@prefix rdf:           <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix rdfs:          <http://www.w3.org/2000/01/rdf-schema#> .
@prefix xsd:           <http://www.w3.org/2001/XMLSchema#> .
@prefix owl:           <http://www.w3.org/2002/07/owl#> .
@prefix skos:          <http://www.w3.org/2004/02/skos/core#> .
@prefix prov:          <http://www.w3.org/ns/prov#> .
@prefix sio:           <http://semanticscience.org/resource/> .
@prefix dcat:          <http://www.w3.org/ns/dcat#> .
@prefix dcterms:       <http://purl.org/dc/terms/> .
@prefix void:          <http://rdfs.org/ns/void#> .
@prefix foaf:          <http://xmlns.com/foaf/0.1/> .
@prefix ov:            <http://open.vocab.org/terms/> .
@prefix setl:          <http://purl.org/twc/vocab/setl/> .
@prefix csvw:          <http://www.w3.org/ns/csvw#> .
@prefix pv:            <http://purl.org/net/provenance/ns#>.

@prefix :              <http://vocab.rpi.edu/whyis/setl/bibtex/> .

@prefix whyis: <http://vocab.rpi.edu/whyis/>.

:bibtex_input a <https://www.iana.org/assignments/media-types/text/plain>;
  prov:wasGeneratedBy [
    a setl:Extract;
    prov:used  [ a <https://www.iana.org/assignments/media-types/application/x-bibtex>]
  ].

:bibtex_entries a owl:Class, prov:SoftwareAgent, setl:PythonScript;
  rdfs:subClassOf prov:Activity;
  prov:qualifiedDerivation [
    a prov:Derivation;
    prov:entity :bibtex_input;
    prov:hadRole [ dcterms:identifier "bibtex_file"]
  ];
  prov:value '''
import bibtexparser

bib_database = bibtexparser.load(bibtex_file)

result = enumerate(bib_database.entries)
'''.

:create_article_uri a owl:Class, prov:SoftwareAgent, setl:PythonScript;
  rdfs:subClassOf prov:Activity;
  prov:value '''
import rdflib
global rdflib

global base_uri
collection = this.graph.value(predicate=rdflib.RDF.type,
                            object=rdflib.URIRef("https://www.iana.org/assignments/media-types/application/x-bibtex"))
base_uri = rdflib.Namespace(collection+'/')
global doi
doi = rdflib.Namespace("http://dx.doi.org/")

def result(entry):
    if 'doi' in entry:
        return doi[entry['doi']]
    else:
        return base_uri[entry['ID']]
result.base_uri = base_uri
result.collection = collection

'''.

:author_fn_ln a owl:Class, prov:SoftwareAgent, setl:PythonScript;
  rdfs:subClassOf prov:Activity;
  prov:value '''
import re
global re

def result(author):
    if ',' in author:
        ln, fn = re.split('\\\\s*,\\\\s*', author)
    else:
        names = re.split('\\\\s+', author)
        fn = names[0]
        ln = ' '.join(names[1:])
    return fn, ln
'''.

:create_date a owl:Class, prov:SoftwareAgent, setl:PythonScript;
  rdfs:subClassOf prov:Activity;
  prov:value '''
import re
global re

def result(row):
    date = row['year']
    months = ['jan','feb','mar','apr','may','jun','jul','aug','sep','oct','nov','dec']
    if 'month' in row:
        for i, m in enumerate(months):
            if m in row['month'].lower():
                return date + '-'+str(i+1)
    return date
'''.


:bib_resource_types a owl:Class, prov:SoftwareAgent, setl:PythonScript;
  rdfs:subClassOf prov:Activity;
  prov:value '''
result = {
    'article' : ["dcterms:BibliographicResource",'bibo:AcademicArticle'],
    'book' : ["dcterms:BibliographicResource",'bibo:Book'],
    'booklet' : ["dcterms:BibliographicResource",'http://purl.org/net/nknouf/ns/bibtex#Booklet'],
    'inbook' : ["dcterms:BibliographicResource",'bibo:Chapter'],
    'incollection' : ["dcterms:BibliographicResource",'bibo:BookSection'],
    'inproceedings' : ["dcterms:BibliographicResource","http://vivoweb.org/ontology/core#ConferencePaper"],
    'manual' : ["dcterms:BibliographicResource","bibo:Manual"],
    'mastersthesis' : ["dcterms:BibliographicResource",'http://purl.org/net/nknouf/ns/bibtex#Mastersthesis', 'bibo:Thesis'],
    'misc' : ["dcterms:BibliographicResource"],
    'phdthesis' :  ["dcterms:BibliographicResource",'http://purl.org/net/nknouf/ns/bibtex#Phdthesis', 'bibo:Thesis'],
    'proceedings' :  ["dcterms:BibliographicResource",'bibo:Proceedings'],
    'techreport' :  ["dcterms:BibliographicResource",'http://purl.org/spar/fabio/TechnicalReport','bibo:Report'],
    'unpublished' :  ["dcterms:BibliographicResource",'bibo:Manuscript'],
}
'''.

