oracle/oci-go-sdk/tree/master/example Optional fields are represented with the `mandatory:"false"` tag on input structs. The SDK will omit all optional fields that are nil when making requests. In the case of enum-type fields, the SDK will omit fields whose value is an empty string. The SDK uses pointers for primitive types in many input structs. To aid in the construction of such structs, the SDK provides functions that return a pointer for a given value. For example: Dedicated endpoints are the endpoint templates defined by the service for a specific realm at client level. OCI Go SDK allows you to enable the use of these realm-specific endpoint templates feature at application level and at client level. The value set at client level takes precedence over the value set at the application level. This feature is disabled by default. For reference, please refer https://github.com/oracle/oci-go-sdk/blob/master/example/example_objectstorage_test.go#L222-L251 The SDK exposes functionality that allows the user to customize any http request before is sent to the service. You can do so by setting the `Interceptor` field in any of the `Client` structs. For example: The Interceptor closure gets called before the signing process, thus any changes done to the request will be properly signed and submitted to the service. The SDK exposes a stand-alone signer that can be used to signing custom requests. Related code can be found here: https://github.com/oracle/oci-go-sdk/blob/master/common/http_signer.go. The example below shows how to create a default signer. The signer also allows more granular control on the headers used for signing. For example: You can combine a custom signer with the exposed clients in the SDK. This allows you to add custom signed headers to the request. Following is an example: Bear in mind that some services have a white list of headers that it expects to be signed. Therefore, adding an arbitrary header can result in authentications errors. To see a runnable example, see https://github.com/oracle/oci-go-sdk/blob/master/example/example_identity_test.go For more information on the signing algorithm refer to: https://docs.cloud.oracle.com/Content/API/Concepts/signingrequests.htm Some operations accept or return polymorphic JSON objects. The SDK models such objects as interfaces. Further the SDK provides structs that implement such interfaces. Thus, for all operations that expect interfaces as input, pass the struct in the SDK that satisfies such interface. For example: In the case of a polymorphic response you can type assert the interface to the expected type. For example: An example of polymorphic JSON request handling can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_test.go#L63 When calling a list operation, the operation will retrieve a page of results. To retrieve more data, call the list operation again, passing in the value of the most recent response's OpcNextPage as the value of Page in the next list operation call. When there is no more data the OpcNextPage field will be nil. An example of pagination using this logic can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_pagination_test.go The SDK has a built-in logging mechanism used internally. The internal logging logic is used to record the raw http requests, responses and potential errors when (un)marshalling request and responses. Built-in logging in the SDK is controlled via the environment variable "OCI_GO_SDK_DEBUG" and its contents. The below are possible values for the "OCI_GO_SDK_DEBUG" variable 1. "info" or "i" enables all info logging messages 2. "debug" or "d" enables all debug and info logging messages 3. "verbose" or "v" or "1" enables all verbose, debug and info logging messages 4. "null" turns all logging messages off. If the value of the environment variable does not match any of the above then default logging level is "info". If the environment variable is not present then no logging messages are emitted. You can also enable logs by code. For example This way you enable debug logs by code. The default destination for logging is Stderr and if you want to output log to a file you can set via environment variable "OCI_GO_SDK_LOG_OUTPUT_MODE". The below are possible values 1. "file" or "f" enables all logging output saved to file 2. "combine" or "c" enables all logging output to both stderr and file You can also customize the log file location and name via "OCI_GO_SDK_LOG_FILE" environment variable, the value should be the path to a specific file If this environment variable is not present, the default location will be the project root path Sometimes you may need to wait until an attribute of a resource, such as an instance or a VCN, reaches a certain state. An example of this would be launching an instance and then waiting for the instance to become available, or waiting until a subnet in a VCN has been terminated. You might also want to retry the same operation again if there's network issue etc... This can be accomplished by using the RequestMetadata.RetryPolicy(request level configuration), alternatively, global(all services) or client level RetryPolicy configration is also possible. You can find the examples here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_retry_test.go If you are trying to make a PUT/POST API call with binary request