Performance improve on Jaydata filter and toArray

Question

I am using JayData 1.3.5 Pro in my HTML5 application to connect to IndexDB.

I have a PRODUCTS (contextEntity) table and I am using the following query to retrieve 15839 records from the same table.

dataContext.onReady(function () {
                contextEntity
                    .filter(function (result) {
                        result.prod_num.toLowerCase.contains(this.prod_num) === -1 ||
                        result.vendor_name.toLowerCase.contains(this.vendor_name) === -1 ||
                        result.vendor_prod_num.toLowerCase.contains(this.vendor_prod_num) === -1 ||
                        result.desc1.toLowerCase.contains(this.desc1) === -1 ||
                        result.desc2.toLowerCase.contains(this.desc2) === -1 ||
                        result.desc3.toLowerCase.contains(this.desc3) === -1 ||
                        result.lookup.toLowerCase.contains(this.lookup) === -1 ||
                        result.icmastx_desc1.toLowerCase.contains(this.icmastx_desc1) === -1 ||
                        result.icmastx_desc5.toLowerCase.contains(this.icmastx_desc5) === -1 ||
                        result.icmastx_desc7.toLowerCase.contains(this.icmastx_desc7) === -1
                    },
                  {
                      prod_num: productPageProperty.search.toLowerCase(), // search input
                      vendor_name: productPageProperty.search.toLowerCase(),
                      vendor_prod_num: productPageProperty.search.toLowerCase(),
                      desc1: productPageProperty.search.toLowerCase(),
                      desc2: productPageProperty.search.toLowerCase(),
                      desc3: productPageProperty.search.toLowerCase(),
                      lookup: productPageProperty.search.toLowerCase(),
                      icmastx_desc1: productPageProperty.search.toLowerCase(),
                      icmastx_desc5: productPageProperty.search.toLowerCase(),
                      icmastx_desc7: productPageProperty.search.toLowerCase()
                  })

                    .toLiveArray(function (result) {
                        if (result != null && result.length > 0) {
                            result.forEach(function (item) {
                                productViewModel.push({
                                    unit_conv: item.unit_conv,
                                    cartid_sequence_num: null,
                                    selling_disc: item.selling_disc,
                                    pcat: item.pcat,
                                    division_ck: item.division_ck,
                                    uom: item.uom,
                                    prod_num: item.prod_num,
                                    desc1: item.desc1,
                                    desc2: item.desc2,
                                    net_avail: item.net_avail,
                                    selling_price: item.selling_price,
                                    cust_price: item.cust_price,
                                    cust_disc: item.cust_disc,
                                    qty_ordered: null,
                                    extension: null
                                })
                            });
                        }
                        else {
                            productViewModel = [];
                        }
});

Now what happens is, it takes 30 seconds to retrive results for the search input. Is there a way to optimize this code so that I can retreive the search results faster like in 3 seconds. I am trying to find an alternate for toArray / toLiveArray.

Please suggest an alternate approach to improve the performance.

Thanks, Sriram

Answer 1

In IndexedDB (or any database), there are only two ways to query. Index-base query and enumerating all records in a objectStore (or table).

Index-base query are very fast and (almost) independent of number of records in the store.

Enumerating all records in the table get slower as number of records increase.

Index-base query require you creating index that match your query. If sorting is not required, most query can be done by indexes.