erroneus std.string

// Written in the D programming language.

/**
String handling functions. Objects of types $(D _string), $(D
wstring), and $(D dstring) are value types and cannot be mutated
element-by-element. For using mutation during building strings, use
$(D char[]), $(D wchar[]), or $(D dchar[]). The $(D *_string) types
are preferable because they don't exhibit undesired aliasing, thus
making code more robust.

Macros: WIKI = Phobos/StdString

Copyright: Copyright Digital Mars 2007-.

License: $(WEB boost.org/LICENSE_1_0.txt, Boost License 1.0).

Authors: $(WEB digitalmars.com, Walter Bright),
         $(WEB erdani.org, Andrei Alexandrescu),
         and Jonathan M Davis

Source:    $(PHOBOSSRC std/_string.d)

*/
module std.string;

debug=string;                 // uncomment to turn on debugging printf's
debug(string) import core.stdc.stdio;

import core.exception : RangeError, onRangeError;
import core.vararg, core.stdc.stdlib, core.stdc.string,
    std.algorithm, std.ascii, std.conv, std.exception, std.format, std.functional,
    std.range, std.traits,
    std.typecons, std.typetuple, std.uni, std.utf;

//Remove when repeat is finally removed. They're only here as part of the
//deprecation of these functions in std.string.
public import std.algorithm : startsWith, endsWith, cmp, count;
public import std.array : join, split;

version(Windows) extern (C)
{
    size_t wcslen(in wchar *);
    int wcscmp(in wchar *, in wchar *);
}

version(unittest) import std.algorithm : filter;

/* ************* Exceptions *************** */

/++
    Exception thrown on errors in std.string functions.
  +/
class StringException : Exception
{
    /++
        Params:
            msg  = The message for the exception.
            file = The file where the exception occurred.
            line = The line number where the exception occurred.
            next = The previous exception in the chain of exceptions, if any.
      +/
    this(string msg,
         string file = __FILE__,
         size_t line = __LINE__,
         Throwable next = null) @safe pure nothrow
    {
        super(msg, file, line, next);
    }
}


/++
    Compares two ranges of characters lexicographically. The comparison is
    case insensitive. Use $(XREF algorithm, cmp) for a case sensitive
    comparison. For details see $(XREF uni, _icmp).

    $(BOOKTABLE,
        $(TR $(TD $(D < 0))  $(TD $(D s1 < s2) ))
        $(TR $(TD $(D = 0))  $(TD $(D s1 == s2)))
        $(TR $(TD $(D > 0))  $(TD $(D s1 > s2)))
     )
+/
alias icmp = std.uni.icmp;

unittest
{
    debug(string) printf("string.icmp.unittest\n");

    assertCTFEable!(
    {
    assert(icmp("Ü", "ü") == 0, "Über failure");
    assert(icmp("abc", "abc") == 0);
    assert(icmp("ABC", "abc") == 0);
    assert(icmp("abc"w, "abc") == 0);
    assert(icmp("ABC", "abc"w) == 0);
    assert(icmp("abc"d, "abc") == 0);
    assert(icmp("ABC", "abc"d) == 0);
    assert(icmp(cast(char[])"abc", "abc") == 0);
    assert(icmp("ABC", cast(char[])"abc") == 0);
    assert(icmp(cast(wchar[])"abc"w, "abc") == 0);
    assert(icmp("ABC", cast(wchar[])"abc"w) == 0);
    assert(icmp(cast(dchar[])"abc"d, "abc") == 0);
    assert(icmp("ABC", cast(dchar[])"abc"d) == 0);
    assert(icmp(cast(string)null, cast(string)null) == 0);
    assert(icmp("", "") == 0);
    assert(icmp("abc", "abcd") < 0);
    assert(icmp("abcd", "abc") > 0);
    assert(icmp("abc", "abd") < 0);
    assert(icmp("bbc", "abc") > 0);
    assert(icmp("abc", "abc"w) == 0);
    assert(icmp("ABC"w, "abc") == 0);
    assert(icmp("", ""w) == 0);
    assert(icmp("abc"w, "abcd") < 0);
    assert(icmp("abcd", "abc"w) > 0);
    assert(icmp("abc", "abd") < 0);
    assert(icmp("bbc"w, "abc") > 0);
    assert(icmp("aaa", "aaaa"d) < 0);
    assert(icmp("aaaa"w, "aaa"d) > 0);
    assert(icmp("aaa"d, "aaa"w) == 0);
    assert(icmp("\u0430\u0411\u0543"d, "\u0430\u0411\u0543") == 0);
    assert(icmp("\u0430\u0411\u0543"d, "\u0431\u0410\u0544") < 0);
    assert(icmp("\u0431\u0411\u0544"d, "\u0431\u0410\u0543") > 0);
    assert(icmp("\u0430\u0410\u0543"d, "\u0430\u0410\u0544") < 0);
    assert(icmp("\u0430\u0411\u0543"d, "\u0430\u0411\u0543\u0237") < 0);
    assert(icmp("\u0430\u0411\u0543\u0237"d, "\u0430\u0411\u0543") > 0);

    assert(icmp("aaa", filter!"true"("aaa")) == 0);
    assert(icmp(filter!"true"("aaa"), "aaa") == 0);
    assert(icmp(filter!"true"("aaa"), filter!"true"("aaa")) == 0);
    assert(icmp(filter!"true"("\u0430\u0411\u0543"d), "\u0430\u0411\u0543") == 0);
    assert(icmp(filter!"true"("\u0430\u0411\u0543"d), "\u0431\u0410\u0544"w) < 0);
    assert(icmp("\u0431\u0411\u0544"d, filter!"true"("\u0431\u0410\u0543"w)) > 0);
    assert(icmp("\u0430\u0410\u0543"d, filter!"true"("\u0430\u0410\u0544")) < 0);
    assert(icmp(filter!"true"("\u0430\u0411\u0543"d), filter!"true"("\u0430\u0411\u0543\u0237")) < 0);
    assert(icmp(filter!"true"("\u0430\u0411\u0543\u0237"d), filter!"true"("\u0430\u0411\u0543")) > 0);
    });
}


/++
    Returns a C-style zero-terminated string equivalent to $(D s). $(D s)
    must not contain embedded $(D '\0')'s as any C function will treat the first
    $(D '\0') that it sees as the end of the string. If $(D s.empty) is
    $(D true), then a string containing only $(D '\0') is returned.

    $(RED Important Note:) When passing a $(D char*) to a C function, and the C
    function keeps it around for any reason, make sure that you keep a reference
    to it in your D code. Otherwise, it may go away during a garbage collection
    cycle and cause a nasty bug when the C code tries to use it.
  +/
immutable(char)* toStringz(const(char)[] s) pure nothrow
in
{
    // The assert below contradicts the unittests!
    //assert(memchr(s.ptr, 0, s.length) == null,
    //text(s.length, ": `", s, "'"));
}
out (result)
{
    if (result)
    {
        auto slen = s.length;
        while (slen > 0 && s[slen-1] == 0) --slen;
        assert(strlen(result) == slen);
        assert(memcmp(result, s.ptr, slen) == 0);
    }
}
body
{
    /+ Unfortunately, this isn't reliable.
     We could make this work if string literals are put
     in read-only memory and we test if s[] is pointing into
     that.

     /* Peek past end of s[], if it's 0, no conversion necessary.
     * Note that the compiler will put a 0 past the end of static
     * strings, and the storage allocator will put a 0 past the end
     * of newly allocated char[]'s.
     */
     char* p = &s[0] + s.length;
     if (*p == 0)
     return s;
     +/

    // Need to make a copy
    auto copy = new char[s.length + 1];
    copy[0..s.length] = s[];
    copy[s.length] = 0;

    return assumeUnique(copy).ptr;
}

/++ Ditto +/
immutable(char)* toStringz(string s) pure nothrow
{
    if (s.empty) return "".ptr;
    /* Peek past end of s[], if it's 0, no conversion necessary.
     * Note that the compiler will put a 0 past the end of static
     * strings, and the storage allocator will put a 0 past the end
     * of newly allocated char[]'s.
     */
    immutable p = s.ptr + s.length;
    // Is p dereferenceable? A simple test: if the p points to an
    // address multiple of 4, then conservatively assume the pointer
    // might be pointing to a new block of memory, which might be
    // unreadable. Otherwise, it's definitely pointing to valid
    // memory.
    if ((cast(size_t) p & 3) && *p == 0)
        return s.ptr;
    return toStringz(cast(const char[]) s);
}

unittest
{
    debug(string) printf("string.toStringz.unittest\n");

    // TODO: CTFEable toStringz is really necessary?
    //assertCTFEable!(
    //{
    auto p = toStringz("foo");
    assert(strlen(p) == 3);
    const(char)[] foo = "abbzxyzzy";
    p = toStringz(foo[3..5]);
    assert(strlen(p) == 2);

    string test = "";
    p = toStringz(test);
    assert(*p == 0);

    test = "\0";
    p = toStringz(test);
    assert(*p == 0);

    test = "foo\0";
    p = toStringz(test);
    assert(p[0] == 'f' && p[1] == 'o' && p[2] == 'o' && p[3] == 0);
    //});
}
///convert a null terminated string in a D slice.No heap allocation is made
T[]fromStringz(T)(T*str,T term=T.init)pure nothrow
{
    return StringzRange(str,term).array;
}
///An input range iterating over a null terminated string.Because it does *not*
///compute the total length of the string it can be faster than fromStringz when
///you need only a input and a Forward range and nothing more.
struct StringzRange(T)
{
public:
    ///the pointer to the begin of the string
    T*ptr;
    ///the terminating character
    immutable T back;
    ///the default constructor.T.init is the null character when T is char
    this(T*pointer,T terminator=T.init)pure nothrow
    {
        ptr=pointer;
        back=terminator;
    }
    ///the input range interface
    @propriety bool empty()pure nothrow
    {
        return*ptr==back;
    }
    ///ditto
    @propriety ref inout(T) front()inout pure nothrow
    {
        return*ptr;
    }
    ///ditto
    void popFront()pure nothrow
    {
        ++ptr;
    }
    ///the Forward range interface
    @propriety inout(StringzRange!T)save()inout pure nothrow
    {
        return this;
    }
    ///convert the null terminated string in a D slice on the fly.No heap allocation is made
    @propriety T[]array()nothrow
    {
        T[]result;
        result.ptr=ptr;
        static if(is(T:const char))
            result.length=std.c.string.strlen(ptr);
        else
            result.length=walkLength(this);
        return result;
    }/*TODO:really needed implicit conversion to array?
    ///Rest of range interface
    alias array this;*/
}
unittest{}//TODO
/**
   Flag indicating whether a search is case-sensitive.
*/
enum CaseSensitive { no, yes }

/++
    Returns the index of the first occurrence of $(D c) in $(D s). If $(D c)
    is not found, then $(D -1) is returned.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t indexOf(S)(S s,ElementType!S c,CaseSensitive cs = CaseSensitive.yes) @safe pure
    if (isSomeString!S||( isInputRange!S && isSomeChar!(ElementType!S)))
{
    alias Char=ElementType!S;
    if (cs == CaseSensitive.yes)
    {
        static if (is(S==Char[]) && Char.sizeof == 1)
        {
            if (std.ascii.isASCII(c) && !__ctfe)
            {                                               // Plain old ASCII
                auto trustedmemchr() @trusted { return cast(Char*)memchr(s.ptr, c, s.length); }
                auto p = trustedmemchr();
                if (p)
                    return p - s.ptr;
                else
                    return -1;
            }
        }

        // c is a universal character
        foreach (ptrdiff_t i, dchar c2; s)
        {
            if (c == c2)
                return i;
        }
    }
    else
    {
        if (std.ascii.isASCII(c))
        {                                                   // Plain old ASCII
            auto c1 = cast(char) std.ascii.toLower(c);

            foreach (ptrdiff_t i, c2; s)
            {
                auto c3 = std.ascii.toLower(c2);
                if (c1 == c3)
                    return i;
            }
        }
        else
        {                                                   // c is a universal character
            auto c1 = std.uni.toLower(c);

            foreach (ptrdiff_t i, dchar c2; s)
            {
                auto c3 = std.uni.toLower(c2);
                if (c1 == c3)
                    return i;
            }
        }
    }
    return -1;
}

unittest
{
    debug(string) printf("string.indexOf.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(string, wstring, dstring))
    {
        assert(indexOf(cast(S)null, cast(dchar)'a') == -1);
        assert(indexOf(to!S("def"), cast(dchar)'a') == -1);
        assert(indexOf(to!S("abba"), cast(dchar)'a') == 0);
        assert(indexOf(to!S("def"), cast(dchar)'f') == 2);

        assert(indexOf(to!S("def"), cast(dchar)'a', CaseSensitive.no) == -1);
        assert(indexOf(to!S("def"), cast(dchar)'a', CaseSensitive.no) == -1);
        assert(indexOf(to!S("Abba"), cast(dchar)'a', CaseSensitive.no) == 0);
        assert(indexOf(to!S("def"), cast(dchar)'F', CaseSensitive.no) == 2);
        assert(indexOf(to!S("ödef"), 'ö', CaseSensitive.no) == 0);

        S sPlts = "Mars: the fourth Rock (Planet) from the Sun.";
        assert(indexOf("def", cast(char)'f', CaseSensitive.no) == 2);
        assert(indexOf(sPlts, cast(char)'P', CaseSensitive.no) == 23);
        assert(indexOf(sPlts, cast(char)'R', CaseSensitive.no) == 2);
    }

    foreach (cs; EnumMembers!CaseSensitive)
    {
        assert(indexOf("hello\U00010143\u0100\U00010143", '\u0100', cs) == 9);
        assert(indexOf("hello\U00010143\u0100\U00010143"w, '\u0100', cs) == 7);
        assert(indexOf("hello\U00010143\u0100\U00010143"d, '\u0100', cs) == 6);
    }
    });
}

/++
    Returns the index of the first occurrence of $(D c) in $(D s) with respect
    to the start index $(D startIdx). If $(D c) is not found, then $(D -1) is
    returned. If $(D c) is found the value of the returned index is at least
    $(D startIdx). $(D startIdx) represents a codeunit index in $(D s). If the
    sequence starting at $(D startIdx) does not represent a well formed codepoint,
    then a $(XREF utf,UTFException) may be thrown.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t indexOf(S)(S s, ElementType!S c, const size_t startIdx,
        CaseSensitive cs = CaseSensitive.yes) @safe pure
    if (isSomeString!S||(isInputRange!S &&  isSomeChar!(ElementType!S)))
{
    static if(hasLenght!S)
    {
        if (startIdx >= s.length)return -1;
        s.popFrontExactly(startIdx);
    }
    else
    {
        s.popFrontN(startIdx);
    }
    ptrdiff_t foundIdx = indexOf(s, c, cs);
    if (foundIdx == -1)return-1;
    return foundIdx + cast(ptrdiff_t)startIdx;
}

unittest
{
    debug(string) printf("string.indexOf(startIdx).unittest\n");

    foreach (S; TypeTuple!(string, wstring, dstring))
    {
        assert(indexOf(cast(S)null, cast(dchar)'a', 1) == -1);
        assert(indexOf(to!S("def"), cast(dchar)'a', 1) == -1);
        assert(indexOf(to!S("abba"), cast(dchar)'a', 1) == 3);
        assert(indexOf(to!S("def"), cast(dchar)'f', 1) == 2);

        assert((to!S("def")).indexOf(cast(dchar)'a', 1,
                CaseSensitive.no) == -1);
        assert(indexOf(to!S("def"), cast(dchar)'a', 1,
                CaseSensitive.no) == -1);
        assert(indexOf(to!S("def"), cast(dchar)'a', 12,
                CaseSensitive.no) == -1);
        assert(indexOf(to!S("AbbA"), cast(dchar)'a', 2,
                CaseSensitive.no) == 3);
        assert(indexOf(to!S("def"), cast(dchar)'F', 2, CaseSensitive.no) == 2);

