[Tutor] Re: Why is this write defined as a tuple, instead of picking list?

[Adam]

On Fri, 16 Apr 2004 20:22:37 +0200
Andrei <project5 at redrival.net> wrote:

> Adam wrote on Friday 16 April 2004 12:16:
> 
> > I'm using this function to cycle through a list and add
> > some formatting to it to prepare it for saving:
> <snip>
> >     for x in article_items:
> >         write = "<h3>", article_items[0], "</h3>"
> >         print "write is type: ", type(write)
> <snip>
> > At the time article_items is passed it is a list -
> > however, when I try to do the formatting with write =,
> > write is declared as a tuple - why is this, and how can
> > I force it to be a string (I believe I need a string for
> > the file saving function?)
> 
> The relevant line looks like this:
> 
>    write = "a", "b", "c'
> 
> This happens to be syntax for tuples, meaning that write
> becomes a tuple. If you want to concatenate strings, you
> should use "+":
> 
>    write = "a" + "b" + "c"
> 
> It is generally not recommended to use this, because it's
> slow. It's better to have a list of substrings like this:
> 
>   write = ["a", "b", "c"]

This is very obvious now - thanks to you and to Danny. My
experience of previous programming (a little java) is that
problems are often difficult to find, but here it seems that
they are very obvious. 

> and then use the string method join() to convert that list
> into a string when you really need it to be a string (e.g.
> before writing to a file):
> 
>   strwrite = "".join(write)
> 
> Have a good look at this syntax, many (or even most)
> people find it counter-intuitive at first. "" is the
> separator char that will be inserted between the string
> parts in the list. In this case it's an empty string, but
> you could also use "\n" (newline) for example to make each
> of the substrings appear on a new line in the concatenated
> string.
> 
> In this particular (partial) case, I'd even recommend
> string formatting:
> 
>   write = "<h3>%s</h3>" % article_items[0]
> 
> But if you plan to add more content to the HTML page, the
> solution with a list which is joined at a later point in
> time is the best one.

I did look at the join to do this, and I agree that it is by
far a better way of doing it. 

> 
> Btw, I don't understand this logic:
> 
> >     for x in article_items:
> >         write = "<h3>", article_items[0], "</h3>"
> 
> Why are you writing the same header for every item in the
> list?

I'm not giving the same header - these are just variables
that were collected earlier in the application and then are
to be saved to disk as a HTML document. A sort of mini
content management system for me to keep notes on
articles that I read. The next part of the code will be
something like: 

write = "<h3>", article_items[0], "</h3>", "<p>",
article_item(1), "</p>" ... and so on with the data. 

I often find that I spend ages looking for an article, and
keeping some kind of electronic record of what I read is an
ideal way for me to search through them all. I thought
python might be useful for this very task. And it is
definitely turning out to be! 

If anyone wants this code once it's finished, I'll be happy
to share it under GPL. 

> Yours,
> 
> Andrei

Thanks again Andrei. 

adam