License:
BSD style: see license.txt
Version:
Initial release: December 2005
author:
Kris
Text is a class for storing and manipulating Unicode characters.
Text maintains a current "selection", controlled via the mark(),
select() and selectPrior() methods. Each of append(), prepend(),
replace() and remove() operate with respect to the selection. The
select() methods operate with respect to the current selection
also, providing a means of iterating across matched patterns. To
set a selection across the entire content, use the mark() method
with no arguments.
Indexes and lengths of content always count code units, not code
points. This is similar to traditional ascii string handling, yet
indexing is rarely used in practice due to the selection idiom:
substring indexing is generally implied as opposed to manipulated
directly. This allows for a more streamlined model with regard to
surrogates.
Strings support a range of functionality, from insert and removal
to utf encoding and decoding. There is also an immutable subset
called TextView, intended to simplify life in a multi-threaded
environment. However, TextView must expose the raw content as
needed and thus immutability depends to an extent upon so-called
"honour" of a callee. D does not enable immutability enforcement
at this time, but this class will be modified to support such a
feature when it arrives - via the slice() method.
The class is templated for use with char[], wchar[], and dchar[],
and should migrate across encodings seamlessly. In particular, all
functions in tango.text.Util are compatible with Text content in
any of the supported encodings. In future, this class will become
the principal gateway to the extensive ICU unicode library.
Note that several common text operations can be constructed through
combining tango.text.Text with tango.text.Util e.g. lines of text
can be processed thusly:
auto source = new Text!(char)("one\ntwo\nthree");
foreach (line; Util.lines(source.slice))
// do something with line
Substituting patterns within text can be implemented simply and
rather efficiently:
auto dst = new Text!(char);
foreach (element; Util.patterns ("all cows eat grass", "eat", "chew"))
dst.append (element);
Speaking a bit like Yoda might be accomplished as follows:
auto dst = new Text!(char);
foreach (element; Util.delims ("all cows eat grass", " "))
dst.prepend (element);
Below is an overview of the API and class hierarchy:
class Text(T) : TextView!(T)
{
// set or reset the content
Text set (T[] chars, bool mutable=true);
Text set (TextView other, bool mutable=true);
// retreive currently selected text
T[] selection ();
// get the index and length of the current selection
Span selectionSpan ();
// mark a selection
Text select (int start=0, int length=int.max);
// move the selection around
bool select (T c);
bool select (T[] pattern);
bool select (TextView other);
bool selectPrior (T c);
bool selectPrior (T[] pattern);
bool selectPrior (TextView other);
// append behind current selection
Text append (TextView other);
Text append (T[] text);
Text append (T chr, int count=1);
Text append (int value, options);
Text append (long value, options);
Text append (double value, options);
// transcode behind current selection
Text encode (char[]);
Text encode (wchar[]);
Text encode (dchar[]);
// insert before current selection
Text prepend (T[] text);
Text prepend (TextView other);
Text prepend (T chr, int count=1);
// replace current selection
Text replace (T chr);
Text replace (T[] text);
Text replace (TextView other);
// remove current selection
Text remove ();
// clear content
Text clear ();
// trim leading and trailing whitespace
Text trim ();
// trim leading and trailing chr instances
Text strip (T chr);
// truncate at point, or current selection
Text truncate (int point = int.max);
// reserve some space for inserts/additions
Text reserve (int extra);
}
class TextView(T) : UniText
{
// hash content
hash_t toHash ();
// return length of content
uint length ();
// compare content
bool equals (T[] text);
bool equals (TextView other);
bool ends (T[] text);
bool ends (TextView other);
bool starts (T[] text);
bool starts (TextView other);
int compare (T[] text);
int compare (TextView other);
int opEquals (Object other);
int opCmp (Object other);
// copy content
T[] copy (T[] dst);
// return content
T[] slice ();
// return data type
typeinfo encoding ();
// replace the comparison algorithm
Comparator comparator (Comparator other);
}
class UniText
{
// convert content
abstract char[] toString (char[] dst = null);
abstract wchar[] toString16 (wchar[] dst = null);
abstract dchar[] toString32 (dchar[] dst = null);
}
$(DDOC_MODULE_MEMBERS
class
Text
(T): TextView!(T);
$(DDOC_DECL_DD The mutable
Text
class actually implements the full API, whereas
the superclasses are purely abstract (could be interfaces instead).
