BibDB

database layout

Heres a list of the existing tables, and what I think their fields mean...

• add_info: paper_id,name,value,ord,hidden,delimiter

  • I can't figure the difference between this and the bib_text table..

• author: id,firstname,middle,lastname,affiliation,homepage,email

  • The "Details about .." Page is generated by this Data. affilations means for who the author works (??can't this differ in several works of the same author?)

• bibtex: paper_id,entry_value,entry_name,ord,delimiter

  • Maybe to add unusual entries to the bibtex file of paper_id in the ordth line with a specific delimiter

• bibtypes id,text,bib

  • Every work is connect to its type by the id. The Type exists as human readable text or as its specified for bibTex

• main_data: paper_id,content_pdf,content_ps

  • connects a paper_id to large blobs calles content_pdf and content_ps

• paper:id,title,label,abstract,year,collection,bibtype,project,bibkey,url_pdf,url_ps,comment,updated,hidden,updated_pdf,updated_ps,bt_order

  • the main part of the database. label or bibkey seems a label used to reference to this work, bibtype references the bibtype table. bt_order might be the implementation of the Gerd_order, updated,hidden,updated_pdf,updated_ps are IMO used for triggering an updated of the static content

• paper_author: paper_id,author_id,ord_author

  • .connects a paper_id to multiple author_ids, ord_author determines the order

• paperkeys: paper_id, paperkey

  • .connects some of the paper_ids to a string paperkey like 'A Virtual Machine for Multi-Language Execution_2002_1_2', whatever it is. Dunno why this isnt a field in paper..

• to_delete: .. is empty and has just one field file, guess these are files marked for later deletion

Tasklist

Tasks done on the working system

1. provide a script to embed ones papers in his homepage http://www.ps.uni-saarland.de/Papers/my_publications_test.php?author=Name_des_Autors

   • The script to include is Papers/my_publications.php
     • beside this (wich should be done manually by someone who knows about the publication) the tasked is
2. remove infos about authors .. only provide search feature
3. mark publications with bibtype 9 (Master thesisis) according to their bibtex property
4. make static pages dynamic
   • to be honest, index.html is still created when the "update static view"-button is pressed, but no necessary. Since you still ne the update-static-view feature to create the files in abstract we cannot totally deactivate this button right now. This will be done on the test system

Todo on the working system

1. cleanup bibtex-entries (Guido)
2. remove html-markup in the following abstracts:
   • per id: 147, 241, 274, 277, 295, 297, 308 They will need some extra care, they make heavy use of htmltags

Tasks on the Test System

The work is to be done testwise in http://www.ps.uni-saarland.de/~robert/papers_new/

1. write scripts so that .bib, .pdf and .ps files are created from blobs in the database DONE

2. remove unnecessary staticview-functions (DONE) and merge show-functions.php and stativview-functions.php so their names make some sense

3. unify cite-keys
   • Question: When I unify the cite-keys with a specific function and so alter the old keys too, do I have to tell every one so they keep their bibtex references up-to-date?
4. alter database layout to reflect the changes above, before that: update the mysqldb from the working database
5. change pdf/ps file handling to something like an attachment handling
6. repair collections feature
7. wizard gui for entering/editing entries
8. code documentation