.
Exceptions which can be raised in python client are just after constants section.
Basic anonymous client in python would look like this:
-
CGImport(metadata, directory, token=None)
Import build from a content generator
metadata can be one of the following
- json encoded string representing the metadata
- a dictionary (parsed metadata)
- a filename containing the metadata
:param metadata: describes the content for this build.
:param str directory: directory on the hub where files are located
:param str token: (optional) a reservation token for this build.
You obtain a token from the CGInitBuild method.
If you specify a token, you must also specify a build_id
in the metadata.
:returns: buildinfo dict
-
CGInitBuild(cg, data)
Create (reserve) a build_id for given data.
If build or reservation already exists, init_build will raise GenericError
:param str cg: content generator name
:param dict data: build data same as for new_build, for given usecase
only name,version,release,epoch keys make sense. Some
other values will be ignored anyway (owner, state, ...)
:return: dict with build_id and token
-
CGRefundBuild(cg, build_id, token, state=3)
If build is reserved and not finished yet, there is an option
to release reservation and mark build either FAILED or CANCELED.
For this calling CG needs to know build_id and reservation token.
Refunded build behaves like any other failed/canceled build. So,
its nvr can be reclaimed again and get_next_release can return
this nvr.
:param str cg: content generator name
:param int build_id: build id
:param str token: token from CGInitBuild
:param int state: new state (koji.BUILD_STATES)
:return: None, on error raises exception
-
_listapi()
-
addArchiveType(name, description, extensions)
Add new archive type.
Use this to tell Koji about new builds' files' extensions before
importing the files.
:param str name: archive type name, eg. "yaml"
:param str description: eg. "YAML Ain't Markup Language"
:param str extensions: space-separated list of descriptions, eg. "yaml yml"
-
addBType(name)
Add a new btype with the given name
-
addChannel(channel_name, description=None)
Add a channel.
:param str channel_name: channel name
:param str description: description of channel
-
addExternalRPM(rpminfo, external_repo, strict=True)
Import an external RPM
This call is mainly for testing. Normal access will be through
a host call
-
addExternalRepoToTag(tag_info, repo_info, priority, merge_mode='koji', arches=None)
Add an external repo to a tag.
:param tag_info: Tag name or ID number
:param repo_info: External repository name or ID number
:param int priority: Priority of this repository for this tag
:param str merge_mode: This must be one of the values of the
koji.REPO_MERGE_MODES set. If unspecified,
the default is "koji".
:param str arches: space-separated list of arches handled by the repo.
-
addGroupMember(group, user, strict=True)
Add user to group
-
addHost(hostname, arches, krb_principal=None, force=False)
Add a builder host to the database.
:param str hostname: name for the host entry (fqdn recommended).
:param list arches: list of architectures this builder supports.
:param str krb_principal: (optional) a non-default kerberos principal
for the host.
:param bool force: override user type
:returns: new host id
If krb_principal is not given then that field will be generated
from the HostPrincipalFormat setting (if available).
-
addHostToChannel(hostname, channel_name, create=False, force=False)
Add the host to the specified channel
Channel must already exist unless create option is specified
-
addRPMSig(an_rpm, data)
Store a signature header for an rpm
data: the signature header encoded as base64
-
addUserKrbPrincipal(user, krb_principal)
Add a Kerberos principal for user
-
addVolume(name, strict=True)
Add a new storage volume in the database
-
applyVolumePolicy(build, strict=False)
Apply the volume policy to a given build
The volume policy is normally applied at import time, but it can
also be applied with this call.
Parameters:
build: the build to apply the policy to
strict: if True, raises on exception on policy issues
-
assignTask(task_id, host, force=False)
Assign a task to a host
Specify force=True to assign a non-free task
-
build(src, target, opts=None, priority=None, channel=None)
Create a build task
priority: the amount to increase (or decrease) the task priority, relative
to the default priority; higher values mean lower priority; only
admins have the right to specify a negative priority here
channel: the channel to allocate the task to
Returns the task id
-
buildImage(name, version, arch, target, ksfile, img_type, opts=None, priority=None)
Create an image using a kickstart file and group package list.
-
buildImageIndirection(opts=None, priority=None)
Create an image using two other images and an indirection template
-
buildImageOz(name, version, arches, target, inst_tree, opts=None, priority=None)
Create an image using a kickstart file and group package list.
-
buildReferences(build, limit=None, lazy=False)
Returns references to a build
This call is used to determine whether a build can be deleted
:param int build_id: numeric build id
:param int limit: If given, only return up to N results of each ref type
:param bool lazy: If true, stop when any reference is found
:returns: dict of reference results for each reference type
-
cancelBuild(buildID)
Cancel the build with the given buildID
If the build is associated with a task, cancel the task as well.
Return True if the build was successfully canceled, False if not.
-
cancelTask(task_id, recurse=True)
Cancel a task
-
cancelTaskChildren(task_id)
Cancel a task's children, but not the task itself
-
cancelTaskFull(task_id, strict=True)
Cancel a task and all tasks in its group
-
chainBuild(srcs, target, opts=None, priority=None, channel=None)
Create a chained build task for building sets of packages in order
srcs: list of pkg lists, ie [[src00, src01, src03],[src20],[src30,src31],...]
where each of the top-level lists gets built and a new repo is created
before the next list is built.
target: build target
priority: the amount to increase (or decrease) the task priority, relative
to the default priority; higher values mean lower priority; only
admins have the right to specify a negative priority here
channel: the channel to allocate the task to
Returns a list of all the dependent task ids
-
chainMaven(builds, target, opts=None, priority=None, channel='maven')
Create a Maven chain-build task
builds: a list of maps defining the parameters for the sequence of builds
target: the build target
priority: the amount to increase (or decrease) the task priority, relative
to the default priority; higher values mean lower priority; only
admins have the right to specify a negative priority here
channel: the channel to allocate the task to (defaults to the "maven" channel)
Returns the task ID
-
changeBuildVolume(build, volume, strict=True)
Move a build to a different storage volume
-
checkTagAccess(tag_id, user_id=None)
Determine if user has access to tag package with tag.
Returns a tuple (access, override, reason)
access: a boolean indicating whether access is allowed
override: a boolean indicating whether access may be forced
reason: the reason access is blocked
-
checkTagPackage(tag, pkg)
Check that pkg is in the list for tag. Returns true/false
-
checkUpload(path, name, verify=None, tail=None, volume=None)
Return basic information about an uploaded file
-
count(methodName, *args, **kw)
Execute the XML-RPC method with the given name and count the results.
A method return value of None will return O, a return value of type "list", "tuple", or
"dict" will return len(value), and a return value of any other type will return 1. An
invalid methodName will raise GenericError.
-
countAndFilterResults(methodName, *args, **kw)
Filter results by a given name and count total results account.
Execute the XML-RPC method with the given name and filter the results
based on the options specified in the keywork option "filterOpts".
The method must return a list of maps. Any other return type will
result in a GenericError.
Args:
offset: the number of elements to trim off the front of the list
limit: the maximum number of results to return
order: the map key to use to sort the list; the list will be sorted
before offset or limit are applied
noneGreatest: when sorting, consider 'None' to be greater than all
other values; python considers None less than all other values,
but Postgres sorts NULL higher than all other values; default
to True for consistency with database sorts
Returns:
Tuple of total results amount and the filtered results.
-
createBuildTarget(name, build_tag, dest_tag)
Create a new build target
-
createEmptyBuild(name, version, release, epoch, owner=None)
Creates empty build entry
:param str name: build name
:param str version: build version
:param str release: release version
:param str epoch: epoch version
:param userInfo: a str (Kerberos principal or name) or an int (user id)
or a dict:
- id: User's ID
- name: User's name
- krb_principal: Kerberos principal
:return: int build ID
-
createExternalRepo(name, url)
Create a new external repo with the given name and url.
Return a map containing the id, name, and url
of the new repo.
-
createImageBuild(build_info)
Associate image metadata with an existing build. When build isn`t existing, creates a
build. The build must not already have associated image metadata.
:param build_info: a str (build name) if build is existing
or a dict:
- name: build name
- version: build version
- release: build release
- epoch: build epoch
:raises: GenericError if type for build_info is not dict, when build isn`t existing.
:raises: GenericError if build info doesn't have mandatory keys.
-
createMavenBuild(build_info, maven_info)
Associate Maven metadata with an existing build. When build isn`t existing, creates a
build. The build must not already have associated Maven metadata. maven_info must be
a dict containing group_id, artifact_id, and version entries.
:param build_info: a str (build name) if build is existing
or a dict:
- name: build name
- version: build version
- release: build release
- epoch: build epoch
:param dict maven_info:
- group_id: Group's ID
- artifact_id: Artifact's ID
- version: version
:raises: GenericError if type for build_info is not dict, when build isn`t existing.
:raises: GenericError if build info doesn't have mandatory keys.
-
createNotification(user_id, package_id, tag_id, success_only)
Create a new notification. If the user_id does not match the currently logged-in user
and the currently logged-in user is not an admin, raise a GenericError.
-
createNotificationBlock(user_id, package_id=None, tag_id=None)
Create notification block. If the user_id does not match the
currently logged-in user and the currently logged-in user is not an
admin, raise a GenericError.
-
createTag(name, parent=None, arches=None, perm=None, locked=False, maven_support=False, maven_include_all=False, extra=None)
Create a new tag
-
createUser(username, status=None, krb_principal=None)
Add a user to the database
:param str username: The username for this Koji user.