body, please make sure the binary request body is resettable, which means the request body should inherit Seeker interface. The Retry behavior Precedence (Highest to lowest) is defined as below:- The OCI Go SDK defines a default retry policy that retries on the errors suitable for retries (see https://docs.oracle.com/en-us/iaas/Content/API/References/apierrors.htm), for a recommended period of time (up to 7 attempts spread out over at most approximately 1.5 minutes). The default retry policy is defined by : Default Retry-able Errors Below is the list of default retry-able errors for which retry attempts should be made. The following errors should be retried (with backoff). HTTP Code Customer-facing Error Code Apart from the above errors, retries should also be attempted in the following Client Side errors : 1. HTTP Connection timeout 2. Request Connection Errors 3. Request Exceptions 4. Other timeouts (like Read Timeout) The above errors can be avoided through retrying and hence, are classified as the default retry-able errors. Additionally, retries should also be made for Circuit Breaker exceptions (Exceptions raised by Circuit Breaker in an open state) Default Termination Strategy The termination strategy defines when SDKs should stop attempting to retry. In other words, it's the deadline for retries. The OCI SDKs should stop retrying the operation after 7 retry attempts. This means the SDKs will have retried for ~98 seconds or ~1.5 minutes have elapsed due to total delays. SDKs will make a total of 8 attempts. (1 initial request + 7 retries) Default Delay Strategy Default Delay Strategy - The delay strategy defines the amount of time to wait between each of the retry attempts. The default delay strategy chosen for the SDK – Exponential backoff with jitter, using: 1. The base time to use in retry calculations will be 1 second 2. An exponent of 2. When calculating the next retry time, the SDK will raise this to the power of the number of attempts 3. A maximum wait time between calls of 30 seconds (Capped) 4. Added jitter value between 0-1000 milliseconds to spread out the requests Configure and use default retry policy You can set this retry policy for a single request: or for all requests made by a client: or for all requests made by all clients: or setting default retry via environment variable, which is a global switch for all services: Some services enable retry for operations by default, this can be overridden using any alternatives mentioned above. To know which service operations have retries enabled by default, look at the operation's description in the SDK - it will say whether that it has retries enabled by default Some resources may have to be replicated across regions and are only eventually consistent. That means the request to create, update, or delete the resource succeeded, but the resource is not available everywhere immediately. Creating, updating, or deleting any resource in the Identity service is affected by eventual consistency, and doing so may cause other operations in other services to fail until the Identity resource has been replicated. For example, the request to CreateTag in the Identity service in the home region succeeds, but immediately using that created tag in another region in a request to LaunchInstance in the Compute service may fail. If you are creating, updating, or deleting resources in the Identity service, we recommend using an eventually consistent retry policy for any service you access. The default retry policy already deals with eventual consistency. Example: This retry policy will use a different strategy if an eventually consistent change was made in the recent past (called the "eventually consistent window", currently defined to be 4 minutes after the eventually consistent change). This special retry policy for eventual consistency will: 1. make up to 9 attempts (including the initial attempt); if an attempt is successful, no more attempts will be made 2. retry at most until (a) approximately the end of the eventually consistent window or (b) the end of the default retry period of about 1.5 minutes, whichever is farther in the future; if an attempt is successful, no more attempts will be made, and the OCI Go SDK will not wait any longer 3. retry on the error codes 400-RelatedResourceNotAuthorizedOrNotFound, 404-NotAuthorizedOrNotFound, and 409-NotAuthorizedOrResourceAlreadyExists, for which the default retry policy does not retry, in addition to the errors the default retry policy retries on (see https://docs.oracle.com/en-us/iaas/Content/API/References/apierrors.htm) If there were no eventually consistent actions within the recent past, then this special retry strategy is not used. If you want a retry policy that does not handle eventual consistency in a special way, for example because you retry on all error responses, you can use DefaultRetryPolicyWithoutEventualConsistency or NewRetryPolicyWithOptions with the common.ReplaceWithValuesFromRetryPolicy(common.DefaultRetryPolicyWithoutEventualConsistency()) option: The NewRetryPolicy function also creates a retry policy without eventual consistency. Circuit Breaker can prevent an application repeatedly trying to execute an operation that is likely to fail, allowing it to continue without waiting for the fault to be rectified or wasting CPU cycles, of course, it also enables an application to detect whether the fault has been resolved. If the problem appears to have been rectified, the application can attempt to invoke the operation. Go SDK intergrates sony/gobreaker solution, wraps in a circuit breaker object, which monitors for failures. Once the failures reach a certain threshold, the circuit breaker trips, and all further calls to the circuit breaker return with an error, this also saves the service from being overwhelmed with network calls in case of an outage. Circuit Breaker Configuration Definitions 1. Failure Rate Threshold - The state of the CircuitBreaker changes from CLOSED to OPEN when the failure rate is equal or greater than a configurable threshold. For example when more than 50% of the recorded calls have failed. 