OCS Archive Library

A base library for the Science Archive and Ingester library to support generalized input file types, generalized data stores, and shared configuration items. This library is configurable via environment variables, but more customization is possible by subclassing the DataFile class for a specific file type, or subclassing the FileStore class for a specific file storage scheme.

Prerequisites

Optional prerequisites may be skipped for reduced functionality.

• Python >= 3.8

Usage

This library is used with the OCS Science Archive and OCS Ingester library. It has built in support for .fits files, compressed .fits files, or custom files if the metadata is supplied separately. It is critical that the Header Mapping environment variables below are set up to map to the proper .fits header keys in your data product definition, so that the information can be automatically extracted and stored in the Science Archive. The FileStore environment variables also need to be set for whichever file storage system you are targetting (S3 preferred). The DataFile environment variables aid in the extracting and ingestion of header values from data. For more information on how to customize your OCS Science Archive, please review the data flow documentation.

Installation

It is highly recommended that you install and run your python code inside a dedicated python virtual environment.

Add the ocs_archive package to your python environment:

(venv) $ pip install ocs_archive

Configuration

Environment Variables

Group	Variable	Description	Default
FileStore	FILESTORE_TYPE	Type of filestorage to use. Options are dummy, local, or s3.	dummy
	FILESYSTEM_STORAGE_ROOT_DIR	If using local file storage, this is the directory on the local filesystem to use as the root of the storage directories	empty string
	FILESYSTEM_STORAGE_BASE_URL	If using local file storage, this is the base URL at which those files will be hosted from	http://0.0.0.0/
Observation Portal	OBSERVATION_PORTAL_BASE_URL	Base URL for the Observation Portal	empty string
	OBSERVATION_PORTAL_API_TOKEN	API Token used to authenticate with the Observation Portal	empty string
AWS	BUCKET	If using s3 file storage; AWS S3 Bucket Name	testbucket
	AWS_ACCESS_KEY_ID	If using s3 file storage; AWS Access Key with write access to the S3 bucket	empty string
	AWS_SECRET_ACCESS_KEY	If using s3 file storage; AWS Secret Access Key	empty string
	AWS_DEFAULT_REGION	If using s3 file storage; AWS S3 Default Region	empty string
	S3_ADDRESSING_STYLE	If using s3 file storage; Sets the addressing_style for generated s3 paths. Can be one of virtual, path, or auto.	virtual
	S3_ENDPOINT_URL	If using s3 file storage; Endpoint url for connecting to s3. This can be modified to connect to a local instance of s3.	"http://s3.us-west-2.amazonaws.com"
	S3_DAYS_TO_IA_STORAGE	If using s3 file storage, this is the age in days after which data will be ingested directly to Infrequent Access (IA) storage vs normal storage.	60
DataFile	FILETYPE_MAPPING_OVERRIDES	A string literal representation of a python dictionary containing a mapping of file extensions to dotpaths to python Classes which subclass the DataFile class. This appends and overrides the default list in the FileFactory class.	"{}"
	HEADER_BLACKLIST	Comma delimited string list of header values that should be removed from the data before storage in the archive. This can be overriden when instantiating a DataFile as well as via environment variable.	HISTORY,COMMENT
	REQUIRED_HEADERS	Comma delimited string list of header values that must be present in the DataFile. This can be overriden when instantiating a DataFile as well as via environment variable
	NULL_HEADER_VALUES	Comma delimited string list of header values that should be turned into None or empty keys. This only applies to the FitsFile class.	N/A,UNSPECIFIED,UNKNOWN
	CALIBRATION_TYPES	Comma delimited string list of configuration types which represent calibration images. This is used to automatically set calibration images public date to be the observation date if it is not present	BIAS,DARK,SKYFLAT,EXPERIMENTAL
	PUBLIC_PROPOSAL_TAGS	A comma delimited string list of Observation Portal proposal tags to denote data from this proposal as public. If public, the public date will be set to the observation date. The ocs_archive will fall back to the list of PUBLIC_PROPOSALS if any of a proposal's tags are not found in this list.	public
	PRIVATE_PROPOSAL_TAGS	A comma delimited string list of Observation Portal proposal tags to denote data from this proposal as private. If private, the public date will be set to 999 years in the future. The ocs_archive will fall back to the list of PRIVATE_PROPOSALS if any of a proposal's tags are not found in this list.	private,internal
	PUBLIC_PROPOSALS	Comma delimited string list of proposal IDs which represent public proposals. This is used to set the public date of observations under those proposals to the observation date if it is not present. The matching is based on if each character group appears anywhere within the proposal ID	EPO,calib,standard,pointing
	PRIVATE_PROPOSALS	A comma delimited string list of proposal IDs which represent private proposals. This is used to set the public date of the observations under those proposals to be 999 years in the future. The matching is based on if each character group appears anywhere within the proposal ID	LCOEngineering
	DAYS_UNTIL_PUBLIC	The number of days until user data becomes public by default. This is added onto the observation date to get the public date if one is not specifed with the data	365
	PRIVATE_FILE_TYPES	A comma delimited string list of fragments of the file name which denote a private data file. If any of the fragments are found within the filename, the public date will be set 999 years in the future for this file	-t00,-x00
Header Mapping	OBSERVATION_DATE_KEY	The key in which to find an iso formatted observation date within the header data	DATE-OBS
	OBSERVATION_DAY_KEY	The key in which to find an iso formatted observation day within the header data	DAY-OBS
	OBSERVATION_END_TIME_KEY	The key in which to find an iso formatted observation end date within the header data	UTSTOP
	REDUCTION_LEVEL_KEY	The key in which to find a numeric reduction level within the header data. Raw is 0, while anything non-zero is some form of processing	RLEVEL
	EXPOSURE_TIME_KEY	The key in which to find the exposure time in fractional seconds in the header data	EXPTIME
	INSTRUMENT_ID_KEY	The key in which to find the instrument ID in the header data	INSTRUME
	SITE_ID_KEY	The key in which to find the site ID in the header data	SITEID
	TELESCOPE_ID_KEY	The key in which to find the telescope ID in the header data	TELID
	OBSERVATION_ID_KEY	The key in which to find the observation ID in the header data	BLKUID
	CONFIGURATION_ID_KEY	The key in which to find the configuration ID in the header data	MOLUID
	PRIMARY_OPTICAL_ELEMENT_KEY	The key in which to find the primary filter value in the header data	FILTER
	TARGET_NAME_KEY	The key in which to find the target object's name in the header data	OBJECT
	REQUEST_ID_KEY	The key in which to find the request ID in the header data	REQNUM
	REQUESTGROUP_ID_KEY	The key in which to find the request group ID in the header data	TRACKNUM
	CONFIGURATION_TYPE_KEY	The key in which to find the configuration type in the header data	OBSTYPE
	PROPOSAL_ID_KEY	The key in which to find the proposal ID in the header data	PROPID
	CATALOG_TARGET_FRAME_KEY	The key in which to find the base filename of the catalog file for the target of this observation in the header data	L1IDCAT
	PUBLIC_DATE_KEY	The key in which to find the iso formatted date in which this data should become available to the public in the header data	L1PUBDAT
	RELATED_FRAME_KEYS	A comma delimited list of keys in the header data to look for related frame base filenames for this observation	L1IDBIAS,L1IDDARK,L1IDFLAT,L1IDSHUT,L1IDMASK,L1IDFRNG,L1IDCAT,L1IDARC,L1ID1D,L1ID2D,L1IDSUM,TARFILE,ORIGNAME,ARCFILE,FLATFILE,GUIDETAR
	THUMBNAIL_FRAME_BASENAME_KEY	The key used to store the original frame (e.g. FITS image) basename for the associated thumbnail	frame_basename
	THUMBNAIL_SIZE_KEY	The key used to store the string description of the thumbnail size (e.g. small, medium, large)	size
	RADIUS_KEY	The key in which to find FOV radius for a circular FOV, used to calculate WCS polygon if specified. Unit of arcseconds	RADIUS
	RA_KEY	The key in which to find FOV center RA for a circular FOV, used to calculate WCS polygon if specified. Unit of hour angle	RA
	DEC_KEY	The key in which to find FOV center DEC for a circular FOV, used to calculate WCS polygon if specified. Unit of decimal degrees	DEC