        S sPlts = "Mars: the fourth Rock (Planet) from the Sun.";
        assert(indexOf("def", cast(char)'f', cast(uint)2,
            CaseSensitive.no) == 2);
        assert(indexOf(sPlts, cast(char)'P', 12, CaseSensitive.no) == 23);
        assert(indexOf(sPlts, cast(char)'R', cast(ulong)1,
            CaseSensitive.no) == 2);
    }

    foreach(cs; EnumMembers!CaseSensitive)
    {
        assert(indexOf("hello\U00010143\u0100\U00010143", '\u0100', 2, cs)
            == 9);
        assert(indexOf("hello\U00010143\u0100\U00010143"w, '\u0100', 3, cs)
            == 7);
        assert(indexOf("hello\U00010143\u0100\U00010143"d, '\u0100', 6, cs)
            == 6);
    }
}

/++
    Returns the index of the first occurrence of $(D sub) in $(D s). If $(D sub)
    is not found, then $(D -1) is returned.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t indexOf(S1, S2)(S1 s,S2 sub,CaseSensitive cs = CaseSensitive.yes)
    if (is(Unqual!(ElementType!S1)==Unqual!(ElementType!S2))
        &&(isSomeString!S1||(isSomeChar!(ElementType!S1)
        &&isForwardRange!S1&&isForwardRange!S2)))
{
    S1 balance;
    if (cs == CaseSensitive.yes)
    {
        balance = std.algorithm.find(s, sub);
    }
    else
    {
        balance = std.algorithm.find!
            ((a, b) => std.uni.toLower(a) == std.uni.toLower(b))
            (s, sub);
    }
    if(balance.empty)return-1;
    static if(is(S1==ElementType!S1[]))
        return s.ptr-balance.ptr;
    return indexOf(s,balance.front,cs);//TODO:implement a more efficient algorithm for this case
}

unittest
{
    debug(string) printf("string.indexOf.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(string, wstring, dstring))
    {
        foreach (T; TypeTuple!(string, wstring, dstring))
        {
            assert(indexOf(cast(S)null, to!T("a")) == -1);
            assert(indexOf(to!S("def"), to!T("a")) == -1);
            assert(indexOf(to!S("abba"), to!T("a")) == 0);
            assert(indexOf(to!S("def"), to!T("f")) == 2);
            assert(indexOf(to!S("dfefffg"), to!T("fff")) == 3);
            assert(indexOf(to!S("dfeffgfff"), to!T("fff")) == 6);

            assert(indexOf(to!S("dfeffgfff"), to!T("a"), CaseSensitive.no) == -1);
            assert(indexOf(to!S("def"), to!T("a"), CaseSensitive.no) == -1);
            assert(indexOf(to!S("abba"), to!T("a"), CaseSensitive.no) == 0);
            assert(indexOf(to!S("def"), to!T("f"), CaseSensitive.no) == 2);
            assert(indexOf(to!S("dfefffg"), to!T("fff"), CaseSensitive.no) == 3);
            assert(indexOf(to!S("dfeffgfff"), to!T("fff"), CaseSensitive.no) == 6);

            S sPlts = "Mars: the fourth Rock (Planet) from the Sun.";
            S sMars = "Who\'s \'My Favorite Maritian?\'";

            assert(indexOf(sMars, to!T("MY fAVe"), CaseSensitive.no) == -1);
            assert(indexOf(sMars, to!T("mY fAVOriTe"), CaseSensitive.no) == 7);
            assert(indexOf(sPlts, to!T("mArS:"), CaseSensitive.no) == 0);
            assert(indexOf(sPlts, to!T("rOcK"), CaseSensitive.no) == 17);
            assert(indexOf(sPlts, to!T("Un."), CaseSensitive.no) == 41);
            assert(indexOf(sPlts, to!T(sPlts), CaseSensitive.no) == 0);

            assert(indexOf("\u0100", to!T("\u0100"), CaseSensitive.no) == 0);

            // Thanks to Carlos Santander B. and zwang
            assert(indexOf("sus mejores cortesanos. Se embarcaron en el puerto de Dubai y",
                           to!T("page-break-before"), CaseSensitive.no) == -1);
        }

        foreach (cs; EnumMembers!CaseSensitive)
        {
            assert(indexOf("hello\U00010143\u0100\U00010143", to!S("\u0100"), cs) == 9);
            assert(indexOf("hello\U00010143\u0100\U00010143"w, to!S("\u0100"), cs) == 7);
            assert(indexOf("hello\U00010143\u0100\U00010143"d, to!S("\u0100"), cs) == 6);
        }
    }
    });
}

/++
    Returns the index of the first occurrence of $(D sub) in $(D s) with
    respect to the start index $(D startIdx). If $(D sub) is not found, then
    $(D -1) is returned. If $(D sub) is found the value of the returned index
    is at least $(D startIdx). $(D startIdx) represents a codeunit index in
    $(D s). If the sequence starting at $(D startIdx) does not represent a well
    formed codepoint, then a $(XREF utf,UTFException) may be thrown.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t indexOf(S1, S2)(S1 s, S2 sub,
        const size_t startIdx, CaseSensitive cs = CaseSensitive.yes)
    if (isSomeString!S||(isForwardRange!S1
        && isForwardRange!S2 && isSomeChar!(ElementType!S1)))
{
    static if(hasLenght!S1)
    {
        if (startIdx >= s.length)return -1;
        s.popFrontExactly(startIdx);
    }
    else
    {
        s.popFrontN(startIdx);
    }
    ptrdiff_t foundIdx = indexOf(s,sub, cs);
    if (foundIdx == -1)return-1;
    return foundIdx + cast(ptrdiff_t)startIdx;
}

unittest
{
    debug(string) printf("string.indexOf(startIdx).unittest\n");

    foreach(S; TypeTuple!(string, wstring, dstring))
    {
        foreach(T; TypeTuple!(string, wstring, dstring))
        {
            assert(indexOf(cast(S)null, to!T("a"), 1337) == -1);
            assert(indexOf(to!S("def"), to!T("a"), 0) == -1);
            assert(indexOf(to!S("abba"), to!T("a"), 2) == 3);
            assert(indexOf(to!S("def"), to!T("f"), 1) == 2);
            assert(indexOf(to!S("dfefffg"), to!T("fff"), 1) == 3);
            assert(indexOf(to!S("dfeffgfff"), to!T("fff"), 5) == 6);

            assert(indexOf(to!S("dfeffgfff"), to!T("a"), 1, CaseSensitive.no) == -1);
            assert(indexOf(to!S("def"), to!T("a"), 2, CaseSensitive.no) == -1);
            assert(indexOf(to!S("abba"), to!T("a"), 3, CaseSensitive.no) == 3);
            assert(indexOf(to!S("def"), to!T("f"), 1, CaseSensitive.no) == 2);
            assert(indexOf(to!S("dfefffg"), to!T("fff"), 2, CaseSensitive.no) == 3);
            assert(indexOf(to!S("dfeffgfff"), to!T("fff"), 4, CaseSensitive.no) == 6);
            assert(indexOf(to!S("dfeffgffföä"), to!T("öä"), 9, CaseSensitive.no) == 9,
                to!string(indexOf(to!S("dfeffgffföä"), to!T("öä"), 9, CaseSensitive.no))
                ~ " " ~ S.stringof ~ " " ~ T.stringof);

            S sPlts = "Mars: the fourth Rock (Planet) from the Sun.";
            S sMars = "Who\'s \'My Favorite Maritian?\'";

            assert(indexOf(sMars, to!T("MY fAVe"), 10,
                CaseSensitive.no) == -1);
            assert(indexOf(sMars, to!T("mY fAVOriTe"), 4, CaseSensitive.no) == 7);
            assert(indexOf(sPlts, to!T("mArS:"), 0, CaseSensitive.no) == 0);
            assert(indexOf(sPlts, to!T("rOcK"), 12, CaseSensitive.no) == 17);
            assert(indexOf(sPlts, to!T("Un."), 32, CaseSensitive.no) == 41);
            assert(indexOf(sPlts, to!T(sPlts), 0, CaseSensitive.no) == 0);

            assert(indexOf("\u0100", to!T("\u0100"), 0, CaseSensitive.no) == 0);

            // Thanks to Carlos Santander B. and zwang
            assert(indexOf("sus mejores cortesanos. Se embarcaron en el puerto de Dubai y",
                           to!T("page-break-before"), 10, CaseSensitive.no) == -1);

            // In order for indexOf with and without index to be consistent
            assert(indexOf(to!S(""), to!T("")) == indexOf(to!S(""), to!T(""), 0));
        }

        foreach(cs; EnumMembers!CaseSensitive)
        {
            assert(indexOf("hello\U00010143\u0100\U00010143", to!S("\u0100"),
                3, cs) == 9);
            assert(indexOf("hello\U00010143\u0100\U00010143"w, to!S("\u0100"),
                3, cs) == 7);
            assert(indexOf("hello\U00010143\u0100\U00010143"d, to!S("\u0100"),
                3, cs) == 6);
        }
    }
}

/++
    Returns the index of the last occurrence of $(D c) in $(D s). If $(D c)
    is not found, then $(D -1) is returned.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t lastIndexOf(S)(S s,ElementType!S c,CaseSensitive cs = CaseSensitive.yes) @safe pure
    if (isSomeString!S||(isBidirectionalRange!S
        &&hasLenght!S&&isSomeChar!(ElementType!S)))
{
    uint i=indexOf(s.reverse(),c,cs);
    return i==-1?-1:s.length-i-1;
}

unittest
{
    debug(string) printf("string.lastIndexOf.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(string, wstring, dstring))
    {
        assert(lastIndexOf(cast(S) null, 'a') == -1);
        assert(lastIndexOf(to!S("def"), 'a') == -1);
        assert(lastIndexOf(to!S("abba"), 'a') == 3);
        assert(lastIndexOf(to!S("def"), 'f') == 2);
        assert(lastIndexOf(to!S("ödef"), 'ö') == 0);

        assert(lastIndexOf(cast(S) null, 'a', CaseSensitive.no) == -1);
        assert(lastIndexOf(to!S("def"), 'a', CaseSensitive.no) == -1);
        assert(lastIndexOf(to!S("AbbA"), 'a', CaseSensitive.no) == 3);
        assert(lastIndexOf(to!S("def"), 'F', CaseSensitive.no) == 2);
        assert(lastIndexOf(to!S("ödef"), 'ö', CaseSensitive.no) == 0);
        assert(lastIndexOf(to!S("i\u0100def"), to!dchar("\u0100"),
            CaseSensitive.no) == 1);

        S sPlts = "Mars: the fourth Rock (Planet) from the Sun.";

        assert(lastIndexOf(to!S("def"), 'f', CaseSensitive.no) == 2);
        assert(lastIndexOf(sPlts, 'M', CaseSensitive.no) == 34);
        assert(lastIndexOf(sPlts, 'S', CaseSensitive.no) == 40);
    }

    foreach (cs; EnumMembers!CaseSensitive)
    {
        assert(lastIndexOf("\U00010143\u0100\U00010143hello", '\u0100', cs) == 4);
        assert(lastIndexOf("\U00010143\u0100\U00010143hello"w, '\u0100', cs) == 2);
        assert(lastIndexOf("\U00010143\u0100\U00010143hello"d, '\u0100', cs) == 1);
    }
    });
}

/++
    Returns the index of the last occurrence of $(D c) in $(D s). If $(D c) is
    not found, then $(D -1) is returned. The $(D startIdx) slices $(D s) in
    the following way $(D s[0 .. startIdx]). $(D startIdx) represents a
    codeunit index in $(D s). If the sequence ending at $(D startIdx) does not
    represent a well formed codepoint, then a $(XREF utf,UTFException) may be
    thrown.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t lastIndexOf(S)(S s, ElementType!S c, const size_t startIdx,
        CaseSensitive cs = CaseSensitive.yes) @safe pure
    if (isSomeString!S||(isBidirectionalRange!S && hasLenght!S && isSomeChar!S))
{
    uint i=indexOf(s.reverse(),c,startIdx,cs);
    return i==-1?-1:s.length-i-1;
}

unittest
{
    debug(string) printf("string.lastIndexOf.unittest\n");

    foreach(S; TypeTuple!(string, wstring, dstring))
    {
        assert(lastIndexOf(cast(S) null, 'a') == -1);
        assert(lastIndexOf(to!S("def"), 'a') == -1);
        assert(lastIndexOf(to!S("abba"), 'a', 3) == 0);
        assert(lastIndexOf(to!S("deff"), 'f', 3) == 2);

        assert(lastIndexOf(cast(S) null, 'a', CaseSensitive.no) == -1);
        assert(lastIndexOf(to!S("def"), 'a', CaseSensitive.no) == -1);
        assert(lastIndexOf(to!S("AbbAa"), 'a', to!ushort(4), CaseSensitive.no) == 3,
                to!string(lastIndexOf(to!S("AbbAa"), 'a', 4, CaseSensitive.no)));
        assert(lastIndexOf(to!S("def"), 'F', 3, CaseSensitive.no) == 2);

        S sPlts = "Mars: the fourth Rock (Planet) from the Sun.";

        assert(lastIndexOf(to!S("def"), 'f', 4, CaseSensitive.no) == -1);
        assert(lastIndexOf(sPlts, 'M', sPlts.length -2, CaseSensitive.no) == 34);
        assert(lastIndexOf(sPlts, 'S', sPlts.length -2, CaseSensitive.no) == 40);
    }

    foreach(cs; EnumMembers!CaseSensitive)
    {
        assert(lastIndexOf("\U00010143\u0100\U00010143hello", '\u0100', cs) == 4);
        assert(lastIndexOf("\U00010143\u0100\U00010143hello"w, '\u0100', cs) == 2);
        assert(lastIndexOf("\U00010143\u0100\U00010143hello"d, '\u0100', cs) == 1);
    }
}

/++
    Returns the index of the last occurrence of $(D sub) in $(D s). If $(D sub)
    is not found, then $(D -1) is returned.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t lastIndexOf(S1,S2)(S1 s,S2 sub,CaseSensitive cs = CaseSensitive.yes) @safe pure
    if(isBidirectionalRange!S1 && isBidirectionalRange!S2
        && isForwardRange!S1 && isForwardRange!S2 &&hasLenght!S1)
{
    uint i=indexOf(s.reverse(),sub.reverse(),cs);
    return i==-1?-1:s.length-i-1;
}
unittest
{
    debug(string) printf("string.lastIndexOf.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(string, wstring, dstring))
    {
        foreach (T; TypeTuple!(string, wstring, dstring))
        {
            enum typeStr = S.stringof ~ " " ~ T.stringof;

            assert(lastIndexOf(cast(S)null, to!T("a")) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("c")) == 6, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("cd")) == 6, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("ef")) == 8, typeStr);
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("c")) == 2, typeStr);
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("cd")) == 2, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("x")) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("xy")) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("")) == 10, typeStr);
            assert(lastIndexOf(to!S("öabcdefcdef"), to!T("ö")) == 0, typeStr);

            assert(lastIndexOf(cast(S)null, to!T("a"), CaseSensitive.no) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("c"), CaseSensitive.no) == 6, typeStr);
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("cD"), CaseSensitive.no) == 6, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("x"), CaseSensitive.no) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("xy"), CaseSensitive.no) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T(""), CaseSensitive.no) == 10, typeStr);
            assert(lastIndexOf(to!S("öabcdefcdef"), to!T("ö"), CaseSensitive.no) == 0, typeStr);

            assert(lastIndexOf(to!S("abcdefcdef"), to!T("c"), CaseSensitive.no) == 6, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("cd"), CaseSensitive.no) == 6, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("def"), CaseSensitive.no) == 7, typeStr);

            assert(lastIndexOf(to!S("ödfeffgfff"), to!T("ö"), CaseSensitive.yes) == 0);

            S sPlts = "Mars: the fourth Rock (Planet) from the Sun.";
            S sMars = "Who\'s \'My Favorite Maritian?\'";

            assert(lastIndexOf(sMars, to!T("RiTE maR"), CaseSensitive.no) == 14, typeStr);
            assert(lastIndexOf(sPlts, to!T("FOuRTh"), CaseSensitive.no) == 10, typeStr);
            assert(lastIndexOf(sMars, to!T("whO\'s \'MY"), CaseSensitive.no) == 0, typeStr);
            assert(lastIndexOf(sMars, to!T(sMars), CaseSensitive.no) == 0, typeStr);
        }

        foreach (cs; EnumMembers!CaseSensitive)
        {
            enum csString = to!string(cs);

            assert(lastIndexOf("\U00010143\u0100\U00010143hello", to!S("\u0100"), cs) == 4, csString);
            assert(lastIndexOf("\U00010143\u0100\U00010143hello"w, to!S("\u0100"), cs) == 2, csString);
            assert(lastIndexOf("\U00010143\u0100\U00010143hello"d, to!S("\u0100"), cs) == 1, csString);
        }
    }
    });
}
unittest{}//TODO
/++
    Returns the index of the last occurrence of $(D sub) in $(D s). If $(D sub)
    is not found, then $(D -1) is returned. The $(D startIdx) slices $(D s) in
    the following way $(D s[0 .. startIdx]). $(D startIdx) represents a
    codeunit index in $(D s). If the sequence ending at $(D startIdx) does not
    represent a well formed codepoint, then a $(XREF utf,UTFException) may be
    thrown.