2. Reset Timeout - The timeout after which an open circuit breaker will attempt a request if a request is made 3. Failure Exceptions - The list of Exceptions that will be regarded as failures for the circuit. 4. Minimum number of calls/ Volume threshold - Configures the minimum number of calls which are required (per sliding window period) before the CircuitBreaker can calculate the error rate. 1. Failure Rate Threshold - 80% - This means when 80% of the requests calculated for a time window of 120 seconds have failed then the circuit will transition from closed to open. 2. Minimum number of calls/ Volume threshold - A value of 10, for the above defined time window of 120 seconds. 3. Reset Timeout - 30 seconds to wait before setting the breaker to halfOpen state, and trying the action again. 4. Failure Exceptions - The failures for the circuit will only be recorded for the retryable/transient exceptions. This means only the following exceptions will be regarded as failure for the circuit. HTTP Code Customer-facing Error Code Apart from the above, the following client side exceptions will also be treated as a failure for the circuit : 1. HTTP Connection timeout 2. Request Connection Errors 3. Request Exceptions 4. Other timeouts (like Read Timeout) Go SDK enable circuit breaker with default configuration for most of the service clients, if you don't want to enable the solution, can disable the functionality before your application running Go SDK also supports customize Circuit Breaker with specified configurations. You can find the examples here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_circuitbreaker_test.go To know which service clients have circuit breakers enabled, look at the service client's description in the SDK - it will say whether that it has circuit breakers enabled by default As a result of the SDK treating responses with a non-2xx HTTP status code as an error, the SDK will produce an error on 3xx responses. This can impact operations which support conditional GETs, such as GetObject() and HeadObject() methods as these can return responses with an HTTP status code of 304 if passed an 'IfNoneMatch' that corresponds to the current etag of the object / bucket. In order to account for this, you should check for status code 304 when an error is produced. For example: The GO SDK uses the net/http package to make calls to OCI services. If your environment requires you to use a proxy server for outgoing HTTP requests then you can set this up in the following ways: 1. Configuring environment variable as described here https://golang.org/pkg/net/http/#ProxyFromEnvironment 2. Modifying the underlying Transport struct for a service client In order to modify the underlying Transport struct in HttpClient, you can do something similar to (sample code for audit service client): The Object Storage service supports multipart uploads to make large object uploads easier by splitting the large object into parts. The Go SDK supports raw multipart upload operations for advanced use cases, as well as a higher level upload class that uses the multipart upload APIs. For links to the APIs used for multipart upload operations, see Managing Multipart Uploads (https://docs.cloud.oracle.com/iaas/Content/Object/Tasks/usingmultipartuploads.htm). Higher level multipart uploads are implemented using the UploadManager, which will: split a large object into parts for you, upload the parts in parallel, and then recombine and commit the parts as a single object in storage. This code sample shows how to use the UploadManager to automatically split an object into parts for upload to simplify interaction with the Object Storage service: https://github.com/oracle/oci-go-sdk/blob/master/example/example_objectstorage_test.go Some response fields are enum-typed. In the future, individual services may return values not covered by existing enums for that field. To address this possibility, every enum-type response field is a modeled as a type that supports any string. Thus if a service returns a value that is not recognized by your version of the SDK, then the response field will be set to this value. When individual services return a polymorphic JSON response not available as a concrete struct, the SDK will return an implementation that only satisfies the interface modeling the polymorphic JSON response. If you are using a version of the SDK released prior to the announcement of a new region, you may need to use a workaround to reach it, depending on whether the region is in the oraclecloud.com