This page describes the components of the Open Siddur technical architecture. All references to directories below are relative to the root directory of the [source code] (https://github.com/opensiddur/opensiddur).
The Open Siddur Project is developing a socially-networked online publishing platform for rendering print-ready digital files from publicly shared and user-contributed content. Textual content is semantically tagged, internally linked, and annotated. Graphical and other content is stored as annotation to text.
The platform will enable users to:
- share original and adapted content between friends, groups, or globally for creative reuse,
- maintain and share changes in the content of any shared work,
- compare and trace variations in the content of any shared work,
- arbitrarily recombine (i.e., remix) content selected from multiple sources into a new linear work,
- design sharable layout templates adjustable to the desired shape and format of the finished work.
Additionally, the platform will preserve and record a chain of metadata for all shared work, including:
- what source(s) it came from
- who wrote it, edited it, and/or proofread it
#Summary Diagram The platform architecture is summarized in the following diagram:
- Source code repository : Serves code and documentation
- Virtual Private Server (VPS)/External hosting : Serves the database, web applications, and demos
These parts provide format specifications and provide the backbone of the Open Siddur Project's publishing platform.
- XML Database : An installation of [eXist] (http://exist-db.org) at <dev.jewishliturgy.org>, which you may build and mirror (see Building). Stores all primary data and acts as an interface to client applications by execution of XQueries through its REST interface.
JLPTEI Specifications : A [TEI] (http://www.tei-c.org/)-based XML schema for representing liturgical texts. Also includes validation software, written in [TEI ODD] (http://www.tei-c.org/Guidelines/Customization/odds.xml ) and [ISO Schematron] (http://www.schematron.com). The source code is in the
ant schemacommand will generate validatable schemas and schema documentation (such as it is) in the
opensiddur-server/build/schemadirectory of the source.
Transforms : Conversion programs to translate JLPTEI into output formats, such as XHTML or PDF. The transform code is stored in git and served off the database. The source code is in the
REST API : XQuery APIs that run off the database and can be accessed via the database's REST interface. Provide basic functions for manipulating data in the database and functions that are used frequently by applications. The source code and inline documentation is in the
The client application is an integrated web-based application to be used to manipulate data on the database. As of now, there are a number of demos that act as a thin wrapper around the REST API and allow a user to do some manipulations of database data. The source code is in the
The end-user client application should be a user-friendly web interface that can display, edit and compile JLPTEI data from the database.
One-time converters : If we are given data in a format that is not JLPTEI, but is consistently applied, then it can be converted to JLPTEI directly via one-time-use conversion applications. As we receive more contributed material, the importance of one-time converters will increase. Some existing converters are in the
Texts and annotated graphics form the basis of the Open Siddur Project's content. They are the reason we are writing all the software. Content falls into the following categories:
- Scanned documents : Books or manuscripts as images. We are no longer considering storing scanned documents on our own. If you scan a siddur or know of an existing online, free scan and want to assist in bringing it into Open Siddur, let us know. We will likely suggest that you upload it to Wikimedia Commons. We also maintain a list of Wanted sources.
- Contributed texts : These are texts that are original works submitted to us (for now, through <opensiddur.org> )
- Original text transcriptions : Transcriptions are currently being collected through the contribute page on <opensiddur.org> or via git.
- Encoded texts : These are the forms of text that can be processed by our transforms. As of now, only one-time converter software produces encoded text. You can also enter encoded texts in JLPTEI XML directly using the Open Siddur client
- Translations : We are not ready yet to accept translations directly into the archive. If you are interested in translating, talk to us. If you already have translated material that you would like us to include, see the contribute page on <opensiddur.org>.
- Instructional material : We are not ready yet to accept instructional material directly into the archive. If you are interested in writing it, talk to us. If you already have instructional material that you would like us to include, see the contribute page on <opensiddur.org>.
- Comments/essays : We are not ready yet to accept commentary directly into the archive. If you are interested in writing it, talk to us. If you already have commentary that you would like us to include, see the contribute page on <opensiddur.org>.
In addition to the technical aspects of the project, there are also some non-technical things we do:
- Membership in the community : While we cannot give legal advice, we are willing to help Jewish content providers assure bidirectional [content compatibility] (http://opensiddur.net/2009/11/jewish-content-free-culture-and-content-compatibility/) with our project. We also maintain an Environmental Scan of projects with similar goals.
- Documentation : We need end-user and mid-level documentation to ease people into contributing to the project and using the materials we already have.
- Project management : We are looking to improve resource management in the project, including volunteer management to make sure anyone who wants to contribute can find a place in the project.