pepsi_max2k

Hmm... not really sure I understand, I'll read up on stuff when it's not past midnight For now, here's a pic of what I meant above. Will eventually be displayed on a page like this: Item: 'name' - linked from 'ID' or directly from table below Price: linked from 'ID' or 'Price' Stock: 'stock' Condition Box: 'conditionA' Sleeve: 'conditionB' Manual: 'conditionC' Disc: 'conditionD' (Or General Contents: 'conditionE' displayed if available and the above are missing) 'comment' [attachment deleted by admin]

Yeah, I thought as much, but... a table purely for "Condtion"? What does it contain? Product ID plus four condition columns, each with an individual rating, each only even relating to the individual product id? Seems to me I wouldn't be cutting down on any replication, just moving out data that relates to a specific product out of one table's entry for that product and in to a seperate table. Edit: Only reason for doing it I can think of - if I had a new table for "Contents", then could add a list of contents and rate their condition, linking them to the product id. If I just assume everything has much the same contents, having five extra columns for each product with a rating between 0 to 5 (or N/A) doesn't seem so bad.

Hey all, I'm about to start coding an online store. I have one database full of various items (with basic info ie. title, format, and prices linked to them) and think it's best if I create another table for the main store, and join in item entries from the main item table when needed. The thing is, not all items will be present in the first table, and I won't always want the price included as it is in the first table (actually, it's in a second table, but it's complicated enough...). I also want to add in extra info (like condition, random comments, etc). So how does the following look? === Table 2 ==== Item ID (link to T1 item ID for details and price) Item Name override (if there's no item ID in T1, ie. Item ID is blank) Price Override (if I want to override the price from T1) Condition 1 (case / box) Condition 2 (sleeve / inserts) Condition 3 (manual) Condition 4 (disc / item) Other Condition (anything that doesn't fit the above) Comments (for random comentary for anything that needs a better description than just item title and condition) Image (maybe, possibly, I seem to recall adding images to a database is horrendously hard, maybe just a link to one then, though keeping track of images that need to be present and those that dont (for sold out items) should be fun...) I think that looks just about ok. Looks like I could possibly split out condition in to a seperate table, though doesn't really seem like enough to warrant it (it'll be literally a choice between 0 and 5 for each condition, relating specifically to the one product entry). I should probably google how to code a store... or just stick to eBay...

Hey all, just wondering about how I should best go about adding an fwrite command to a small loop... I want to write one line to a file every time I pass through one loop. I could loop up to 20,000 times, around 50 bytes per line, around 1 MB if doing 20,000 entries. So, do I fwrite for every entry (which I presume is taking up extra processing time) or do I continuously append lines to one super large up-to 1 MB string then perform a single fwrite once the loop has finished? Thinking about it, that doesn't sound so great, I'm only fwriting to debug if something does wrong during the loop... one fwrite = no output at all if something goes wrong... I think I found my answer :|

Figured it out eventually. Add this in to replace the current date sorting code. Works fine for DD-MM-YY. sort.date.formats = [ // DD-MM-YY[YY] { re:/(\d{1,2})-(\d{1,2})-(\d{2,4})/ , f:function(x){ return (new Date(sort.date.fixYear(x[3]),+x[2],+x[1])).getTime(); } } // Any catch-all format that new Date() can handle. This is not reliable except for long formats, for example: 31 Jan 2000 01:23:45 GMT ,{ re:/(.*\d{4}.*\d+:\d+\d+.*)/, f:function(x){ var d=new Date(x[1]); if(d){return d.getTime();} } } ];