realm. A region is a localized geographic area. For more information on regions and how to identify them, see Regions and Availability Domains(https://docs.cloud.oracle.com/iaas/Content/General/Concepts/regions.htm). A realm is a set of regions that share entities. You can identify your realm by looking at the domain name at the end of the network address. For example, the realm for xyz.abc.123.oraclecloud.com is oraclecloud.com. oraclecloud.com Realm: For regions in the oraclecloud.com realm, even if common.Region does not contain the new region, the forward compatibility of the SDK can automatically handle it. You can pass new region names just as you would pass ones that are already defined. For more information on passing region names in the configuration, see Configuring (https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring). For details on common.Region, see (https://github.com/oracle/oci-go-sdk/blob/master/common/common.go). Other Realms: For regions in realms other than oraclecloud.com, you can use the following workarounds to reach new regions with earlier versions of the SDK. NOTE: Be sure to supply the appropriate endpoints for your region. You can overwrite the target host with client.Host: If you are authenticating via instance principals, you can set the authentication endpoint in an environment variable: In order to use a custom CA bundle, you can set the environment variable OCI_DEFAULT_CERTS_PATH to point to the path of custom CA Bundle you want the OCI GO SDK to use while making API calls to the OCI services If you additionally want to set custom leaf/client certs, then you can use the the environment variables OCI_DEFAULT_CLIENT_CERTS_PATH and OCI_DEFAULT_CLIENT_CERTS_PRIVATE_KEY_PATH to set the path of the custom client/leaf cert and the private key respectively. The default refresh interval for custom CA bundle or client certs is 30 minutes. If you want to modify this, then you can configure the refresh interval in minutes by using either the Global property OciGlobalRefreshIntervalForCustomCerts defined in the common package or set the environment variable OCI_DEFAULT_REFRESH_INTERVAL_FOR_CUSTOM_CERTS to set it instead. Please note, that the property OciGlobalRefreshIntervalForCustomCerts has a higher precedence than the environment variable OCI_DEFAULT_REFRESH_INTERVAL_FOR_CUSTOM_CERTS. If this value is negative, then it would be assumed that it is unset. If it is set to 0, then the SDK would disable the custom ca bundle and client cert refresh Got a fix for a bug, or a new feature you'd like to contribute? The SDK is open source and accepting pull requests on GitHub https://github.com/oracle/oci-go-sdk Licensing information available at: https://github.com/oracle/oci-go-sdk/blob/master/LICENSE.txt To be notified when a new version of the Go SDK is released, subscribe to the following feed: https://github.com/oracle/oci-go-sdk/releases.atom Please refer to this link: https://github.com/oracle/oci-go-sdk#help

This is the official Go SDK for Oracle Cloud Infrastructure Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#installing for installation instructions. Refer to https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring for configuration instructions. The following example shows how to get started with the SDK. The example belows creates an identityClient struct with the default configuration. It then utilizes the identityClient to list availability domains and prints them out to stdout More examples can be found in the SDK Github repo: https://github.com/oracle/oci-go-sdk/tree/master/example Optional fields are represented with the `mandatory:"false"` tag on input structs. The SDK will omit all optional fields that are nil when making requests. In the case of enum-type fields, the SDK will omit fields whose value is an empty string. The SDK uses pointers for primitive types in many input structs. To aid in the construction of such structs, the SDK provides functions that return a pointer for a given value. For example: The SDK exposes functionality that allows the user to customize any http request before is sent to the service. You can do so by setting the `Interceptor` field in any of the `Client` structs. For example: The Interceptor closure gets called before the signing process, thus any changes done to the request will be properly signed and submitted to the service. The SDK exposes a stand-alone signer that can be used to signing custom requests. Related code can be found here: https://github.com/oracle/oci-go-sdk/blob/master/common/http_signer.go. The example below shows how to create a default signer. The signer also allows more granular control on the headers used for signing. For example: You can combine a custom signer with the exposed clients in the SDK. This allows you to add custom signed headers to the request. Following is an example: Bear in mind that some services have a white list of headers that it expects to be signed. Therefore, adding an arbitrary header can result in authentications errors. To see a runnable example, see https://github.com/oracle/oci-go-sdk/blob/master/example/example_identity_test.go For more information on the signing algorithm refer to: https://docs.cloud.oracle.com/Content/API/Concepts/signingrequests.htm Some operations accept or return polymorphic JSON