Metacat

Remove the organization configuration stuff for the moment. Add geoserver configuration.

Allow script to be run from command line, more token changes including all path changes for URLs and directories.

use context url to get cgi url

Change discoverTomcatDir to discoverDeployDir

Fix error where maxconnectionnumber was being read into maxconnectionage.

Create data and inline data directories

Consolidation of property names in register-dataset.cgi; returned runConfiguration to default true

Fully qualify spatial and replication properties

Qualify harvester properties

Replace tomcatDir with deployDir

Use qualified properties

use the docid, not the filename when updating this node.
fixes bug: http://bugzilla.ecoinformatics.org/show_bug.cgi?id=3260

include leading slash for path to the spatial files (context is given without a trailing slash)

initial doc id list from original conversion

script used to update keywords for existing kruger documents when upgrade to new SANParks/SAEON skin. committing before additional work to update the package ids for these files (incorrectly set to the wrong rev number)

Add support for separate LDAP and organization level configurations

Fix the db script suffix.

Add functionality to allow developer to bypass configuration utility

FIx comments for getContextDir to reflect that we no long append webapps to the path

Add sql debug statements

Use SystemUtil.getContextDir() instead of constructing directory path.

fixed bug where the searchmode param was put into the search fields

Differentiate between tomcat directory and deployment directory.

Replace Options class with PropertyService

Create PropertyService using log directory instead of servlet context.

Append context url onto system id instead of server url.

Move references to sql scripts into the WEB-INF directory.

Add method to return a map of db script suffixes from metacat.properties

Fix backup file pathing issues. Modify checkAndSetProperty to get skin specific values from request.

implement getPropertiesByGroup() method

Implement form submittal processing.

use metacat.properties to determine script suffixes

replace <script>.sql files with <script>-oracle.sql files since the base files are implicitly oracle scripts

Change default-style property to application.default-style

Add addProperty and addPropertyNoPersist for new properties

Backup properties should use addProperty, not setProperty

Add GeneralPropertyException to property setters

Merge 1.9 changes into Head

fixed bug 3403

Merge METACAT_1_8_1 to cvs head.

added support to upload an xml document using multipart encoding. the enables better support for web interfaces that wish to use multipart to upload xml documents. the action is 'insertmultipart'

made it so that the error and success messages from a multipart upload can be transformed using xslt. also added upload functionality to the sms skin

Check into cvs head. Comment out an obsoleted code.

Check into CVS head. Add code to handle that referral server is down.

Check into cvs head. Remove "nodedata='document'" in xml_nodes and xml_nodes_revisions deleting query.

Check into cvs head. Modify the query of deleting xml_nodes_revisions table.

Check into cvs head. Add a debug statement for printing out eml document

Check into cvs head. In handle replication list method, parser will be intialized every time when it handles a list from another server.

Check into head. Add a new method to get server name.

Check in to cvs head. Don't close db connection before returning the connection pool.

Add to cvs headd. Add code to correct invalidated eml 201 doucments.

Add to cvs head. Change the run method signature.

Checkin to head. Remove this class from Runnable and add sql to delete extral nodes in xml_nodes_revision table.

Modify the delete xml_index table query.

Add code to execute sql command and finish the task.

A class to correct invalid eml201 documents. Metacat 1.8.0 and ealier version points incorrect eml-resource.xsd. We need fix eml documents.
Haven't done yet.

Add a new class will build query result able in advance. It has been done yet.

Fixed a bug pathexpr was modified if pathexpr is an attribute.

Changes to remove ANT tokens for the turnkey install project

Enable file naming when returning a zip file of the requested documents.
File names for each docid are provided as additional parameters named by the actual docid value that they are providing a name for. An example query string might look like:

action=read&docid=doc.1.1&docid=doc.2.1&qformat=zip&doc.1.1=fileName1.txt&doc.2.1=fileName2.jpg...

Fixed bug that lsid couldn't be shown up in result page of kepler skin. The bug is that in extended query, lsid attribute was treated as a element, but it is attribute.
I also optimized the query by deleting xml_documents table from query. The table is not needed.

Fixed bug that kepler skin wouldn't return actor name in query. The reason is there was no separate vector to store returnfield with "]". So those fields with or without "]" are handled
together (which shouldn't). When seperate those return fields, the bug was fixed.

Include the contextURL as a parameter for the transformer. Remove leading / from resultset paths

Add a new string buffer to store normalized query xml itself.

Fixes to record sies of ecogrid files, and to flag html files.

Don't transfer character which is greater than 128 or less thank 32 to numerical entity symbol.

Modified the normalize subroutine - don't use 26 (url encoding) to replace x%x

Using reponse.getOutputStream to replace reponse.getWriter in query and read method. So metacat can send correct special character back.

Modify normolize method by adding code to handle < > ' and "

Remove an obsolete class.

Remove an obsolete private method getPartNodeRecordList.

Remove an obsolete class.

Add new code for xml_nodes table.

Some script to generate text search vector.

Add timeout parameter for searching user name. If some exception happens, it wouldnot throw it, just print it out.

if there is no clientViewBean in the session, insert a new one - saves us from nullPointerException when using this in skins that may not be aware of the bean uses.

ignore commented out DOCTYPE declarations (don't force validation on them)

add debug statements so it's clear what transformer parameters are being passed

Use dev=1 replace dev like '1'. since postgres 8.3 doesn't support it.

include contextURL and cgi-prefix as parameters in all the transforms (initially used for nceas skin to remove ant tokens in the XSL files....but will be used by other skins as they are converted away from at tokens)

Add debug info for special charater.

Remove the relation query.

Add server url to system id.

Skeleton for data insertion script. Also added documentation to Metacat
library.

MOdified query passing, and default cache location.

Initial script for downloading and caching data entities associated
with an EML document that is stored in Metacat.

Improved the javadoc documentation for the upload() methods to clarify what upload() does and contrast it to insert() and update().

prevent null pointer exception when there are no nodes found (just an empty list)

Modification for changes to the field "MetacatUtil.skinconfigs" (bug 3057 fix).
Also changes reading skin.configs properties/files (bug 3058 fix).

Modification for changes to the field "MetacatUtil.skinconfigs".

Changes to the field "skinconfigs":
1) Made it private
2) Created setter and getter methods
3) Created a 'hasSkinConfig' method

fix mistake where there is an attempt to select nodetype column from table xml_path_index table (it's not there!)

changes for fix to bug #2999 (only one attribute of an element is returned in resultset doc)
-added merging for large TEXT nodes when they have been broken up.
-allow multiple attributes to be returned in the resultset doc (uses newly-available field type to determine text vs attribute node).

add retrieval of field type for each node (TEXT, ATTRIBUTE, etc).
this is to support a fix for bug#2999

Workaround for a bizzare submission issue stemming from the debug sub. Modified the NCEAS project list to be robust and rely on the presence of the AdminDB module and a working connection to the DB for listing projects.

Fixed a couple of bugs: 1) Threw an error when updating a packages public access (lost the data files doc id's) and 2) returns a message after access is changed.
Both were fixed by not attempting to get a child text node from a text node.

This file was renamed from upgrade-db-to-1.7.1.sql

Rename this file to upgrade-db-to-1.8.sql

A file rename from 1.7.1 since we decide to release 1.8 rather than 1.7.1

Rename this file to upgrade-db-to-1.8-postgres.sql