Tool to download all datasets from a Socrata data portal and keep revisions on S3.
Specifically, Catalog It maintains a list of all "datasets" (this is a specific type on Socrata), looks for the last modified date from the dataset headers, and then uploads a copy of that data if it has changed.
Install: npm install catalog-it -g
Use catalog-it --help
to show this help page.
Usage: catalog-it [options] [command]
Commands:
update|u Fetch the list of datasets.
changed|c Look at each dataset to see what has changed.
data|d Archive data that has changed.
one|o <id> Get one dataset; more useful for testing.
Options:
-h, --help output usage information
-V, --version output the version number
-c, --catalog <catalog> The domain where the data catalog is location, for instance: "nycopendata.socrata.com"
-b, --bucket <bucket> S3 bucket
-p, --profile [profile] Name of the profile in .aws/credentials to use. Overrides key options
--access-key-id [key] AWS access key to use. It is suggest to use the AWS_ACCESS_KEY_ID instead
--secret-access-key [key] AWS secret key to use. It is suggest to use the AWS_SECRET_ACCESS_KEY instead
-a, --acl [acl] ACL of items, defaults to "private"; should be "private", "public-read", "public-read-write", or "authenticated-read"
-C, --concurrency [int] Concurrency of fetching datasets and uploading to S3, defaults to 5
-f, --format [format] Format of data download, defaults to "csv"; should be "csv" or "json"
-r, --prefix [path] The path prefix for S3 items; for example: "backups/catalogs"
-t, --timeout [milleseconds] Timeout for uploads and downloads in milliseconds, defaults to 100000 (Socrata can be slow)
-P, --path [path] Path to directory where cache will live, defauls to "HOME/.catalog-it"
-o, --config [path] Path to environment variable file as defined by the "dotenv" module.
--compress Compress with gzip when uploading to S3 (saves space); this is the default;
--no-compress Do not compress with gzip when uploading to S3 (saves space).
--bucket-create Attempt to create the bucket on start; this is the default.
--no-bucket-create Do not attempt to create the bucket on start.
The command line will also look for environment variables and use those if they are not provided by the command line. Each environment variable should be prefixed with CATALOG_IT_
. The command line will also try to read in an .env
file if it exists, or the file provided by the --config
option.
Install: npm install catalog-it --save
Catalog It can also be used as a JS module. Simply include and use the same options as the command line except that the option properties are camelcase and the config
option is not used.
const CatalogIt = require('catalog-it');
let c = new CatalogIt({
catalog: 'something.catalog.com',
bucket: 'my-s3.bucket',
profile: 'default'
});
The module also supports using environment variables but will not read in any file.
catalog
: Current catalog in memory.updateCatalog(function(error, catalog) { ... })
: Gets new catalog and updates existing.scanHeaders(function(errors, updates) { ... })
: Gets info from headers to determine what needs changing. This function will skip network errors.scanData(function(errors, updates) { ... })
: Gets new data and archives to S3. This function will skip network errors.
Scheduling can happen however you like, but here is a cron example that runs once a week on Sunday at 2AM. This updates the list of datasets, determines which ones have changed, and then archives the data.
0 2 * * 0 catalog-it u && catalog-it c && catalog-it d;