    $(D cs) indicates whether the comparisons are case sensitive.
  +/
ptrdiff_t lastIndexOf(S1, S2)(S1 s, S2 sub,
        const size_t startIdx, CaseSensitive cs = CaseSensitive.yes) @safe pure
    if (isSomeString!S1||(isBidirectionalRange!S1 && isForwardRange!S2 && isForwardRange!S1))
{
    static if(hasLenght!S1)
    {
        if(str.length<=startIdx)return-1;
        str.popBackExactly(startIdx);
    }
    else
        str.popBackN(startIdx);
    return lastIndexOf(s,sub,cs);
}

unittest
{
    debug(string) printf("string.lastIndexOf.unittest\n");

    foreach(S; TypeTuple!(string, wstring, dstring))
    {
        foreach(T; TypeTuple!(string, wstring, dstring))
        {
            enum typeStr = S.stringof ~ " " ~ T.stringof;

            assert(lastIndexOf(cast(S)null, to!T("a")) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("c"), 5) == 2, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("cd"), 3) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("ef"), 6) == 4, typeStr ~
                format(" %u", lastIndexOf(to!S("abcdefcdef"), to!T("ef"), 6)));
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("c"), 5) == 2, typeStr);
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("cd"), 3) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdefx"), to!T("x"), 1) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdefxy"), to!T("xy"), 6) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T(""), 8) == 8, typeStr);
            assert(lastIndexOf(to!S("öafö"), to!T("ö"), 3) == 0, typeStr ~
                    to!string(lastIndexOf(to!S("öafö"), to!T("ö"), 3))); //BUG 10472

            assert(lastIndexOf(cast(S)null, to!T("a"), 1, CaseSensitive.no) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("c"), 5, CaseSensitive.no) == 2, typeStr);
            assert(lastIndexOf(to!S("abcdefCdef"), to!T("cD"), 4, CaseSensitive.no) == 2, typeStr ~
                " " ~ to!string(lastIndexOf(to!S("abcdefCdef"), to!T("cD"), 3, CaseSensitive.no)));
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("x"),3 , CaseSensitive.no) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdefXY"), to!T("xy"), 4, CaseSensitive.no) == -1, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T(""), 7, CaseSensitive.no) == 7, typeStr);

            assert(lastIndexOf(to!S("abcdefcdef"), to!T("c"), 4, CaseSensitive.no) == 2, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("cd"), 4, CaseSensitive.no) == 2, typeStr);
            assert(lastIndexOf(to!S("abcdefcdef"), to!T("def"), 6, CaseSensitive.no) == 3, typeStr);
            assert(lastIndexOf(to!S(""), to!T(""), 0) == lastIndexOf(to!S(""), to!T("")), typeStr);
        }

        foreach(cs; EnumMembers!CaseSensitive)
        {
            enum csString = to!string(cs);

            assert(lastIndexOf("\U00010143\u0100\U00010143hello", to!S("\u0100"), 6, cs) == 4, csString);
            assert(lastIndexOf("\U00010143\u0100\U00010143hello"w, to!S("\u0100"), 6, cs) == 2, csString);
            assert(lastIndexOf("\U00010143\u0100\U00010143hello"d, to!S("\u0100"), 3, cs) == 1, csString);
        }
    }
}


/**
 * Returns the representation of a string, which has the same type
 * as the string except the character type is replaced by $(D ubyte),
 * $(D ushort), or $(D uint) depending on the character width.
 */
auto representation(Char)(Char[] s) pure nothrow
    if (isSomeChar!Char)
{
    // Get representation type
    alias U = TypeTuple!(ubyte, ushort, uint)[Char.sizeof / 2];

    // const and immutable storage classes
    static if (is(Char == immutable))
        alias T = immutable(U);
    else static if (is(Char == const))
        alias T = const(U);
    else
        alias T = U;

    // shared storage class (because shared(const(T)) is possible)
    static if (is(Char == shared))
        alias ST = shared(T);
    else
        alias ST = T;

    return cast(ST[]) s;
}
///ditto
alias Representation(Range)=typeof([Range.init.front].representation());
///
unittest
{
    string s = "hello";
    static assert(is(typeof(representation(s)) == immutable(ubyte)[]));
    assert(representation(s) is cast(immutable(ubyte)[]) s);
    assert(representation(s) == [0x68, 0x65, 0x6c, 0x6c, 0x6f]);
}
unittest
{
    assertCTFEable!(
    {
    void test(Char, T)(Char[] str)
    {
        static assert(is(typeof(representation(str)) == T[]));
        assert(representation(str) is cast(T[]) str);
    }

    foreach (Type; TypeTuple!(Tuple!(char , ubyte ),
                              Tuple!(wchar, ushort),
                              Tuple!(dchar, uint  )))
    {
        alias Char = FieldTypeTuple!Type[0];
        alias Int  = FieldTypeTuple!Type[1];
        enum immutable(Char)[] hello = "hello";

        test!(   immutable Char,    immutable Int)(hello);
        test!(       const Char,        const Int)(hello);
        test!(             Char,              Int)(hello.dup);
        test!(      shared Char,       shared Int)(cast(shared) hello.dup);
        test!(const shared Char, const shared Int)(hello);
    }
    });
}


/++
    Returns a string which is identical to $(D s) except that all of its
    characters are converted to lowercase (by preforming Unicode lowercase mapping).
    If none of $(D s) characters were affected, then $(D s) itself is returned.
  +/
alias toLower = std.uni.toLower;
/++
    Converts $(D s) to lowercase (by performing Unicode lowercase mapping) in place.
    For a few characters string length may increase after the transformation,
    in such a case the function reallocates exactly once.
    If $(D s) does not have any uppercase characters, then $(D s) is unaltered.
 +/
alias toLowerInPlace = std.uni.toLowerInPlace;

/++
    Returns a string which is identical to $(D s) except that all of its
    characters are converted to uppercase (by preforming Unicode uppercase mapping).
    If none of $(D s) characters were affected, then $(D s) itself is returned.
  +/
alias toUpper = std.uni.toUpper;

/++
    Converts $(D s) to uppercase  (by performing Unicode uppercase mapping) in place.
    For a few characters string length may increase after the transformation,
    in such a case the function reallocates exactly once.
    If $(D s) does not have any lowercase characters, then $(D s) is unaltered.
 +/
alias toUpperInPlace = std.uni.toUpperInPlace;

/++
    Capitalize the first character of $(D s) and convert the rest of $(D s)
    to lowercase.
 +/
S capitalize(S)(S s) @trusted pure
    if (isSomeString!S)
{
    Unqual!(typeof(s[0]))[] retval;
    bool changed = false;

    foreach (i, dchar c; s)
    {
        dchar c2;

        if (i == 0)
        {
            c2 = std.uni.toUpper(c);
            if (c != c2)
                changed = true;
        }
        else
        {
            c2 = std.uni.toLower(c);
            if (c != c2)
            {
                if (!changed)
                {
                    changed = true;
                    retval = s[0 .. i].dup;
                }
            }
        }

        if (changed)
            std.utf.encode(retval, c2);
    }

    return changed ? cast(S)retval : s;
}

unittest
{
    assertCTFEable!(
    {
    foreach (S; TypeTuple!(string, wstring, dstring, char[], wchar[], dchar[]))
    {
        S s1 = to!S("FoL");
        S s2;

        s2 = capitalize(s1);
        assert(cmp(s2, "Fol") == 0);
        assert(s2 !is s1);

        s2 = capitalize(s1[0 .. 2]);
        assert(cmp(s2, "Fo") == 0);
        assert(s2.ptr == s1.ptr);

        s1 = to!S("fOl");
        s2 = capitalize(s1);
        assert(cmp(s2, "Fol") == 0);
        assert(s2 !is s1);
        s1 = to!S("\u0131 \u0130");
        s2 = capitalize(s1);
        assert(cmp(s2, "I \u0130") == 0);
        assert(s2 !is s1);

        s1 = to!S("\u017F \u0049");
        s2 = capitalize(s1);
        assert(cmp(s2, "\u0053 \u0069") == 0);
        assert(s2 !is s1);
    }
    });
}
///ditto
struct Capitalize(R)
{
private:
    typeof(R.init.map!std.uni.toLower())rg;
    bool isFirst=true;
public:
    alias rg this;
    this(R r)pure
    {
        rg=r.map!std.uni.toLower();
    }
    @propriety auto front()
    {
        if(isFirst)return rg.front.toUpper();
        else return rg.front;
    }
    void popFront()
    {
        isFirst=false;
        rg.popFront();
    }
    static if(isRandomAccessRange!R)
    {
        auto opIndex(size_t i)
        {
            if(i==0)return rg.front.toUpper();
            else return rg[i];
        }
    }
    static if(hasSlicing!R)
        Capitalize!R opSlice(size_t i,size_t j)
        {
            Capitalize!R result;
            result.rg=rg[i..j];
            result.isFirst= i==0;
            return result;
        }
}
///ditto
Capitalize!R capitalize(R)(R r)@safe pure if(isInputRange!R)
{
    return typeof(return)(r);
}
unittest{}//TODO
/++
    Split $(D s) into an array of lines using $(D '\r'), $(D '\n'),
    $(D "\r\n"), $(XREF uni, lineSep), and $(XREF uni, paraSep) as delimiters.
    If $(D keepTerm) is set to $(D KeepTerminator.yes), then the delimiter
    is included in the strings returned.
  +/
enum KeepTerminator : bool { no, yes }
/// ditto
S[] splitLines(S)(S s, KeepTerminator keepTerm = KeepTerminator.no) @safe pure
    if (isSomeString!S)
{
    size_t iStart = 0;
    size_t nextI = 0;
    auto retval = appender!(S[])();

    for (size_t i; i < s.length; i = nextI)
    {
        immutable c = decode(s, nextI);

        if (c == '\r' || c == '\n' || c == lineSep || c == paraSep)
        {
            immutable isWinEOL = c == '\r' && i + 1 < s.length && s[i + 1] == '\n';
            auto iEnd = i;

            if (keepTerm == KeepTerminator.yes)
            {
                iEnd = isWinEOL? nextI + 1 : nextI;
            }

            retval.put(s[iStart .. iEnd]);
            iStart = nextI;

            if (isWinEOL)
            {
                ++nextI;
                ++iStart;
            }
        }
    }

    if (iStart != nextI)
        retval.put(s[iStart .. $]);

    return retval.data;
}

unittest
{
    debug(string) printf("string.splitLines.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
    {
        auto s = to!S("\rpeter\n\rpaul\r\njerry\u2028ice\u2029cream\n\nsunday\n");

        auto lines = splitLines(s);
        assert(lines.length == 9);
        assert(lines[0] == "");
        assert(lines[1] == "peter");
        assert(lines[2] == "");
        assert(lines[3] == "paul");
        assert(lines[4] == "jerry");
        assert(lines[5] == "ice");
        assert(lines[6] == "cream");
        assert(lines[7] == "");
        assert(lines[8] == "sunday");

        lines = splitLines(s, KeepTerminator.yes);
        assert(lines.length == 9);
        assert(lines[0] == "\r");
        assert(lines[1] == "peter\n");
        assert(lines[2] == "\r");
        assert(lines[3] == "paul\r\n");
        assert(lines[4] == "jerry\u2028");
        assert(lines[5] == "ice\u2029");
        assert(lines[6] == "cream\n");
        assert(lines[7] == "\n");
        assert(lines[8] == "sunday\n");

        s.popBack(); // Lop-off trailing \n
        lines = splitLines(s);
        assert(lines.length == 9);
        assert(lines[8] == "sunday");

        lines = splitLines(s, KeepTerminator.yes);
        assert(lines.length == 9);
        assert(lines[8] == "sunday");
    }
    });
}
/+
///ditto
auto splitLines(R)(R r,KeepTerminator keepTerm=KeepTerminator.no)@safe pure
    if(isInputRange!R&&hasLvalueElements!R)
{
    alias Ret=typeof(return);
    immutable nl="\n";//TODO:newline based on os' own newline
    static auto sanitize(Ror)(Ror ror,bool keep)
    {
        if(keep){
            static if(isBidirectionalRange!Ror&& isBidirectionalRange!Ret)
                return ror.filterBidirectional!(a=>a!="")();
            else static if(!isBidirectionalRange!Ret)
                return ror.filter!(a=>a!="")();
            else static assert(0,"can't give a bidirectional range from a non-bidirectional one")
        }
        else
        {
            if(hasLvalueElements!Ror)assert(0);
            typeof(ror.front)*prev;
            foreach(el;ror)
            {
                if(el.empty)*prev~=nl;
                else prev=&el;
            }
            return ror;
        }
    }
    return r.splitter(nl).sanitize(keepTerm==KeepTerminator.yes);
}
unittest{}//TODO
+/
/++
    Strips leading whitespace (as defined by $(XREF uni, isWhite)).

    Returns: $(D str) stripped of leading whitespace.

    Postconditions: $(D str) and the returned value
    will share the same tail (see $(XREF array, sameTail)).
  +/
auto stripLeft(S)(S str) @safe pure
    if (isInputRange!S &&isSomeChar!(ElementType!S))
{
    return str.stripLeft!std.uni.isWhite();
}

///
@safe pure unittest
{
    assert(stripLeft("     hello world     ") ==
           "hello world     ");
    assert(stripLeft("\n\t\v\rhello world\n\t\v\r") ==
           "hello world\n\t\v\r");
    assert(stripLeft("hello world") ==
           "hello world");
    assert(stripLeft([lineSep] ~ "hello world" ~ lineSep) ==
           "hello world" ~ [lineSep]);
    assert(stripLeft([paraSep] ~ "hello world" ~ paraSep) ==
           "hello world" ~ [paraSep]);
}


/++
    Strips trailing whitespace (as defined by $(XREF uni, isWhite)).

    Returns: $(D str) stripped of trailing whitespace.

    Postconditions: $(D str) and the returned value
    will share the same head (see $(XREF array, sameHead)).
  +/
auto stripRight(S)(S str) @safe pure
    if (isBidirectionalRange!S&&isSomeChar!(ElementType!S))
{
    return str.stripRight!std.uni.isWhite();
}

///
@safe pure unittest
{
    assert(stripRight("     hello world     ") ==
           "     hello world");
    assert(stripRight("\n\t\v\rhello world\n\t\v\r") ==
           "\n\t\v\rhello world");
    assert(stripRight("hello world") ==
           "hello world");
    assert(stripRight([lineSep] ~ "hello world" ~ lineSep) ==
           [lineSep] ~ "hello world");
    assert(stripRight([paraSep] ~ "hello world" ~ paraSep) ==
           [paraSep] ~ "hello world");
}


/++
    Strips both leading and trailing whitespace (as defined by
    $(XREF uni, isWhite)).

