#! /usr/bin/env guile
!#

(use-modules (srfi srfi-1)
             (srfi srfi-26)
             (ice-9 getopt-long)
             (ice-9 match)
             (ice-9 regex)
             (transform strings)
             (transform sql)
             (transform triples)
             (transform special-forms))


;; One email ID in the Investigators table has spaces in it. This
;; function fixes that.
(define (fix-email-id email)
  (string-delete #\space email))

(define (investigator-attributes->id first-name last-name email)
  ;; There is just one record corresponding to "Evan Williams" which
  ;; does not have an email ID. To accommodate that record, we
  ;; construct the investigator ID from not just the email ID, but
  ;; also the first and the last names. It would be preferable to just
  ;; find Evan Williams' email ID and insert it into the database.
  (string->identifier "investigator"
                      (string-join
                       (list first-name last-name (fix-email-id email))
                       "_")
                      #:separator "_"))


(define-transformer investigators
  ;; There are a few duplicate entries. We group by email to
  ;; deduplicate.
  (tables (Investigators)
          "GROUP BY Email")
  (triples (investigator-attributes->id (field Investigators FirstName)
                                        (field Investigators LastName)
                                        "")
    (set rdf:type 'foaf:Person)
    (set foaf:name (string-append (field Investigators FirstName) " "
                                  (field Investigators LastName)))
    (set foaf:givenName
         (field Investigators FirstName))
    (set foaf:familyName
         (field Investigators LastName))
    (set foaf:homepage (field Investigators Url))
    (set v:adr (field Investigators Address))
    (set v:locality (field Investigators City))
    (set v:region (field Investigators State))
    (set v:postal-code (field Investigators ZipCode))
    (set v:country-name (field Investigators Country))))


(let* ((option-spec
        '((settings (single-char #\s) (value #t))
          (output (single-char #\o) (value #t))
          (documentation (single-char #\d) (value #t))))
       (options (getopt-long (command-line) option-spec))
       (settings (option-ref options 'settings #f))
       (output (option-ref options 'output #f))
       (documentation (option-ref options 'documentation #f))
       (%connection-settings
        (call-with-input-file settings
          read)))
  (with-documentation
   (name "Info files / Investigators Metadata")
   (connection %connection-settings)
   (table-metadata? #f)
   (prefixes
    '(
      ("dcat:" "<http://www.w3.org/ns/dcat#>")
      ("dct:" "<http://purl.org/dc/terms/>")
      ("foaf:" "<http://xmlns.com/foaf/0.1/#term_>")
      ("geoSeries:" "<http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=>")
      ("gn:" "<http://rdf.genenetwork.org/v1/id/>")
      ("gnc:" "<http://rdf.genenetwork.org/v1/category/>")
      ("gnt:" "<http://rdf.genenetwork.org/v1/term/>")
      ("owl:" "<http://www.w3.org/2002/07/owl#>")
      ("rdf:" "<http://www.w3.org/1999/02/22-rdf-syntax-ns#>")
      ("rdfs:" "<http://www.w3.org/2000/01/rdf-schema#>")
      ("skos:" "<http://www.w3.org/2004/02/skos/core#>")
      ("taxon:" "<http://purl.uniprot.org/taxonomy/>")
      ("v:" "<http://www.w3.org/2006/vcard/ns#>")
      ("xkos:" "<http://rdf-vocabulary.ddialliance.org/xkos#>")
      ("xsd:" "<http://www.w3.org/2001/XMLSchema#>")
      ))
   (inputs
    (list investigators))
   (outputs
    `(#:documentation ,documentation
      #:rdf ,output))))