:param int status: This must be one of the values of the
koji.USER_STATUS enum. If unspecified,
the default is koji.USER_STATUS['NORMAL'].
:param str krb_principal: a custom Kerberos principal, or None.
:raises: GenericError if the user or Kerberos principal already
exists.
-
createWinBuild(build_info, win_info)
Associate Windows metadata with an existing build. When build isn`t existing, creates
a build. The build must not already have associated Windows metadata. win_info must be
a dict containing a platform entry.
:param build_info: a str (build name) if build is existing
or a dict:
- name: build name
- version: build version
- release: build release
- epoch: build epoch
:param dict win_info:
- platform: build platform
:raises: GenericError if type for build_info is not dict, when build isn`t existing.
:raises: GenericError if build info doesn't have mandatory keys.
-
deleteBuild(build, strict=True, min_ref_age=604800)
delete a build, if possible
Attempts to delete a build. A build can only be deleted if it is
unreferenced.
If strict is true (default), an exception is raised if the build cannot
be deleted.
Note that a deleted build is not completely gone. It is marked deleted and some
data remains in the database. Mainly, the rpms are removed.
Note in particular that deleting a build DOES NOT free any NVRs (or NVRAs) for
reuse.
Returns True if successful, False otherwise
-
deleteBuildTarget(buildTargetInfo)
Delete the build target with the given name. If no build target
exists, raise a GenericError.
-
deleteExternalRepo(info)
Remove an external repository for any tags and delete it.
:param info: external repository name or ID number
:raises: GenericError if the repository does not exist.
-
deleteNotification(id)
Delete the notification with the given ID. If the currently logged-in
user is not the owner of the notification or an admin, raise a GenericError.
-
deleteNotificationBlock(id)
Delete the notification block with the given ID. If the currently logged-in
user is not the owner of the notification or an admin, raise a GenericError.
-
deleteRPMSig(rpminfo, sigkey=None, all_sigs=False)
Delete rpm signature
Only use this method in extreme situations, because it goes against
Koji's design of immutable, auditable data.
This call requires ``admin`` permission (``sign`` is not sufficient).
:param dict/str/id rpm: map containing 'name', 'version', 'release', and 'arch'
string N-V-R.A
int ID
:param str sigkey: Signature key.
:param bool all_sigs: Delete all signed copies for specified RPM.
-
deleteTag(tagInfo)
Delete the specified tag.
-
disableChannel(channelname, comment=None)
Mark a channel as disabled
-
disableHost(hostname)
Mark a host as disabled
-
disableUser(username)
Disable logins by the specified user
-
distRepo(tag, keys, **task_opts)
Create a dist-repo task. returns task id
-
downloadTaskOutput(taskID, fileName, offset=0, size=-1, volume=None)
Download the file with the given name, generated by the task with the
given ID.
-
dropGroupMember(group, user)
Drop user from group
-
echo(*args)
-
editBuildTarget(buildTargetInfo, name, build_tag, dest_tag)
Set the build_tag and dest_tag of an existing build_target to new values
-
editChannel(channelInfo, **kw)
Edit information for an existing channel.
:param str/int channelInfo: channel name or ID
:param str name: new channel name
:param str description: description of channel
:param str comment: comment about channel
-
editExternalRepo(info, name=None, url=None)
Edit an existing external repo
-
editHost(hostInfo, **kw)
Edit information for an existing host.
hostInfo specifies the host to edit, either as an integer (id)
or a string (name).
fields to be changed are specified as keyword parameters:
- arches (a space-separated string)
- capacity (float or int)
- description (string)
- comment (string)
Returns True if changes are made to the database, False otherwise.
-
editPermission(permission, description)
Edit a permission description
-
editTag(tagInfo, name, arches, locked, permissionID, extra=None)
Edit information for an existing tag.
-
editTag2(tagInfo, **kwargs)
Edit information for an existing tag.
The tagInfo argument is the only required argument. After the tagInfo
argument, specify any tag changes with additional keyword arguments.
:param tagInfo: koji tag ID or name to edit (required).
:type tagInfo: int or str
:param str name: rename the tag.
:param str arches: a space-separated list of arches for this tag.
:param bool locked: whether this tag is locked or not.
:param perm: the permission ID or name for this tag.
:type perm: int, str, or None
:param bool maven_support: whether Maven repos should be generated for the
tag.
:param bool maven_include_all: include every build in this tag (including
multiple versions of the same package) in
the Maven repo.
:param dict extra: add or update extra tag parameters.
:param list remove_extra: remove extra tag parameters.
:param list block_extra: block inherited extra tag parameters.
-
editTagExternalRepo(tag_info, repo_info, priority=None, merge_mode=None, arches=None)
Edit a tag<->external repo association
This allows you to update the priority and merge_mode without removing/adding the repo.
Note that None value of priority and merge_mode means no change on it
-
editUser(userInfo, name=None, krb_principal_mappings=None)
Edit information for an existing user.
Use this method to rename a user, or to add/remove/modify Kerberos
principal(s) for this account.
Example krb_principal_mappings values:
To add a new Kerberos principal to a user account:
[{'old': None, 'new': 'myuser@NEW.EXAMPLE.COM'}]
To remove an old Kerberos principal from a user account:
[{'old': 'myuser@OLD.EXAMPLE.COM', 'new': None}]
To modify a user's old Kerberos principal to a new one:
[{'old': 'myuser@OLD.EXAMPLE.NET', 'new': 'myuser@NEW.EXAMPLE.NET'}]
:param userInfo: username (str) or ID (int)
:param str name: new name for this user account
:param list krb_principal_mappings: List of changes to make for this
user's Kerberos principal. Each change
is a dict of "old" and "new"
Kerberos principals.
:raises: GenericError if the user does not exist, or if there were
problems in the krb_principal_mappings.
-
enableChannel(channelname, comment=None)
Mark a channel as enabled
-
enableHost(hostname)
Mark a host as enabled
-
enableUser(username)
Enable logins by the specified user
-
error()
debugging. raise an error
-
evalPolicy(name, data)
Evaluate named policy with given data and return the result
:param str name: the policy name
:param dict data: the policy data
:returns the action as a string
:raises koji.GenericError if the policy is empty or not found
-
exclusiveSession(*args, **opts)
Make this session exclusive
-
fault()
debugging. raise an error
-
filterResults(methodName, *args, **kw)
Execute the XML-RPC method with the given name and filter the results
based on the options specified in the keywork option "filterOpts". The method
must return a list of maps. Any other return type will result in a GenericError.
Currently supported options are:
- offset: the number of elements to trim off the front of the list
- limit: the maximum number of results to return
- order: the map key to use to sort the list; the list will be sorted before
offset or limit are applied
- noneGreatest: when sorting, consider 'None' to be greater than all other values;
python considers None less than all other values, but Postgres sorts
NULL higher than all other values; default to True for consistency
with database sorts
-
findBuildID(X, strict=False)
gets build ID for various inputs
:param int|str|dict X: build ID | NVR | dict with name, version and release values
:returns int: build ID
-
freeTask(task_id)
Free a task
-
getAPIVersion()
-
getActiveRepos()
Get data on all active repos
This is a list of all the repos that the repo daemon needs to worry about.
-
getAllArches()
Return a list of all (canonical) arches available from hosts
-
getAllPerms()
Get a list of all permissions in the system. Returns a list of maps. Each
map contains the following keys:
- id
- name
- description
-
getArchive(archive_id, strict=False)
Get information about the archive with the given ID. Returns a map
containing the following keys:
id: unique id of the archive file (integer)
type_id: id of the archive type (Java jar, Solaris pkg, Windows exe, etc.) (integer)
build_id: id of the build that generated this archive (integer)
buildroot_id: id of the buildroot where this archive was built (integer)
filename: name of the archive (string)
size: size of the archive (integer)
checksum: checksum of the archive (string)
checksum_type: type of the checksum (integer)
If the archive is part of a Maven build, the following keys will be included:
group_id
artifact_id
version
If the archive is part of a Windows builds, the following keys will be included:
relpath
platforms
flags
If the archive is part of an image build, and it is the image file that
contains the root partitioning ('/'), there will be a additional fields:
rootid
arch
-
getArchiveFile(archive_id, filename, strict=False)
Get information about a file with the given filename
contained in the archive with the given ID.
Returns a map with with following keys:
archive_id: id of the archive the file is contained in (integer)
name: name of the file (string)
size: uncompressed size of the file (integer)
If strict is True, raise GenericError if:
- this file is not found in the archive
- build btype of this archive belong to is not maven, win or image
- archive_type is not that we are able to expand
Regardless of strict, an error will be raised if the archive_id is invalid
-
getArchiveType(filename=None, type_name=None, type_id=None, strict=False)
Get the archive type for the given filename, type_name, or type_id.
-
getArchiveTypes()
Return a list of all supported archive types.
-
getAverageBuildDuration(package, age=None)
Get the average duration of a build of the given package.
:param int|str package: Package name or id
:param int age: length of history in months
:return float|None: average number of seconds - If package wasn't built
during past age months (or never), None is returned
-
getBuild(buildInfo, strict=False)
Return information about a build.
buildInfo may be either a int ID, a string NVR, or a map containing
'name', 'version' and 'release'.