:BibTeXSETL rdfs:subClassOf setl:SemanticETLScript;
  setl:hasTemplatePrefix "http://vocab.rpi.edu/whyis/setl/bibtex/";
  rdfs:subClassOf [ a owl:Restriction;
    owl:onProperty prov:used;
    owl:someValuesFrom <https://www.iana.org/assignments/media-types/application/x-bibtex>
  ].

:bibtex_collection a void:Dataset, dcat:Dataset;
  prov:wasGeneratedBy [a :BibTeXSETL], [
    a setl:Transform, setl:JSLDT;
    prov:used :bibtex_entries;
    prov:qualifiedUsage [ a prov:Usage; prov:entity :create_article_uri; prov:hadRole [ dcterms:identifier "create_article_uri"]];
    prov:qualifiedUsage [ a prov:Usage; prov:entity :bib_resource_types; prov:hadRole [ dcterms:identifier "bib_resource_types"]];
    prov:qualifiedUsage [ a prov:Usage; prov:entity :author_fn_ln; prov:hadRole [ dcterms:identifier "author_fn_ln"]];
    prov:qualifiedUsage [ a prov:Usage; prov:entity :create_date; prov:hadRole [ dcterms:identifier "create_date"]];
    setl:hasContext '''{
  "@vocab" :  "http://hadatac.org/ont/chear#",
  "dataset" : "https://hbgd.tw.rpi.edu/dataset/",
  "sio" :     "http://semanticscience.org/resource/",
  "skos" :    "http://www.w3.org/2004/02/skos/core#",
  "dcterms" :    "http://purl.org/dc/terms/",
  "prov" :    "http://www.w3.org/ns/prov#",
  "bibo" :    "http://purl.org/ontology/bibo/",
  "rdfs" :    "http://www.w3.org/2000/01/rdf-schema#",
  "foaf" :    "http://xmlns.com/foaf/0.1/",