objects. The SDK models such objects as interfaces. Further the SDK provides structs that implement such interfaces. Thus, for all operations that expect interfaces as input, pass the struct in the SDK that satisfies such interface. For example: In the case of a polymorphic response you can type assert the interface to the expected type. For example: An example of polymorphic JSON request handling can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_test.go#L63 When calling a list operation, the operation will retrieve a page of results. To retrieve more data, call the list operation again, passing in the value of the most recent response's OpcNextPage as the value of Page in the next list operation call. When there is no more data the OpcNextPage field will be nil. An example of pagination using this logic can be found here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_core_pagination_test.go The SDK has a built-in logging mechanism used internally. The internal logging logic is used to record the raw http requests, responses and potential errors when (un)marshalling request and responses. Built-in logging in the SDK is controlled via the environment variable "OCI_GO_SDK_DEBUG" and its contents. The below are possible values for the "OCI_GO_SDK_DEBUG" variable 1. "info" or "i" enables all info logging messages 2. "debug" or "d" enables all debug and info logging messages 3. "verbose" or "v" or "1" enables all verbose, debug and info logging messages 4. "null" turns all logging messages off. If the value of the environment variable does not match any of the above then default logging level is "info". If the environment variable is not present then no logging messages are emitted. The default destination for logging is Stderr and if you want to output log to a file you can set via environment variable "OCI_GO_SDK_LOG_OUTPUT_MODE". The below are possible values 1. "file" or "f" enables all logging output saved to file 2. "combine" or "c" enables all logging output to both stderr and file You can also customize the log file location and name via "OCI_GO_SDK_LOG_FILE" environment variable, the value should be the path to a specific file If this environment variable is not present, the default location will be the project root path Sometimes you may need to wait until an attribute of a resource, such as an instance or a VCN, reaches a certain state. An example of this would be launching an instance and then waiting for the instance to become available, or waiting until a subnet in a VCN has been terminated. You might also want to retry the same operation again if there's network issue etc... This can be accomplished by using the RequestMetadata.RetryPolicy(request level configuration), alternatively, global(all services) or client level RetryPolicy configration is also possible. You can find the examples here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_retry_test.go If you are trying to make a PUT/POST API call with binary request body, please make sure the binary request body is resettable, which means the request body should inherit Seeker interface. The Retry behavior Precedence (Highest to lowest) is defined as below:- The OCI Go SDK defines a default retry policy that retries on the errors suitable for retries (see https://docs.oracle.com/en-us/iaas/Content/API/References/apierrors.htm), for a recommended period of time (up to 7 attempts spread out over at most approximately 1.5 minutes). The default retry policy is defined by : Default Retry-able Errors Below is the list of default retry-able errors for which retry attempts should be made. The following errors should be retried (with backoff). HTTP Code Customer-facing Error Code Apart from the above errors, retries should also be attempted in the following Client Side errors : 1. HTTP Connection timeout 2. Request Connection Errors 3. Request Exceptions 4. Other timeouts (like Read Timeout) The above errors can be avoided through retrying and hence, are classified as the default retry-able errors. Additionally, retries should also be made for Circuit Breaker exceptions (Exceptions raised by Circuit Breaker in an open state) Default Termination Strategy The termination strategy defines when SDKs should stop attempting to retry. In other words, it's the deadline for retries. The OCI SDKs should stop retrying the operation after 7 retry attempts. This means the SDKs will have retried for ~98 seconds or ~1.5 minutes have elapsed due to total delays. SDKs will make a total of 8 attempts. (1 initial request + 7 retries) Default Delay Strategy Default Delay Strategy - The delay strategy defines the amount of time to wait between each of the retry attempts. The default delay strategy chosen for the SDK – Exponential backoff with jitter, using: 1. The base time to use in retry calculations will be 1 second 2. An exponent of 2. When calculating the next retry time, the SDK will raise this to the power of the number of attempts 3. A maximum wait time between calls of 30 seconds (Capped) 4. Added jitter value between 0-1000 milliseconds to spread out the requests Configure and use default retry policy You can set this retry policy for a single request: or for all requests made by a client: or for all requests made by all clients: or setting default retry via environment varaible, which is