We assume that an editor wants to input new data to the ISMI database on the basis of a certain manuscript source. Here is the procedure that would be normally followed once all the data has been gathered and secured (i.e. the authorship and identification of the text has been made, and all the codicological data interpreted). This should be done on the basis of the two data sheets distributed by Sally Ragep.

NB: In the following bold case is used to refer to the abstract object entities in the database

1. The user interface (swing or web client) should be in the modus “input new manuscript source”.
2. Look whether the author is already present as a person object in the database.
   • A control mechanism should be implemented to minimize the risk of entering one person twice under a different alias.
     • For example if one types “Isfahani” or “Iṣfahānī” or “اصفهاني” all person objects (and associated aliases) that match (or are very similar) to those strings would be presented as suggestions to the data editor.¹⁾ Each suggestion would also provide minimal biographical info (place and dates) if available, in order to avoid the converse problem of merging two different persons who happen to share the same name.

1. If it is not present, add it (e.g. in a separate window).
2. Identify the author by linking to an appropriate person.
3. Same thing (steps 2-4 above) for text.
   • date of composition
   • place of composition
   • language of text
   • subject of text (hay'a, zīj, mīqāt, astrolabe, etc)
     • → a canonical list of subjects should be present in the database
   • type/genre of text (matn, sharḥ, taḥrīr, …)
   • format/structure of text (number of books, parts, chapters, sections)
   • Table of contents
   • dedication
   • incipit
   • explicit
   • information whether the work has been published, translated, studied, with refs to literature db
4. Same for library:
   • name of library
   • city/country
5. collection:
   • name of coll.
6. codex:
   • Usually the codex will not be present in the db, unless it is a collection of several works.
   • codex number.
   • number in codex
   • foliation/pagination
   • has codex been microfilmed? (if yes give ref number)
   • has codex been digitized? (if yes give ref number)
   • should microfilm/scans be ordered?
     • information about contact person at library/institution, with email, purchase price, how long it will take to order and mail
7. Enter all codicological details
   1. on codex
      • dimensions
      • material
      • distinguishing features (e.g. presentation copy, decoration)
      • table of contents (with titles and folio numbers)
      • other information (including language(s) used in codex)
      • owner(s)
      • patron(s)
      • reader(s)
        • for the above three: locations and dates if available
   2. on manuscript copy
      • folios
      • lines per page
      • script
        • include comments on legibility, ink, etc
      • distinguishing features (e.g. state of preservation)
      • diagrams
        • include number of diagrams, folios, color of ink, brief description
      • tables
        • include number of tables if possible, folios, color of ink, brief description
   3. on content of manuscript
      • Copyist
      • autograph?
      • Place of copy
      • Date of copy
      • Dedication of copy
      • Colophon
      • Other information
        • content
        • passages
        • marginal notes
        • if studied
        • annotations
        • etc

• To be determined…

• The database should keep information on editors:
  • Name, City/Country
  • Date/time work began (for data collection)
  • Date/time work ended