[BalletDystopia]

(Original post by ldvico)
Edited it cause I wrote it the wrong way round haha. The above is <integer><operator><integer> so that obeys the expression

(I haven't exactly been on point because my head's been bombarded with a lot of info today D: )

Ah i get it now thanks. They are 2 separate things! Integer Operator and Integer, which makes sense because the book didn't allow one with a minus in front of the integer which would be <operator><integer>. Cheers!

[SecondHand]

(Original post by BalletDystopia)
You beat me to answering him I need some help too though.

A person who collects coins has decided to record the details of his collection in a
database. He has decided that the following details should be stored for each coin.
CoinID
Type (for example 10p)
Year
Condition (for example “very fine”)
Value
Date purchased
SupplierID
Supplier’s Name
Supplier’s Address
Supplier’s Telephone Number
Price paid

When the data is stored in a normalised database two relations, Coin and Supplier, are used. For each of these, complete the following Data Definition Language (DDL)
statements to create the tables, including all key fields. The two ddl headings for this are coin and supplier. I don't understand why SupplierID goes in the Coin table instead of CoinID being a Foreign key in the Supplier table.

And Backus Naur Form. I still don't understand this and we only glossed over it at college and the book isn't helping me.

<expression> ::= <integer> | <integer> <operator> <expression>
<integer> ::= 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9
<operator> ::= + | - | * | /

I don't really understand the first lineof it means. Is it saying the operator has to come AFTER the integer?

It depends if each coin can have many suppliers and if each supplier can supply many coins.

Case 1: A coin has 1 supplier and suppliers can supply many coins - Supplier ID in the Coin table
Case2: A coin can have many suppliers but each supplier can only supply 1 coin - CoinID in supplier table
Case 3: A coin can have many suppliers and suppliers can store more than 1 coin - You need a table in the middle or the Coin table and the Supplier Table which only holds the link between the two tables (as to break the many to many relationship).

[Amirrryy]

FFS i am absolutely rubbish at entity relationships man! just cant get my head round the question properly 9/10 times i get them wrong -.- >:O

[exe]

(Original post by Amirrryy)
FFS i am absolutely rubbish at entity relationships man! just cant get my head round the question properly 9/10 times i get them wrong -.- >:O

yeah man I totally feel you there. Good thing they're only a couple of marks

[Amirrryy]

(Original post by exe)
yeah man I totally feel you there. Good thing they're only a couple of marks

haha yea that made me feel better since i know i'm not the only one :P
yknow when you are writing a query in SQL for a relation etc (SELECT-FROM-WHERE etc) in the condition section (WHERE) do you always have to have something.condition = true/false/or a value where "something" is the name of a table(relation)? say for example it says write a query for the surnames of students where lifecyclephaseID= 1. LifeCyclephaseID is in multiple tables. In mark scheme it has chosen one of these tables (markAwarded.LifeCyclePhaseID = 1 instead of just writing LifeCyclephaseID = 1....why? is it wrong to write it in the other form?
i know this probs doesnt make sense so just incase i attached the link to the q. scroll down to "2005 winter CPT5" - part bii
thanks in advance!

http://www.starteractivity.com/ictle...cc2qa6.html#21

[SoulReaper1470]

(Original post by Amirrryy)
~

SELECT <attributes> FROM <tables> is complete on it's own. WHERE and ORDER BY are optional if you need them. Let's say you wanted all the attributes from table Clients and you have no conditions and want to sort the result in ascending order by ClientID, you'll use the query:

SELECT * FROM Clients ORDER BY ClientID

[gowans07]

(Original post by Amirrryy)
is it possible we get a question asking us to write relations in fully normalised form for an un-normalised relational database? i am not good with them and pretty unconfident since our teacher had no clue about 1NF, 2NF and 3NF :L

Fairly simple
1NF = No repeating attributes (fields)
2NF = No partial non key dependencies
3NF = No Non key dependencies


Normalisation improves the design of the database allowing searches to be quicker.

[SoulReaper1470]

(Original post by gowans07)
Fairly simple
1NF = No repeating attributes (fields)
2NF = No partial non key dependencies
3NF = No Non key dependencies


Normalisation improves the design of the database allowing searches to be quicker.

I might be wrong, but I remember my teacher telling us to say repeating groups instead of repeating attributes. Apparently, a mark scheme said reject repeating attributes.

[gowans07]

(Original post by BalletDystopia)
And Backus Naur Form. I still don't understand this and we only glossed over it at college and the book isn't helping me.