a global switch for all services: Some services enable retry for operations by default, this can be overridden using any alternatives mentioned above. To know which service operations have retries enabled by default, look at the operation's description in the SDK - it will say whether that it has retries enabled by default Some resources may have to be replicated across regions and are only eventually consistent. That means the request to create, update, or delete the resource succeeded, but the resource is not available everywhere immediately. Creating, updating, or deleting any resource in the Identity service is affected by eventual consistency, and doing so may cause other operations in other services to fail until the Identity resource has been replicated. For example, the request to CreateTag in the Identity service in the home region succeeds, but immediately using that created tag in another region in a request to LaunchInstance in the Compute service may fail. If you are creating, updating, or deleting resources in the Identity service, we recommend using an eventually consistent retry policy for any service you access. The default retry policy already deals with eventual consistency. Example: This retry policy will use a different strategy if an eventually consistent change was made in the recent past (called the "eventually consistent window", currently defined to be 4 minutes after the eventually consistent change). This special retry policy for eventual consistency will: 1. make up to 9 attempts (including the initial attempt); if an attempt is successful, no more attempts will be made 2. retry at most until (a) approximately the end of the eventually consistent window or (b) the end of the default retry period of about 1.5 minutes, whichever is farther in the future; if an attempt is successful, no more attempts will be made, and the OCI Go SDK will not wait any longer 3. retry on the error codes 400-RelatedResourceNotAuthorizedOrNotFound, 404-NotAuthorizedOrNotFound, and 409-NotAuthorizedOrResourceAlreadyExists, for which the default retry policy does not retry, in addition to the errors the default retry policy retries on (see https://docs.oracle.com/en-us/iaas/Content/API/References/apierrors.htm) If there were no eventually consistent actions within the recent past, then this special retry strategy is not used. If you want a retry policy that does not handle eventual consistency in a special way, for example because you retry on all error responses, you can use DefaultRetryPolicyWithoutEventualConsistency or NewRetryPolicyWithOptions with the common.ReplaceWithValuesFromRetryPolicy(common.DefaultRetryPolicyWithoutEventualConsistency()) option: The NewRetryPolicy function also creates a retry policy without eventual consistency. Circuit Breaker can prevent an application repeatedly trying to execute an operation that is likely to fail, allowing it to continue without waiting for the fault to be rectified or wasting CPU cycles, of course, it also enables an application to detect whether the fault has been resolved. If the problem appears to have been rectified, the application can attempt to invoke the operation. Go SDK intergrates sony/gobreaker solution, wraps in a circuit breaker object, which monitors for failures. Once the failures reach a certain threshold, the circuit breaker trips, and all further calls to the circuit breaker return with an error, this also saves the service from being overwhelmed with network calls in case of an outage. Circuit Breaker Configuration Definitions 1. Failure Rate Threshold - The state of the CircuitBreaker changes from CLOSED to OPEN when the failure rate is equal or greater than a configurable threshold. For example when more than 50% of the recorded calls have failed. 2. Reset Timeout - The timeout after which an open circuit breaker will attempt a request if a request is made 3. Failure Exceptions - The list of Exceptions that will be regarded as failures for the circuit. 4. Minimum number of calls/ Volume threshold - Configures the minimum number of calls which are required (per sliding window period) before the CircuitBreaker can calculate the error rate. 1. Failure Rate Threshold - 80% - This means when 80% of the requests calculated for a time window of 120 seconds have failed then the circuit will transition from closed to open. 2. Minimum number of calls/ Volume threshold - A value of 10, for the above defined time window of 120 seconds. 3. Reset Timeout - 30 seconds to wait before setting the breaker to halfOpen state, and trying the action again. 