Thread: Saving?!

a .txt file would be fine to do what you want... I would load the entire file into memory, then edit what you need while it's still in memory. then dump the memory into the file using ios::trunc.

>i cant figure out how to make a database(?)
For a simple database you can use a single file (or several files that you choose from based on a search key) to hold the records, then read them into a suitable data structure in memory. Any changes made to the database would be done in memory, then written back to the file to save them. For a larger and more professional database, you're looking at writing an external data structure such as some variant of a B-tree to give you efficient lookup and modification without having to read every record into memory. However, I wouldn't recommend that particular path unless it is absolutely necessary, external B-trees are not trivial to write.

Ok, i dont think you get what i am asking... What i am tring to do is write to a file then pull up only part of it hers an example
Stuff a
blaw bla blaw

stuff b
blaw blaw
blaw
blawww

stuff C
blaw blaw blaw blaw
blaw

and so on... so what i want to be able to do is pick a spot in a file and only read that block.... not sure if it is possible tough because i tired everything.

-KJ

>i dont think you get what i am asking...
I'm well aware of what you want to do. However, I'm also pretty sure you don't want to do what you think you do.

>so what i want to be able to do is pick a spot in a file and only read that block....
This is simple provided you use binary files and have a strict indexing scheme as well as records that will never vary in size. For most applications, this is too restrictive, so we usually recommend processing whole files instead of individual records.

>>so what i want to be able to do is pick a spot in a file and only read that block.... not sure if it is possible tough because i tired everything.

if all you want to do is read, I suggest using an XML file... it looks prettier, but it's a little more code to parse the XML file than just search for a name in a list... you could set it (txt file) up this way:

Code:

NAME
(XXX)XXX-XXXX
ADDRESS
ADDRESS
ADDRESS

NAME
(XXX)XXX-XXXX
ADDRESS
ADDRESS
ADDRESS

and an XML file would look something like:

Code:

<PROGRAM NAME>
   <BLOCK>
      <NAME>name</NAME>
      <PHONE>(XXX)XXX-XXXX</PHONE>
      <ADDY1>address</ADDY1>
      <ADDY2>address</ADDY2>
      <ADDY3>address</ADDY3>
   <BLOCK>

and then search for the name and only load the next four lines... that works good for one search per run of the program and no inserting... if you're planning on having it search the file multiple times, it's easier to put the file in a linked list or even an array of structs... either way, if you want to insert something in a specific place, it's much easier to go with an array of structs or linked list...