  "void" :    "http://rdfs.org/ns/void#"
}''';
  prov:value '''[{
  "@context" : {
    "authorList" : {"@id" : "bibo:authorList", "@type": "@list"}
  },
  "@id" : "{{create_article_uri(row)}}",
  "@type" : [ { "@for" : "t in bib_resource_types[row['ENTRYTYPE']]", "@do" : "{{t}}"}],
  "dcterms:creator" : [
    {
        "@if" : "'author' in row",
        "@for" : "author in re.split('\\\\s*and\\\\s*', row['author'])",
        "@do" : {
          "@with" : "author_fn_ln(author) as fn, ln",
          "@do" : {
            "@id" : "{{create_article_uri.base_uri}}author/{{slugify(' '.join([fn, ln]))}}",
            "@type" : "foaf:Person",
            "foaf:familyName" : "{{ln}}",
            "foaf:givenName" : "{{fn}}",
            "foaf:name" : "{{fn}} {{ln}}"
          }
        }
    }
  ],
  "authorList" : [
    {
        "@if" : "'author' in row",
        "@for" : "author in re.split('\\\\s*and\\\\s*', row['author'])",
        "@do" : {
          "@with" : "author_fn_ln(author) as fn, ln",
          "@do" : {
            "@id" : "{{create_article_uri.base_uri}}author/{{slugify(' '.join([fn, ln]))}}"
          }
        }
    }
  ],
  "dcterms:abstract" : [
    {
      "@if" : "'abstract' in row",
      "@value" : "{{row['abstract']}}"
    }
  ],
  "skos:editorialNote" : [
    {
      "@if" : "'annote' in row",
      "@value" : "{{row['annote']}}"
    },
    {
      "@if" : "'note' in row",
      "@value" : "{{row['note']}}"
    }
  ],
  "dcterms:title" : [
    {
      "@if" : "'title' in row",
      "@value" : "{{row['title']}}"
    }
  ],
  "dcterms:publisher" : [
    {
      "@if" : "'publisher' in row",
      "@id" : "{{create_article_uri.base_uri}}publisher/{{slugify(row['publisher'])}}",
      "rdfs:label" : "{{row['publisher']}}",
      "http://schemas.talis.com/2005/address/schema#localityName" : [
        {
          "@if" : "'address' in row",
          "@value" : "{{row['address']}}"
        }
      ]
    },
    {
      "@if" : "'institution' in row",
      "@id" : "{{create_article_uri.base_uri}}institution/{{slugify(row['institution'])}}",
      "rdfs:label" : "{{row['institution']}}",
      "http://schemas.talis.com/2005/address/schema#localityName" : [
        {
          "@if" : "'address' in row",
          "@value" : "{{row['address']}}"
        }
      ]
    },
    {
      "@if" : "'school' in row",
      "@id" : "{{create_article_uri.base_uri}}institution/{{slugify(row['school'])}}",
      "rdfs:label" : "{{row['school']}}",
      "http://schemas.talis.com/2005/address/schema#localityName" : [
        {
          "@if" : "'address' in row",
          "@value" : "{{row['address']}}"
        }
      ]
    }
  ],
  "dcterms:date" : [
    {
      "@if" : "'year' in row",
      "@value" : "{{create_date(row)}}",
      "@type" : "xsd:dateTime"
    }
  ],
  "http://prismstandard.org/namespaces/basic/2.1/doi" : [
    {
      "@if" : "'doi' in row",
      "@value" : "{{row['doi']}}"
    }
  ],
  "dcterms:identifier" : [
    {
      "@if" : "'doi' in row",
      "@value" : "{{row['doi']}}"
    }
  ],
  "http://prismstandard.org/namespaces/basic/2.1/startingPage" : [
    {
      "@if" : "'pages' in row",
      "@value" : "{{row['pages'].split('-')[0]}}"
    }
  ],
  "bibo:pageStart" : [
    {
      "@if" : "'pages' in row",
      "@value" : "{{row['pages'].split('-')[0]}}"
    }
  ],
  "http://prismstandard.org/namespaces/basic/2.1/endingPage" : [
    {
      "@if" : "'pages' in row",
      "@value" : "{{row['pages'].split('-')[-1]}}"
    }
  ],
  "bibo:pageEnd" : [
    {
      "@if" : "'pages' in row",
      "@value" : "{{row['pages'].split('-')[-1]}}"
    }
  ],
  "http://prismstandard.org/namespaces/basic/2.1/volume" : [
    {
      "@if" : "'volume' in row",
      "@value" : "{{row['volume']}}"
    }
  ],
  "bibo:chapter" : [
    {
      "@if" : "'chapter' in row",
      "@value" : "{{row['chapter']}}"
    }
  ],
  "@reverse" : {
    "dcterms:hasPart" : {
      "@id" : "{{create_article_uri.collection}}"
    }
  },
  "rdfs:seeAlso" : [
    {
      "@if" : "'url' in row",
      "@id" : "{{row['url']}}"
    }
  ],
  "http://www.w3.org/ns/dcat#downloadURL" : [
    {
      "@if" : "'url' in row",
      "@id" : "{{row['url']}}"
    }
  ],
  "dcterms:isPartOf" : [
    {
      "@id" : "{{create_article_uri.collection}}",
      "@type" : "http://purl.org/dc/dcmitype/Collection"
    },
    {
      "@if" : "'journal' in row",
      "@id" : "{{create_article_uri.base_uri}}journal/{{slugify(row['journal'])}}",
      "@type" : "bibo:Journal",
      "dcterms:title" : "{{row['journal']}}",
      "dcterms:publisher" : [
        {
          "@if" : "'publisher' in row",
          "@id" : "{{create_article_uri.base_uri}}publisher/{{slugify(row['publisher'])}}"
        }
      ]
    },
    {
      "@if" : "'booktitle' in row",
      "@id" : "{{create_article_uri.base_uri}}book/{{slugify(row['booktitle'])}}",
      "@type" : "bibo:Book",
      "dcterms:title" : "{{row['booktitle']}}",
      "bibo:edition" : [
        {
          "@if" : "'edition' in row",
          "@value" : "{{row['edition']}}"
        }
      ],
      "dcterms:isPartOf" : [
        {
          "@if" : "'series' in row",
          "@id" : "{{create_article_uri.base_uri}}series/{{slugify(row['series'])}}",
          "@type" : "bibo:Series",
          "dcterms:title" : "{{row['series']}}"
        }
      ],
      "dcterms:contributor" : [
        {
            "@if" : "'editor' in row",
            "@for" : "editor in re.split('\\\\s*and\\\\s*', row['editor'])",
            "@do" : {
              "@with" : "author_fn_ln(editor) as ln, fn",
              "@do" : {
                "@id" : "{{create_article_uri.base_uri}}author/{{slugify(' '.join([fn, ln]))}}",
                "@type" : "foaf:Person",
                "foaf:familyName" : "{{ln}}",
                "foaf:givenName" : "{{fn}}",
                "foaf:name" : "{{fn}} {{ln}}"
              }
            }
        }
      ],
      "prov:wasAttributedTo" : [
        {
          "@if" : "'organization' in row",
          "@id" : "{{create_article_uri.base_uri}}organization/{{slugify(row['organization'])}}",
          "rdfs:label" : "{{row['organization']}}",
          "@type" : ["skos:Organization", "prov:Organization"]
        }
      ],
      "dcterms:publisher" : [
        {
          "@if" : "'publisher' in row",
          "@id" : "{{create_article_uri.base_uri}}publisher/{{slugify(row['publisher'])}}"
        }
      ]
    }
  ]
}]'''
  ].