REST Resources Provided By: Bitbucket Server - Repository Ref Sync

This is the reference document for the Atlassian Bitbucket REST API. The REST API is for developers who want to:

You can read more about developing Bitbucket plugins in the Bitbucket Developer Documentation.

Getting started

Because the REST API is based on open standards, you can use any web development language or command line tool capable of generating an HTTP request to access the API. See the developer documentation for a basic usage example.

If you're already working with the Atlassian SDK, the REST API Browser is a great tool for exploring and experimenting with the Bitbucket REST API.

Structure of the REST URIs

Bitbucket's REST APIs provide access to resources (data entities) via URI paths. To use a REST API, your application will make an HTTP request and parse the response. The Bitbucket REST API uses JSON as its communication format, and the standard HTTP methods like GET, PUT, POST and DELETE. URIs for Bitbucket's REST API resource have the following structure:

    http://host:port/context/rest/api-name/api-version/path/to/resource

For example, the following URI would retrieve a page of the latest commits to the jira repository in the JIRA project on https://stash.atlassian.com.

    https://stash.atlassian.com/rest/api/1.0/projects/JIRA/repos/jira/commits

See the API descriptions below for a full list of available resources.

Alternatively we also publish a list of resources in WADL format. It is available here.

Paged APIs

Bitbucket uses paging to conserve server resources and limit response size for resources that return potentially large collections of items. A request to a paged API will result in a values array wrapped in a JSON object with some paging metadata, like this:

    {
        "size": 3,
        "limit": 3,
        "isLastPage": false,
        "values": [
            { /* result 0 */ },
            { /* result 1 */ },
            { /* result 2 */ }
        ],
        "start": 0,
        "filter": null,
        "nextPageStart": 3
    }

Clients can use the limit and start query parameters to retrieve the desired number of results.

The limit parameter indicates how many results to return per page. Most APIs default to returning 25 if the limit is left unspecified. This number can be increased, but note that a resource-specific hard limit will apply. These hard limits can be configured by server administrators, so it's always best practice to check the limit attribute on the response to see what limit has been applied. The request to get a larger page should look like this:

    http://host:port/context/rest/api-name/api-version/path/to/resource?limit={desired size of page}

For example:

    https://stash.atlassian.com/rest/api/1.0/projects/JIRA/repos/jira/commits?limit=1000

The start parameter indicates which item should be used as the first item in the page of results. All paged responses contain an isLastPage attribute indicating whether another page of items exists.

Important: If more than one page exists (i.e. the response contains "isLastPage": false), the response object will also contain a nextPageStart attribute which must be used by the client as the start parameter on the next request. Identifiers of adjacent objects in a page may not be contiguous, so the start of the next page is not necessarily the start of the last page plus the last page's size. A client should always use nextPageStart to avoid unexpected results from a paged API. The request to get a subsequent page should look like this:

    http://host:port/context/rest/api-name/api-version/path/to/resource?start={nextPageStart from previous response}

For example:

    https://stash.atlassian.com/rest/api/1.0/projects/JIRA/repos/jira/commits?start=25

Authentication

Any authentication that works against Bitbucket will work against the REST API. The preferred authentication methods are HTTP Basic (when using SSL) and OAuth. Other supported methods include: HTTP Cookies and Trusted Applications.

You can find OAuth code samples in several programming languages at bitbucket.org/atlassian_tutorial/atlassian-oauth-examples.

The log-in page uses cookie-based authentication, so if you are using Bitbucket in a browser you can call REST from JavaScript on the page and rely on the authentication that the browser has established.

