Difference between revisions of "Technical problems encountered during the application development"
From Biowikifarm Metawiki
(3 intermediate revisions by the same user not shown) | |||
Line 1: | Line 1: | ||
− | Some technical problems | + | Some technical problems have been successfully overfulfilled during the application development: |
+ | |||
+ | * Respect the special requirement from WP4 leader, dr. Gregor Hagedorn to not use more the 512MB memory for ingest application. This requirement force the XSLT validation tool and FOXML generator tool to adopt a low-memory strategy in development and re-thinking of the initial approach. | ||
+ | * Soap API-A exposed by Fedora Commons with Search service, doesn't allow " ' " (apostrophe) for a search string. | ||
+ | * Mapping between work database and temporary working folders for collection, specially on Windows. A MD5 algorithm have been used for mappings. | ||
+ | * Mapping between physically metadata and logical level required by Fedora Commons Content Model Architecture. | ||
+ | * Problems with old collection retrieving. The application try 3 kind of search in order to decide if a collection already exists or not in repository. We used: Fedora search service, GSearch service and RiSearch service. | ||
+ | |||
− | |||
− | |||
− | |||
− | |||
(Return to [[Implementation_details_of_Java-based_Fedora_Ingestion_code/Technical_documentation]]) | (Return to [[Implementation_details_of_Java-based_Fedora_Ingestion_code/Technical_documentation]]) |
Latest revision as of 18:35, 27 September 2009
Some technical problems have been successfully overfulfilled during the application development:
- Respect the special requirement from WP4 leader, dr. Gregor Hagedorn to not use more the 512MB memory for ingest application. This requirement force the XSLT validation tool and FOXML generator tool to adopt a low-memory strategy in development and re-thinking of the initial approach.
- Soap API-A exposed by Fedora Commons with Search service, doesn't allow " ' " (apostrophe) for a search string.
- Mapping between work database and temporary working folders for collection, specially on Windows. A MD5 algorithm have been used for mappings.
- Mapping between physically metadata and logical level required by Fedora Commons Content Model Architecture.
- Problems with old collection retrieving. The application try 3 kind of search in order to decide if a collection already exists or not in repository. We used: Fedora search service, GSearch service and RiSearch service.
(Return to Implementation_details_of_Java-based_Fedora_Ingestion_code/Technical_documentation)