A map will be returned containing the following keys*:
id: build ID
package_id: ID of the package built
package_name: name of the package built
name: same as package_name
version
release
epoch
nvr
state
task_id: ID of the task that kicked off the build
owner_id: ID of the user who kicked off the build
owner_name: name of the user who kicked off the build
volume_id: ID of the storage volume
volume_name: name of the storage volume
creation_event_id: id of the create_event
creation_time: time the build was created (text)
creation_ts: time the build was created (epoch)
start_time: time the build was started (may be null)
start_ts: time the build was started (epoch, may be null)
completion_time: time the build was completed (may be null)
completion_ts: time the build was completed (epoch, may be null)
source: the SCM URL of the sources used in the build -
dereferenced git hash is stored here
extra: dictionary with extra data about the build
- source:
- original_url: while build.source contains concrete
SCM hash, this field can contain SCM url which was
used when launching build (e.g. git_url#master)
cg_id: ID of CG which reserved or imported this build
cg_name: name of CG which reserved or imported this build
If there is no build matching the buildInfo given, and strict is specified,
raise an error. Otherwise return None.
[*] Not every build will have data for all keys. E.g. not all builds will
associated task ids, and not all import methods provide source info.
-
getBuildConfig(tag, event=None)
Return build configuration associated with a tag
-
getBuildLogs(build)
Return a list of log files for the given build
This method will only return logs for builds that are complete.
If a build is in progress, failed, or canceled, you must look at the
build's task logs instead (see listTaskOutput).
:param build: A build ID (int), a NVR (string), or a dict containing
"name", "version" and "release".
:returns: a possibly-empty list of log file entries. Each entry is a dict
with three keys:
"name" (log file name)
"dir" (immediate parent subdirectory)
"path" (the full path under koji's topdir)
-
getBuildNotification(id, strict=False)
Get the build notification with the given ID.
If there is no notification with the given ID, when strict is True,
raise GenericError, else return None.
-
getBuildNotificationBlock(id, strict=False)
Get the build notification with the given ID.
If there is no notification with the given ID, when strict is True,
raise GenericError, else return None.
-
getBuildNotificationBlocks(userID=None)
Get build notifications for the user with the given ID, name or
Kerberos principal. If no user is specified, get the notifications for
the currently logged-in user. If there is no currently logged-in user,
raise a GenericError.
-
getBuildNotifications(userID=None)
Get build notifications for the user with the given ID, name or
Kerberos principal. If no user is specified, get the notifications for
the currently logged-in user. If there is no currently logged-in user,
raise a GenericError.
-
getBuildTarget(info, event=None, strict=False)
Return the build target with the given name or ID.
If there is no matching build target, return None.
-
getBuildTargets(info=None, event=None, buildTagID=None, destTagID=None, queryOpts=None)
Return data on all the build targets
provide event to query at a different time
-
getBuildType(buildInfo, strict=False)
Return type info about the build
buildInfo should be a valid build specification
Returns a dictionary whose keys are type names and whose values are
the type info corresponding to that type
-
getBuildroot(buildrootID, strict=False)
Return information about a buildroot. buildrootID must be an int ID.
-
getBuildrootListing(id)
Return a list of packages in the buildroot
-
getChangelogEntries(buildID=None, taskID=None, filepath=None, author=None, before=None, after=None, queryOpts=None, strict=False)
Get changelog entries for the build with the given ID,
or for the rpm generated by the given task at the given path
- author: only return changelogs with a matching author
- before: only return changelogs from before the given date (in UTC)
(a datetime object, a string in the 'YYYY-MM-DD HH24:MI:SS format, or integer
seconds since the epoch)
- after: only return changelogs from after the given date (in UTC)
(a datetime object, a string in the 'YYYY-MM-DD HH24:MI:SS format, or integer
seconds since the epoch)
- queryOpts: query options used by the QueryProcessor
- strict: if srpm doesn't exist raise an error, otherwise return empty list
If "order" is not specified in queryOpts, results will be returned in reverse chronological
order.
Results will be returned as a list of maps with 'date', 'author', and 'text' keys.
If there are no results, an empty list will be returned.
-
getChannel(channelInfo, strict=False)
Look up the ID number and name for a channel.
:param channelInfo: channel ID or name
:type channelInfo: int or str
:param bool strict: If True, raise an error if we found no matching
channel. If False, simply return None if we found no
matching channel. If unspecified, the default value is
False.
:returns: dict of the channel ID and name, or None.
For example, {'id': 20, 'name': 'container'}
-
getEvent(id)
Get information about the event with the given id.
A map will be returned with the following keys:
- id (integer): id of the event
- ts (float): timestamp the event was created, in
seconds since the epoch
If no event with the given id exists, an error will be raised.
-
getExternalRepo(info, strict=False, event=None)
Get information about a single external repository.
:param info: a string (name) or an integer (id).
:param bool strict: If True, raise an error if we found no matching
repository. If False, simply return None if we found
no matching repository. If unspecified, the default
value is False.
:param int event: The event ID at which to search. If unspecified, the
default behavior is to search for the "active" repo
settings.
:returns: a map containing the id, name, and url of the repository.
-
getExternalRepoList(tag_info, event=None)
Get an ordered list of all external repos associated with the tags in the
hierarchy rooted at the specified tag. External repos will be returned
depth-first, and ordered by priority for each tag. Duplicates will be
removed. Returns a list of maps containing the following fields:
tag_id
tag_name
external_repo_id
external_repo_name
url
merge_mode
priority
-
getFullInheritance(tag, event=None, reverse=False, **kwargs)
:param int|str tag: tag ID | name
:param int event: event ID
:param bool reverse: return reversed tree (descendants instead of
parents)
:param dict stops: SHOULD NOT BE USED, BACKWARDS COMPATIBLE ONLY
:param dict jumps: SHOULD NOT BE USED, BACKWARDS COMPATIBLE ONLY
:returns: list of node dicts
-
getGroupMembers(group)
Get the members of a group
-
getHost(hostInfo, strict=False, event=None)
Get information about the given host. hostInfo may be
either a string (hostname) or int (host id). A map will be returned
containing the following data:
- id
- user_id
- name
- arches
- task_load
- capacity
- description
- comment
- ready
- enabled
-
getImageArchive(archive_id, strict=False)
Retrieve image-specific information about an archive.
Returns a map containing the following keys:
archive_id: id of the build (integer)
arch: the architecture of the image
rootid: True if this image has the root '/' partition
-
getImageBuild(buildInfo, strict=False)
Retrieve image-specific information about a build.
buildInfo can be either a string (n-v-r) or an integer
(build ID). This function really only exists to verify a build
is an image build; there is no additional data.
Returns a map containing the following keys:
build_id: id of the build
-
getInheritanceData(tag, event=None)
Return inheritance data for tag
-
getKojiVersion()
-
getLastEvent(before=None)
Get the id and timestamp of the last event recorded in the system.
Events are usually created as the result of a configuration change
in the database.
If "before" (int or float) is specified, return the last event
that occurred before that time (in seconds since the epoch).
If there is no event before the given time, an error will be raised.
Note that due to differences in precision between the database and python,
this method can return an event with a timestamp the same or slightly higher
(by a few microseconds) than the value of "before" provided. Code using this
method should check that the timestamp returned is in fact lower than the parameter.
When trying to find information about a specific event, the getEvent() method
should be used.
-
getLastHostUpdate(hostID, ts=False)
Return the latest update timestamp for the host
The timestamp represents the last time the host with the given
ID contacted the hub. Returns None if the host has never contacted
the hub.
-
getLatestBuilds(tag, event=None, package=None, type=None)
List latest builds for tag (inheritance enabled)
-
getLatestMavenArchives(tag, event=None, inherit=True)
Return a list of the latest Maven archives in the tag, as of the given event
(or now if event is None). If inherit is True, follow the tag hierarchy
and return a list of the latest archives for all tags in the tree.
-
getLatestRPMS(tag, package=None, arch=None, event=None, rpmsigs=False, type=None)
List latest RPMS for tag (inheritance enabled)
-
getLoggedInUser()
Return information about the currently logged-in user. Returns data
in the same format as getUser(), plus the authtype. If there is no
currently logged-in user, return None.
-
getMavenArchive(archive_id, strict=False)
Retrieve Maven-specific information about an archive.
Returns a map containing the following keys:
archive_id: id of the build (integer)
group_id: Maven groupId (string)
artifact_id: Maven artifact_Id (string)
version: Maven version (string)
-
getMavenBuild(buildInfo, strict=False)
Retrieve Maven-specific information about a build.
buildInfo can be either a string (n-v-r) or an integer
(build ID).
Returns a map containing the following keys:
build_id: id of the build (integer)
group_id: Maven groupId (string)
artifact_id: Maven artifact_Id (string)
version: Maven version (string)
-
getNextRelease(build_info, incr=1)
Find the next release for a package's version.
This method searches the latest building, successful, or deleted build and
returns the "next" release value for that version.
Examples:
None becomes "1"
"123" becomes "124"
"123.el8" becomes "124.el8"
"123.snapshot.456" becomes "123.snapshot.457"
All other formats will raise koji.BuildError.
:param dict build_info: a dict with two keys: a package "name" and
"version" of the builds to search. For example,
{"name": "bash", "version": "4.4.19"}
:param int incr: value which should be added to latest found release
(it is used for solving race-condition conflicts)
:returns: a release string for this package, for example "15.el8".
:raises: BuildError if the latest build uses a release value that Koji
does not know how to increment.
-
getPackage(info, strict=False, create=False)
Get the id,name for package
-
getPackageConfig(tag, pkg, event=None)
Get config for package in tag
-
getPackageID(name, strict=False)
Get package ID by name.