Errors & Validation

If a request fails due to client error, the resource will return an HTTP response code in the 40x range. These can be broadly categorised into:

HTTP Code Description
400 (Bad Request) One or more of the required parameters or attributes:
  • were missing from the request;
  • incorrectly formatted; or
  • inappropriate in the given context.
401 (Unauthorized) Either:
  • Authentication is required but was not attempted.
  • Authentication was attempted but failed.
  • Authentication was successful but the authenticated user does not have the requisite permission for the resource.
See the individual resource documentation for details of required permissions.
403 (Forbidden) Actions are usually "forbidden" if they involve breaching the licensed user limit of the server, or degrading the authenticated user's permission level. See the individual resource documentation for more details.
404 (Not Found) The entity you are attempting to access, or the project or repository containing it, does not exist.
405 (Method Not Allowed) The request HTTP method is not appropriate for the targeted resource. For example an HTTP GET to a resource that only accepts an HTTP POST will result in a 405.
409 (Conflict) The attempted update failed due to some conflict with an existing resource. For example:
  • Creating a project with a key that already exists
  • Merging an out-of-date pull request
  • Deleting a comment that has replies
  • etc.
See the individual resource documentation for more details.
415 (Unsupported Media Type) The request entity has a Content-Type that the server does not support. Almost all of the Bitbucket REST API accepts application/json format, but check the individual resource documentation for more details. Additionally, double-check that you are setting the Content-Type header correctly on your request (e.g. using -H "Content-Type: application/json" in cURL).

For 400 HTTP codes the response will typically contain one or more validation error messages, for example:

    {
        "errors": [
            {
                "context": "name",
                "message": "The name should be between 1 and 255 characters.",
                "exceptionName": null
            },
            {
                "context": "email",
                "message": "The email should be a valid email address.",
                "exceptionName": null
            }
        ]
    }
    

The context attribute indicates which parameter or request entity attribute failed validation. Note that the context may be null.

For 401, 403, 404 and 409 HTTP codes, the response will contain one or more descriptive error messages:

    {
        "errors": [
            {
                "context": null,
                "message": "A detailed error message.",
                "exceptionName": null
            }
        ]
    }
    

A 500 (Server Error) HTTP code indicates an incorrect resource url or an unexpected server error. Double-check the URL you are trying to access, then report the issue to your server administrator or Atlassian Support if problems persist.

Personal Repositories

Bitbucket allows users to manage their own repositories, called personal repositories. These are repositories associated with the user and to which they always have REPO_ADMIN permission.

Accessing personal repositories via REST is achieved through the normal project-centric REST URLs using the user's slug prefixed by tilde as the project key. E.g. to list personal repositories for a user with slug "johnsmith" you would make a GET to:

http://example.com/rest/api/1.0/projects/~johnsmith/repos

In addition to this, Bitbucket allows access to these repositories through an alternate set of user-centric REST URLs beginning with:

http://example.com/rest/api/1.0/users/~{userSlug}/repos
E.g. to list the forks of the repository with slug "nodejs" in the personal project of user with slug "johnsmith" using the regular REST URL you would make a GET to:
http://example.com/rest/api/1.0/projects/~johnsmith/repos/nodejs/forks
Using the alternate URL, you would make a GET to:
http://example.com/rest/api/1.0/users/johnsmith/repos/nodejs/forks

Index

Adds support for ref synchronization between forks and their upstream repositories

Resources

/rest/sync/1.0/projects/{projectKey}/repos/{repositorySlug}

Methods

POST

Enables or disables synchronization for the specified repository. When synchronization is enabled, branches within the repository are immediately synchronized and the status is updated with the outcome. That initial synchronization is performed before the REST request returns, allowing it to return the updated status.

The authenticated user must have REPO_ADMIN permission for the specified repository. Anonymous users cannot manage synchronization, even on public repositories. Additionally, synchronization must be available for the specified repository. Synchronization is only available if:

  • The repository is a fork, since its origin is used as upstream
  • The owning user still has access to the fork's origin, if the repository is a personal fork

Example request representations:

  • application/json [expand]

    Example
    {"enabled":true}

Example response representations:

  • 200 - application/json (status) [expand]

    Example
    {"available":true,"enabled":true,"lastSync":1331038800000,"aheadRefs":[],"divergedRefs":[{"id":"refs/heads/master","displayId":"master","type":"BRANCH","state":"DIVERGED","tag":false}],"orphanedRefs":[]}

    The updated synchronization status for the repository, after enabling synchronization. 204 NO CONTENT is returned instead after disabling synchronization.

  • 204 - application/json [expand]

    Synchronization has successfully been disabled. 200 OK, with updated status information, is returned instead after enabling synchronization.

