Gerrit Code Review comes with a REST like API available over HTTP. The API is suitable for automated tools to build upon, as well as supporting some ad-hoc scripting use cases.
See also: REST API Developers' Notes.
Endpoints
- /access/
-
Access Right related REST endpoints
- /accounts/
-
Account related REST endpoints
- /changes/
-
Change related REST endpoints
- /config/
-
Config related REST endpoints
- /groups/
-
Group related REST endpoints
- /plugins/
-
Plugin related REST endpoints
- /projects/
-
Project related REST endpoints
- /Documentation/
-
Documentation related REST endpoints
Protocol Details
Authentication
By default all REST endpoints assume anonymous access and filter results to correspond to what anonymous users can read (which may be nothing at all).
Users (and programs) can authenticate with HTTP passwords by prefixing
the endpoint URL with /a/
. For example to authenticate to
/projects/
, request the URL /a/projects/
. Gerrit will use HTTP basic
authentication with the HTTP password from the user’s account settings
page. This form of authentication bypasses the need for XSRF tokens.
An authorization cookie may be presented in the request URL inside the
access_token
query parameter. XSRF tokens are not required when a
valid access_token
is used in the URL.
CORS
Cross-site scripting may be supported if the administrator has configured site.allowOriginRegex.
Approved web applications running from an allowed origin can rely on
CORS preflight to authorize requests requiring cookie based
authentication, or mutations (POST, PUT, DELETE). Mutations require a
valid XSRF token in the X-Gerrit-Auth
request header.
Alternatively applications can use access_token
in the URL (see
above) to authorize requests. Mutations sent as POST with a request
content type of text/plain
can skip CORS preflight. Gerrit accepts
additional query parameters $m
to override the correct method (PUT,
POST, DELETE) and $ct
to specify the actual content type, such as
application/json; charset=UTF-8
. Example:
POST /changes/42/topic?$m=PUT&$ct=application/json%3B%20charset%3DUTF-8&access_token=secret HTTP/1.1 Content-Type: text/plain Content-Length: 23 {"topic": "new-topic"}
Preconditions
Clients can request PUT to create a new resource and not overwrite
an existing one by adding If-None-Match: *
to the request HTTP
headers. If the named resource already exists the server will respond
with HTTP 412 Precondition Failed.
Output Format
JSON responses are encoded using UTF-8 and use content type
application/json
.
By default most APIs return pretty-printed JSON, which uses extra whitespace to make the output more readable for humans.
Compact JSON can be requested by setting the pp=0
query parameter,
or by setting the Accept
HTTP request header to include application/json
:
GET /projects/ HTTP/1.0 Accept: application/json
Producing (and parsing) the non-pretty compact format is more efficient, so tools should request it whenever possible.
To prevent against Cross Site Script Inclusion (XSSI) attacks, the JSON response body starts with a magic prefix line that must be stripped before feeding the rest of the response body to a JSON parser:
)]}' [ ... valid JSON ... ]
Responses will be gzip compressed by the server if the HTTP
Accept-Encoding
request header is set to gzip
. This may
save on network transfer time for larger responses.
Input Format
Unknown JSON parameters will simply be ignored by Gerrit without causing an exception. This also applies to case-sensitive parameters, such as map keys.
Timestamp
Timestamps are given in UTC and have the format "'yyyy-mm-dd hh:mm:ss.fffffffff'" where "'ffffffffff'" represents nanoseconds.
Encoding
All IDs that appear in the URL of a REST call (e.g. project name, group name) must be URL encoded.
Response Codes
The Gerrit REST endpoints use HTTP status codes as described in the HTTP specification.
In most cases, the response body of an error response will be a plaintext, human-readable error message.
Here are examples that show how HTTP status codes are used in the context of the Gerrit REST API.
400 Bad Request
“400 Bad Request” is returned if the request is not understood by the server due to malformed syntax.
E.g. “400 Bad Request” is returned if JSON input is expected but the 'Content-Type' of the request is not 'application/json' or the request body doesn’t contain valid JSON.
“400 Bad Request” is also returned if required input fields are not set or if options are set which cannot be used together.
403 Forbidden
“403 Forbidden” is returned if the operation is not allowed because the calling user does not have sufficient permissions.
E.g. some REST endpoints require that the calling user has certain global capabilities assigned.
“403 Forbidden” is also returned if self
is used as account ID and the
REST call was done without authentication.
404 Not Found
“404 Not Found” is returned if the resource that is specified by the URL is not found or is not visible to the calling user. A resource cannot be found if the URL contains a non-existing ID or view.
405 Method Not Allowed
“405 Method Not Allowed” is returned if the resource exists but doesn’t support the operation.
E.g. some of the /groups/
endpoints are only supported for Gerrit
internal groups; if they are invoked for an external group the response
is “405 Method Not Allowed”.
409 Conflict
“409 Conflict” is returned if the request cannot be completed because the current state of the resource doesn’t allow the operation.
E.g. if you try to submit a change that is abandoned, this fails with “409 Conflict” because the state of the change doesn’t allow the submit operation.
“409 Conflict” is also returned if you try to create a resource but the name is already occupied by an existing resource.
412 Precondition Failed
“412 Precondition Failed” is returned if a precondition from the request header fields is not fulfilled, as described in the Preconditions section.
422 Unprocessable Entity
“422 Unprocessable Entity” is returned if the ID of a resource that is specified in the request body cannot be resolved.
Request Tracing
For each REST endpoint tracing can be enabled by setting the
trace=<trace-id>
request parameter. It is recommended to use the ID
of the issue that is being investigated as trace ID.
GET /changes/myProject~master~I8473b95934b5732ac55d26311a706c9c2bde9940/suggest_reviewers?trace=issue/123&q=J
It is also possible to omit the trace ID and get a unique trace ID generated.
GET /changes/myProject~master~I8473b95934b5732ac55d26311a706c9c2bde9940/suggest_reviewers?trace&q=J
Alternatively request tracing can also be enabled by setting the
X-Gerrit-Trace
header:
GET /changes/myProject~master~I8473b95934b5732ac55d26311a706c9c2bde9940/suggest_reviewers?q=J X-Gerrit-Trace: issue/123
Enabling tracing results in additional logs with debug information that
are written to the error_log
. All logs that correspond to the traced
request are associated with the trace ID. The trace ID is returned with
the REST response in the X-Gerrit-Trace
header.
HTTP/1.1 200 OK Content-Disposition: attachment Content-Type: application/json; charset=UTF-8 X-Gerrit-Trace: 1533885943749-8257c498 )]}' ... <json> ...
Given the trace ID an administrator can find the corresponding logs and investigate issues more easily.
Part of Gerrit Code Review