    Returns: $(D str) stripped of trailing whitespace.
  +/
auto strip(S)(S str) @safe pure
    if (isBidirectionalRange!S&&isSomeChar!(ElementType!S))
{
    return stripRight(stripLeft(str));
}

///
@safe pure unittest
{
    assert(strip("     hello world     ") ==
           "hello world");
    assert(strip("\n\t\v\rhello world\n\t\v\r") ==
           "hello world");
    assert(strip("hello world") ==
           "hello world");
    assert(strip([lineSep] ~ "hello world" ~ [lineSep]) ==
           "hello world");
    assert(strip([paraSep] ~ "hello world" ~ [paraSep]) ==
           "hello world");
}

unittest
{
    debug(string) printf("string.strip.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!( char[], const  char[],  string,
                           wchar[], const wchar[], wstring,
                           dchar[], const dchar[], dstring))
    {
        assert(equal(stripLeft(to!S("  foo\t ")), "foo\t "));
        assert(equal(stripLeft(to!S("\u2008  foo\t \u2007")), "foo\t \u2007"));
        assert(equal(stripLeft(to!S("\u0085 μ \u0085 \u00BB \r")), "μ \u0085 \u00BB \r"));
        assert(equal(stripLeft(to!S("1")), "1"));
        assert(equal(stripLeft(to!S("\U0010FFFE")), "\U0010FFFE"));
        assert(equal(stripLeft(to!S("")), ""));

        assert(equal(stripRight(to!S("  foo\t ")), "  foo"));
        assert(equal(stripRight(to!S("\u2008  foo\t \u2007")), "\u2008  foo"));
        assert(equal(stripRight(to!S("\u0085 μ \u0085 \u00BB \r")), "\u0085 μ \u0085 \u00BB"));
        assert(equal(stripRight(to!S("1")), "1"));
        assert(equal(stripRight(to!S("\U0010FFFE")), "\U0010FFFE"));
        assert(equal(stripRight(to!S("")), ""));

        assert(equal(strip(to!S("  foo\t ")), "foo"));
        assert(equal(strip(to!S("\u2008  foo\t \u2007")), "foo"));
        assert(equal(strip(to!S("\u0085 μ \u0085 \u00BB \r")), "μ \u0085 \u00BB"));
        assert(equal(strip(to!S("\U0010FFFE")), "\U0010FFFE"));
        assert(equal(strip(to!S("")), ""));
    }
    });
}

@safe pure unittest
{
    assertCTFEable!(
    {
    wstring s = " ";
    assert(s.sameTail(s.stripLeft()));
    assert(s.sameHead(s.stripRight()));
    });
}


/++
    If $(D str) ends with $(D delimiter), then $(D str) is returned without
    $(D delimiter) on its end. If it $(D str) does $(I not) end with
    $(D delimiter), then it is returned unchanged.

    If no $(D delimiter) is given, then one trailing  $(D '\r'), $(D '\n'),
    $(D "\r\n"), $(XREF uni, lineSep), or $(XREF uni, paraSep) is removed from
    the end of $(D str). If $(D str) does not end with any of those characters,
    then it is returned unchanged.
  +/
S chomp(S)(S str) @safe pure
    if (isBidirectionalRange!S && isSomeChar!(ElementType!S))
{
    if (str.empty)
        return str;

    switch (str.back)
    {
        case '\n':
        {
            str.popBack();
            if(str.back!='\r')return str;
            goto case;
        }
        case '\r':
        {
            str.popBack();
            return str;
        }

        //Pops off the last character if it's lineSep or paraSep.
        static if (is(ElementType!S : const char)&&isForwardRange!S)
        {
            //In UTF-8, lineSep and paraSep are [226, 128, 168], and
            //[226, 128, 169] respectively, so their first two bytes are the same.
            case 168: //Last byte of lineSep
            case 169: //Last byte of paraSep
            {
                S orig=str.save;
                str.popBack();
                if(str.back==128)
                {
                    str.popBack();
                    if(str.back==266){
                        str.popBack();
                        return str;
                    }
                }
                return orig;
            }
        }
        else
        {
            case lineSep:
            case paraSep:
                goto case '\r';
        }
        default:
            return str;
    }
}

/// Ditto
S1 chomp(S1, S2)(S1 str, S2 delimiter) @safe pure
    if (isSomeChar!(ElementType!S1) && isSomeChar!(ElementType!S2)&&
    isBidirectionalRange!S1&&isBidirectionalRange!S2
    &&is(Unqual!(ElementType!S1) == Unqual!(ElementType!S2))&&hasLenght!S2)
{
    if (delimiter.empty)
        return chomp(str);

    if (str.endsWith(delimiter))
        str.popBackExactly(delimiter.length);
    return str;
}

/// Ditto
S1 chomp(S1, S2)(S1 str, S2 delimiter) @safe pure
    if (isSomeChar!(ElementType!S1) && isSomeChar!(ElementType!S2)&&
    isBidirectionalRange!S1&&isBidirectionalRange!S2&&isForwardRange!S1
    (!is(Unqual!(ElementType!S1) == Unqual!(ElementType!S2))|| !hasLenght!S2) )
{
    if (delimiter.empty)
        return chomp(str);
    auto orig = str.save;
    foreach_reverse (dchar c; delimiter)
    {
        if (str.empty || str.back != c)
        {
            return orig;
        }
        str.popBack();
    }
    return str;
    }

///
@safe pure unittest
{
    assert(chomp(" hello world  \n\r") == " hello world  \n");
    assert(chomp(" hello world  \r\n") == " hello world  ");
    assert(chomp(" hello world  \n\n") == " hello world  \n");
    assert(chomp(" hello world  \n\n ") == " hello world  \n\n ");
    assert(chomp(" hello world  \n\n" ~ [lineSep]) == " hello world  \n\n");
    assert(chomp(" hello world  \n\n" ~ [paraSep]) == " hello world  \n\n");
    assert(chomp(" hello world") == " hello world");
    assert(chomp("") == "");

    assert(chomp(" hello world", "orld") == " hello w");
    assert(chomp(" hello world", " he") == " hello world");
    assert(chomp("", "hello") == "");

    // Don't decode pointlessly
    assert(chomp("hello\xFE", "\r") == "hello\xFE");
}

unittest
{
    debug(string) printf("string.chomp.unittest\n");
    string s;

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
    {
        // @@@ BUG IN COMPILER, MUST INSERT CAST
        assert(chomp(cast(S)null) is null);
        assert(chomp(to!S("hello")) == "hello");
        assert(chomp(to!S("hello\n")) == "hello");
        assert(chomp(to!S("hello\r")) == "hello");
        assert(chomp(to!S("hello\r\n")) == "hello");
        assert(chomp(to!S("hello\n\r")) == "hello\n");
        assert(chomp(to!S("hello\n\n")) == "hello\n");
        assert(chomp(to!S("hello\r\r")) == "hello\r");
        assert(chomp(to!S("hello\nxxx\n")) == "hello\nxxx");
        assert(chomp(to!S("hello\u2028")) == "hello");
        assert(chomp(to!S("hello\u2029")) == "hello");
        assert(chomp(to!S("hello\u2028\u2028")) == "hello\u2028");
        assert(chomp(to!S("hello\u2029\u2029")) == "hello\u2029");

        foreach (T; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
        {
            // @@@ BUG IN COMPILER, MUST INSERT CAST
            assert(chomp(cast(S)null, cast(T)null) is null);
            assert(chomp(to!S("hello\n"), cast(T)null) == "hello");
            assert(chomp(to!S("hello"), to!T("o")) == "hell");
            assert(chomp(to!S("hello"), to!T("p")) == "hello");
            // @@@ BUG IN COMPILER, MUST INSERT CAST
            assert(chomp(to!S("hello"), cast(T) null) == "hello");
            assert(chomp(to!S("hello"), to!T("llo")) == "he");
            assert(chomp(to!S("\uFF28ello"), to!T("llo")) == "\uFF28e");
            assert(chomp(to!S("\uFF28el\uFF4co"), to!T("l\uFF4co")) == "\uFF28e");
        }
    }
    });
}


/++
    If $(D str) starts with $(D delimiter), then the part of $(D str) following
    $(D delimiter) is returned. If it $(D str) does $(I not) start with
    $(D delimiter), then it is returned unchanged.
 +/
S1 chompPrefix(S1,S2)(S1 str, S2 delimiter) @safe pure
    if (isInputRange!S1 && isInputRange!S2 && hasLenght!S2
        &&is(Unqual!(ElementType!S1) == Unqual!(ElementType!S2)))
{
    if (str.startsWith(delimiter))
        str.popFrontExactly(delimiter.length);
    return str;
}
///ditto
S1 chompPrefix(S1,S2)(S1 str, S2 delimiter) @safe pure
    if (isSomeChar!(ElementType!S1) && isSomeChar!(ElementType!S2) &&
     isInputRange!S2 && isForwardRange!S1)
{
    S1 orig = str.save;
    size_t index = 0;

    foreach (dchar c; delimiter)
    {
        if (index >= str.length || decode(str, index) != c)
            return orig;
    }
    str.popFrontExactly(index);
    return str;
}
///
@safe pure unittest
{
    assert(chompPrefix("hello world", "he") == "llo world");
    assert(chompPrefix("hello world", "hello w") == "orld");
    assert(chompPrefix("hello world", " world") == "hello world");
    assert(chompPrefix("", "hello") == "");
}

/* @safe */ pure unittest
{
    assertCTFEable!(
    {
    foreach (S; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
    {
        foreach (T; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
        {
            assert(equal(chompPrefix(to!S("abcdefgh"), to!T("abcde")), "fgh"));
            assert(equal(chompPrefix(to!S("abcde"), to!T("abcdefgh")), "abcde"));
            assert(equal(chompPrefix(to!S("\uFF28el\uFF4co"), to!T("\uFF28el\uFF4co")), ""));
            assert(equal(chompPrefix(to!S("\uFF28el\uFF4co"), to!T("\uFF28el")), "\uFF4co"));
            assert(equal(chompPrefix(to!S("\uFF28el"), to!T("\uFF28el\uFF4co")), "\uFF28el"));
        }
    }
    });
}


/++
    Returns $(D str) without its last character, if there is one. If $(D str)
    ends with $(D "\r\n"), then both are removed. If $(D str) is empty, then
    then it is returned unchanged.
 +/
S chop(S)(S str) @safe pure
    if (isBidirectionalRange!S &&isSomeChar!(ElementType!S))
{
    if (str.empty)
        return str;
    auto last=str.back;
    str.popBack();
    if(last=='\n'&&str.back=='\r')str.popBack();
    return str;
}

///
@safe pure unittest
{
    assert(chop("hello world") == "hello worl");
    assert(chop("hello world\n") == "hello world");
    assert(chop("hello world\r") == "hello world");
    assert(chop("hello world\n\r") == "hello world\n");
    assert(chop("hello world\r\n") == "hello world");
    assert(chop("Walter Bright") == "Walter Brigh");
    assert(chop("") == "");
}

unittest
{
    debug(string) printf("string.chop.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
    {
        assert(chop(cast(S) null) is null);
        assert(equal(chop(to!S("hello")), "hell"));
        assert(equal(chop(to!S("hello\r\n")), "hello"));
        assert(equal(chop(to!S("hello\n\r")), "hello\n"));
        assert(equal(chop(to!S("Verité")), "Verit"));
        assert(equal(chop(to!S(`さいごの果実`)), "さいごの果"));
        assert(equal(chop(to!S(`ミツバチと科学者`)), "ミツバチと科学"));
    }
    });
}


/++
    Left justify $(D s) in a field $(D width) characters wide. $(D fillChar)
    is the character that will be used to fill up the space in the field that
    $(D s) doesn't fill.
  +/
S leftJustify(S)(S s, size_t width, dchar fillChar = ' ') @trusted pure
    if (isSomeString!S)
{
    alias C = ElementEncodingType!S;

    if (canSearchInCodeUnits!C(fillChar))
    {
        immutable len = s.walkLength();
        if (len >= width)
            return s;

        auto retval = new Unqual!(C)[width - len + s.length];
        retval[0 .. s.length] = s[];
        retval[s.length .. $] = cast(C)fillChar;
        return cast(S)retval;
    }
    else
    {
        auto dstr = to!dstring(s);
        if (dstr.length >= width)
            return s;

        auto retval = new dchar[](width);
        retval[0 .. dstr.length] = dstr[];
        retval[dstr.length .. $] = fillChar;
        return to!S(retval);
    }
}


/++
    Right justify $(D s) in a field $(D width) characters wide. $(D fillChar)
    is the character that will be used to fill up the space in the field that
    $(D s) doesn't fill.
  +/
S rightJustify(S)(S s, size_t width, dchar fillChar = ' ') @trusted pure
    if (isSomeString!S)
{
    alias C = ElementEncodingType!S;

    if (canSearchInCodeUnits!C(fillChar))
    {
        immutable len = s.walkLength();
        if (len >= width)
            return s;

        auto retval = new Unqual!C[width - len + s.length];
        retval[0 .. $ - s.length] = cast(C)fillChar;
        retval[$ - s.length .. $] = s[];
        return cast(S)retval;
    }
    else
    {
        auto dstr = to!dstring(s);
        if (dstr.length >= width)
            return s;

        auto retval = new dchar[](width);
        retval[0 .. $ - dstr.length] = fillChar;
        retval[$ - dstr.length .. $] = dstr[];
        return to!S(retval);
    }
}


/++
    Center $(D s) in a field $(D width) characters wide. $(D fillChar)
    is the character that will be used to fill up the space in the field that
    $(D s) doesn't fill.
  +/
S center(S)(S s, size_t width, dchar fillChar = ' ') @trusted pure
    if (isSomeString!S)
{
    alias C = ElementEncodingType!S;

    if (canSearchInCodeUnits!C(fillChar))
    {
        immutable len = s.walkLength();
        if (len >= width)
            return s;

        auto retval = new Unqual!C[width - len + s.length];
        immutable left = (retval.length - s.length) / 2;
        retval[0 .. left] = cast(C)fillChar;
        retval[left .. left + s.length] = s[];
        retval[left + s.length .. $] = cast(C)fillChar;
        return to!S(retval);
    }
    else
    {
        auto dstr = to!dstring(s);
        if (dstr.length >= width)
            return s;

        auto retval = new dchar[](width);
        immutable left = (retval.length - dstr.length) / 2;
        retval[0 .. left] = fillChar;
        retval[left .. left + dstr.length] = dstr[];
        retval[left + dstr.length .. $] = fillChar;
        return to!S(retval);
    }
}

unittest
{
    debug(string) printf("string.justify.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
    {
        S s = to!S("hello");

        assert(leftJustify(s, 2) == "hello");
        assert(rightJustify(s, 2) == "hello");
        assert(center(s, 2) == "hello");

        assert(leftJustify(s, 7) == "hello  ");
        assert(rightJustify(s, 7) == "  hello");
        assert(center(s, 7) == " hello ");

        assert(leftJustify(s, 8) == "hello   ");
        assert(rightJustify(s, 8) == "   hello");
        assert(center(s, 8) == " hello  ");

        assert(leftJustify(s, 8, '\u0100') == "hello\u0100\u0100\u0100");
        assert(rightJustify(s, 8, '\u0100') == "\u0100\u0100\u0100hello");
        assert(center(s, 8, '\u0100') == "\u0100hello\u0100\u0100");

        assert(leftJustify(s, 8, 'ö') == "helloööö");
        assert(rightJustify(s, 8, 'ö') == "öööhello");
        assert(center(s, 8, 'ö') == "öhelloöö");
    }
    });
}


/++
    Replace each tab character in $(D s) with the number of spaces necessary
    to align the following character at the next tab stop where $(D tabSize)
    is the distance between tab stops.
  +/
S detab(S)(S s, size_t tabSize = 8) @trusted pure
    if (isSomeString!S)
{
    assert(tabSize > 0);
    alias C = Unqual!(typeof(s[0]));
    bool changes = false;
    C[] result;
    int column;
    size_t nspaces;

    foreach (size_t i, dchar c; s)
    {
        switch (c)
        {
        case '\t':
            nspaces = tabSize - (column % tabSize);
            if (!changes)
            {
                changes = true;
                result = null;
                result.length = s.length + nspaces - 1;
                result.length = i + nspaces;
                result[0 .. i] = s[0 .. i];
                result[i .. i + nspaces] = ' ';
            }
            else
            {
                ptrdiff_t j = result.length;
                result.length = j + nspaces;
                result[j .. j + nspaces] = ' ';
            }
            column += nspaces;
            break;

        case '\r':
        case '\n':
        case paraSep:
        case lineSep:
            column = 0;
            goto L1;

        default:
            column++;
        L1:
            if (changes)
            {
                std.utf.encode(result, c);
            }
            break;
        }
    }

    return changes ? cast(S) result : s;
}

unittest
{
    debug(string) printf("string.detab.unittest\n");

    assertCTFEable!(
    {
    foreach (S; TypeTuple!(char[], wchar[], dchar[], string, wstring, dstring))
    {
        S s = to!S("This \tis\t a fofof\tof list");
        assert(cmp(detab(s), "This    is       a fofof        of list") == 0);

        assert(detab(cast(S)null) is null);
        assert(detab("").empty);
        assert(detab("a") == "a");
        assert(detab("\t") == "        ");
        assert(detab("\t", 3) == "   ");
        assert(detab("\t", 9) == "         ");
        assert(detab(  "  ab\t asdf ") == "  ab     asdf ");
        assert(detab(  "  \U00010000b\tasdf ") == "  \U00010000b    asdf ");
    }
    });
}

/++
    Replaces spaces in $(D s) with the optimal number of tabs.
    All spaces and tabs at the end of a line are removed.