4. Failure Exceptions - The failures for the circuit will only be recorded for the retryable/transient exceptions. This means only the following exceptions will be regarded as failure for the circuit. HTTP Code Customer-facing Error Code Apart from the above, the following client side exceptions will also be treated as a failure for the circuit : 1. HTTP Connection timeout 2. Request Connection Errors 3. Request Exceptions 4. Other timeouts (like Read Timeout) Go SDK enable circuit breaker with default configuration for most of the service clients, if you don't want to enable the solution, can disable the functionality before your application running Go SDK also supports customize Circuit Breaker with specified configurations. You can find the examples here: https://github.com/oracle/oci-go-sdk/blob/master/example/example_circuitbreaker_test.go To know which service clients have circuit breakers enabled, look at the service client's description in the SDK - it will say whether that it has circuit breakers enabled by default The GO SDK uses the net/http package to make calls to OCI services. If your environment requires you to use a proxy server for outgoing HTTP requests then you can set this up in the following ways: 1. Configuring environment variable as described here https://golang.org/pkg/net/http/#ProxyFromEnvironment 2. Modifying the underlying Transport struct for a service client In order to modify the underlying Transport struct in HttpClient, you can do something similar to (sample code for audit service client): The Object Storage service supports multipart uploads to make large object uploads easier by splitting the large object into parts. The Go SDK supports raw multipart upload operations for advanced use cases, as well as a higher level upload class that uses the multipart upload APIs. For links to the APIs used for multipart upload operations, see Managing Multipart Uploads (https://docs.cloud.oracle.com/iaas/Content/Object/Tasks/usingmultipartuploads.htm). Higher level multipart uploads are implemented using the UploadManager, which will: split a large object into parts for you, upload the parts in parallel, and then recombine and commit the parts as a single object in storage. This code sample shows how to use the UploadManager to automatically split an object into parts for upload to simplify interaction with the Object Storage service: https://github.com/oracle/oci-go-sdk/blob/master/example/example_objectstorage_test.go Some response fields are enum-typed. In the future, individual services may return values not covered by existing enums for that field. To address this possibility, every enum-type response field is a modeled as a type that supports any string. Thus if a service returns a value that is not recognized by your version of the SDK, then the response field will be set to this value. When individual services return a polymorphic JSON response not available as a concrete struct, the SDK will return an implementation that only satisfies the interface modeling the polymorphic JSON response. If you are using a version of the SDK released prior to the announcement of a new region, you may need to use a workaround to reach it, depending on whether the region is in the oraclecloud.com realm. A region is a localized geographic area. For more information on regions and how to identify them, see Regions and Availability Domains(https://docs.cloud.oracle.com/iaas/Content/General/Concepts/regions.htm). A realm is a set of regions that share entities. You can identify your realm by looking at the domain name at the end of the network address. For example, the realm for xyz.abc.123.oraclecloud.com is oraclecloud.com. oraclecloud.com Realm: For regions in the oraclecloud.com realm, even if common.Region does not contain the new region, the forward compatibility of the SDK can automatically handle it. You can pass new region names just as you would pass ones that are already defined. For more information on passing region names in the configuration, see Configuring (https://github.com/oracle/oci-go-sdk/blob/master/README.md#configuring). For details on common.Region, see (https://github.com/oracle/oci-go-sdk/blob/master/common/common.go). Other Realms: For regions in realms other than oraclecloud.com, you can use the following workarounds to reach new regions with earlier versions of the SDK. NOTE: Be sure to supply the appropriate endpoints for your region. You can overwrite the target host with client.Host: If you are authenticating via instance principals, you can set the authentication endpoint in an environment variable: Got a fix for a bug, or a new feature you'd like to contribute? The SDK is open source and accepting pull requests on GitHub https://github.com/oracle/oci-go-sdk Licensing information available at: https://github.com/oracle/oci-go-sdk/blob/master/LICENSE.txt To be notified when a new version of the Go SDK is released, subscribe to the following feed: https://github.com/oracle/oci-go-sdk/releases.atom Please refer to this link: https://github.com/oracle/oci-go-sdk#help

Package rpcclient implements a websocket-enabled Decred JSON-RPC client. This client provides a robust and easy to use client for interfacing with a Decred RPC server that uses a mostly btcd/bitcoin core style Decred JSON-RPC API. This client has been tested with dcrd (https://github.com/decred/dcrd) and dcrwallet (https://github.com/decred/dcrwallet). In addition to the compatible standard HTTP POST JSON-RPC API, dcrd and dcrwallet provide a websocket interface that is more efficient than the standard HTTP POST method of accessing RPC. The section below discusses the differences between HTTP POST and websockets. By default, this client assumes the RPC server supports websockets and has TLS enabled. In practice, this currently means it assumes you are talking to dcrd or dcrwallet by default. However, configuration options are provided to fall back to HTTP POST and disable TLS to support talking with inferior bitcoin core style RPC servers. In HTTP POST-based JSON-RPC, every request