uqac-websem-project/backend/WebSem/WebApiWebSem/DataAccess/DbPediaDataAccess.cs

using System;
using System.Collections.Generic;
using System.Linq;
using Luticate2.Utils.Dbo.Basic;
using Luticate2.Utils.Dbo.PaginatedRequest;
using Luticate2.Utils.Dbo.Result;
using Microsoft.Extensions.Logging;
using VDS.RDF.Nodes;
using VDS.RDF.Query;
using WebApiWebSem.Dbo.Articles;
using WebApiWebSem.Dbo.ArticlesFields;

namespace WebApiWebSem.DataAccess
{
    public class DbPediaDataAccess
    {
        private readonly SparqlRemoteEndpoint _sparqlRemoteEndpoint;
        private readonly ILogger<DbPediaDataAccess> _logger;

        protected string Prefixes = @"
PREFIX owl: <http://www.w3.org/2002/07/owl#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX foaf: <http://xmlns.com/foaf/0.1/>
PREFIX dc: <http://purl.org/dc/elements/1.1/>
PREFIX : <http://dbpedia.org/resource/>
PREFIX dbpedia2: <http://dbpedia.org/property/>
PREFIX dbpedia: <http://dbpedia.org/>
PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
PREFIX dbpedia-owl: <http://dbpedia.org/ontology/>
";

        public DbPediaDataAccess(SparqlRemoteEndpoint sparqlRemoteEndpoint, ILogger<DbPediaDataAccess> logger)
        {
            _sparqlRemoteEndpoint = sparqlRemoteEndpoint;
            _logger = logger;
        }

        protected string Escape(string str)//TODO
        {
            return str.Replace('\'', ' ').Replace('"', ' ');
        }

        protected LuResult<SparqlResultSet> Query(string query)
        {
            _logger.LogInformation(query);
            try
            {
                var results = _sparqlRemoteEndpoint.QueryWithResultSet(query);
                return LuResult<SparqlResultSet>.Ok(results);
            }
            catch (Exception e)
            {
                return LuResult<SparqlResultSet>.Error(LuStatus.BackendError, e);
            }
        }

        protected ArticlesDbo ModelToDbo(SparqlResult result)
        {
            var dbo = new ArticlesDbo
            {
                Fields = new List<ArticlesFieldsDbo>()
            };
            foreach (var variable in result.Variables)
            {
                string data = null;
                var v = result[variable];
                if (v != null)
                {
                    data = v.AsValuedNode().AsString();
                }
                if (variable == "id")
                {
                    dbo.Id = data.Split('/').Last();
                }
                else if (variable == "abstract")
                {
                    dbo.Text = data;
                }
                else if (variable == "pictureUrl")
                {
                    dbo.PictureUrl = data;
                }
                else if (variable == "pictureDesc")
                {
                    dbo.PictureCaption = data;
                }
                else
                {
                    dbo.Fields.Add(new ArticlesFieldsDbo
                    {
                        Property = variable,
                        Type = "text/plain",
                        Value = data
                    });
                }
            }
            return dbo;
        }

        public LuResult<LuPaginatedDbo<ArticlesDbo>> GetMultiple(LuPaginatedRequestDbo request)
        {
            var type = ArticlesDataAccess.GetFilterType(request.Filter);
            string query = null;
            string querySelect = null;
            string queryWhereId = null;
            string queryWhereVars = null;
            if (type == "home")
            {
                querySelect = "?id ?abstract ?pictureUrl ?pictureDesc";
                queryWhereId = @"?id foaf:isPrimaryTopicOf wikipedia-en:World_War_II .";
                queryWhereVars = @"
OPTIONAL { ?id dbo:abstract ?abstract FILTER(lang(?abstract) = 'en') }
OPTIONAL { ?id foaf:depiction ?pictureUrl }
OPTIONAL { ?id dbpedia-owl:depictionDescription ?pictureDesc }
BIND ((?abstract) AS ?search)";
            }
            else if (type == "persons")
            {
                querySelect = "?id ?abstract ?pictureUrl ?pictureDesc (MIN(?name) AS ?name) (str(?birthDate) AS ?birthDate) (str(?deathDate) AS ?deathDate)";
                queryWhereId = @"?id rdf:type dbo:Person ;
?type dbr:World_War_II .";
                queryWhereVars = @"
OPTIONAL { ?id dbo:abstract ?abstract FILTER(lang(?abstract) = 'en') }
OPTIONAL { ?id foaf:depiction ?pictureUrl }
OPTIONAL { ?id dbpedia-owl:depictionDescription ?pictureDesc }
OPTIONAL { ?id dbp:name ?name . FILTER (lang(?name) = 'en') }
OPTIONAL { ?id dbp:birthDate ?birthDate }
OPTIONAL { ?id dbp:deathDate ?deathDate }
BIND (CONCAT(?abstract, ' ', ?id) AS ?search)";
            }

            var filter = $"FILTER regex(?search, '{request.Filter.Query}', 'i')";

            var count = Query(string.Format(@"{0}
SELECT (COUNT(?id) AS ?count) WHERE {{
{1}
{2}
{3}
}}
", Prefixes, queryWhereId, queryWhereVars, filter));
            if (!count)
            {
                return count.To<LuPaginatedDbo<ArticlesDbo>>();
            }

            var res = Query(string.Format(@"{0}
SELECT DISTINCT {1} WHERE {{
{2}
{3}
{4}
}}
ORDER BY {5}
LIMIT {6}
OFFSET {7}", Prefixes, querySelect, queryWhereId, queryWhereVars, filter, request.OrderBy.ToDbPediaModel(),
                request.PerPage, request.Page * request.PerPage));
            if (!res)
            {
                return res.To<LuPaginatedDbo<ArticlesDbo>>();
            }
            var results = new LuPaginatedDbo<ArticlesDbo>
            {
                Count = count.Data.Results.First()["count"].AsValuedNode().AsInteger(),
                Data = res.Data.Results.Select(ModelToDbo).Select(dbo =>
                {
                    dbo.Type = type;
                    return dbo;
                }).ToList()
            };
            return LuResult<LuPaginatedDbo<ArticlesDbo>>.Ok(results);
        }
    }
}