    Params:
        s       = String to convert.
        tabSize = Tab columns are $(D tabSize) spaces apart.
 +/
S entab(S)(S s, size_t tabSize = 8) @trusted pure
    if (isSomeString!S)
{
    bool changes = false;
    alias C = Unqual!(typeof(s[0]));
    C[] result;

    int nspaces = 0;
    int nwhite = 0;
    size_t column = 0;         // column number

    foreach (size_t i, dchar c; s)
    {

        void change()
        {
            changes = true;
            result = null;
            result.length = s.length;
            result.length = i;
            result[0 .. i] = s[0 .. i];
        }

        switch (c)
        {
        case '\t':
            nwhite++;
            if (nspaces)
            {
                if (!changes)
                    change();

                ptrdiff_t j = result.length - nspaces;
                auto ntabs = (((column - nspaces) % tabSize) + nspaces) / tabSize;
                result.length = j + ntabs;
                result[j .. j + ntabs] = '\t';
                nwhite += ntabs - nspaces;
                nspaces = 0;
            }
            column = (column + tabSize) / tabSize * tabSize;
            break;

        case '\r':
        case '\n':
        case paraSep:
        case lineSep:
            // Truncate any trailing spaces or tabs
            if (nwhite)
            {
                if (!changes)
                    change();
                result = result[0 .. result.length - nwhite];
            }
            break;

        default:
            if (nspaces >= 2 && (column % tabSize) == 0)
            {
                if (!changes)
                    change();

                auto j = result.length - nspaces;
                auto ntabs = (nspaces + tabSize - 1) / tabSize;
                result.length = j + ntabs;
                result[j .. j + ntabs] = '\t';
                nwhite += ntabs - nspaces;
                nspaces = 0;
            }
            if (c == ' ')
            {   nwhite++;
                nspaces++;
            }
            else
            {   nwhite = 0;
                nspaces = 0;
            }
            column++;
            break;
        }
        if (changes)
        {
            std.utf.encode(result, c);
        }
    }

    // Truncate any trailing spaces or tabs
    if (nwhite)
    {
        if (changes)
            result = result[0 .. result.length - nwhite];
        else
            s = s[0 .. s.length - nwhite];
    }
    return changes ? assumeUnique(result) : s;
}

unittest
{
    debug(string) printf("string.entab.unittest\n");

    assertCTFEable!(
    {
    assert(entab(cast(string) null) is null);
    assert(entab("").empty);
    assert(entab("a") == "a");
    assert(entab("        ") == "");
    assert(entab("        x") == "\tx");
    assert(entab("  ab    asdf ") == "  ab\tasdf");
    assert(entab("  ab     asdf ") == "  ab\t asdf");
    assert(entab("  ab \t   asdf ") == "  ab\t   asdf");
    assert(entab("1234567 \ta") == "1234567\t\ta");
    assert(entab("1234567  \ta") == "1234567\t\ta");
    assert(entab("1234567   \ta") == "1234567\t\ta");
    assert(entab("1234567    \ta") == "1234567\t\ta");
    assert(entab("1234567     \ta") == "1234567\t\ta");
    assert(entab("1234567      \ta") == "1234567\t\ta");
    assert(entab("1234567       \ta") == "1234567\t\ta");
    assert(entab("1234567        \ta") == "1234567\t\ta");
    assert(entab("1234567         \ta") == "1234567\t\t\ta");

    assert(entab("a               ") == "a");
    assert(entab("a\v") == "a\v");
    assert(entab("a\f") == "a\f");
    assert(entab("a\n") == "a\n");
    assert(entab("a\n\r") == "a\n\r");
    assert(entab("a\r\n") == "a\r\n");
    assert(entab("a\u2028") == "a\u2028");
    assert(entab("a\u2029") == "a\u2029");
    assert(entab("a  ") == "a");
    assert(entab("a\t") == "a");
    assert(entab("\uFF28\uFF45\uFF4C\uFF4C567      \t\uFF4F \t") ==
                 "\uFF28\uFF45\uFF4C\uFF4C567\t\t\uFF4F");
    });
}


/++
    Replaces the characters in $(D str) which are keys in $(D transTable) with
    their corresponding values in $(D transTable). $(D transTable) is an AA
    where its keys are $(D dchar) and its values are either $(D dchar) or some
    type of string. Also, if $(D toRemove) is given, the characters in it are
    removed from $(D str) prior to translation. $(D str) itself is unaltered.
    A copy with the changes is returned.

    See_Also:
        $(LREF tr)
        $(XREF array, replace)

    Params:
        str        = The original string.
        transTable = The AA indicating which characters to replace and what to
                     replace them with.
        toRemove   = The characters to remove from the string.
  +/
C1[] translate(C1, C2 = immutable char)(C1[] str,
                                        dchar[dchar] transTable,
                                        const(C2)[] toRemove = null) @safe pure
    if (isSomeChar!C1 && isSomeChar!C2)
{
    auto buffer = appender!(C1[])();
    translateImpl(str, transTable, toRemove, buffer);
    return buffer.data;
}

///
@safe pure unittest
{
    dchar[dchar] transTable1 = ['e' : '5', 'o' : '7', '5': 'q'];
    assert(translate("hello world", transTable1) == "h5ll7 w7rld");

    assert(translate("hello world", transTable1, "low") == "h5 rd");

    string[dchar] transTable2 = ['e' : "5", 'o' : "orange"];
    assert(translate("hello world", transTable2) == "h5llorange worangerld");
}

/* @safe */ pure unittest
{
    assertCTFEable!(
    {
    foreach (S; TypeTuple!( char[], const( char)[], immutable( char)[],
                           wchar[], const(wchar)[], immutable(wchar)[],
                           dchar[], const(dchar)[], immutable(dchar)[]))
    {
        assert(translate(to!S("hello world"), cast(dchar[dchar])['h' : 'q', 'l' : '5']) ==
               to!S("qe55o wor5d"));
        assert(translate(to!S("hello world"), cast(dchar[dchar])['o' : 'l', 'l' : '\U00010143']) ==
               to!S("he\U00010143\U00010143l wlr\U00010143d"));
        assert(translate(to!S("hello \U00010143 world"), cast(dchar[dchar])['h' : 'q', 'l': '5']) ==
               to!S("qe55o \U00010143 wor5d"));
        assert(translate(to!S("hello \U00010143 world"), cast(dchar[dchar])['o' : '0', '\U00010143' : 'o']) ==
               to!S("hell0 o w0rld"));
        assert(translate(to!S("hello world"), cast(dchar[dchar])null) == to!S("hello world"));

        foreach (T; TypeTuple!( char[], const( char)[], immutable( char)[],
                               wchar[], const(wchar)[], immutable(wchar)[],
                               dchar[], const(dchar)[], immutable(dchar)[]))
        {
            assert(translate(to!S("hello world"),
                             cast(dchar[dchar])['h' : 'q', 'l' : '5'],
                             to!T("r")) ==
                   to!S("qe55o wo5d"));
            assert(translate(to!S("hello world"),
                             cast(dchar[dchar])['h' : 'q', 'l' : '5'],
                             to!T("helo")) ==
                   to!S(" wrd"));
            assert(translate(to!S("hello world"),
                             cast(dchar[dchar])['h' : 'q', 'l' : '5'],
                             to!T("q5")) ==
                   to!S("qe55o wor5d"));
            assert(translate(to!S("hello \U00010143 world"),
                             cast(dchar[dchar])['o' : '0', '\U00010143' : 'o'],
                             to!T("\U00010143 ")) ==
                   to!S("hell0w0rld"));
        }

        auto s = to!S("hello world");
        dchar[dchar] transTable = ['h' : 'q', 'l' : '5'];
        static assert(is(typeof(s) == typeof(translate(s, transTable))));
    }
    });
}

/++ Ditto +/
C1[] translate(C1, S, C2 = immutable char)(C1[] str,
                                           S[dchar] transTable,
                                           const(C2)[] toRemove = null) @safe pure
    if (isSomeChar!C1 && isSomeString!S && isSomeChar!C2)
{
    auto buffer = appender!(C1[])();
    translateImpl(str, transTable, toRemove, buffer);
    return buffer.data;
}

/* @safe */ pure unittest
{
    assertCTFEable!(
    {
    foreach (S; TypeTuple!( char[], const( char)[], immutable( char)[],
                           wchar[], const(wchar)[], immutable(wchar)[],
                           dchar[], const(dchar)[], immutable(dchar)[]))
    {
        assert(translate(to!S("hello world"), ['h' : "yellow", 'l' : "42"]) ==
               to!S("yellowe4242o wor42d"));
        assert(translate(to!S("hello world"), ['o' : "owl", 'l' : "\U00010143\U00010143"]) ==
               to!S("he\U00010143\U00010143\U00010143\U00010143owl wowlr\U00010143\U00010143d"));
        assert(translate(to!S("hello \U00010143 world"), ['h' : "yellow", 'l' : "42"]) ==
               to!S("yellowe4242o \U00010143 wor42d"));
        assert(translate(to!S("hello \U00010143 world"), ['o' : "owl", 'l' : "\U00010143\U00010143"]) ==
               to!S("he\U00010143\U00010143\U00010143\U00010143owl \U00010143 wowlr\U00010143\U00010143d"));
        assert(translate(to!S("hello \U00010143 world"), ['h' : ""]) ==
               to!S("ello \U00010143 world"));
        assert(translate(to!S("hello \U00010143 world"), ['\U00010143' : ""]) ==
               to!S("hello  world"));
        assert(translate(to!S("hello world"), cast(string[dchar])null) == to!S("hello world"));

        foreach (T; TypeTuple!( char[], const( char)[], immutable( char)[],
                               wchar[], const(wchar)[], immutable(wchar)[],
                               dchar[], const(dchar)[], immutable(dchar)[]))
        {
            assert(translate(to!S("hello world"), ['h' : "yellow", 'l' : "42"], to!T("r")) ==
                   to!S("yellowe4242o wo42d"));
            assert(translate(to!S("hello world"), ['h' : "yellow", 'l' : "42"], to!T("helo")) ==
                   to!S(" wrd"));
            assert(translate(to!S("hello world"), ['h' : "yellow", 'l' : "42"], to!T("y42")) ==
                   to!S("yellowe4242o wor42d"));
            assert(translate(to!S("hello \U00010143 world"),
                             ['o' : "owl", '\U00010143' : "\n"],
                             to!T("\U00010143 ")) ==
                   to!S("hellowlwowlrld"));
            assert(translate(to!S("hello world"), ['h' : "yellow", 'l' : "42"], to!T("hello world")) ==
                   to!S(""));
            assert(translate(to!S("hello world"), ['h' : "yellow", 'l' : "42"], to!T("42")) ==
                   to!S("yellowe4242o wor42d"));
        }

        auto s = to!S("hello world");
        string[dchar] transTable = ['h' : "silly", 'l' : "putty"];
        static assert(is(typeof(s) == typeof(translate(s, transTable))));
    }
    });
}

/++
    This is an overload of $(D translate) which takes an existing buffer to write the contents to.

    Params:
        str        = The original string.
        transTable = The AA indicating which characters to replace and what to
                     replace them with.
        toRemove   = The characters to remove from the string.
        buffer     = An output range to write the contents to.
  +/
void translate(C1, C2 = immutable char, Buffer)(C1[] str,
                                        dchar[dchar] transTable,
                                        const(C2)[] toRemove,
                                        Buffer buffer)
    if (isSomeChar!C1 && isSomeChar!C2 && isOutputRange!(Buffer, C1))
{
    translateImpl(str, transTable, toRemove, buffer);
}

///
@safe pure unittest
{
    dchar[dchar] transTable1 = ['e' : '5', 'o' : '7', '5': 'q'];
    auto buffer = appender!(dchar[])();
    translate("hello world", transTable1, null, buffer);
    assert(buffer.data == "h5ll7 w7rld");

    buffer.clear();
    translate("hello world", transTable1, "low", buffer);
    assert(buffer.data == "h5 rd");

    buffer.clear();
    string[dchar] transTable2 = ['e' : "5", 'o' : "orange"];
    translate("hello world", transTable2, null, buffer);
    assert(buffer.data == "h5llorange worangerld");
}

/++ Ditto +/
void translate(C1, S, C2 = immutable char, Buffer)(C1[] str,
                                                   S[dchar] transTable,
                                                   const(C2)[] toRemove,
                                                   Buffer buffer)
    if (isSomeChar!C1 && isSomeString!S && isSomeChar!C2 && isOutputRange!(Buffer, S))
{
    translateImpl(str, transTable, toRemove, buffer);
}

private void translateImpl(C1, T, C2, Buffer)(C1[] str,
                                      T transTable,
                                      const(C2)[] toRemove,
                                      Buffer buffer)
{
    bool[dchar] removeTable;

    foreach (dchar c; toRemove)
        removeTable[c] = true;

    foreach (dchar c; str)
    {
        if (c in removeTable)
            continue;

        auto newC = c in transTable;

        if (newC)
            put(buffer, *newC);
        else
            put(buffer, c);
    }
}

/++
    This is an $(I $(RED ASCII-only)) overload of $(LREF _translate). It
    will $(I not) work with Unicode. It exists as an optimization for the
    cases where Unicode processing is not necessary.

    Unlike the other overloads of $(LREF _translate), this one does not take
    an AA. Rather, it takes a $(D string) generated by $(LREF makeTrans).

    The array generated by $(D makeTrans) is $(D 256) elements long such that
    the index is equal to the ASCII character being replaced and the value is
    equal to the character that it's being replaced with. Note that translate
    does not decode any of the characters, so you can actually pass it Extended
    ASCII characters if you want to (ASCII only actually uses $(D 128)
    characters), but be warned that Extended ASCII characters are not valid
    Unicode and therefore will result in a $(D UTFException) being thrown from
    most other Phobos functions.

    Also, because no decoding occurs, it is possible to use this overload to
    translate ASCII characters within a proper UTF-8 string without altering the
    other, non-ASCII characters. It's replacing any code unit greater than
    $(D 127) with another code unit or replacing any code unit with another code
    unit greater than $(D 127) which will cause UTF validation issues.