  • 400 - application/json (errors) [expand]

    Example
    {"errors":[{"context":"field_a","message":"A detailed validation error message for field_a.","exceptionName":null},{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The JSON payload for the request did not define the "enabled" property.

  • 401 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The currently authenticated user has insufficient permissions to manage synchronization in the specified repository.

  • 404 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The specified repository does not exist.

GET

/rest/sync/1.0/projects/{projectKey}/repos/{repositorySlug}?at

Retrieves the synchronization status for the specified repository. In addition to listing refs which cannot be synchronized, if any, the status also provides the timestamp for the most recent synchronization and indicates whether synchronization is available and enabled. If "?at" is specified in the URL, the synchronization status for the specified ref is returned, rather than the complete repository status.

The authenticated user must have REPO_READ permission for the repository, or it must be public if the request is anonymous. Additionally, after synchronization is enabled for a repository, meaning synchronization was available at that time, permission changes and other actions can cause it to become unavailable. Even when synchronization is enabled, if it is no longer available for the repository it will not be performed.

request query parameters
parametervaluedescription

at

string

retrieves the synchronization status for the specified ref within the repository, rather than for the entire repository

Example response representations:

  • 200 - application/json (status) [expand]

    Example
    {"available":true,"enabled":true,"lastSync":1331038800000,"aheadRefs":[],"divergedRefs":[{"id":"refs/heads/master","displayId":"master","type":"BRANCH","state":"DIVERGED","tag":false}],"orphanedRefs":[]}

    Synchronization status for the specified repository, or specific ref within that repository.

  • 401 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The currently authenticated user has insufficient permissions to view the repository, or the repository is not public if the request is anonymous.

  • 404 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The specified repository does not exist.

/rest/sync/1.0/projects/{projectKey}/repos/{repositorySlug}/synchronize

Methods

POST

Allows developers to apply a manual operation to bring a ref back in sync with upstream when it becomes out of sync due to conflicting changes. The following actions are supported:

  • MERGE: Merges in commits from the upstream ref. After applying this action, the synchronized ref will be AHEAD (as it still includes commits that do not exist upstream.
    • This action is only supported for DIVERGED refs
    • If a "commitMessage" is provided in the context, it will be used on the merge commit. Otherwise a default message is used.
  • DISCARD: Throws away local changes in favour of those made upstream. This is a destructive operation where commits in the local repository are lost.
    • No context entries are supported for this action
    • If the upstream ref has been deleted, the local ref is deleted as well
    • Otherwise, the local ref is updated to reference the same commit as upstream, even if the update is not fast-forward (similar to a forced push)

The authenticated user must have REPO_WRITE permission for the specified repository. Anonymous users cannot synchronize refs, even on public repositories. Additionally, synchronization must be enabled and available for the specified repository.

Example request representations:

  • application/json [expand]

    Example
    {"refId":"refs/heads/master","action":"MERGE","context":{"commitMessage":"Merging in latest from upstream."}}

Example response representations:

  • 200 - application/json (status) [expand]

    Example
    {"id":"refs/heads/master","displayId":"master","type":"BRANCH","state":"AHEAD","tag":false}

    The requested action was successfully performed, and has updated the ref's state, but the ref if is still not in sync with upstream. For example, after applying the MERGE action, the ref will still be AHEAD of upstream. If the action brings the ref in sync with upstream, 204 NO CONTENT is returned instead.

  • 204 - application/json [expand]

    The requested action was successfully performed and the ref is now in sync with upstream. If the action updates the ref but does not bring it in sync with upstream, 200 OK is returned instead.

  • 400 - application/json (errors) [expand]

    Example
    {"errors":[{"context":"field_a","message":"A detailed validation error message for field_a.","exceptionName":null},{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The requested synchronization action was not understood.

  • 401 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The currently authenticated user has insufficient permissions to update refs within the specified repository.

  • 404 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The specified repository does not exist.

  • 409 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    Synchronization is not available or enabled for the specified repository, or the ref is already in sync with upstream.

  • 501 - application/json (errors) [expand]

    Example
    {"errors":[{"context":null,"message":"A detailed error message.","exceptionName":null}]}

    The requested synchronization action was understood by the server, but the mechanism to apply it has not been implemented.