Hey all, I'm currently trying to edit a small bit of javascript to get it working with UK dates, and it's doing my head in The code is a table sorting program from http://www.javascripttoolbox.com/lib/table/source.php , the part in question is as follows, and it's set up to use US dates but I'm trying to use it with UK dates like 30-01-09 (30th Jan 09). // Date sort // --------- sort.date = function(a,b) { return sort.numeric(sort.date.convert(a),sort.date.convert(b)); }; // Convert 2-digit years to 4 sort.date.fixYear=function(yr) { yr = +yr; if (yr<50) { yr += 2000; } else if (yr<100) { yr += 1900; } return yr; }; sort.date.formats = [ // YY[YY]-MM-DD { re:/(\d{2,4})-(\d{1,2})-(\d{1,2})/ , f:function(x){ return (new Date(sort.date.fixYear(x[1]),+x[2],+x[3])).getTime(); } } // MM/DD/YY[YY] or MM-DD-YY[YY] ,{ re:/(\d{1,2})[\/-](\d{1,2})[\/-](\d{2,4})/ , f:function(x){ return (new Date(sort.date.fixYear(x[3]),+x[1],+x[2])).getTime(); } } // Any catch-all format that new Date() can handle. This is not reliable except for long formats, for example: 31 Jan 2000 01:23:45 GMT ,{ re:/(.*\d{4}.*\d+:\d+\d+.*)/, f:function(x){ var d=new Date(x[1]); if(d){return d.getTime();} } } ]; sort.date.convert = function(val) { var m,v, f = sort.date.formats; for (var i=0,L=f.length; i<L; i++) { if (m=val.match(f[i].re)) { v=f[i].f(m); if (typeof(v)!="undefined") { return v; } } } return 9999999999999; // So non-parsed dates will be last, not first }; To start with I've no idea how javascript work, so I've just been hacking away to see what happens. The best I have is: sort.date.formats = [ // YY[YY]-MM-DD { re:/(\d{1,2})-(\d{1,2})-(\d{2,4})/ , f:function(x){ return (new Date(sort.date.fixYear(+x[1],+x[2]),+x[3])).getTime(); } } which seems to sort the years and the months properly, but the days are now in reverse, eg: 09-02-09 10-02-09 11-02-09 09-01-09 10-01-09 11-01-09 09-12-09 Anyone know how to edit it to use with uk dates? I don't really wanna use US ones but it's looking like the simplest option.

Hey all, I have a question about parsing HTML pages. I have a script that grabs one, stores it as $result, then runs numerous preg_match_all commands on $result (without changing it). First I run a preg_match to find whether there's any values worth further parsing (looks for a single line). Then I run one preg_match_all to find all negative occurances of a value, then another preg_match_all to find all positive results of one type, if non are found then again to find them of a different type, then again to find them of a third type if there's none of a second. That's one preg_match and up to four preg_match_alls on the same page, much of which contains data that's irrelevant to the parsing. So I'm just wondering whether I should run a preg_replace to strip half of the text in $result away, in the hope it'll speed up the latter preg_match_alls? Or is the result not gonna be worth it - maybe preg_replace will take up so much extra time, I may as well leave it as it is? Oh and while I'm here - if statements. Is using lots really gonna cause any speed issues? I have a few things like this in my code: if ($type=="Game") { // Alter search url based on console. if ($console=="Xbox") { do something } elseif ($console=="PS1") { do something else } elseif ($console=="PS2") { do something else } elseif ($console=="PS3") { do something else } elseif ($console=="GameCube") { do something else } elseif ($console=="Xbox 360") { do something else } elseif ($console=="Wii") { do something else } elseif ($console=="Nintendo DS") { do something else } } elseif ($type=="Accessory") { if ($console=="PS1") { do something else } elseif ($console=="PS2") { do something else } else { do something else } } else { do something else } Seems a bit, erm... slow to read though. Just seems there'd be a quicker way...

Well that wasn't too hard. "OR prices.date IS NULL". Full query is: SELECT products.*, prices.* FROM products LEFT JOIN prices ON products.id = prices.id WHERE prices_.date = (select max(prices.date) from prices where products.id = prices.id) OR prices.date IS NULL ORDER BY products.id

SELECT products.*, prices.* FROM products LEFT JOIN prices ON products.id = prices.id WHERE prices.date = (select max(prices.date) from prices where prices.id = products.id)"; That WHERE part helps a lot, outputting only one row with the latest date, but now it doesn't output anything with no price entries at all which I'd rather it did.