    See_Also:
        $(LREF tr)
        $(XREF array, replace)

    Params:
        str        = The original string.
        transTable = The string indicating which characters to replace and what
                     to replace them with. It is generated by $(LREF makeTrans).
        toRemove   = The characters to remove from the string.
  +/
C[] translate(C = immutable char)(in char[] str, in char[] transTable, in char[] toRemove = null) @trusted pure nothrow
    if (is(Unqual!C == char))
in
{
    assert(transTable.length == 256);
}
body
{
    bool[256] remTable = false;

    foreach (char c; toRemove)
        remTable[c] = true;

    size_t count = 0;
    foreach (char c; str)
    {
        if (!remTable[c])
            ++count;
    }

    auto buffer = new char[count];
    translateImplAscii(str, transTable, remTable, buffer, toRemove);
    return cast(C[])(buffer);
}


/++ Ditto +/
string makeTrans(in char[] from, in char[] to) @trusted pure nothrow
in
{
    assert(from.length == to.length);
    assert(from.length <= 256);
    foreach (char c; from)
        assert(std.ascii.isASCII(c));
    foreach (char c; to)
        assert(std.ascii.isASCII(c));
}
body
{
    char[] transTable = new char[256];

    foreach (i; 0 .. transTable.length)
        transTable[i] = cast(char)i;
    foreach (i; 0 .. from.length)
        transTable[from[i]] = to[i];

    return assumeUnique(transTable);
}

///
@safe pure nothrow unittest
{
    auto transTable1 = makeTrans("eo5", "57q");
    assert(translate("hello world", transTable1) == "h5ll7 w7rld");

    assert(translate("hello world", transTable1, "low") == "h5 rd");
}

@safe pure unittest
{
    assertCTFEable!(
    {
    foreach (C; TypeTuple!(char, const char, immutable char))
    {
        assert(translate!C("hello world", makeTrans("hl", "q5")) == to!(C[])("qe55o wor5d"));

        auto s = to!(C[])("hello world");
        auto transTable = makeTrans("hl", "q5");
        static assert(is(typeof(s) == typeof(translate!C(s, transTable))));
    }

    foreach (S; TypeTuple!(char[], const(char)[], immutable(char)[]))
    {
        assert(translate(to!S("hello world"), makeTrans("hl", "q5")) == to!S("qe55o wor5d"));
        assert(translate(to!S("hello \U00010143 world"), makeTrans("hl", "q5")) ==
               to!S("qe55o \U00010143 wor5d"));
        assert(translate(to!S("hello world"), makeTrans("ol", "1o")), to!S("heool wlrdd"));
        assert(translate(to!S("hello world"), makeTrans("", "")) == to!S("hello world"));
        assert(translate(to!S("hello world"), makeTrans("12345", "67890")) == to!S("hello world"));
        assert(translate(to!S("hello \U00010143 world"), makeTrans("12345", "67890")) ==
               to!S("hello \U00010143 world"));

        foreach (T; TypeTuple!(char[], const(char)[], immutable(char)[]))
        {
            assert(translate(to!S("hello world"), makeTrans("hl", "q5"), to!T("r")) ==
                   to!S("qe55o wo5d"));
            assert(translate(to!S("hello \U00010143 world"), makeTrans("hl", "q5"), to!T("r")) ==
                   to!S("qe55o \U00010143 wo5d"));
            assert(translate(to!S("hello world"), makeTrans("hl", "q5"), to!T("helo")) ==
                   to!S(" wrd"));
            assert(translate(to!S("hello world"), makeTrans("hl", "q5"), to!T("q5")) ==
                   to!S("qe55o wor5d"));
        }
    }
    });
}

/++
    This is an $(I $(RED ASCII-only)) overload of $(D translate) which takes an existing buffer to write the contents to.

    Params:
        str        = The original string.
        transTable = The string indicating which characters to replace and what
                     to replace them with. It is generated by $(LREF makeTrans).
        toRemove   = The characters to remove from the string.
        buffer     = An output range to write the contents to.
  +/
void translate(C = immutable char, Buffer)(in char[] str, in char[] transTable, in char[] toRemove, Buffer buffer)
    if (is(Unqual!C == char) && isOutputRange!(Buffer, char))
in
{
    assert(transTable.length == 256);
}
body
{
    bool[256] remTable = false;

    foreach (char c; toRemove)
        remTable[c] = true;

    translateImplAscii(str, transTable, remTable, buffer, toRemove);
}

///
@safe pure unittest
{
    auto buffer = appender!(char[])();
    auto transTable1 = makeTrans("eo5", "57q");
    translate("hello world", transTable1, null, buffer);
    assert(buffer.data == "h5ll7 w7rld");

    buffer.clear();
    translate("hello world", transTable1, "low", buffer);
    assert(buffer.data == "h5 rd");
}

private void translateImplAscii(C = immutable char, Buffer)(in char[] str, in char[] transTable, ref bool[256] remTable, Buffer buffer, in char[] toRemove = null)
{
    static if (isOutputRange!(Buffer, char))
    {
        foreach (char c; str)
        {
            if (!remTable[c])
                put(buffer, transTable[c]);
        }
    }
    else
    {
        size_t i = 0;
        foreach (char c; str)
        {
            if (!remTable[c])
                buffer[i++] = transTable[c];
        }
    }
}

/*****************************************************
 * Format arguments into a string.
 *
 * Params: fmt  = Format string. For detailed specification, see $(XREF format,formattedWrite).
 *         args = Variadic list of arguments to format into returned string.
 *
 *  $(RED format's current implementation has been replaced with $(LREF xformat)'s
 *        implementation. in November 2012.
 *        This is seamless for most code, but it makes it so that the only
 *        argument that can be a format string is the first one, so any
 *        code which used multiple format strings has broken. Please change
 *        your calls to format accordingly.
 *
 *        e.g.:
 *        ----
 *        format("key = %s", key, ", value = %s", value)
 *        ----
 *        needs to be rewritten as:
 *        ----
 *        format("key = %s, value = %s", key, value)
 *        ----
 *   )
 */
string format(Char, Args...)(in Char[] fmt, Args args)
{
    auto w = appender!string();
    auto n = formattedWrite(w, fmt, args);
    version (all)
    {
        // In the future, this check will be removed to increase consistency
        // with formattedWrite
        enforce(n == args.length, new FormatException(
            text("Orphan format arguments: args[", n, "..", args.length, "]")));
    }
    return w.data;
}

unittest
{
    debug(string) printf("std.string.format.unittest\n");

    assertCTFEable!(
    {
//  assert(format(null) == "");
    assert(format("foo") == "foo");
    assert(format("foo%%") == "foo%");
    assert(format("foo%s", 'C') == "fooC");
    assert(format("%s foo", "bar") == "bar foo");
    assert(format("%s foo %s", "bar", "abc") == "bar foo abc");
    assert(format("foo %d", -123) == "foo -123");
    assert(format("foo %d", 123) == "foo 123");

    assertThrown!FormatException(format("foo %s"));
    assertThrown!FormatException(format("foo %s", 123, 456));

    assert(format("hel%slo%s%s%s", "world", -138, 'c', true) ==
                  "helworldlo-138ctrue");
    });
}


/*****************************************************
 * Format arguments into buffer <i>buf</i> which must be large
 * enough to hold the result. Throws RangeError if it is not.
 * Returns: The slice of $(D buf) containing the formatted string.
 *
 *  $(RED sformat's current implementation has been replaced with $(LREF xsformat)'s
 *        implementation. in November 2012.
 *        This is seamless for most code, but it makes it so that the only
 *        argument that can be a format string is the first one, so any
 *        code which used multiple format strings has broken. Please change
 *        your calls to sformat accordingly.
 *
 *        e.g.:
 *        ----
 *        sformat(buf, "key = %s", key, ", value = %s", value)
 *        ----
 *        needs to be rewritten as:
 *        ----
 *        sformat(buf, "key = %s, value = %s", key, value)
 *        ----
 *   )
 */
char[] sformat(Char, Args...)(char[] buf, in Char[] fmt, Args args)
{
    size_t i;

    struct Sink
    {
        void put(dchar c)
        {
            char[4] enc;
            auto n = encode(enc, c);

            if (buf.length < i + n)
                onRangeError("std.string.sformat", 0);

            buf[i .. i + n] = enc[0 .. n];
            i += n;
        }
        void put(const(char)[] s)
        {
            if (buf.length < i + s.length)
                onRangeError("std.string.sformat", 0);

            buf[i .. i + s.length] = s[];
            i += s.length;
        }
        void put(const(wchar)[] s)
        {
            for (; !s.empty; s.popFront())
                put(s.front);
        }
        void put(const(dchar)[] s)
        {
            for (; !s.empty; s.popFront())
                put(s.front);
        }
    }
    auto n = formattedWrite(Sink(), fmt, args);
    version (all)
    {
        // In the future, this check will be removed to increase consistency
        // with formattedWrite
        enforce(n == args.length, new FormatException(
            text("Orphan format arguments: args[", n, "..", args.length, "]")));
    }
    return buf[0 .. i];
}

unittest
{
    debug(string) printf("std.string.sformat.unittest\n");

    assertCTFEable!(
    {
    char[10] buf;

    assert(sformat(buf[], "foo") == "foo");
    assert(sformat(buf[], "foo%%") == "foo%");
    assert(sformat(buf[], "foo%s", 'C') == "fooC");
    assert(sformat(buf[], "%s foo", "bar") == "bar foo");
    assertThrown!RangeError(sformat(buf[], "%s foo %s", "bar", "abc"));
    assert(sformat(buf[], "foo %d", -123) == "foo -123");
    assert(sformat(buf[], "foo %d", 123) == "foo 123");

    assertThrown!FormatException(sformat(buf[], "foo %s"));
    assertThrown!FormatException(sformat(buf[], "foo %s", 123, 456));

    assert(sformat(buf[], "%s %s %s", "c"c, "w"w, "d"d) == "c w d");
    });
}

// Explicitly undocumented. It will be removed in July 2014.
deprecated("Please use std.string.format instead.") alias xformat = format;

deprecated unittest
{
    debug(string) printf("std.string.xformat.unittest\n");

    assertCTFEable!(
    {
//  assert(xformat(null) == "");
    assert(xformat("foo") == "foo");
    assert(xformat("foo%%") == "foo%");
    assert(xformat("foo%s", 'C') == "fooC");
    assert(xformat("%s foo", "bar") == "bar foo");
    assert(xformat("%s foo %s", "bar", "abc") == "bar foo abc");
    assert(xformat("foo %d", -123) == "foo -123");
    assert(xformat("foo %d", 123) == "foo 123");

    assertThrown!FormatException(xformat("foo %s"));
    assertThrown!FormatException(xformat("foo %s", 123, 456));
    });
}

// Explicitly undocumented. It will be removed in July 2014.
deprecated("Please use std.string.sformat instead.") alias xsformat = sformat;

deprecated unittest
{
    debug(string) printf("std.string.xsformat.unittest\n");

    assertCTFEable!(
    {
    char[10] buf;

    assert(xsformat(buf[], "foo") == "foo");
    assert(xsformat(buf[], "foo%%") == "foo%");
    assert(xsformat(buf[], "foo%s", 'C') == "fooC");
    assert(xsformat(buf[], "%s foo", "bar") == "bar foo");
    assertThrown!RangeError(xsformat(buf[], "%s foo %s", "bar", "abc"));
    assert(xsformat(buf[], "foo %d", -123) == "foo -123");
    assert(xsformat(buf[], "foo %d", 123) == "foo 123");

    assertThrown!FormatException(xsformat(buf[], "foo %s"));
    assertThrown!FormatException(xsformat(buf[], "foo %s", 123, 456));

    assert(xsformat(buf[], "%s %s %s", "c"c, "w"w, "d"d) == "c w d");
    });
}


/***********************************************
 * See if character c is in the pattern.
 * Patterns:
 *
 *  A <i>pattern</i> is an array of characters much like a <i>character
 *  class</i> in regular expressions. A sequence of characters
 *  can be given, such as "abcde". The '-' can represent a range
 *  of characters, as "a-e" represents the same pattern as "abcde".
 *  "a-fA-F0-9" represents all the hex characters.
 *  If the first character of a pattern is '^', then the pattern
 *  is negated, i.e. "^0-9" means any character except a digit.
 *  The functions inPattern, <b>countchars</b>, <b>removeschars</b>,
 *  and <b>squeeze</b>
 *  use patterns.
 *
 * Note: In the future, the pattern syntax may be improved
 *  to be more like regular expression character classes.
 */

bool inPattern(S)(dchar c, in S pattern) @safe pure if (isSomeString!S)
{
    bool result = false;
    int range = 0;
    dchar lastc;

    foreach (size_t i, dchar p; pattern)
    {
        if (p == '^' && i == 0)
        {
            result = true;
            if (i + 1 == pattern.length)
                return (c == p);    // or should this be an error?
        }
        else if (range)
        {
            range = 0;
            if (lastc <= c && c <= p || c == p)
                return !result;
        }
        else if (p == '-' && i > result && i + 1 < pattern.length)
        {
            range = 1;
            continue;
        }
        else if (c == p)
            return !result;
        lastc = p;
    }
    return result;
}


unittest
{
    debug(string) printf("std.string.inPattern.unittest\n");

    assertCTFEable!(
    {
    assert(inPattern('x', "x") == 1);
    assert(inPattern('x', "y") == 0);
    assert(inPattern('x', string.init) == 0);
    assert(inPattern('x', "^y") == 1);
    assert(inPattern('x', "yxxy") == 1);
    assert(inPattern('x', "^yxxy") == 0);
    assert(inPattern('x', "^abcd") == 1);
    assert(inPattern('^', "^^") == 0);
    assert(inPattern('^', "^") == 1);
    assert(inPattern('^', "a^") == 1);
    assert(inPattern('x', "a-z") == 1);
    assert(inPattern('x', "A-Z") == 0);
    assert(inPattern('x', "^a-z") == 0);
    assert(inPattern('x', "^A-Z") == 1);
    assert(inPattern('-', "a-") == 1);
    assert(inPattern('-', "^A-") == 0);
    assert(inPattern('a', "z-a") == 1);
    assert(inPattern('z', "z-a") == 1);
    assert(inPattern('x', "z-a") == 0);
    });
}


/***********************************************
 * See if character c is in the intersection of the patterns.
 */

bool inPattern(S)(dchar c, S[] patterns) @safe pure if (isSomeString!S)
{
    foreach (string pattern; patterns)
    {
        if (!inPattern(c, pattern))
        {
            return false;
        }
    }
    return true;
}


/********************************************
 * Count characters in s that match pattern.
 */

size_t countchars(S, S1)(S s, in S1 pattern) @safe pure if (isSomeString!S && isSomeString!S1)
{
    size_t count;
    foreach (dchar c; s)
    {
        count += inPattern(c, pattern);
    }
    return count;
}

unittest
{
    debug(string) printf("std.string.count.unittest\n");

    assertCTFEable!(
    {
    assert(countchars("abc", "a-c") == 3);
    assert(countchars("hello world", "or") == 3);
    });
}


/********************************************
 * Return string that is s with all characters removed that match pattern.
 */

S removechars(S)(S s, in S pattern) @safe pure if (isSomeString!S)
{
    Unqual!(typeof(s[0]))[] r;
    bool changed = false;

    foreach (size_t i, dchar c; s)
    {
        if (inPattern(c, pattern))
        {
            if (!changed)
            {
                changed = true;
                r = s[0 .. i].dup;
            }
            continue;
        }
        if (changed)
        {
            std.utf.encode(r, c);
        }
    }
    if (changed)
        return r;
    else
        return s;
}

unittest
{
    debug(string) printf("std.string.removechars.unittest\n");

    assertCTFEable!(
    {
    assert(removechars("abc", "a-c").length == 0);
    assert(removechars("hello world", "or") == "hell wld");
    assert(removechars("hello world", "d") == "hello worl");
    assert(removechars("hah", "h") == "a");
    });
}


/***************************************************
 * Return string where sequences of a character in s[] from pattern[]
 * are replaced with a single instance of that character.
 * If pattern is null, it defaults to all characters.
 */

S squeeze(S)(S s, in S pattern = null)
{
    Unqual!(typeof(s[0]))[] r;
    dchar lastc;
    size_t lasti;
    int run;
    bool changed;

    foreach (size_t i, dchar c; s)
    {
        if (run && lastc == c)
        {
            changed = true;
        }
        else if (pattern is null || inPattern(c, pattern))
        {
            run = 1;
            if (changed)
            {
                if (r is null)
                    r = s[0 .. lasti].dup;
                std.utf.encode(r, c);
            }
            else
                lasti = i + std.utf.stride(s, i);
            lastc = c;
        }
        else
        {
            run = 0;
            if (changed)
            {
                if (r is null)
                    r = s[0 .. lasti].dup;
                std.utf.encode(r, c);
            }
        }
    }
    return changed ? ((r is null) ? s[0 .. lasti] : cast(S) r) : s;
}

unittest
{
    debug(string) printf("std.string.squeeze.unittest\n");

    assertCTFEable!(
    {
    string s;

    assert(squeeze("hello") == "helo");

    s = "abcd";
    assert(squeeze(s) is s);
    s = "xyzz";
    assert(squeeze(s).ptr == s.ptr); // should just be a slice

    assert(squeeze("hello goodbyee", "oe") == "hello godbye");
    });
}

/***************************************************************
 Finds the position $(D_PARAM pos) of the first character in $(D_PARAM
 s) that does not match $(D_PARAM pattern) (in the terminology used by
 $(LINK2 std_string.html,inPattern)). Updates $(D_PARAM s =
 s[pos..$]). Returns the slice from the beginning of the original
 (before update) string up to, and excluding, $(D_PARAM pos).