If package doesn't exist, return None, unless strict is True in which
case an exception is raised.
-
getPerms()
Get a list of the permissions granted to the currently logged-in user.
-
getRPM(rpminfo, strict=False, multi=False)
Get information about the specified RPM
rpminfo may be any one of the following:
- a int ID
- a string N-V-R.A
- a string N-V-R.A@location
- a map containing 'name', 'version', 'release', and 'arch'
(and optionally 'location')
If specified, location should match the name of an external repo
A map will be returned, with the following keys:
- id
- name
- version
- release
- arch
- epoch
- payloadhash
- size
- buildtime
- build_id
- buildroot_id
- external_repo_id
- external_repo_name
- metadata_only
- extra
If there is no RPM with the given ID, None is returned, unless strict
is True in which case an exception is raised
If more than one RPM matches, and multi is True, then a list of results is
returned. If multi is False, a single match is returned (an internal one if
possible).
-
getRPMDeps(rpmID, depType=None, queryOpts=None, strict=False)
Return dependency information about the RPM with the given ID.
If depType is specified, restrict results to dependencies of the given type.
Otherwise, return all dependency information. A list of maps will be returned,
each with the following keys:
- name
- version
- flags
- type
If there is no *internal* RPM with the given ID, or no RPM file found,
an empty list will be returned, unless strict is True in which case a
GenericError is raised.
If the RPM has no dependency information, an empty list will be returned.
-
getRPMFile(rpmID, filename, strict=False)
Get info about the file in the given RPM with the given filename.
A map will be returned with the following keys:
- rpm_id
- name
- digest
- md5 (alias for digest)
- digest_algo
- size
- flags
- user
- group
- mtime
- mode
If there is no *internal* RPM with the given ID, or no RPM file found,
an empty map will be returned, unless strict is True in which case a
GenericError is raised.
If no such file exists, an empty map will be returned, unless strict is
True in which case a GenericError is raised.
-
getRPMHeaders(rpmID=None, taskID=None, filepath=None, headers=None, strict=False)
Get the requested headers from the rpm, specified either by rpmID or taskID + filepath
If the specified ID is not valid or the rpm does not exist on the file system an empty
map will be returned, unless strict=True is given.
Header names are case-insensitive. If a header is requested that does not exist an
exception will be raised (regardless of strict option).
:param int|str rpmID: query the specified rpm
:param int taskID: query a file from the specified task (filepath must also be passed)
:param str filepath: the rpm path relative to the task directory
:param list headers: a list of rpm header names (as strings)
:param bool strict: raise an exception for invalid or missing rpms/paths
:returns dict: a map of header names to values
-
getRepo(tag, state=None, event=None, dist=False)
Get individual repository data based on tag and additional filters.
If more repos fits, most recent is returned.
:param int|str tag: tag ID or name
:param int state: value from koji.REPO_STATES
:param int event: event ID
:param bool dist: True = dist repo, False = regular repo
:returns: dict with repo data (id, state, create_event, time, dist)
-
getSessionInfo()
Return session info for current user
-
getTag(tagInfo, strict=False, event=None, blocked=False)
Get tag information based on the tagInfo. tagInfo may be either
a string (the tag name) or an int (the tag ID).
Returns a map containing the following keys:
- id : unique id for the tag
- name : name of the tag
- perm_id : permission id (may be null)
- perm : permission name (may be null)
- arches : tag arches (string, may be null)
- locked : lock setting (boolean)
- maven_support : maven support flag (boolean)
- maven_include_all : maven include all flag (boolean)
- extra : extra tag parameters (dictionary)
- query_event : return "event" parameter for current call
if something was passed in
If there is no tag matching the given tagInfo, and strict is False,
return None. If strict is True, raise a GenericError.
Note that in order for a tag to 'exist', it must have an active entry
in tag_config. A tag whose name appears in the tag table but has no
active tag_config entry is considered deleted.
event option can be either event_id or "auto" which will pick last
recorded create_event (option for getting deleted tags)
-
getTagExternalRepos(tag_info=None, repo_info=None, event=None)
Get a list of tag<->external repo associations.
The list of associations is ordered by the priority field.
Each map containing the following fields:
tag_id
tag_name
external_repo_id
external_repo_name
url
merge_mode
priority
:param tag_info: Tag name or ID number. This field is optional. If you
specify a value here, Koji will only return
repo association information for this single tag.
:param repo_info: External repository name or ID number. This field is
optional. If you specify a value here, Koji will only
return tag association information for this single
repository.
:param int event: The event ID at which to search. If unspecified, the
default behavior is to search for the "active" tag and
repo settings.
-
getTagGroups(tag, event=None, inherit=True, incl_pkgs=True, incl_reqs=True, incl_blocked=False)
Return group data for the tag with blocked entries removed
Also scrubs data into an xmlrpc-safe format (no integer keys)
Blocked packages/groups can alternatively also be listed if incl_blocked is set to True
-
getTagID(info, strict=False, create=False)
Get the id for tag
-
getTaskChildren(task_id, request=False, strict=False)
Return a list of the children
of the Task with the given ID.
-
getTaskDescendents(task_id, request=False)
Get all descendents of the task with the given ID.
Return a map of task_id -> list of child tasks. If the given
task has no descendents, the map will contain a single elements
mapping the given task ID to an empty list. Map keys will be strings
representing integers, due to limitations in xmlrpclib. If "request"
is true, the parameters sent with the xmlrpc request will be decoded and
included in the map.
-
getTaskInfo(task_id, request=False, strict=False)
Get information about a task
:param int task_id: Task id
:param bool request: if True, return also task's request
:param bool strict: raise exception, if task is not found
:returns dict: task info
-
getTaskRequest(taskId)
Return original task request as a list. Content depends on task type
:param int taskId: id of task queried
:returns list: request
-
getTaskResult(taskId, raise_fault=True)
Returns task results depending on task type. For buildArch it is a dict with build info,
for newRepo list with two items, etc.
:param int taskId: id of task queried
:param bool raise_fault: if task's result is a fault, raise it also here, otherwise
just get dict with error code/message
:returns any: dict/list/etc. with task result
-
getUser(userInfo=None, strict=False, krb_princs=True)
Return information about a user.
:param userInfo: a str (Kerberos principal or name) or an int (user id)
or a dict:
- id: User's ID
- name: User's name
- krb_principal: Kerberos principal
:param bool strict: whether raising Error when no user found
:param bool krb_princs: whether show krb_principals in result
:return: a dict as user's information:
id: user id
name: user name
status: user status (int), may be null
usertype: user type (int), 0 person, 1 for host, may be null
krb_principals: the user's Kerberos principals (list)
-
getUserPerms(userID=None)
Get a list of the permissions granted to the user with the given ID/name.
Options:
- userID: User ID or username. If no userID provided, current login user's
permissions will be listed.
-
getVolume(volume, strict=False)
Lookup the given volume
Returns a dictionary containing the name and id of the matching
volume, or None if no match.
If strict is true, raises an error if no match.
-
getWinArchive(archive_id, strict=False)
Retrieve Windows-specific information about an archive.
Returns a map containing the following keys:
archive_id: id of the build (integer)
relpath: the relative path where the file is located (string)
platforms: space-separated list of platforms the file is suitable for use on (string)
flags: space-separated list of flags used when building the file (fre, chk) (string)
-
getWinBuild(buildInfo, strict=False)
Retrieve Windows-specific information about a build.
buildInfo can be either a string (n-v-r) or an integer
(build ID).
Returns a map containing the following keys:
build_id: id of the build (integer)
platform: the platform the build was performed on (string)
-
grantCGAccess(user, cg, create=False)
Grant user access to act as the given content generator
:param user: koji userid or username
:type user: int or str
:param cg: content generator id or name
:type cg: int or str
:param bool create: If True, and the requested cg name entry does not
already exist, then Koji will create the content
generator entry. In such a case, the cg parameter
must be a string. The default is False.
-
grantPermission(userinfo, permission, create=False, description=None)
Grant a permission to a user
-
groupListAdd(taginfo, grpinfo, block=False, force=False, **opts)
Add to (or update) group list for tag
-
groupListBlock(taginfo, grpinfo)
Block the group in tag
-
groupListRemove(taginfo, grpinfo, force=False)
Remove group from the list for tag
Permission required: admin
:param taginfo: tag id or name which group is removed from
:type taginfo: int or str
:param grpinfo: group id or name which is removed
:type grpinfo: int or str
:param bool force: If False(default), GenericException will be raised when
no group found in the list for tag. If True, revoking
will be force to execute, no matter if the relationship
exists.
Really this shouldn't be used except in special cases
Most of the time you really want to use the block or unblock functions
-
groupListUnblock(taginfo, grpinfo)
Unblock the group in tag
If the group is blocked in this tag, then simply remove the block.
Otherwise, raise an error
-
groupPackageListAdd(taginfo, grpinfo, pkg_name, block=False, force=False, **opts)
Add package to group for tag
-
groupPackageListBlock(taginfo, grpinfo, pkg_name)
Block the package in group-tag
-
groupPackageListRemove(taginfo, grpinfo, pkg_name, force=None)
Remove package from the list for group-tag
Really this shouldn't be used except in special cases
Most of the time you really want to use the block or unblock functions
-
groupPackageListUnblock(taginfo, grpinfo, pkg_name)
Unblock the package in group-tag
If blocked (directly) in this tag, then simply remove the block.
