Migration from eDoc to PubMan
=============================

Components
----------
* EDocImport.java : Executable that transforms exported eDoc data into eSciDoc 
  item list.
* PubManImport.java : Exceutable that validates an item list and imports it 
  into PubMan.
* CreatePurgeScript : Executable that generates a Fedora purge script to remove 
  the imported items from the repository.

* edoc-to-escidoc.xslt : XSLT transformation from eDoc to publication item list
* mpipl_authors.xml, mpipl_collections.xml, mpipl_ous.xml : data includes for 
  the transformation specific for MPIPL

Workflow
--------

1) In the admin tool, the context for the imported publications has to be 
   created and opened. Remember the id.
2) The ID of this context has to be entered in edoc-to-escidoc.xslt (line 70).
3) If needed, a special user has to be entered in PubMan granted the right to 
   create and release items in the specific context.
4) In PubMan, the organizational structure has to be setup, i.e. the OUs have 
   to be entered.
5) This structure has to be transferred to a file like mpipl_ous.xml. When new 
   files are created, these have to be imported into edoc-to-escidoc.xslt 
   (line 75).
6) The authors of the MPI have to be mapped in a file like mpipl_authors.xml. 
   The aliasing is only needed when a mapping to one authorized name is desired.
7) If there should be a mapping of the authors to the eDoc collection, then a
   file like mpipl_collections.xml has to be created.
8) eDoc data have to be exported into an XML file.
9) The data are transformed with the following execution
     java EDocImport -Dfile.encoding=UTF-8 -Xmx512m <eDoc-data-file>
   where <eDoc-data-file> is the path and filename of the exported data. A new 
   file is created in the project root directory named "edoc_export_out.xml"
10) Check in your settings.xml file if you point to the correct coreservice instance.
11) The data are imported with the following execution
      java PubManImport -Dfile.encoding=UTF-8 -Xmx512m "edoc_export_out.xml" <eSciDoc-user-name> <eSciDoc-password>
12) CHECK THE IMPORTED DATA

If the imported data shall be removed from the repository, do the following

1) Set the context id in CreatePurgeScript2.java (line 56)
2) Execute
      java -Xmx512m CreatePurgeScript
   A file is created in the project root directory name "purge.sh"
3) Copy this file to the following path of your fedora installation
     FEDORA_HOME/client/bin
4) Set the execution flags of the file
     chmod 755 purge.sh
5) Execute it
     ./purge.sh
   This might take a while.
6) Recache the repository
7) Reindex the repository
8) Have a drink