creates a new HTTP connection, issues the call, waits for the response, and closes the connection. This adds quite a bit of overhead to every call and lacks flexibility for features such as notifications. In contrast, the websocket-based JSON-RPC interface provided by dcrd and dcrwallet only uses a single connection that remains open and allows asynchronous bi-directional communication. The websocket interface supports all of the same commands as HTTP POST, but they can be invoked without having to go through a connect/disconnect cycle for every call. In addition, the websocket interface provides other nice features such as the ability to register for asynchronous notifications of various events. The client provides both a synchronous (blocking) and asynchronous API. The synchronous (blocking) API is typically sufficient for most use cases. It works by issuing the RPC and blocking until the response is received. This allows straightforward code where you have the response as soon as the function returns. The asynchronous API works on the concept of futures. When you invoke the async version of a command, it will quickly return an instance of a type that promises to provide the result of the RPC at some future time. In the background, the RPC call is issued and the result is stored in the returned instance. Invoking the Receive method on the returned instance will either return the result immediately if it has already arrived, or block until it has. This is useful since it provides the caller with greater control over concurrency. The first important part of notifications is to realize that they will only work when connected via websockets. This should intuitively make sense because HTTP POST mode does not keep a connection open! All notifications provided by dcrd require registration to opt-in. For example, if you want to be notified when funds are received by a set of addresses, you register the addresses via the NotifyReceived (or NotifyReceivedAsync) function. Notifications are exposed by the client through the use of callback handlers which are setup via a NotificationHandlers instance that is specified by the caller when creating the client. It is important that these notification handlers complete quickly since they are intentionally in the main read loop and will block further reads until they complete. This provides the caller with the flexibility to decide what to do when notifications are coming in faster than they are being handled. In particular this means issuing a blocking RPC call from a callback handler will cause a deadlock as more server responses won't be read until the callback returns, but the callback would be waiting for a response. Thus, any additional RPCs must be issued an a completely decoupled manner. By default, when running in websockets mode, this client will automatically keep trying to reconnect to the RPC server should the connection be lost. There is a back-off in between each connection attempt until it reaches one try per minute. Once a connection is re-established, all previously registered notifications are automatically re-registered and any in-flight commands are re-issued. This means from the caller's perspective, the request simply takes longer to complete. The caller may invoke the Shutdown method on the client to force the client to cease reconnect attempts and return ErrClientShutdown for all outstanding commands. The automatic reconnection can be disabled by setting the DisableAutoReconnect flag to true in the connection config when creating the client. Minor RPC Server Differences and Chain/Wallet Separation Some of the commands are extensions specific to a particular RPC server. For example, the DebugLevel call is an extension only provided by dcrd (and dcrwallet passthrough). Therefore if you call one of these commands against an RPC server that doesn't provide them, you will get an unimplemented error from the server. An effort has been made to call out which commmands are extensions in their documentation. Also, it is important to realize that dcrd intentionally separates the wallet functionality into a separate process named dcrwallet. This means if you are connected to the dcrd RPC server directly, only the RPCs which are related to chain services will be available. Depending on your application, you might only need chain-related RPCs. In contrast, dcrwallet provides pass through treatment for chain-related RPCs, so it supports them in addition to wallet-related RPCs. There are 3 categories of errors that will be returned throughout this package: The first category of errors are typically one of ErrInvalidAuth, ErrInvalidEndpoint, ErrClientDisconnect, or ErrClientShutdown. NOTE: The ErrClientDisconnect will not be returned unless the DisableAutoReconnect flag is set since the client automatically handles reconnect by default as previously described. The second category of errors typically indicates a programmer error and as such the type can vary, but usually will be best handled by simply showing/logging it. The third category of errors, that is errors returned by the server, can be detected by type asserting the error in a *dcrjson.RPCError. For example, to detect if a command is unimplemented by the remote RPC server: The following full-blown client examples are in the examples directory: