added some params to the loop that segments the listObjects resultset
allow any nonquote characters in the schemaLocation attribute, otherwise the whole line is matched and this prevents schemaLocation from appearing anywhere but at the end of the line.
finished the crud part of listObjects. need to implement it in the rest interface now.
implementing getObjects in CrudService.
listObjects is almost working. need to get the result set parsed when returned to CrudService.
setting up the framework for the listObjects api call
changed the detailCode for DataOne exceptions to a string
added getSystemMetadata to ResourceHandler and changed the url format to be meta/guid. added a new servlet reponse handler in the web.xml file to handle the new urls and send them to the RestServlet.
implemented update in ResourceHandler and fixed a bug in CrudService where data files were not handled correctly on update
amost have update working. still need to get unit test squared away.
added exception testing to the crud test
getting getSystemMetadata to work
refactored XMLSchemaService to not have static methods. made the CrudServiceTest more robust.
fixed schema location bug. the dataone schemas are now correclty found
removed CrudService dependency on servlet params. CrudService is now a singleton. I'm getting an error from metacat saying it can't find teh systemmetadata schema, even though it is, in fact, registered with metacat. need to identify why this is happening.
refactoring to remove duplicate code
adding a system to track system metadata documents for dataone
Added initial implementation of getSystemMetadata and its associated REST service. Current implementation is returning a hardcoded system metadata document -- need to look up the real document for each guid and return that.
Completed main parts of CrudService.create(). Now the method writes both data and metadata objects along with their system metadata, and handles the mapping between global identifiers and local identifiers. Access control and logging still need to be dealt with to properly authenticate and then log activities.
Modify CrudService to write SystemMetadata to disk with an autogenerated localId and an autogenerated GUID. Validation depends on the DataONE schemas being setup in xml_catalog properly, so still need to check upgrade scripts to be sure these new schemas are added. Still need to handle the metadata document insert, but should be same as system metadata insert.
Added comment about MIME form processing.
Refactored schema service to call doRefresh() in constructor to eliminate code duplication.
Added note on TODO for access bug.
replication of guids now works. tested this for both forced replication and update/insert/delete triggered replication
fixed bug where guid end tag wasn't getting printed
added a method in IdentifierManager to get a guid from a docid and rev. added fields in the documentinfo replication document to pass the guid. now need to handle the guid and insert it into the table if its found
Refactored metacat to use dataone-service-api-0.3.0 jar file, with the new JIBX generated types and serialization. Included JIBX runtime jar in metacat to handle serialization and deserialization of SystemMetadata objects. Removed the old API jar file.
Modifications to support the DataONE service API version 0.1.0. For DataONE, the get() andcreate() services are partially complete. Several more functions and checks need to be added tocreate() before it is viable. This DataONE support is not complete, and the current support breaks the MetacatRestClientTest for the time being (this client will eventually be removed).
Merge 1.9.2 changes back into the trunk
Modified ResourceHandler to return error codes in XML format for the get() method.
Modifications to metacat rest service to use the new DataONE CrudService forget() operation. Minor issue with permissions errors propagating up thestack, but otherwise this is the first implementation of a full DataONEmethod.
Modified readFromMetacat() to pass most exceptions up the call stack, which allows creation of new entry points for calling reads. Still need tocontinue factoring out the HTTPServletResponse that is passed in in order tomake entrypoints that are not servlet based possible. Problem now is in...
when looking up group membership, skip referrals that have errorshttp://bugzilla.ecoinformatics.org/show_bug.cgi?id=4900
skip referrals that have errors (connection timeout, host down, etc)http://bugzilla.ecoinformatics.org/show_bug.cgi?id=4900
Refactored putObject method to separate the create() and update() portions in order to match the method signatures needed for DataONE.
Modified ResourceHandler.deleteObject() to take a globally unique id (guid)as input rather than a metacat-conformant docid. The guid is translatedinto a metacat docid which is then used in the handleDelete() call. So nowMetacat supports use of arbitrary identifiers for delete operations, but...
Added support to ResourceHandler to allow the putObject method to usearbitrary guid strings as input. These strings are examined, and if theymatch the Metacat docid format (scope.id.rev), they are used directly. Ifthe id is a string in another format, a new localId is generated based on...
Fixed bug in getNextRevision() where "error" documents were not beingdetected properly before searching for the next identifier.
Added initial support for arbitrary string identifiers in metacat (referred to as a guid). The new identifier table is used to map arbitrary string identifiers to Metacat's current docid format (referred to as the localId). Added a new IdentifierManager class to manage this table, adding new mappings as objects are added to the system. Modified the MetacatTest service to utilize this mapping table to look up a localId from a guid. IdentiferManagerTest is working with these guids now, as is the MetacatTest service get() method, but other parts of the system are unaware of them (e.g., create, update, delete operations are unaware, and no mapping is created when new documents are created). As a consequence, the MetacatRestClientTest is not working (although it also had lots of hardcoded dependencies that need to be fixed as well).
Added copyright, license, and class documentation.
Modified ResourceHandler to change the GET url from /objects to /object tomatch the DataONE method name.
Merging in REST interface implementation that was created by Serhan Akin. Main change isa refactored MetaCatServlet.java class, in which all of the handle* methods were movedinto a separate MetacatHandler.java class. This allows both the standard MetaCatServlet...
Adding missing brackets in else clause. Need to verify with Daigle that this is what he meant.
Move the chunking of large test element data to centralized location in DBSAXNode.writeChildNodeToDB(). Beff up logging
Beef up log messages
Add debug log statements
make sure querygroup is not null before trying to print it.
Get the latest revision from metacat when a modify call docid does not have a revision number.
Send appropriate message when trying to update a docid that was previously deleted.
Reformat file. No code change.
do a quick retry if building path index fails with a SQL error
Make sure buildIndex throws an exception if it has a sqlexception. That way the indexing object will be added to the indexing queue and reprocessed.
Pass the doc xml as a string to docImpl.write and writeRepication. This is so a reader can be create for the parsing and for the write to disk. Also created a db access class for xml query result deletion.
make constants final
Fix the bug of http://bugzilla.ecoinformatics.org/show_bug.cgi?id=4645. handleGetRevisionAndDocTypeAction will search both xml_documents and xml_revisions table.It also changed some constrain in AccessionNumber when user update a document, of which all previous versions are in xml_revisions table.
Add a sleep statement between getting the dtd from source and checking for it via url. Add method names to log output.
Log doc and rev query counts and times. Fix mis-spellings.
remove call to print busy db connections
Add debug to show sql
Fix for Bug #4637 - Metacat Harvester fails to catch some insert and update failures. As per comments in the bug entry, the Metacat Harvester logic has been modified to examine the Metacat client return string to confirm that it contains the substring "<success>" following an insert or update operation. It no longer considers just the absence of an exception as indicative of a successful operation.
log connection warnings only when thresholds are exceeded. Add method names to log output
add method name to log statements
fix spelling of word additional
add method name to log messages and create configured thresholds for warning messages
Change log levels where appropriate and add class/method name to output.
Change warn messages to info
beef up error log messages
Make sure session data exists before extracting information
Separate code to do tls and non-tls authentication. Introduce AuthTLSException to make error handling easier.
Beefed up the logging
Change add sql to use a prepared statement. Only try to download a cert if a url was provided.
change unregister error message to be an info message
Fix errors in workflow scheduler unit test
Do not register the session service
special handling for RDF namespace documents (semtools project)
add connection serial number where it was missing.
Add ability to schedule daily, weekly and monthly jobs.
Format and add comments.
add getaccesscontrol and setaccess api calls
Handle the case of trying to insert a conflicting permission order for a document
move dbconnection object out of base class and create it for each db access so it can be released
Add success message to metacat response for setaccess action
handle multiple <principal> elements in the allow/deny blocks
Verify we are in the permission and principal sections before setting values on dao object.
Use OR to add permissions, not AND
Change access section in getaccesscontrol and getdocumentinfo apis to be more eml 2.1.0 compliant. Add a block access option to setaccess api.
change AccessControlForSingleFile to only be instantiated for one file. move ACL methods to AccessControlForSingleFile. Change format of access sections returned to EML 2.1.0.
Removed some old commented out code
Remove unneeded references to Timer
SMS-related addition: retrieve docids for a given doctype
QueryTerm.java will produce an incorrect SQL statement when there is an attribute in the XPath being used to search on. Regardless of the presence of the attribute, the clause:
(SELECT nodeid FROM xml_index WHERE path LIKE " + "'" + path + "') "
would be added to the query....
Create access dao and centralized db access classes. Update create access code to combine access records for principal/doc/accesstype when multiple exist.
Move access control source to it's own directory.