<expression> ::= <integer> | <integer> <operator> <expression>
<integer> ::= 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9
<operator> ::= + | - | * | /

I don't really understand the first lineof it means. Is it saying the operator has to come AFTER the integer?

okay as you see the words located in <> are sort of variables. A vertical line such as | means or. ::= simply means is defined as, so an <integer> is defined as ::= a 0 or a 1 or a 2 etc

Hope this answers your question

[Amirrryy]

When do you use/ how do you know when to use two attributes as a single primary key to a relation? i cant figure out why you do it and how you know which attributes in a table/relation to pick for the primary key and its pissing me off now -.- can someone please explain?
many thanks

[gowans07]

(Original post by SoulReaper1470)
I might be wrong, but I remember my teacher telling us to say repeating groups instead of repeating attributes. Apparently, a mark scheme said reject repeating attributes.

hmm first i've heard of that, i've learnt it in Computing and ICT and it has always been attributes. couldnt of been a new mark scheme so was probably of the old spec. Double check but i've always said attributes and its been fine.

[SoulReaper1470]

(Original post by gowans07)
hmm first i've heard of that, i've learnt it in Computing and ICT and it has always been attributes. couldnt of been a new mark scheme so was probably of the old spec. Double check but i've always said attributes and its been fine.

Yea I think it was from a very old mark scheme and he is ancient as well, just a minor thing :P

[gowans07]

(Original post by Amirrryy)
When do you use/ how do you know when to use two attributes as a single primary key to a relation? i cant figure out why you do it and how you know which attributes in a table/relation to pick for the primary key and its pissing me off now -.- can someone please explain?
many thanks

We usually create a ID field which is simply either a number or a short string. For example Teacher ID = 1, Supplier ID = KELLOG

We simply choose a field which is unique, so for example there is a chance that two teachers could have the same first name so we cant take that chance and make First_Name as the primary key. The same with Supplier ID we may get a situation where two companies share the same postcode (random but go with the flow)

Basically choose a field or create one that is unique

[gowans07]

(Original post by SoulReaper1470)
Yea I think it was from a very old mark scheme and he is ancient as well, just a minor thing :P

No worries, had me worried there for a bit

[Amirrryy]

(Original post by gowans07)
We usually create a ID field which is simply either a number or a short string. For example Teacher ID = 1, Supplier ID = KELLOG

We simply choose a field which is unique, so for example there is a chance that two teachers could have the same first name so we cant take that chance and make First_Name as the primary key. The same with Supplier ID we may get a situation where two companies share the same postcode (random but go with the flow)

Basically choose a field or create one that is unique

yea i know how to create a primary key, but when does the primary key compose of TWO or even THREE attributes rather than the usual one? (i.e. composite keys)

[exam2k10]

Not really looking forward to this exam, and I feel that abstraction/information hiding and intractable problems may also come up

[BalletDystopia]

(Original post by SecondHand)
It depends if each coin can have many suppliers and if each supplier can supply many coins.

Case 1: A coin has 1 supplier and suppliers can supply many coins - Supplier ID in the Coin table
Case2: A coin can have many suppliers but each supplier can only supply 1 coin - CoinID in supplier table
Case 3: A coin can have many suppliers and suppliers can store more than 1 coin - You need a table in the middle or the Coin table and the Supplier Table which only holds the link between the two tables (as to break the many to many relationship).

I understand you, and i can probably use this example to wing it in the exam. But 'why' is it like that? Sorry to be a pain. Am I missing something? Is it really simple?

[SecondHand]

(Original post by BalletDystopia)
I understand you, and i can probably use this example to wing it in the exam. But 'why' is it like that? Sorry to be a pain. Am I missing something? Is it really simple?

Think about what happens in each case if some information about the coin changes or if some information about the supplier changes. Think about it first and then I will explain if you are still stuck.

[gowans07]

(Original post by exam2k10)
Not really looking forward to this exam, and I feel that abstraction/information hiding and intractable problems may also come up

could do although intractable problem came up in either 2010 or 2011, as far as information hiding and abstraction i agree could come up. Also those 2 tie in OS's

[exam2k10]

Even though Im going to be doing Computer Science at university, I am not looking forward to this module and will be so glad once it is over. It has eaten up a lot of my time, and now I am quite nervous for the exam.

What do you guys think of this revision structure for today and tomorrow
Today

• Do the COMP3 June 2011 paper
• Do some old CPT4 and CPT5 papers
• Read over some of the topics again

Tomorrow

• Read over all of my notes again

Tuesday Morning

• Read over all of my notes again

Would appreciate it, if you could comment on this structure.

BTW, how screwed am I and am I the only one feeling a bit nervous?

Thanks