Otherwise, raise an error
-
groupReqListAdd(taginfo, grpinfo, reqinfo, block=False, force=False, **opts)
Add group requirement to group for tag
-
groupReqListBlock(taginfo, grpinfo, reqinfo)
Block the group requirement in group-tag
-
groupReqListRemove(taginfo, grpinfo, reqinfo, force=None)
Remove group requirement from the list for group-tag
Really this shouldn't be used except in special cases
Most of the time you really want to use the block or unblock functions
-
groupReqListUnblock(taginfo, grpinfo, reqinfo)
Unblock the group requirement in group-tag
If blocked (directly) in this tag, then simply remove the block.
Otherwise, raise an error
-
hasPerm(perm, strict=False)
Check if the logged-in user has the given permission. Return False if
they do not have the permission, or if they are not logged-in.
-
hello(*args)
Simple testing call returning a string
-
host.assertPolicy(name, data, default='deny')
-
host.checkPolicy(name, data, default='deny', strict=False)
-
host.closeTask(task_id, response)
-
host.completeBuild(task_id, build_id, srpm, rpms, brmap=None, logs=None)
Import final build contents into the database
-
host.completeImageBuild(task_id, build_id, results)
Set an image build to the COMPLETE state
-
host.completeMavenBuild(task_id, build_id, maven_results, rpm_results)
Complete the Maven build.
-
host.completeWinBuild(task_id, build_id, results, rpm_results)
Complete a Windows build
-
host.createMavenBuild(build_info, maven_info)
Associate Maven metadata with an existing build. Used
by the rpm2maven plugin.
-
host.distRepoMove(repo_id, uploadpath, arch)
Move one arch of a dist repo into its final location
Unlike normal repos, dist repos have all their content linked (or
copied) into place.
repo_id - the repo to move
uploadpath - where the uploaded files are
arch - the arch of the repo
uploadpath should contain a repo_manifest file
The uploaded files should include:
- kojipkgs: json file with information about the component rpms
- repo metadata files
-
host.evalPolicy(name, data)
Evaluate named policy with given data and return the result
-
host.failBuild(task_id, build_id)
Mark the build as failed. If the current state is not
'BUILDING', or the current completion_time is not null, a
GenericError will be raised.
-
host.failTask(task_id, response)
-
host.freeTasks(tasks)
-
host.getHost()
Return information about this host
-
host.getHostTasks()
-
host.getID()
-
host.getLoadData()
-
host.importArchive(filepath, buildinfo, type, typeInfo)
Import an archive file and associate it with a build. The archive can
be any non-rpm filetype supported by Koji. Used by the rpm2maven plugin.
-
host.importImage(task_id, build_info, results)
Import a built image, populating the database with metadata and
moving the image to its final location.
-
host.importWrapperRPMs(task_id, build_id, rpm_results)
Import the wrapper rpms and associate them with the given build. The build
must not have any existing rpms associated with it.
-
host.initBuild(data)
Create a stub (rpm) build entry.
This is done at the very beginning of the build to inform the
system the build is underway.
This function is only called for rpm builds, other build types
have their own init function
-
host.initImageBuild(task_id, build_info)
create a new in-progress image build
-
host.initMavenBuild(task_id, build_info, maven_info)
Create a new in-progress Maven build
Synthesize the release number by taking the (integer) release of the
last successful build and incrementing it.
-
host.initWinBuild(task_id, build_info, win_info)
Create a new in-progress Windows build.
-
host.isEnabled()
-
host.moveBuildToScratch(task_id, srpm, rpms, logs=None)
Move a completed scratch build into place (not imported)
-
host.moveImageBuildToScratch(task_id, results)
move a completed image scratch build into place
-
host.moveMavenBuildToScratch(task_id, results, rpm_results)
Move a completed Maven scratch build into place (not imported)
-
host.moveWinBuildToScratch(task_id, results, rpm_results)
Move a completed Windows scratch build into place (not imported)
-
host.newBuildRoot(repo, arch, task_id=None)
-
host.openTask(task_id)
-
host.repoDone(repo_id, data, expire=False)
Finalize a repo
repo_id: the id of the repo
data: a dictionary of repo files in the form:
{ arch: [uploadpath, [file1, file2, ...]], ...}
expire: if set to true, mark the repo expired immediately [*]
Actions:
* Move uploaded repo files into place
* Mark repo ready
* Expire earlier repos
* Move/create 'latest' symlink
For dist repos, the move step is skipped (that is handled in
distRepoMove).
* This is used when a repo from an older event is generated
-
host.repoInit(tag, task_id=None, with_src=False, with_debuginfo=False, event=None, with_separate_src=False)
Initialize a new repo for tag
-
host.setBuildRootList(brootid, rpmlist, task_id=None)
-
host.setBuildRootState(brootid, state, task_id=None)
-
host.setTaskWeight(task_id, weight)
-
host.subtask(method, arglist, parent, **opts)
-
host.subtask2(_HostExports__parent, _HostExports__taskopts, _HostExports__method, *args, **opts)
A wrapper around subtask with optional signature
Parameters:
__parent: task id of the parent task
__taskopts: dictionary of task options
__method: the method to be invoked
Remaining args are passed on to the subtask
-
host.tagBuild(task_id, tag, build, force=False, fromtag=None)
Tag a build (host version)
This tags as the user who owns the task
If fromtag is specified, also untag the package (i.e. move in a single
transaction)
No return value
-
host.tagNotification(is_successful, tag_id, from_id, build_id, user_id, ignore_success=False, failure_msg='')
Create a tag notification message.
Handles creation of tagNotification tasks for hosts.
-
host.taskSetWait(parent, tasks)
-
host.taskWait(parent)
-
host.taskWaitResults(parent, tasks, canfail=None)
-
host.updateBuildRootList(brootid, rpmlist, task_id=None)
-
host.updateBuildrootArchives(brootid, task_id, archives, project=False)
-
host.updateHost(task_load, ready)
-
host.updateMavenBuildRootList(brootid, task_id, mavenlist, ignore=None, project=False, ignore_unknown=False, extra_deps=None)
-
host.writeSignedRPM(an_rpm, sigkey, force=False)
Write a signed copy of the rpm
-
importArchive(filepath, buildinfo, type, typeInfo)
Import an archive file and associate it with a build. The archive can
be any non-rpm filetype supported by Koji.
filepath: path to the archive file (relative to the Koji workdir)
buildinfo: information about the build to associate the archive with
May be a string (NVR), integer (buildID), or dict (containing keys: name,
version, release)
type: type of the archive being imported. Currently supported archive types: maven, win
typeInfo: dict of type-specific information
-
importRPM(path, basename)
Import an RPM into the database.
The file must be uploaded first.
-
krbLogin(*args, **opts)
-
listArchiveFiles(archive_id, queryOpts=None, strict=False)
Get information about the files contained in the archive with the given ID.
Returns a list of maps with with following keys:
archive_id: id of the archive the file is contained in (integer)
name: name of the file (string)
size: uncompressed size of the file (integer)
If strict is True, raise GenericError if archive_type is not one that we
are able to expand
Regardless of strict, an error will be raised if the archive_id is invalid
-
listArchives(buildID=None, buildrootID=None, componentBuildrootID=None, hostID=None, type=None, filename=None, size=None, checksum=None, checksum_type=None, typeInfo=None, queryOpts=None, imageID=None, archiveID=None, strict=False)
Retrieve information about archives.
If buildID is not null it will restrict the list to archives built by the build with that ID.
If buildrootID is not null it will restrict the list to archives built in the buildroot with
that ID.
If componentBuildrootID is not null it will restrict the list to archives that were present in
the buildroot with that ID.
If hostID is not null it will restrict the list to archives built on the host with that ID.
If filename, size, checksum and/or checksum_type are not null it will filter
the results to entries matching the provided values.
Returns a list of maps containing the following keys:
id: unique id of the archive file (integer)
type_id: id of the archive type (Java jar, Solaris pkg, Windows exe, etc.) (integer)
type_name: name of the archive type
type_description: description of the archive
type_extensions: valid extensions for the type
build_id: id of the build that generated this archive (integer)
buildroot_id: id of the buildroot where this archive was built (integer)
filename: name of the archive (string)
size: size of the archive (integer)
checksum: checksum of the archive (string)
checksum_type: the checksum type (integer)
Koji only stores one checksum type per archive. Some content generators
store md5 checksums in Koji, and others store sha256 checksums, and this may
change for older and newer archives as different content generators upgrade
to sha256.
If you search for an archive by its sha256 checksum, and Koji has only
stored an md5 checksum record for that archive, then Koji will not return a
result for that archive. You may need to call listArchives multiple times,
once for each checksum_type you want to search.
If componentBuildrootID is specified, then the map will also contain the following key:
project: whether the archive was pulled in as a project dependency, or as part of the
build environment setup (boolean)
If 'type' is specified, then the archives listed will be limited
those associated with additional metadata of the given type.
Supported types are "maven", "win", "image", or the btype name of any
other content generator (eg. "remote-sources").
If 'maven' is specified as a type, each returned map will contain
these additional keys:
group_id: Maven groupId (string)
artifact_id: Maven artifactId (string)
version: Maven version (string)
if 'win' is specified as a type, each returned map will contain
these additional keys:
relpath: the relative path where the file is located (string)
platforms: space-separated list of platforms the file is suitable for use on (string)
flags: space-separated list of flags used when building the file (fre, chk) (string)
if 'image' is specified as a type, each returned map will contain an
additional key:
arch: The architecture if the image itself, which may be different from the
task that generated it
typeInfo is a dict that can be used to filter the output by type-specific info.
