pcsoft.us.windevCopyright 2026, PC SOFT17 Jul 2015 12:08:53 Z15 Jul 2015 22:25:00 ZHi. These days I'm trying to speedup my data searches reducing the access to my server side MySql DB and increasing the memory workload on client side. The goal is to keep the performance as possible when accessing data outside company lan (on lan queries go like rockets, obviously). When you perform a browsing query you can perform a single query for getting 100/200 items and all their linked data (6/7 linked mysql tables). So 10 queries = 2000 full items... Quite fast anyway even outside lan! Instead, when you start from a given list of items and the search is one-per-row (you cannot start from a set of items delivered by a filtering query, you have a scattered random items list with maybe nothing in common each other) the performance falls down. 1000 rows = 1000 SELECT queries, doesn't it? Now I'm going to import from a list (txt) a set of items which compose a provider's catalog. Weekly catalogs often contains only an item code and a price, all the other fields/columns of each item/record have to be filled searching in the "history" of catalogs imported by the company, stored in MySql DB. The match between a "row item" and the "history items" is by item code. You can have N past history records matching the same single row, in case you choose the latest (I semplified, there are more complex criteria). So I can execute a query for getting the past catalog records of each item/row, applying the needed filters/sorts by sql query in order to get the lastest history match or execute a unique query (that is fast) saving all the past catalog of the provider in a recordset in memory. The first option is trivial but slow outside lan (N rows = N queries), the second could allow me to work all the most in memory but it cannot succed in sorting/filtering the resulting data source. As I read on the forum, a way could be the FileToArray from the data source but having a different item on each imported row I have to pass (FOR EACH) all the array each time to find the match, didn't it? I'm afraid that even using WHERE in FOR EACH you have to scroll all the array (it seems so by seeing the cronostart info). I thought even to save the catalog on a local datafile and then execute local queries on that but I'm quite newbie and didn't succeded in creating it. Can you help me to make the best choice in WD for working (sort/filter) on data sources or derived arrays? Thank you and sorry for being a little verbose in this post. Stefano30WEBDEVen_UShttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset/read.awpmoderateur@pcsoft.fr (El moderador)webmaster@pcsoft.fr (El webmaster)fromwebpcsoft.us.windevhttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53105/read.awp17 Jul 2015 12:08:53 ZHi again, another way of improving your results would be to maintain one extra file on you sever, a link file between the e…https://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53105/read.awphttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53105/read.awpQuery on a resultset...guestpcsoft.us.windevhttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53089/read.awp16 Jul 2015 12:48:00 ZIs one of the options, but if you have enough memory the array solution is faster. The other option is to copy the txt file t…https://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53089/read.awphttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53089/read.awpQuery on a resultset...guestpcsoft.us.windevhttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53088/read.awp16 Jul 2015 11:41:00 ZOk, thank you. So trying to copy the "history" of the provider X in a local temp datafile and query on that is not an option to …https://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53088/read.awphttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53088/read.awpQuery on a resultset...guestpcsoft.us.windevhttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53086/read.awp16 Jul 2015 11:05:00 ZYou don't need FOR EACH in the array if you use one array of structures you can do a lot with them and they are very fast. Ju…https://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53086/read.awphttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53086/read.awpQuery on a resultset...guestpcsoft.us.windevhttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53083/read.awp15 Jul 2015 23:42:00 ZYes, I think you understood. I thought about that for getting in one shot data from items database. The problem is that i sempl…https://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53083/read.awphttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53083/read.awpQuery on a resultset...guestpcsoft.us.windevhttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53082/read.awp15 Jul 2015 22:51:00 ZHi Stefano, I'm not sure I followed the whole thing, but I'll try anyway... What about: - you import all your IDs from yo…https://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53082/read.awphttps://forum.pcsoft.fr/es-ES/pcsoft.us.windev/53076-query-resultset-53082/read.awpQuery on a resultset...