Input File Format Configuration

The library is designed to be configured mostly through environment variables, but custom DataFile subclasses can be included and specified via an environment variable in order to support new and more complicated data formats. All data files must contain the minimum set of metadata in order to ingested into the archive. This metadata is used to provide filtering and querying support within the archive. The pieces of file metadata that should be specified have their mappings defined in the Header Data section of the environment variables below. The FitsFile class provided will work for normal or funpacked fits files, provided you set up the Header Data environment variables with the correct mapping of observation concepts to header keys in your data format.

File Storage Format Configuration

The library supports three types of file storage by default, that can be selected via environment variable. The dummy type is just used for testing and development and doesn't actually store any file. The local storage just saves the files into a locally mounted directory. It requires you to run a separate file server on that directory so it knows how to direct links to download the files. This can be accomplished as simply as running python -m http.server --directory=/my/root/dir. It could alternatively be served using any other file server, like node's http-server. The third option is s3, and expects to connect to Amazon's S3 or something with that same interface like minio. S3 file storage requires BUCKET, AWS_*, and S3_* environment variables to be set. More storage types can be added via forking the library and subclassing the FileStore class. S3 storage is the preferred method for a production and public-facing archive, since it is the only option to support versioning and secure data access by generating presigned data URLs.

Development

Poetry

We use Poetry for package management. If you already have Poetry installed, you can skip this section.

You can install Poetry using one of the many options listed at https://python-poetry.org/docs/#installation. One simple option is using Pipx:

python3 -m pip install --user pipx
python3 -m pipx ensurepath
pipx install poetry

Install

Install the project and its Python dependencies:

poetry install

This will install the project in a Poetry managed virtual environment. To run commands in that environment either use poetry run ... or start a shell in that environment with poetry shell

Test

poetry run pytest