For the 'maven' type, this dict may contain one or more of group_id, artifact_id, or version,
and the output will be restricted to archives with matching attributes.
If there are no archives matching the selection criteria, if strict is False,
an empty list is returned, otherwise GenericError is raised.
-
listBTypes(query=None, queryOpts=None)
List btypes matching query
:param dict query: Select a particular btype by "name" or "id".
Example: {"name": "image"}.
If this parameter is None (default), Koji returns all
btypes.
:param dict queryOpts: additional options for this query.
:returns: a list of btype dicts. If you specify a query parameter for a
btype name or id that does not exist, Koji will return an empty
list.
-
listBuildRPMs(build)
Get information about all the RPMs generated by the build with the given
ID. A list of maps is returned, each map containing the following keys:
- id
- name
- version
- release
- arch
- epoch
- payloadhash
- size
- buildtime
- build_id
- buildroot_id
If no build has the given ID, or the build generated no RPMs, an empty list is returned.
-
listBuildroots(hostID=None, tagID=None, state=None, rpmID=None, archiveID=None, taskID=None, buildrootID=None, queryOpts=None)
Return a list of matching buildroots
Optional args:
hostID - only buildroots on host.
tagID - only buildroots for tag.
state - only buildroots in state (may be a list)
rpmID - only buildroots the specified rpm was used in
archiveID - only buildroots the specified archive was used in
taskID - only buildroots associated with task.
buildrootID - only the specified buildroot
queryOpts - query options
-
listBuilds(packageID=None, userID=None, taskID=None, prefix=None, state=None, volumeID=None, source=None, createdBefore=None, createdAfter=None, completeBefore=None, completeAfter=None, type=None, typeInfo=None, queryOpts=None, pattern=None, cgID=None)
Return a list of builds that match the given parameters
Filter parameters
:param int|str packageID: only builds of the specified package
:param int|str userID: only builds owned by the given user
:param int taskID: only builds with the given task ID
If taskID is -1, only builds with a non-null task id
:param int volumeID: only builds stored on the given volume
:param str source: only builds where the source field matches (glob pattern)
:param str prefix: only builds whose package name starts with that prefix
:param str pattern: only builds whose nvr matches the glob pattern
:param int state: only builds in the given state
:param int|str cgID: only build from given content generator
Timestamp filter parameters
- these limit the results to builds where the corresponding
timestamp is before or after the given time
- the time value may be specified as seconds since the epoch or
in ISO format ('YYYY-MM-DD HH24:MI:SS')
:param str|timestamp createdBefore: filter for creation_time
:param str|timestamp createdAfter: filter for creation_time
:param str|timestamp completeBefore: filter for completion_time
:param str|timestamp completeAfter: filter for completion_time
Build type parameters:
:param str type: only builds of the given btype (such as maven or image)
:param dict typeInfo: only builds with matching type-specific info (given
as a dictionary). Can only be used in conjunction with the
type parameter. Only limited types are supported.
For type=maven, the provided group_id, artifact_id, and/or version
fields are matched
For type=win, the provided platform fields are matched
:returns: Returns a list of maps. Each map contains the following keys:
- build_id
- version
- release
- epoch
- state
- package_id
- package_name
- name (same as package_name)
- nvr (synthesized for sorting purposes)
- owner_id
- owner_name
- volume_id
- volume_name
- source
- creation_event_id
- creation_time
- creation_ts
- start_time
- start_ts
- completion_time
- completion_ts
- task_id
- extra
If type == 'maven', each map will also contain the following keys:
- maven_group_id
- maven_artifact_id
- maven_version
If type == 'win', each map will also contain the following key:
- platform
If no builds match, an empty list is returned.
-
listCGs()
List all available content generators in the system
:returns: A map of content generators, like {"name": data}. The data map
for each content generator has an "id" key for the content
generator ID, and a "users" key for the a list usernames who
are permitted to import for this content generator.
-
listChannels(hostID=None, event=None, enabled=None)
List builder channels.
:param hostID: Koji builder host ID or hostname. If specified, Koji will
return only the channels associated with this builder host.
If unspecified, Koji will return all channels.
:type hostID: int or str
:param int event: The event ID at which to search. If unspecified, the
default behavior is to search for the "active" host
settings. You must specify a hostID parameter with this
option.
:param bool enabled: Enabled/disabled list of channels
:returns: list of dicts, one per channel. For example,
[{'comment': 'test channel', 'description': 'container channel',
'enabled': True, 'id': 20, 'name': 'container', 'container channel' }]
-
listExternalRepos(info=None, url=None, event=None, queryOpts=None)
Get a list of external repos. If info is not None it may be a
string (name) or an integer (id).
If url is not None, filter the list of repos to those matching the
given url.
-
listHosts(arches=None, channelID=None, ready=None, enabled=None, userID=None, queryOpts=None)
List builder hosts.
All parameters are optional.
:param list arches: a list of string architecture names, e.g.
['x86_64', 'ppc64le']. If one of the arches
associated with a given host appears in the list,
it will be included in the results.
:param int|str channelID: Channel name or ID to search. If specified,
Koji will return only the builder hosts
associated with this channel.
:param bool ready: If specified, only include hosts that are ready
(True) or not ready (False).
:param bool enabled: If specified, only include hosts that are enabled
(True) or not enabled (False).
:param int|str userID: If specified, only include hosts corresponding
to a builder "user" account name or ID number.
:param queryOpts: query options used by the QueryProcessor.
:returns: A list of maps containing information about each builder
host. If no matches are found, this method returns an empty
list.
-
listPackages(tagID=None, userID=None, pkgID=None, prefix=None, inherited=False, with_dups=False, event=None, queryOpts=None, with_owners=True, with_blocked=True)
Returns a list of packages in Koji.
All parameters are optional.
Note that the returned data includes blocked entries
:param int|str tag: filter on tag ID or name
:param int|str userID: filter on package owner
:param int|str pkgID: filter on package
:param str prefix: filter package names that start with a
case-insensitive string.
:param bool inherited: return also inherited packages
:param bool with_dups: possible duplicates from inheritance, makes no
sense with inherited=False
:param int event: filter on event
:param dict queryOpts: Options to order or filter the results. For
example: {'order': 'name'}, or {'limit': 5}.
Valid query options are "countOnly", "order",
"offset", and "limit".
:param bool with_owners: Return owner_id and owner_name in the list of
packages. Default is True. This cannot be
False if userID is not None.
:returns [dict]: List of dicts with "package_id" and "package_name"
keys. If tagID, userID, or pkgID are specified, the
dicts will also contain the following keys.
- tag_id
- tag_name
- owner_id
- owner_name
- extra_arches
- blocked
-
listPackagesSimple(prefix=None, queryOpts=None)
list packages that starts with prefix and are filted
and ordered by queryOpts.
Args:
prefix: default is None. If is not None will filter out
packages which name doesn't start with the prefix.
queryOpts: query options used by the QueryProcessor.
Returns:
A list of maps is returned, and each map contains key
'package_name' and 'package_id'.
-
listRPMFiles(rpmID, queryOpts=None)
List files associated with the RPM with the given ID. A list of maps
will be returned, each with the following keys:
- name
- digest
- md5 (alias for digest)
- digest_algo
- size
- flags
If there is no RPM with the given ID, or that RPM contains no files,
an empty list will be returned.
-
listRPMs(buildID=None, buildrootID=None, imageID=None, componentBuildrootID=None, hostID=None, arches=None, queryOpts=None)
List RPMS. If buildID, imageID and/or buildrootID are specified,
restrict the list of RPMs to only those RPMs that are part of that
build, or were built in that buildroot. If componentBuildrootID is specified,
restrict the list to only those RPMs that will get pulled into that buildroot
when it is used to build another package. A list of maps is returned, each map
containing the following keys:
- id
- name
- version
- release
- nvr (synthesized for sorting purposes)
- arch
- epoch
- payloadhash
- size
- buildtime
- build_id
- buildroot_id
- external_repo_id
- external_repo_name
- metadata_only
- extra
If componentBuildrootID is specified, two additional keys will be included:
- component_buildroot_id
- is_update
If no build has the given ID, or the build generated no RPMs,
an empty list is returned.
-
listTagged(tag, event=None, inherit=False, prefix=None, latest=False, package=None, owner=None, type=None, strict=True, extra=False)
List builds tagged with tag.
:param int|str tag: tag name or ID number
:param int event: event ID
:param bool inherit: If inherit is True, follow the tag hierarchy and return
a list of tagged builds for all tags in the tree
:param str prefix: only builds whose package name starts with that prefix
:param bool|int latest: True for latest build per package,
N to get N latest builds per package.
:param str package: only builds of the specified package
:param owner: only builds of the specified owner
:param str type: only builds of the given btype (such as maven or image)
:param bool strict: If tag doesn't exist, an exception is raised,
unless strict is False in which case returns an empty list.
:param bool extra: Set to "True" to get the build extra info
-
listTaggedArchives(tag, event=None, inherit=False, latest=False, package=None, type=None, strict=True, extra=True)
List archives and builds within a tag.
:param int|str tag: tag name or ID number
:param int event: event ID
:param bool inherit: If inherit is True, follow the tag hierarchy and return
a list of tagged archives for all tags in the tree
:param bool|int latest: Set to "True" to get tagged archives just from latest build.