Hey all, I have the following data: Table A) ID, Name, etc Table B) ID, date, price Table A has individual products with unique IDs. Table B has multiple entries for a single ID, primary key is split between ID and date. eg: Table A) 0001, Fish 0002, Beans Table B) 0001, 2009-01-05, 1.5 0002, 2009-04-02, 5.2 0002, 2009-05-03, 5.7 0001, 2009-04-05, 1.2 How do I select all entries from table A, while left joining table B and being able to output just the latest price from table B for each ID in table A? Eg: 0001, Fish, 1.2 0002, Beans. 5.7 So far my code is this: SELECT products.*, prices.* FROM products LEFT JOIN prices ON products.id = prices.id AND prices_ebay.date = (select max(date) from prices) GROUP BY products.id"; $num=mysql_numrows($result); $i=0; while ($i < $num) { $id=mysql_result($result,$i,"id"); $pricemin= mysql_result($result,$i,"prices.min"); } Without the AND and GROUP BY, it outputs each product X times for every entry in price with the same ID. With the GROUP BY, it outputs each product only once, but with only the first entry for prices, not the last. With the AND, it only outputs a single price entry for a single product - whatever was the last entry overall in the prices table (not the last one for each product/ID). Thanks for any help.

Oh hang on... I'm doing: $i=0; while ($i < $num) { $name=mysql_result($result,$i,"name"); $publisher=mysql_result($result,$i,"publisher"); echo $name; echo $publisher; $i++; } Which would output twice for one product with two publishers. But if I only have one product... $name=mysql_result($result,$i,"name"); echo $name; $i=0; while ($i < $num) { $publisher=mysql_result($result,$i,"publisher"); echo $publisher; $i++; } One product output, multiple publishers, correct? Silly me... Except... What if I have one product, two publishers, and three... say colours (in a seperate linked table). How do I loop through the publishers to output twice, but then colours to output three times? The above would output one product but three publishers and colours... I think... maybe I should try it...

Hey all, I've got a mysql/php question about database normalization. I've got the following tables: Products - inc. product ID and data. Publishers - inc. publisher ID and data. Products/Publishers - inc. product ID and publisher ID, can be multiple entries for each product. What code would I need to grab and output just a single product from the products table, but include all publishers for that product? I can already do the output from a single row by using "$num=mysql_numrows($result);" and a while loop for $num, but as soon as I start joining tables that have multiple entries in a secondary table, it'll then decide there's X rows and output all data from the first table X times, and just change the publisher each time. Thanks, pepsi_max2k.

Not sure if you understood me.... I mean don't allow both types to be used for logins. Either set it to just use user names for everyone, or to just use emails for everyone, not allow both of them to be used. Then you don't have to worry about what to name it

just... give everyone either emails *or* user names, and have it check the login to see if the name/email belongs to a store or shopper account?

Hey all, I very new to database design, trying to figure out how best to store some postage rates and link them to products. Here's the data I want to stick in a database, it's inland postage rates from the Royal Mail (which change frequently so storing set postage costs is a no no): http://www.royalmail.com/portal/rm/content1?catId=400030&mediaId=51000697 http://www.royalmail.com/portal/rm/content1?mediaId=50800722&catId=400029 http://www.royalmail.com/portal/rm/content1?mediaId=51000711&catId=400031 That's 3 different services, each with different size classes, each size class with different weigh classes, each with different prices. And I need some way of adding postage costs not covered by these too. Here's my current design: Postage ID Service Class (First, Second, SP, Other) Size (L, LL, P) Max Weight Weight (100, 250, 1000) Price (0.40, 1.04) Product ID (0000001) EAN (711719906520, 5055060923825) Postage Weight Class (100, 250, 1000) Postage Size Class (L, LL, P, SP) I'm sure it's not perfect though. For any product, I just want to list a single second class service price, or if there is non (decided by weight) then first, and if non then standard parcels. Seems like it'll be a lot of work to sort that out. But if I just reference a single postage entry by ID then I can't go grab a first class cost for it too if needed. My head hurts UPDATE: Any product L, LLs or Ps can show 1st or 2nd prices. And SPs (standard parcels, no first/second avail) can only show SP prices. Not perfect, and deffinately not the way a human would look at the services, but a little better for the database methinks.