 Example:
 ---
 string s = "123abc";
 string t = munch(s, "0123456789");
 assert(t == "123" && s == "abc");
 t = munch(s, "0123456789");
 assert(t == "" && s == "abc");
 ---

The $(D_PARAM munch) function is mostly convenient for skipping
certain category of characters (e.g. whitespace) when parsing
strings. (In such cases, the return value is not used.)
 */

S1 munch(S1, S2)(ref S1 s, S2 pattern)
{
    size_t j = s.length;
    foreach (i, dchar c; s)
    {
        if (!inPattern(c, pattern))
        {
            j = i;
            break;
        }
    }
    scope(exit) s = s[j .. $];
    return s[0 .. j];
}

@safe pure unittest
{
    string s = "123€abc";
    string t = munch(s, "0123456789");
    assert(t == "123" && s == "€abc");
    t = munch(s, "0123456789");
    assert(t == "" && s == "€abc");
    t = munch(s, "£$€¥");
    assert(t == "€" && s == "abc");
}


/**********************************************
 * Return string that is the 'successor' to s[].
 * If the rightmost character is a-zA-Z0-9, it is incremented within
 * its case or digits. If it generates a carry, the process is
 * repeated with the one to its immediate left.
 */

S succ(S)(S s) @safe pure if (isSomeString!S)
{
    if (s.length && std.ascii.isAlphaNum(s[$ - 1]))
    {
        auto r = s.dup;
        size_t i = r.length - 1;

        while (1)
        {
            dchar c = s[i];
            dchar carry;

            switch (c)
            {
            case '9':
                c = '0';
                carry = '1';
                goto Lcarry;
            case 'z':
            case 'Z':
                c -= 'Z' - 'A';
                carry = c;
            Lcarry:
                r[i] = cast(char)c;
                if (i == 0)
                {
                    auto t = new typeof(r[0])[r.length + 1];
                    t[0] = cast(char) carry;
                    t[1 .. $] = r[];
                    return t;
                }
                i--;
                break;

            default:
                if (std.ascii.isAlphaNum(c))
                    r[i]++;
                return r;
            }
        }
    }
    return s;
}

unittest
{
    debug(string) printf("std.string.succ.unittest\n");

    assertCTFEable!(
    {
    assert(succ(string.init) is null);
    assert(succ("!@#$%") == "!@#$%");
    assert(succ("1") == "2");
    assert(succ("9") == "10");
    assert(succ("999") == "1000");
    assert(succ("zz99") == "aaa00");
    });
}


/++
    Replaces the characters in $(D str) which are in $(D from) with the
    the corresponding characters in $(D to) and returns the resulting string.

    $(D tr) is based on
    $(WEB pubs.opengroup.org/onlinepubs/9699919799/utilities/_tr.html, Posix's tr),
    though it doesn't do everything that the Posix utility does.

    Params:
        str       = The original string.
        from      = The characters to replace.
        to        = The characters to replace with.
        modifiers = String containing modifiers.

    Modifiers:
        $(BOOKTABLE,
        $(TR $(TD Modifier) $(TD Description))
        $(TR $(TD $(D 'c')) $(TD Complement the list of characters in $(D from)))
        $(TR $(TD $(D 'd')) $(TD Removes matching characters with no corresponding
                              replacement in $(D to)))
        $(TR $(TD $(D 's')) $(TD Removes adjacent duplicates in the replaced
                              characters))
        )

    If the modifier $(D 'd') is present, then the number of characters in
    $(D to) may be only $(D 0) or $(D 1).

    If the modifier $(D 'd') is $(I not) present, and $(D to) is empty, then
    $(D to) is taken to be the same as $(D from).

    If the modifier $(D 'd') is $(I not) present, and $(D to) is shorter than
    $(D from), then $(D to) is extended by replicating the last character in
    $(D to).

    Both $(D from) and $(D to) may contain ranges using the $(D '-') character
    (e.g. $(D "a-d") is synonymous with $(D "abcd").) Neither accept a leading
    $(D '^') as meaning the complement of the string (use the $(D 'c') modifier
    for that).
  +/
C1[] tr(C1, C2, C3, C4 = immutable char)
       (C1[] str, const(C2)[] from, const(C3)[] to, const(C4)[] modifiers = null)
{
    bool mod_c;
    bool mod_d;
    bool mod_s;

    foreach (char c; modifiers)
    {
        switch (c)
        {
        case 'c':   mod_c = 1; break;   // complement
        case 'd':   mod_d = 1; break;   // delete unreplaced chars
        case 's':   mod_s = 1; break;   // squeeze duplicated replaced chars
        default:    assert(0);
        }
    }

    if (to.empty && !mod_d)
        to = std.conv.to!(typeof(to))(from);

    auto result = appender!(C1[])();
    bool modified;
    dchar lastc;

    foreach (dchar c; str)
    {
        dchar lastf;
        dchar lastt;
        dchar newc;
        int n = 0;

        for (size_t i = 0; i < from.length; )
        {
            dchar f = std.utf.decode(from, i);
            if (f == '-' && lastf != dchar.init && i < from.length)
            {
                dchar nextf = std.utf.decode(from, i);
                if (lastf <= c && c <= nextf)
                {
                    n += c - lastf - 1;
                    if (mod_c)
                        goto Lnotfound;
                    goto Lfound;
                }
                n += nextf - lastf;
                lastf = lastf.init;
                continue;
            }

            if (c == f)
            {   if (mod_c)
                    goto Lnotfound;
                goto Lfound;
            }
            lastf = f;
            n++;
        }
        if (!mod_c)
            goto Lnotfound;
        n = 0;          // consider it 'found' at position 0

      Lfound:

        // Find the nth character in to[]
        dchar nextt;
        for (size_t i = 0; i < to.length; )
        {   dchar t = std.utf.decode(to, i);
            if (t == '-' && lastt != dchar.init && i < to.length)
            {
                nextt = std.utf.decode(to, i);
                n -= nextt - lastt;
                if (n < 0)
                {
                    newc = nextt + n + 1;
                    goto Lnewc;
                }
                lastt = dchar.init;
                continue;
            }
            if (n == 0)
            {   newc = t;
                goto Lnewc;
            }
            lastt = t;
            nextt = t;
            n--;
        }
        if (mod_d)
            continue;
        newc = nextt;

      Lnewc:
        if (mod_s && modified && newc == lastc)
            continue;
        result.put(newc);
        assert(newc != dchar.init);
        modified = true;
        lastc = newc;
        continue;

      Lnotfound:
        result.put(c);
        lastc = c;
        modified = false;
    }

    return result.data;
}

unittest
{
    debug(string) printf("std.string.tr.unittest\n");
    import std.algorithm;

    // Complete list of test types; too slow to test'em all
    // alias TestTypes = TypeTuple!(
    //          char[], const( char)[], immutable( char)[],
    //         wchar[], const(wchar)[], immutable(wchar)[],
    //         dchar[], const(dchar)[], immutable(dchar)[]);

    // Reduced list of test types
    alias TestTypes = TypeTuple!(char[], const(wchar)[], immutable(dchar)[]);

    assertCTFEable!(
    {
    foreach (S; TestTypes)
    {
        foreach (T; TestTypes)
        {
            foreach (U; TestTypes)
            {
                assert(equal(tr(to!S("abcdef"), to!T("cd"), to!U("CD")), "abCDef"));
                assert(equal(tr(to!S("abcdef"), to!T("b-d"), to!U("B-D")), "aBCDef"));
                assert(equal(tr(to!S("abcdefgh"), to!T("b-dh"), to!U("B-Dx")), "aBCDefgx"));
                assert(equal(tr(to!S("abcdefgh"), to!T("b-dh"), to!U("B-CDx")), "aBCDefgx"));
                assert(equal(tr(to!S("abcdefgh"), to!T("b-dh"), to!U("B-BCDx")), "aBCDefgx"));
                assert(equal(tr(to!S("abcdef"), to!T("ef"), to!U("*"), to!S("c")), "****ef"));
                assert(equal(tr(to!S("abcdef"), to!T("ef"), to!U(""), to!T("d")), "abcd"));
                assert(equal(tr(to!S("hello goodbye"), to!T("lo"), to!U(""), to!U("s")), "helo godbye"));
                assert(equal(tr(to!S("hello goodbye"), to!T("lo"), to!U("x"), "s"), "hex gxdbye"));
                assert(equal(tr(to!S("14-Jul-87"), to!T("a-zA-Z"), to!U(" "), "cs"), " Jul "));
                assert(equal(tr(to!S("Abc"), to!T("AAA"), to!U("XYZ")), "Xbc"));
            }
        }

        auto s = to!S("hello world");
        static assert(is(typeof(s) == typeof(tr(s, "he", "if"))));
    }
    });
}


/* ************************************************
 * Version       : v0.3
 * Author        : David L. 'SpottedTiger' Davis
 * Date Created  : 31.May.05 Compiled and Tested with dmd v0.125
 * Date Modified : 01.Jun.05 Modified the function to handle the
 *               :           imaginary and complex float-point
 *               :           datatypes.
 *               :
 * Licence       : Public Domain / Contributed to Digital Mars
 */

/**
 * [in] string s can be formatted in the following ways:
 *
 * Integer Whole Number:
 * (for byte, ubyte, short, ushort, int, uint, long, and ulong)
 * ['+'|'-']digit(s)[U|L|UL]
 *
 * examples: 123, 123UL, 123L, +123U, -123L
 *
 * Floating-Point Number:
 * (for float, double, real, ifloat, idouble, and ireal)
 * ['+'|'-']digit(s)[.][digit(s)][[e-|e+]digit(s)][i|f|L|Li|fi]]
 *      or [nan|nani|inf|-inf]
 *
 * examples: +123., -123.01, 123.3e-10f, 123.3e-10fi, 123.3e-10L
 *
 * (for cfloat, cdouble, and creal)
 * ['+'|'-']digit(s)[.][digit(s)][[e-|e+]digit(s)][+]
 *         [digit(s)[.][digit(s)][[e-|e+]digit(s)][i|f|L|Li|fi]]
 *      or [nan|nani|nan+nani|inf|-inf]
 *
 * examples: nan, -123e-1+456.9e-10Li, +123e+10+456i, 123+456
 *
 * [in] bool bAllowSep
 * False by default, but when set to true it will accept the
 * separator characters "," and "_" within the string, but these
 * characters should be stripped from the string before using any
 * of the conversion functions like toInt(), toFloat(), and etc
 * else an error will occur.
 *
 * Also please note, that no spaces are allowed within the string
 * anywhere whether it's a leading, trailing, or embedded space(s),
 * thus they too must be stripped from the string before using this
 * function, or any of the conversion functions.
 */

bool isNumeric(const(char)[] s, in bool bAllowSep = false) @safe pure
{
    immutable iLen = s.length;
    if (iLen == 0)
        return false;

    // Check for NaN (Not a Number) and for Infinity
    if (s.among!((a, b) => icmp(a, b) == 0)
            ("nan", "nani", "nan+nani", "inf", "-inf"))
        return true;

    immutable j = s[0].among!('-', '+') != 0;
    bool bDecimalPoint, bExponent, bComplex, sawDigits;

    for (size_t i = j; i < iLen; i++)
    {
        immutable c = s[i];

        // Digits are good, continue checking
        // with the popFront character... ;)
        if (c >= '0' && c <= '9')
        {
            sawDigits = true;
            continue;
        }

        // Check for the complex type, and if found
        // reset the flags for checking the 2nd number.
        if (c == '+')
        {
            if (!i)
                return false;
            bDecimalPoint = false;
            bExponent = false;
            bComplex = true;
            sawDigits = false;
            continue;
        }

        // Allow only one exponent per number
        if (c.among!('e', 'E'))
        {
            // A 2nd exponent found, return not a number
            if (bExponent || i + 1 >= iLen)
                return false;
            // Look forward for the sign, and if
            // missing then this is not a number.
            if (!s[i + 1].among!('-', '+'))
                return false;
            bExponent = true;
            i++;
            continue;
        }
        // Allow only one decimal point per number to be used
        if (c == '.' )
        {
            // A 2nd decimal point found, return not a number
            if (bDecimalPoint)
                return false;
            bDecimalPoint = true;
            continue;
        }
        // Check for ending literal characters: "f,u,l,i,ul,fi,li",
        // and whether they're being used with the correct datatype.
        if (i == iLen - 2)
        {
            if (!sawDigits)
                return false;
            // Integer Whole Number
            if (icmp(s[i..iLen], "ul") == 0 &&
                    (!bDecimalPoint && !bExponent && !bComplex))
                return true;
            // Floating-Point Number
            if (s[i..iLen].among!((a, b) => icmp(a, b) == 0)("fi", "li") &&
                    (bDecimalPoint || bExponent || bComplex))
                return true;
            if (icmp(s[i..iLen], "ul") == 0 &&
                    (bDecimalPoint || bExponent || bComplex))
                return false;
            // Could be a Integer or a Float, thus
            // all these suffixes are valid for both
            return s[i..iLen].among!((a, b) => icmp(a, b) == 0)
                ("ul", "fi", "li") != 0;
        }
        if (i == iLen - 1)
        {
            if (!sawDigits)
                return false;
            // Integer Whole Number
            if (c.among!('u', 'l', 'U', 'L') &&
                   (!bDecimalPoint && !bExponent && !bComplex))
                return true;
            // Check to see if the last character in the string
            // is the required 'i' character
            if (bComplex)
                return c.among!('i', 'I') != 0;
            // Floating-Point Number
            return c.among!('l', 'L', 'f', 'F', 'i', 'I') != 0;
        }
        // Check if separators are allowed to be in the numeric string
        if (!bAllowSep || !c.among!('_', ','))
            return false;
    }

    return sawDigits;
}

unittest
{
    assert(!isNumeric("F"));
    assert(!isNumeric("L"));
    assert(!isNumeric("U"));
    assert(!isNumeric("i"));
    assert(!isNumeric("fi"));
    assert(!isNumeric("ul"));
    assert(!isNumeric("li"));
    assert(!isNumeric("."));
    assert(!isNumeric("-"));
    assert(!isNumeric("+"));
    assert(!isNumeric("e-"));
    assert(!isNumeric("e+"));
    assert(!isNumeric(".f"));
    assert(!isNumeric("e+f"));
}


unittest
{
    debug(string) printf("isNumeric(in string, bool = false).unittest\n");

    assertCTFEable!(
    {
    // Test the isNumeric(in string) function
    assert(isNumeric("1") == true );
    assert(isNumeric("1.0") == true );
    assert(isNumeric("1e-1") == true );
    assert(isNumeric("12345xxxx890") == false );
    assert(isNumeric("567L") == true );
    assert(isNumeric("23UL") == true );
    assert(isNumeric("-123..56f") == false );
    assert(isNumeric("12.3.5.6") == false );
    assert(isNumeric(" 12.356") == false );
    assert(isNumeric("123 5.6") == false );
    assert(isNumeric("1233E-1+1.0e-1i") == true );

    assert(isNumeric("123.00E-5+1234.45E-12Li") == true);
    assert(isNumeric("123.00e-5+1234.45E-12iL") == false);
    assert(isNumeric("123.00e-5+1234.45e-12uL") == false);
    assert(isNumeric("123.00E-5+1234.45e-12lu") == false);

    assert(isNumeric("123fi") == true);
    assert(isNumeric("123li") == true);
    assert(isNumeric("--123L") == false);
    assert(isNumeric("+123.5UL") == false);
    assert(isNumeric("123f") == true);
    assert(isNumeric("123.u") == false);