Set latest=N to get only the N latest tagged RPMs.
:param str package: only archives of the specified package
:param str type: only archives of the given btype (such as maven or image)
:param bool strict: If tag doesn't exist, an exception is raised,
unless strict is False in which case returns an empty list.
:param bool extra: Set to "False" to skip the archive extra info
-
listTaggedRPMS(tag, event=None, inherit=False, latest=False, package=None, arch=None, rpmsigs=False, owner=None, type=None, strict=True, extra=True)
List rpms and builds within tag.
:param int|str tag: tag name or ID number
:param int event: event ID
:param bool inherit: If inherit is True, follow the tag hierarchy and return
a list of tagged RPMs for all tags in the tree
:param bool|int latest: Set to "True" to get the single latest build. Set
to an int "N" to get the latest "N" builds. If
unspecified, the default value is "False", and
Koji will list all builds in the tag.
:param str package: only rpms of the specified package
:param str arch: only rpms of the specified arch
:param bool rpmsigs: query will return one record per rpm/signature combination
:param str owner: only rpms of the specified owner
:param str type: only rpms of the given btype (such as maven or image)
:param bool strict: If tag doesn't exist, an exception is raised,
unless strict is False in which case returns an empty list.
:param bool extra: Set to "False" to skip the rpms extra info
-
listTags(build=None, package=None, perms=True, queryOpts=None, pattern=None)
List tags according to filters
:param int|str build: If build is specified, only return tags associated with
the given build. Build can be either an integer ID or
a string N-V-R.
:param int|str package: If package is specified, only return tags associated with the
specified package. Package can be either an integer ID or a
string name.
In this case resulting map will have additional keys:
- owner_id
- owner_name
- blocked
- extra_arches
:param bool perms: If perms is True, perm_id and perm is added to resulting maps.
:param dict queryOpts: hash with query options for QueryProcessor
:param pattern: If glob pattern is specified, only return tags matching that pattern.
:returns list of dicts: Each map contains id, name, arches and locked keys and
additional keys as specified via package or perms options.
-
listTaskOutput(taskID, stat=False, all_volumes=False, strict=False)
List the files generated by the task with the given ID. This
will usually include one or more RPMs, and one or more log files.
If the task did not generate any files, or the output directory
for the task no longer exists, return an empty list.
If stat is True, return a map of filename -> stat_info where stat_info
is a map containing the values of the st_* attributes returned by
os.stat().
If all_volumes is set, results are extended to deal with files in same
relative paths on different volumes.
With all_volumes=True, stat=False, return a map of filename -> list_of_volumes,
{'stdout.log': ['DEFAULT']}
With all_volumes=True, stat=True, return a map of
filename -> map_of_volumes -> stat_info,
{'stdout.log':
{'DEFAULT': {
{
'st_atime': 1488902587.2141163,
'st_ctime': 1488902588.2281106,
'st_mtime': 1488902588.2281106,
'st_size': '526'
}
}
}
If strict is set, function will raise a GenericError if task doesn't
exist. Allows user to distinguish between empty output and non-existent task.
-
listTasks(opts=None, queryOpts=None)
Return list of tasks filtered by options
Options(dictionary):
option[type]: meaning
arch[list]: limit to tasks for given arches
not_arch[list]: limit to tasks without the given arches
state[list]: limit to tasks of given state
not_state[list]: limit to tasks not of the given state
owner[int|list]: limit to tasks owned by the user with the given ID
not_owner[int|list]: limit to tasks not owned by the user with the given ID
host_id[int|list]: limit to tasks running on the host with the given ID
not_host_id[int|list]: limit to tasks running on the hosts with IDs other than the
given ID
channel_id[int|list]: limit to tasks in the specified channel
not_channel_id[int|list]: limit to tasks not in the specified channel
parent[int|list]: limit to tasks with the given parent
not_parent[int|list]: limit to tasks without the given parent
decode[bool]: whether or not xmlrpc data in the 'request' and 'result'
fields should be decoded; defaults to False
method[str]: limit to tasks of the given method
createdBefore[float or str]: limit to tasks whose create_time is before the
given date, in either float (seconds since the epoch)
or str (ISO) format
createdAfter[float or str]: limit to tasks whose create_time is after the
given date, in either float (seconds since the epoch)
or str (ISO) format
startedBefore[float or str]: limit to tasks whose start_time is before the
given date, in either float (seconds since the epoch)
or str (ISO) format
startedAfter[float or str]: limit to tasks whose start_time is after the
given date, in either float (seconds since the epoch)
or str (ISO) format
completeBefore[float or str]: limit to tasks whose completion_time is before
the given date, in either float (seconds since the epoch)
or str (ISO) format
completeAfter[float or str]: limit to tasks whose completion_time is after
the given date, in either float (seconds since the epoch)
or str (ISO) format
-
listUsers(userType=0, prefix=None, queryOpts=None)
List all users in the system.
userType can be an integer value from koji.USERTYPES (defaults to 0,
i.e. normal users). Returns a list of maps with the following keys:
- id
- name
- status
- usertype
- krb_principals
If no users of the specified
type exist, return an empty list.
-
listVolumes()
List storage volumes
-
login(*args, **opts)
Create a login session with plain user/password credentials.
:param str user: username
:param str password: password
:param dict opts: curently can contain only 'host_ip' key for overriding client IP address
:returns dict: session info
-
logout()
expire a login session
-
logoutChild(session_id)
expire a subsession
:param int subsession_id: subsession ID (for current session)
-
makeTask(*args, **opts)
Creates task manually. This is mainly for debugging, only an admin
can make arbitrary tasks. You need to supply all *args and **opts
accordingly to the task.
-
mavenBuild(url, target, opts=None, priority=None, channel='maven')
Create a Maven build task
url: The url to checkout the source from. May be a CVS, SVN, or GIT repository.
target: the build target
priority: the amount to increase (or decrease) the task priority, relative
to the default priority; higher values mean lower priority; only
admins have the right to specify a negative priority here
channel: the channel to allocate the task to (defaults to the "maven" channel)
Returns the task ID
-
mavenEnabled()
Get status of maven support
-
mergeScratch(task_id)
Import the rpms generated by a scratch build, and associate
them with an existing build.
To be eligible for import, the build must:
- be successfully completed
- contain at least one arch-specific rpm
The task must:
- be a 'build' task
- be successfully completed
- use the exact same SCM URL as the build
- contain at least one arch-specific rpm
- have no overlap between the arches of the rpms it contains and
the rpms contained by the build
- contain a .src.rpm whose filename exactly matches the .src.rpm
of the build
Only arch-specific rpms will be imported. noarch rpms and the src
rpm will be skipped. Build logs and buildroot metadata from the
scratch build will be imported along with the rpms.
This is useful for bootstrapping a new arch. RPMs can be built
for the new arch using a scratch build and then merged into an
existing build, incrementally expanding arch coverage and avoiding
the need for a mass-rebuild to support the new arch.
-
moveAllBuilds(tag1, tag2, package, force=False)
Move all builds of a package from tag1 to tag2 in the correct order
Returns the task id of the task performing the move
-
moveBuild(tag1, tag2, build, force=False)
Move a build from tag1 to tag2
Returns the task id of the task performing the move
-
newGroup(name)
Add a user group to the database
-
newRepo(tag, event=None, src=False, debuginfo=False, separate_src=False)
Create a newRepo task. returns task id
-
packageListAdd(taginfo, pkginfo, owner=None, block=None, extra_arches=None, force=False, update=False)
Add to (or update) package list for tag
-
packageListBlock(taginfo, pkginfo, force=False)
Block the package in tag
-
packageListRemove(taginfo, pkginfo, force=False)
Remove a package from a tag's package list
One reason to remove an entry like this is to remove an override so that
the package data can be inherited from elsewhere.
Alternatively you can use the block or unblock functions.
This method always returns None, even if the package does not exist in the
tag.
:param int|str taginfo: tag id or name to remove the package
:param int|str pkginfo: package id or name to remove
:param bool force: If False (default), Koji will check this
operation against the package_list hub policy. If hub
policy does not allow the current user to remove
packages from this tag, then this method will raise
an error.
If True, then this method will bypass hub policy
settings. Only admin users can set force to True.
-
packageListSetArches(taginfo, pkginfo, arches, force=False)
Set extra_arches for package in tag
-
packageListSetOwner(taginfo, pkginfo, owner, force=False)
Set the owner for package in tag
-
packageListUnblock(taginfo, pkginfo, force=False)
Unblock the package in tag
Generally this just adds a unblocked duplicate of the blocked entry.
However, if the block is actually in tag directly (not through inheritance),
the blocking entry is simply removed
-
queryHistory(tables=None, **kwargs)
Returns history data from various tables that support it
tables: list of versioned tables to search, no value implies all tables
valid entries: user_perms, user_groups, tag_inheritance, tag_config,
build_target_config, external_repo_config, tag_external_repos,
tag_listing, tag_packages, tag_package_owners, group_config,
group_req_listing, group_package_listing
- Time options -
times are specified as an integer event or a string timestamp
time options are valid for all record types
before: either created or revoked before timestamp
after: either created or revoked after timestamp
beforeEvent: either created or revoked before event id
afterEvent: either created or revoked after event id
- other versioning options-
active: select by active status
editor: record created or revoked by user
- table-specific search options -
use of these options will implicitly limit the search to applicable tables
package: only for given package
build: only for given build
tag: only for given tag
user: only affecting a given user
permission: only relating to a given permission
external_repo: only relateing to an external repo
build_target: only relating to a build target
group: only relating to a (comps) group
cg: only relating to a content generator
-
queryRPMSigs(rpm_id=None, sigkey=None, queryOpts=None)
Queries db for rpm signatures
:param rpm_id: a int RPM ID,
a string N-V-R.A,
a map containing 'name', 'version', 'release', and 'arch'
:param int sigkey: signature key hash
:param queryOpts: query options used by the QueryProcessor.