- struct
Span
;
- Selection span
- uint
begin
;
- index of selection point
- uint
length
;
-
length
of selection
- this(uint space = 0);
- Create an empty Text with the specified available
space
Note:
A character like 'a' will be implicitly converted to
uint and thus will be accepted for this constructor, making
it appear like you can initialize a Text instance with a
single character, something which is not supported.
- this(T[] content, bool copy = true);
- Create a Text upon the provided content. If said
content is immutable (read-only) then you might consider
setting the 'copy' parameter to false. Doing so will
avoid allocating heap-space for the content until it is
modified via Text methods. This can be useful when
wrapping an array "temporarily" with a stack-based Text
- this(TextViewT other, bool copy = true);
- Create a Text via the content of another. If said
content is immutable (read-only) then you might consider
setting the 'copy' parameter to false. Doing so will avoid
allocating heap-space for the content until it is modified
via Text methods. This can be useful when wrapping an array
temporarily with a stack-based Text
- Text
set
(T[] chars, bool copy = true);
- Set the content to the provided array. Parameter 'copy'
specifies whether the given array is likely to change. If
not, the array is aliased until such time it is altered via
this class. This can be useful when wrapping an array
"temporarily" with a stack-based Text
- Text
set
(TextViewT other, bool copy = true);
- Replace the content of this Text. If the new content
is immutable (read-only) then you might consider setting the
'copy' parameter to false. Doing so will avoid allocating
heap-space for the content until it is modified via one of
these methods. This can be useful when wrapping an array
"temporarily" with a stack-based Text
- Text
select
(int start = 0, int length = (int).max);
- Explicitly set the current selection
- T[]
selection
();
- Return the currently selected content
- Span
selectionSpan
();
- Return the index and length of the current selection
- bool
select
(T c);
- Find and
select
the next occurrence of a BMP code point
in a string. Returns true if found, false otherwise
- bool
select
(TextViewT other);
- Find and
select
the next substring occurrence. Returns
true if found, false otherwise
- bool
select
(T[] chars);
- Find and
select
the next substring occurrence. Returns
true if found, false otherwise
- bool
selectPrior
(T c);
- Find and select a prior occurrence of a BMP code point
in a string. Returns true if found, false otherwise
- bool
selectPrior
(TextViewT other);
- Find and select a prior substring occurrence. Returns
true if found, false otherwise
- bool
selectPrior
(T[] chars);
- Find and select a prior substring occurrence. Returns
true if found, false otherwise
- Text
append
(TextViewT other);
- Append text to this Text
- Text
append
(T[] chars);
- Append text to this Text
- Text
append
(T chr, int count = 1);
- Append a count of characters to this Text
- Text
append
(int v, T[] fmt = null);
- Append an integer to this Text
- Text
append
(long v, T[] fmt = null);
- Append a long to this Text
- Text
append
(double v, uint decimals = 2, int e = 10);
- Append a double to this Text
- Text
prepend
(T chr, int count = 1);
- Insert characters into this Text
- Text
prepend
(T[] other);
- Insert text into this Text
- Text
prepend
(TextViewT other);
- Insert another Text into this Text
- Text
encode
(char[] s);
Text
encode
(wchar[] s);
Text
encode
(dchar[] s);
Text
encode
(Object o);
- Append encoded text at the current selection point. The text
is converted as necessary to the appropritate utf encoding.
- Text
replace
(T chr);
- Replace a section of this Text with the specified
character
- Text
replace
(T[] chars);
- Replace a section of this Text with the specified
array
- Text
replace
(TextViewT other);
- Replace a section of this Text with another
- Text
remove
();
- Remove the selection from this Text and reset the
selection to zero length
- Text
remove
(int start, int count);
- Remove the selection from this Text
- Text
truncate
(int index = (int).max);
- Truncate this string at an optional index. Default behaviour
is to
truncate
at the current append point. Current selection
is moved to the truncation point, with length 0
- Text
clear
();
- Clear the string content
- Text
trim
();
- Remove leading and trailing whitespace from this Text,
and reset the selection to the trimmed content
- Text
strip
(T matches);
- Remove leading and trailing matches from this Text,
and reset the selection to the stripped content
- Text
reserve
(uint extra);
- Reserve some extra room
- TypeInfo
encoding
();
- Get the
encoding
type
- Comparator
comparator
(Comparator other);
- Set the
comparator
delegate. Where other is null, we behave
as a getter only
- hash_t
toHash
();
- Hash this Text
- uint
length
();
- Return the
length
of the valid content
- bool
equals
(TextViewT other);
- Is this Text equal to another?
