Adding a String library

[EricIO]

(Tobias) I am changing this discussion. The new topic of this issue will be "String library".

[TobiasWrigstad]

For anyone following this ticket, @EliasC and I are working on this. We have a fork where string literals actually give rise to new string objects with a semi-rich interface. This is the current, mostly self-explanatory interface of Encore strings. If you are missing a feature, please write here.

  def init(s:string) : void
  def concatenate(b:String) : String
  def copy() : String
  def contains(b:String) : bool
  def contains_ignore_case(b:String) : bool
  def compare(b:String) : int
  def compare_ignore_case(b:String) : int
  def to_upper() : String
  def to_lower() : String
  def length() : int
  def size() : int
  def get_() : String
  def substring(from:int, to:int) : String
  def format(b:[String]) : String

This work is intended to att Strings as the first library class which is loaded automatically by all programs.

[kikofernandez]

I think the equals would be good to have

[TheGrandmother]

having a trim method (removes leading and trailing white-spaces) would be quite nice!

[TobiasWrigstad]

@kikofernandez: is this the intended semantics of equals?

def equals(s:string) : bool
  this.compare(s) == 0

?

[TobiasWrigstad]

@TheGrandmother
Just whitespace? ' ' or also '\n' etc.? (Give me a spec!)

Non-destructive, I guess too?

[kikofernandez]

Yes, that would work.

@kikofernandez: is this the intended semantics of equals?

def equals(s:string) : bool
this.compare(s) == 0
?

[TobiasWrigstad]

Thanks!

[TheGrandmother]

@TobiasWrigstad
Any whitespace character. Tabs and newlines included.
Where is this fork?

split_at(string,substring) and join_with(string,substring) are also nice to have.
Like so:

split_at("bob kjell glenn kjell otto", " kjell ") = ["bob","glenn","otto"]
join_with(["bob","glenn","otto"], " kjell ") = "bob kjell glenn kjell otto"

[TobiasWrigstad]

Tack!

[TobiasWrigstad]

FYI This is the extended String interface:

  def init(s:string) : void
  def concatenate(b:String) : String
  def copy() : String
  def contains(b:String) : bool
  def contains_ignore_case(b:String) : bool
  def compare(b:String) : int
  def compare_ignore_case(b:String) : int
  def to_upper() : String
  def to_lower() : String
  def length() : int
  def size() : int
  def get_() : String
  def substring(from:int, to:int) : String
  def equals(s:String) : bool
  def occurrences(s:String) : int
  def split(s:String) : [String]
  def join(strings:[String]) : String
  def trim() : String
  def replace(a:String, b:String) : String
  def find(a:String) : int
  def find_from(a:String, b:int) : int
  def delete(s:String) : String
  def format(b:[String]) : String

[supercooldave]

What does delete do?
Can concatenate be called concat?

[TobiasWrigstad]

Delete removes occurrences of a string from another. It is like replace, but more efficiently implemented.
I am fine with renaming any of these.

The get_ should go — it is not needed anymore (was for getting to underlying C string).

[supercooldave]

All occurrences?

[TheGrandmother]

What is the difference between length and size ??

[TobiasWrigstad]

@supercooldave All occurrences.
@TheGrandmother nothing.

[TobiasWrigstad]

@supercooldave
Slow:

  def replace(a:String, b:String) : String
    b.join(this.split(a))

Fast (might not even leave the stack for intermediate strings):

  def delete(s:String) : String
      new String(embed string
        a ton of C code
      end)

[EliasC]

Implemented as of 42c83e3