:returns: list of dicts (rpm_id, sigkey, sighash)
-
removeChannel(channel_name, force=False)
Remove a channel.
:param str channel_name: channel name
:param bool force: remove channel which has hosts
Channel must have no hosts, unless force is set to True
If a channel has associated tasks, it cannot be removed
and an exception will be raised.
Removing channel will remove also remove complete history
for that channel.
-
removeExternalRepoFromTag(tag_info, repo_info)
Remove an external repo from a tag
:param tag_info: Tag name or ID number
:param repo_info: External repository name or ID number
:raises: GenericError if this external repo is not associated
with this tag.
-
removeHostFromChannel(hostname, channel_name)
Remove the host from the specified channel
:param str hostname: host name
:param str channel_name: channel name
-
removeUserKrbPrincipal(user, krb_principal)
remove a Kerberos principal for user
-
removeVolume(volume)
Remove unused storage volume from the database
-
renameChannel(old, new)
Rename a channel
-
repoDelete(repo_id)
Attempt to mark repo deleted, return number of references
If the number of references is nonzero, no change is made
Does not remove from disk
-
repoExpire(repo_id)
mark repo expired
-
repoInfo(repo_id, strict=False)
Get repo information
:param int repo_id: repo ID
:param bool strict: raise an error on non-existent repo
:returns: dict (id, state, create_event, creation_time, tag_id, tag_name,
dist)
-
repoProblem(repo_id)
mark repo as broken
-
resetBuild(build)
Reset a build so that it can be reimported
WARNING: this function is highly destructive. use with care.
nulls task_id
sets state to CANCELED
clears all referenced data in other tables, including buildroot and
archive component tables
after reset, only the build table entry is left
-
restartHosts(priority=5, options=None)
Spawns restartHosts task
:param int priority: task priority
:param dict options: additional task arguments (see restartHosts task)
:returns: task ID
-
resubmitTask(taskID)
Retry a canceled or failed task, using the same parameter as the original task.
The logged-in user must be the owner of the original task or an admin.
-
revokeCGAccess(user, cg)
Revoke a user's access to act as the given content generator
:param user: koji userid or username
:type user: int or str
:param cg: content generator id or name
:type cg: int or str
-
revokePermission(userinfo, permission)
Revoke a permission from a user
-
runroot(tagInfo, arch, command, channel=None, **opts)
Create a runroot task
-
search(terms, type, matchType, queryOpts=None)
Search for an item in the database matching "terms".
:param str terms: Search for items in the database that match this
value.
:param str type: What object type to search for. Must be one of
"package", "build", "tag", "target", "user", "host",
"rpm", "maven", or "win".
:param str matchType: The type of search to perform:
- If you specify "glob", Koji will treat "terms"
as a case-insensitive glob.
- If you specify "regexp", Koji will treat
"terms" as a case-insensitive regular
expression.
- Any other value here will cause to Koji to
search for an exact string match for "terms".
:param dict queryOpts: Options to pass into the database query. Use
this to limit or order the results of the
search. For example: {'order': 'name'},
or {'limit': 5, 'order': '-build_id'}, etc.
:returns: A list of maps containing "id" and "name". If no matches
are found, this method returns an empty list.
-
setBuildOwner(build, user)
Sets owner of a build
:param int|str|dict build: build ID, NVR or dict with name, version and release
:param user: a str (Kerberos principal or name) or an int (user id)
or a dict:
- id: User's ID
- name: User's name
- krb_principal: Kerberos principal
:returns: None
-
setBuildTimestamp(build, ts)
Set the completion time for a build
build should a valid nvr or build id
ts should be # of seconds since epoch or optionally an
xmlrpc DateTime value
-
setInheritanceData(tag, data, clear=False)
Set inheritance relationships for a tag.
This tag will be the "child" that inherits from a list of "parents".
:param tag: The koji tag that will inherit from parent tags.
:type tag: int or str
:param list data: Inheritance rules to set for this child tag. This is
a list of rules (dicts) for parent tags and
priorities. If any rule dict in the list has a
special "delete link": True key and value, Koji will
remove this inheritance rule instead of adding it.
:param bool clear: Wipe out all existing inheritance rules and only
apply the ones you submit here. If unspecified,
this defaults to False.
-
setTaskPriority(task_id, priority, recurse=True)
Set task priority
-
sharedSession()
Drop out of exclusive mode
-
showOpts(as_string=True)
Returns hub options
:param bool as_string: if True (default) returns hub options as raw string,
if False, returns hub options as dict
-
showSession()
Return string representation of session for current user
-
sslLogin(*args, **opts)
Login via SSL certificate
:param str proxyuser: proxy username
:returns dict: session info
-
subsession()
Create a subsession
-
system.listMethods()
-
system.methodHelp(method)
-
system.methodSignature(method)
-
tagBuild(tag, build, force=False, fromtag=None)
Request that a build be tagged
The force option will attempt to force the action in the event of:
- tag locked
- missing permission
- package not in list for tag
- policy violation
The force option is really only effective for admins
If fromtag is specified, this becomes a move operation.
This call creates a task that was originally intended to perform more
extensive checks, but never has. We're stuck with this task system until
we're ready to break the api.
The return value is the task id
-
tagBuildBypass(tag, build, force=False, notify=False)
Tag a build without running post checks
This is a short circuit function for imports.
Admin or tag permission required.
Tagging with a locked tag is not allowed unless force is true.
Retagging is not allowed unless force is true. (retagging changes the order
of entries will affect which build is the latest)
-
tagChangedSinceEvent(event, taglist)
Report whether any changes since event affect any of the tags in list
The function is used by the repo daemon to determine which of its repos
are up to date.
This function does not figure inheritance, the calling function should
expand the taglist to include any desired inheritance.
Returns: True or False
-
taskFinished(taskId)
Returns True if task is finished
:param int task: id of task queried
:returns bool: task not/finished
-
untagBuild(tag, build, strict=True, force=False)
Untag a build
Unlike tagBuild, this does not create a task.
:param int|str tag: tag name or ID
:param int|str build: build name or ID
:param bool strict: If True (default), Koji will raise a TagError if
this build is not in the tag. If False, Koji will
not raise TagError when the build is not in the
tag.
:param bool force: If False (default), Koji will check this operation
against the tag hub policy. If hub policy does not
allow the current user to untag packages from this
tag, then this method will raise an error.
If True, then this method will bypass hub policy
settings. Only admin users can set force to True.
:returns: None
-
untagBuildBypass(tag, build, strict=True, force=False, notify=False)
Untag a build without any checks
Admin and tag permission only. Intended for syncs/imports.
Unlike tagBuild, this does not create a task
No return value
-
untaggedBuilds(name=None, queryOpts=None)
Returns the list of untagged builds
-
updateNotification(id, package_id, tag_id, success_only)
Update an existing build notification with new data. If the notification
with the given ID doesn't exist, or the currently logged-in user is not the
owner or the notification or an admin, raise a GenericError.
-
uploadFile(path, name, size, md5sum, offset, data, volume=None, checksum=None)
upload file to the hub
Files can be uploaded in chunks, if so the hash and size describe the
chunk rather than the whole file.
:param str path: the relative path to upload to
:param str name: the name of the file
:param int size: size of contents (bytes)
:param checksum: MD5 hex digest (see md5sum) or a tuple (hash_type, digest) of contents
:type checksum: str or tuple
:param str data: base64 encoded file contents
:param int offset: The offset indicates where the chunk belongs.
The special offset -1 is used to indicate the final
chunk.
:param str md5sum: legacy param name of checksum. md5sum name is misleading,
but it is here for backwards compatibility
:returns: True
-
winBuild(vm, url, target, opts=None, priority=None, channel='vm')
Create a Windows build task
vm: the name of the VM to run the build in
url: The url to checkout the source from. May be a CVS, SVN, or GIT repository.
opts: task options
target: the build target
priority: the amount to increase (or decrease) the task priority, relative
to the default priority; higher values mean lower priority; only
admins have the right to specify a negative priority here
channel: the channel to allocate the task to (defaults to the "vm" channel)
Returns the task ID
-
winEnabled()
Get status of windows support
-
wrapperRPM(build, url, target, priority=None, channel='maven', opts=None)
Create a top-level wrapperRPM task
build: The build to generate wrapper rpms for. Must be in the COMPLETE state and have no
rpms already associated with it.
url: SCM URL to a specfile fragment
target: The build target to use when building the wrapper rpm.
The build_tag of the target will be used to populate the buildroot in which the
rpms are built.
priority: the amount to increase (or decrease) the task priority, relative
to the default priority; higher values mean lower priority; only
admins have the right to specify a negative priority here
channel: the channel to allocate the task to (defaults to the "maven" channel)
returns the task ID
-
writeSignedRPM(an_rpm, sigkey, force=False)
Write a signed copy of the rpm
