Data export formats for leaving or changing supplier #

This page is for project managers who must name the export formats in a contract or exit plan, and for the administrators who produce them. Every tool and format below is documented in Migrating away from Exponential and Exponential's import and export tools. Commands are run in the installation root.

Contents #

1. Which export for which receiver
2. The site archive
3. Single classes
4. Packages
5. The database
6. Files
7. Layouts, forms, consent and audit
8. Content over HTTP
9. What to name in the contract

Which export for which receiver #

The site archive #

./console ext:xrowextract:archive --set=everything --format=zip --output=<dir>

• Per class a CSV with object id, remote id, main node, parent node, URL alias, dates and every attribute.
• A manifest.json that describes the archive and a README.txt.
• ZIP, TAR.GZ, TAR.BZ2 or TAR.XZ.
• Password hashes are never included.

This is the most portable export: plain CSV that spreadsheets, scripts and most importers read.

Single classes #

./console ext:xrowextract:csv --class=<identifier> --node=<id> --output=<file>

Also writes XML and JSON, with a typed manifest (datatype, format, language per column). Options such as --list-classes, --list-columns and --preview=10 help to look before exporting. ./console exp:ezcsvexport writes one CSV per class for a subtree, with --storage-dir for files.

Packages #

php ezpm.php export or ./console ext:xrowextract:package --export write .ezpkg packages of classes and content, for another Exponential installation. See Packages for what a package carries.

The database #

• ./console exp:ezsqldumpschema --output-types=all ... writes schema and data in Exponential's neutral format or SQL, for any of the supported engines.
• Or the database's own dump tool (Database backups per engine).
• sevenx_dse, the administration database editor, exports single tables as SQL, CSV or TSV.

A database dump is the most complete copy but only useful to someone who runs Exponential or reads its schema. Rich text in ezxmltext fields is stored as Exponential's XML: render it or transform it for another system (Leaving Exponential, section 3).

Files #

Images, documents and media are in var/storage/ and var/<site>/storage/: the originals. Copy them with every export that refers to files. On a database cluster, run exp:clusterize -u first so the files are on disk.

Layouts, forms, consent and audit #

Content over HTTP #

The REST API v1 and v2 return nodes, children and fields as JSON, so another system can pull content over HTTP (REST API v1 and v2). Version 2 can return rich text as XHTML.

What to name in the contract #

On request, and at the end of the contract, the Supplier delivers within <n> working days:
- a site archive of all classes (xrowextract, <ZIP|TAR.XZ>) with manifest;
- a full database dump in <engine dump format | Exponential neutral format>;
- the storage directories of every site;
- layouts, form submissions, consent records and the audit trail export with signed manifest;
- settings and keys, transferred privately;
- the source of all custom code in Git with history.

Related pages #

• Exit plans: overview
• Exit plan template
• Migrating away from Exponential
• Free documentation: this page is free to copy, translate and improve under the GNU FDL 1.3.