Global Performing Arts Database Paul Houle Library Systems

Global Performing Arts Database Paul Houle Library Systems Cornell University Library Thanks To James Reidy Joshua Young Chip Goines Susan Specter Karen Brazell Ron Rice Derek Messie

www. glopad. org

Vietnamese Water Puppets

Gertrude Stein Repertory Theatre

Circus Arts

Meyerhold Theatre

www. glopac. org

Individuals Museums PA Organizations University Departments GLOPAD JPARC Teen. PARC …

Three Generations of Glo. PAD Filemaker JSP / Postgre. SQL PHP / Postgre. SQL

digdoc component source object performance performing art production performing arts group piece person collection place date part of body function instrument role named genre era language frame rate country sound field type The Glopad Galaxy

Component Types (25 of 106)

Multilingual text strategies • Multiple columns in one table (title_en, title_ja, title_ru, …) • Multiple tables with hardcoded names (d_digdoc_en) • Satellite table (d_digdoc_ml) with (language_id) column • RDF store (see http: //vivo. library. cornell. edu/)

rd Tools used for 3 generation

Finding the text Digdoc 2 Title: Notes: Country: Japan Performance 2 Production Notes: Name: Piece Title: Digdoc Title: Notes: Performance Place Notes: Name: Component Name: Material: Notes: Function Person Function: Author Title: Notes: Component_type Type: Costume Country 2 Place: Country: United Kingdom Name:

Full-text system: objects glo_searcher glo_search_indexer glo_path_builder glo_path glo_node glo_digdoc_policy glo_traversal_policy glo_friends_policy glo_source_object_policy glo_component_policy glo_table glo_link glo_direct_link glo_performance_policy glo_indirect_link

Move to production • • • Standardize file organization Reorganize CVS repository Ability to run multiple instances Develop db dump & restore procedure Fix showstopper bugs Friendly URLs

Standard File Layout for Virtual Hosts /arts-lib/ /sites/ /www. glopad. org/ /bin /conf /docs /htdocs /lib /logs /templates /var

Friendly URLs http: //www. glopad. org/pi/record/digdoc/2122 http: //www. glopad. org/pi/full_record. php? obj=digdoc&id=2122 Based on Apache URL rewriting Rewrite. Engine on Rewrite. Rule ^/$ /pi/ [R] Rewrite. Rule ^/ei/$ /ei/php [R] Rewrite. Rule ^/pi/record/(digdoc)/([0 -9]*)$ /pi/full_record. php? obj=$1&id=$2

Clone tool Bulk Upload glopad_batch. txt upload. zip

Project Initiation Plan • Change History • Project Purpose • Success criteria/Requirements • Scope and Dependencies • Key Deliverables and Time Frames • Stakeholders • Resources • Risks • Communication Plan

Requirements format (1) Facility (i) will involve adding a button on the bottom of the digital object record screen that will clone an object. (2) The cloning process will create a copy of the metadata records of the object, but will not copy the associated media file(s); it will copy the contents of the d_digdoc row and any rows in the r_ tables associated with other fields (other than performance and component) (3) The user will be left at the object edit screen for the new object, where they have the option of uploading the files for the new DO and of editing selected fields (4) The user will have the option of doing a “deep copy” of the associated Performance and Component records – this will copy the fields contained in the d_performance and d_component records, along with any r_ table rows, and create the r_table records linking the new Digdoc to the new Performance and Component records. (5) If the user chooses not to “deep copy” these records, there will be no linked Performance and Component records. (6) The source object record associated with a digital object will be shallow copied: new digital objects will be linked to the same source object as the template digital object

Release 3. 1 HEAD Paul Houle Bulk Upload Jim Reidy Google Sitemaps branch: release_3_1 Clone tool Tag: release_3_1_0 branch: release_3_0 Move to production Tag: release_3_0_0

$ cvs commit –m “ 0000647: give english error message when user forgets to select file” upload. php

Status Report The following items from the PIP are not yet implemented: . (16) Access to the batch upload facility will be controlled by Apache access control (17 e) Check for files not listed in glopad_batch. txt (24) Shallow copy Source Object records. (29) A note will be inserted into the d_digdoc. notes field documenting the bulk import. I am proposing that we remove the following items: (27) All new records will be tagged with a unique bulk upload id in a new field of the d_digdoc record. (28) The new field will be indexed by the full-text search, so that editors will be able to pull out all of the records linked to the digdoc. And add one more item: (31) Create a cron job that removes old files left by the batch upload process

Release 3. 2 HEAD branch: release_3_2 Bulk upload Tag: release_3_2_0 Tag: release_3_1_1 Google Sitemaps branch: release_3_1 Clone tool Tag: release_3_1_0 branch: release_3_0 Move to production Tag: release_3_0_0

Global Performing Arts Database Any Questions? Image credits: Meyerhold Theatre, Russia; San Francisco Performing Arts Library & Museum; Thang Long Water Puppet Theatre of Hanoi, Vinh Qui ; Gertrude Stein Repertory Theatre