- bool
equals
(T[] other);
- Is this Text equal to the provided text?
- bool
ends
(TextViewT other);
- Does this Text end with another?
- bool
ends
(T[] chars);
- Does this Text end with the specified string?
- bool
starts
(TextViewT other);
- Does this Text start with another?
- bool
starts
(T[] chars);
- Does this Text start with the specified string?
- int
compare
(TextViewT other);
- Compare this Text start with another. Returns 0 if the
content matches, less than zero if this Text is "less"
than the other, or greater than zero where this Text
is "bigger".
- int
compare
(T[] chars);
- Compare this Text start with an array. Returns 0 if the
content matches, less than zero if this Text is "less"
than the other, or greater than zero where this Text
is "bigger".
- T[]
copy
(T[] dst);
- Return content from this Text
A slice of dst is returned, representing a
copy
of the
content. The slice is clipped to the minimum of either
the length of the provided array, or the length of the
content minus the stipulated start point
- T[]
slice
();
- Return an alias to the content of this TextView. Note
that you are bound by honour to leave this content wholly
unmolested. D surely needs some way to enforce immutability
upon array references
- char[]
toString
(char[] dst = null);
wchar[]
toString16
(wchar[] dst = null);
dchar[]
toString32
(dchar[] dst = null);
- Convert to the UniText types. The optional argument
dst will be resized as required to house the conversion.
To minimize heap allocation during subsequent conversions,
apply the following pattern:
_Text string;
wchar[] buffer;
wchar[] result = string.utf16 (buffer);
if (result.length > buffer.length)
buffer = result;
You can also provide a buffer from the stack, but the output
will be moved to the heap if said buffer is not large enough
- int
opCmp
(Object o);
- Compare this Text to another. We compare against other
Strings only. Literals and other objects are not supported
- int
opEquals
(Object o);
int
opEquals
(T[] s);
- Is this Text equal to the text of something else?
- void
pinIndex
(ref int x);
- Pin the given index to a valid position.
- void
pinIndices
(ref int start, ref int length);
- Pin the given index and length to a valid position.
- int
simpleComparator
(T[] a, T[] b);
- Compare two arrays. Returns 0 if the content matches, less
than zero if A is "less" than B, or greater than zero where
A is "bigger". Where the substrings match, the shorter is
considered "less".
- void
expand
(uint index, uint count);
- Make room available to insert or append something
- Text
set
(T chr, uint start, uint count);
- Replace a section of this Text with the specified
character
- void
realloc
(uint count = 0);
- Allocate memory due to a change in the content. We handle
the distinction between mutable and immutable here.
- Text
append
(T* chars, uint count);
- Internal method to support Text appending
- class
TextView
(T): UniText;
- Immutable string
- uint
length
();
- Return the
length
of the valid content
- bool
equals
(TextView other);
- Is this Text equal to another?
- bool
equals
(T[] other);
- Is this Text equal to the the provided text?
- bool
ends
(TextView other);
- Does this Text end with another?
- bool
ends
(T[] chars);
- Does this Text end with the specified string?
- bool
starts
(TextView other);
- Does this Text start with another?
- bool
starts
(T[] chars);
- Does this Text start with the specified string?
- int
compare
(TextView other);
- Compare this Text start with another. Returns 0 if the
content matches, less than zero if this Text is "less"
than the other, or greater than zero where this Text
is "bigger".
- int
compare
(T[] chars);
- Compare this Text start with an array. Returns 0 if the
content matches, less than zero if this Text is "less"
than the other, or greater than zero where this Text
is "bigger".
- T[]
copy
(T[] dst);
- Return content from this Text. A slice of dst is
returned, representing a
copy
of the content. The
slice is clipped to the minimum of either the length
of the provided array, or the length of the content
minus the stipulated start point
- int
opCmp
(Object o);
- Compare this Text to another
- int
opEquals
(Object other);
- Is this Text equal to another?
- int
opEquals
(T[] other);
- Is this Text equal to another?
- TypeInfo
encoding
();
- Get the
encoding
type
- Comparator
comparator
(Comparator other);
- Set the
comparator
delegate
- hash_t
toHash
();
- Hash this Text
- T[]
slice
();
- Return an alias to the content of this TextView. Note
that you are bound by honour to leave this content wholly
unmolested. D surely needs some way to enforce immutability
upon array references
- class
UniText
;
- A string abstraction that converts to anything
|