  // @@@BUG@@ to!string(float) is not CTFEable.
  // Related: formatValue(T) if (is(FloatingPointTypeOf!T))
  if (!__ctfe)
  {
    assert(isNumeric(to!string(real.nan)) == true);
    assert(isNumeric(to!string(-real.infinity)) == true);
    assert(isNumeric(to!string(123e+2+1234.78Li)) == true);
  }

    string s = "$250.99-";
    assert(isNumeric(s[1..s.length - 2]) == true);
    assert(isNumeric(s) == false);
    assert(isNumeric(s[0..s.length - 1]) == false);
    });

    assert(!isNumeric("-"));
    assert(!isNumeric("+"));
}


/*****************************
 * Soundex algorithm.
 *
 * The Soundex algorithm converts a word into 4 characters
 * based on how the word sounds phonetically. The idea is that
 * two spellings that sound alike will have the same Soundex
 * value, which means that Soundex can be used for fuzzy matching
 * of names.
 *
 * Params:
 *  string = String to convert to Soundex representation.
 *  buffer = Optional 4 char array to put the resulting Soundex
 *      characters into. If null, the return value
 *      buffer will be allocated on the heap.
 * Returns:
 *  The four character array with the Soundex result in it.
 *  Returns null if there is no Soundex representation for the string.
 *
 * See_Also:
 *  $(LINK2 http://en.wikipedia.org/wiki/Soundex, Wikipedia),
 *  $(LUCKY The Soundex Indexing System)
 *
 * Bugs:
 *  Only works well with English names.
 *  There are other arguably better Soundex algorithms,
 *  but this one is the standard one.
 */

char[] soundex(const(char)[] string, char[] buffer = null) @safe pure nothrow
in
{
    assert(!buffer.ptr || buffer.length >= 4);
}
out (result)
{
    if (result.ptr)
    {
        assert(result.length == 4);
        assert(result[0] >= 'A' && result[0] <= 'Z');
        foreach (char c; result[1 .. 4])
            assert(c >= '0' && c <= '6');
    }
}
body
{
    static immutable dex =
        // ABCDEFGHIJKLMNOPQRSTUVWXYZ
        "01230120022455012623010202";

    int b = 0;
    char lastc;
    foreach (char cs; string)
    {   auto c = cs;        // necessary because cs is final

        if (c >= 'a' && c <= 'z')
            c -= 'a' - 'A';
        else if (c >= 'A' && c <= 'Z')
        {
        }
        else
        {
            lastc = lastc.init;
            continue;
        }
        if (b == 0)
        {
            if (!buffer.ptr)
                buffer = new char[4];
            buffer[0] = c;
            b++;
            lastc = dex[c - 'A'];
        }
        else
        {
            if (c == 'H' || c == 'W')
                continue;
            if (c == 'A' || c == 'E' || c == 'I' || c == 'O' || c == 'U')
                lastc = lastc.init;
            c = dex[c - 'A'];
            if (c != '0' && c != lastc)
            {
                buffer[b] = c;
                b++;
                lastc = c;
            }
        }
        if (b == 4)
            goto Lret;
    }
    if (b == 0)
        buffer = null;
    else
        buffer[b .. 4] = '0';
  Lret:
    return buffer;
}

@safe pure nothrow unittest
{
    assertCTFEable!(
    {
    char[4] buffer;

    assert(soundex(null) == null);
    assert(soundex("") == null);
    assert(soundex("0123^&^^**&^") == null);
    assert(soundex("Euler") == "E460");
    assert(soundex(" Ellery ") == "E460");
    assert(soundex("Gauss") == "G200");
    assert(soundex("Ghosh") == "G200");
    assert(soundex("Hilbert") == "H416");
    assert(soundex("Heilbronn") == "H416");
    assert(soundex("Knuth") == "K530");
    assert(soundex("Kant", buffer) == "K530");
    assert(soundex("Lloyd") == "L300");
    assert(soundex("Ladd") == "L300");
    assert(soundex("Lukasiewicz", buffer) == "L222");
    assert(soundex("Lissajous") == "L222");
    assert(soundex("Robert") == "R163");
    assert(soundex("Rupert") == "R163");
    assert(soundex("Rubin") == "R150");
    assert(soundex("Washington") == "W252");
    assert(soundex("Lee") == "L000");
    assert(soundex("Gutierrez") == "G362");
    assert(soundex("Pfister") == "P236");
    assert(soundex("Jackson") == "J250");
    assert(soundex("Tymczak") == "T522");
    assert(soundex("Ashcraft") == "A261");

    assert(soundex("Woo") == "W000");
    assert(soundex("Pilgrim") == "P426");
    assert(soundex("Flingjingwaller") == "F452");
    assert(soundex("PEARSE") == "P620");
    assert(soundex("PIERCE") == "P620");
    assert(soundex("Price") == "P620");
    assert(soundex("CATHY") == "C300");
    assert(soundex("KATHY") == "K300");
    assert(soundex("Jones") == "J520");
    assert(soundex("johnsons") == "J525");
    assert(soundex("Hardin") == "H635");
    assert(soundex("Martinez") == "M635");
    });
}


/***************************************************
 * Construct an associative array consisting of all
 * abbreviations that uniquely map to the strings in values.
 *
 * This is useful in cases where the user is expected to type
 * in one of a known set of strings, and the program will helpfully
 * autocomplete the string once sufficient characters have been
 * entered that uniquely identify it.
 * Example:
 * ---
 * import std.stdio;
 * import std.string;
 *
 * void main()
 * {
 *    static string[] list = [ "food", "foxy" ];
 *
 *    auto abbrevs = std.string.abbrev(list);
 *
 *    foreach (key, value; abbrevs)
 *    {
 *       writefln("%s => %s", key, value);
 *    }
 * }
 * ---
 * produces the output:
 * <pre>
 * fox =&gt; foxy
 * food =&gt; food
 * foxy =&gt; foxy
 * foo =&gt; food
 * </pre>
 */

string[string] abbrev(string[] values) @safe pure
{
    string[string] result;

    // Make a copy when sorting so we follow COW principles.
    values = values.dup.sort;   // @@@BUG@@@ not CTFEable

    size_t values_length = values.length;
    size_t lasti = values_length;
    size_t nexti;

    string nv;
    string lv;

    for (size_t i = 0; i < values_length; i = nexti)
    {
        string value = values[i];

        // Skip dups
        for (nexti = i + 1; nexti < values_length; nexti++)
        {
            nv = values[nexti];
            if (value != values[nexti])
                break;
        }

        for (size_t j = 0; j < value.length; j += std.utf.stride(value, j))
        {
            string v = value[0 .. j];

            if ((nexti == values_length || j > nv.length || v != nv[0 .. j]) &&
                (lasti == values_length || j > lv.length || v != lv[0 .. j]))
            {
                result[v] = value;
            }
        }
        result[value] = value;
        lasti = i;
        lv = value;
    }

    return result;
}

unittest
{
    debug(string) printf("string.abbrev.unittest\n");

    // @@@BUG@@@ Built-in arr.sort is not CTFEable
    //assertCTFEable!(
    //{
    string[] values;
    values ~= "hello";
    values ~= "hello";
    values ~= "he";

    string[string] r;

    r = abbrev(values);
    auto keys = r.keys.dup;
    keys.sort;

    assert(keys.length == 4);
    assert(keys[0] == "he");
    assert(keys[1] == "hel");
    assert(keys[2] == "hell");
    assert(keys[3] == "hello");

    assert(r[keys[0]] == "he");
    assert(r[keys[1]] == "hello");
    assert(r[keys[2]] == "hello");
    assert(r[keys[3]] == "hello");
    //});
}


/******************************************
 * Compute column number after string if string starts in the
 * leftmost column, which is numbered starting from 0.
 */

size_t column(S)(S str, size_t tabsize = 8) @safe pure if (isSomeString!S)
{
    size_t column;

    foreach (dchar c; str)
    {
        switch (c)
        {
        case '\t':
            column = (column + tabsize) / tabsize * tabsize;
            break;

        case '\r':
        case '\n':
        case paraSep:
        case lineSep:
            column = 0;
            break;

        default:
            column++;
            break;
        }
    }
    return column;
}

unittest
{
    debug(string) printf("string.column.unittest\n");

    assertCTFEable!(
    {
    assert(column(string.init) == 0);
    assert(column("") == 0);
    assert(column("\t") == 8);
    assert(column("abc\t") == 8);
    assert(column("12345678\t") == 16);
    });
}

/******************************************
 * Wrap text into a paragraph.
 *
 * The input text string s is formed into a paragraph
 * by breaking it up into a sequence of lines, delineated
 * by \n, such that the number of columns is not exceeded
 * on each line.
 * The last line is terminated with a \n.
 * Params:
 *  s = text string to be wrapped
 *  columns = maximum number of _columns in the paragraph
 *  firstindent = string used to _indent first line of the paragraph
 *  indent = string to use to _indent following lines of the paragraph
 *  tabsize = column spacing of tabs
 * Returns:
 *  The resulting paragraph.
 */

S wrap(S)(S s, size_t columns = 80, S firstindent = null,
        S indent = null, size_t tabsize = 8) @safe pure if (isSomeString!S)
{
    typeof(s.dup) result;
    int spaces;
    bool inword;
    bool first = true;
    size_t wordstart;

    result.length = firstindent.length + s.length;
    result.length = firstindent.length;
    result[] = firstindent[];
    auto col = column(result.idup, tabsize);
    foreach (size_t i, dchar c; s)
    {
        if (std.uni.isWhite(c))
        {
            if (inword)
            {
                if (first)
                {
                }
                else if (col + 1 + (i - wordstart) > columns)
                {
                    result ~= '\n';
                    result ~= indent;
                    col = column(indent, tabsize);
                }
                else
                {
                    result ~= ' ';
                    col += 1;
                }
                result ~= s[wordstart .. i];
                col += i - wordstart;
                inword = false;
                first = false;
            }
        }
        else
        {
            if (!inword)
            {
                wordstart = i;
                inword = true;
            }
        }
    }

    if (inword)
    {
        if (col + 1 + (s.length - wordstart) >= columns)
        {
            result ~= '\n';
            result ~= indent;
        }
        else if (result.length != firstindent.length)
            result ~= ' ';
        result ~= s[wordstart .. s.length];
    }
    result ~= '\n';

    return result;
}

unittest
{
    debug(string) printf("string.wrap.unittest\n");

    assertCTFEable!(
    {
    assert(wrap(string.init) == "\n");
    assert(wrap(" a b   df ") == "a b df\n");
    assert(wrap(" a b   df ", 3) == "a b\ndf\n");
    assert(wrap(" a bc   df ", 3) == "a\nbc\ndf\n");
    assert(wrap(" abcd   df ", 3) == "abcd\ndf\n");
    assert(wrap("x") == "x\n");
    assert(wrap("u u") == "u u\n");
    });
}

/******************************************
 * Removes indentation from a multi-line string or an array of single-line strings.
 *
 * This uniformly outdents the text as much as possible.
 * Whitespace-only lines are always converted to blank lines.
 *
 * A StringException will be thrown if inconsistent indentation prevents
 * the input from being outdented.
 *
 * Works at compile-time.
 *
 * Example:
 * ---
 * writeln(q{
 *     import std.stdio;
 *     void main() {
 *         writeln("Hello");
 *     }
 * }.outdent());
 * ---
 *
 * Output:
 * ---
 *
 * import std.stdio;
 * void main() {
 *     writeln("Hello");
 * }
 *
 * ---
 *
 */

S outdent(S)(S str) @safe pure if(isSomeString!S)
{
    return str.splitLines(KeepTerminator.yes).outdent().join();
}

/// ditto
S[] outdent(S)(S[] lines) @safe pure if(isSomeString!S)
{
    if (lines.empty)
    {
        return null;
    }

    static S leadingWhiteOf(S str)
    {
        return str[ 0 .. $-find!(not!(std.uni.isWhite))(str).length ];
    }

    S shortestIndent;
    foreach (i, line; lines)
    {
        auto stripped = __ctfe? line.ctfe_strip() : line.strip();

        if (stripped.empty)
        {
            lines[i] = line[line.chomp().length..$];
        }
        else
        {
            auto indent = leadingWhiteOf(line);

            // Comparing number of code units instead of code points is OK here
            // because this function throws upon inconsistent indentation.
            if (shortestIndent is null || indent.length < shortestIndent.length)
            {
                if (indent.empty)
                    return lines;
                shortestIndent = indent;
            }
        }
    }

    foreach (i; 0..lines.length)
    {
        auto stripped = __ctfe? lines[i].ctfe_strip() : lines[i].strip();

        if (stripped.empty)
        {
            // Do nothing
        }
        else if (lines[i].startsWith(shortestIndent))
        {
            lines[i] = lines[i][shortestIndent.length..$];
        }
        else
        {
            if (__ctfe)
                assert(false, "outdent: Inconsistent indentation");
            else
                throw new StringException("outdent: Inconsistent indentation");
        }
    }

    return lines;
}

// TODO: Remove this and use std.string.strip when retro() becomes ctfe-able.
private S ctfe_strip(S)(S str) if(isSomeString!(Unqual!S))
{
    return str.stripLeft().ctfe_stripRight();
}

// TODO: Remove this and use std.string.strip when retro() becomes ctfe-able.
private S ctfe_stripRight(S)(S str) if(isSomeString!(Unqual!S))
{
    size_t endIndex = 0;
    size_t prevIndex = str.length;

    foreach_reverse (i, dchar ch; str)
    {
        if (!std.uni.isWhite(ch))
        {
            endIndex = prevIndex;
            break;
        }
        prevIndex = i;
    }

    return str[0..endIndex];
}

unittest
{
    debug(string) printf("string.outdent.unittest\n");

    template outdent_testStr(S)
    {
        enum S outdent_testStr =
"
 \t\tX
 \t\U00010143X
 \t\t

 \t\t\tX
\t ";
    }

    template outdent_expected(S)
    {
        enum S outdent_expected =
"
\tX
\U00010143X


\t\tX
";
    }

    assertCTFEable!(
    {
    static assert(ctfe_strip(" \tHi \r\n") == "Hi");
    static assert(ctfe_strip(" \tHi&copy;\u2028 \r\n") == "Hi&copy;");
    static assert(ctfe_strip("Hi")         == "Hi");
    static assert(ctfe_strip(" \t \r\n")   == "");
    static assert(ctfe_strip("")           == "");

    foreach (S; TypeTuple!(string, wstring, dstring))
    {
        enum S blank = "";
        assert(blank.outdent() == blank);
        static assert(blank.outdent() == blank);

        enum S testStr1  = " \n \t\n ";
        enum S expected1 = "\n\n";
        assert(testStr1.outdent() == expected1);
        static assert(testStr1.outdent() == expected1);

        assert(testStr1[0..$-1].outdent() == expected1);
        static assert(testStr1[0..$-1].outdent() == expected1);

        enum S testStr2  = "a\n \t\nb";
        assert(testStr2.outdent() == testStr2);
        static assert(testStr2.outdent() == testStr2);

        enum S testStr3 =
"
 \t\tX
 \t\U00010143X
 \t\t

 \t\t\tX
\t ";

        enum S expected3 =
"
\tX
\U00010143X


\t\tX
";
        assert(testStr3.outdent() == expected3);
        static assert(testStr3.outdent() == expected3);

        enum testStr4 = "  X\r  X\n  X\r\n  X\u2028  X\u2029  X";
        enum expected4 = "X\rX\nX\r\nX\u2028X\u2029X";
        assert(testStr4.outdent() == expected4);
        static assert(testStr4.outdent() == expected4);

        enum testStr5  = testStr4[0..$-1];
        enum expected5 = expected4[0..$-1];
        assert(testStr5.outdent() == expected5);
        static assert(testStr5.outdent() == expected5);

        enum testStr6 = "  \r  \n  \r\n  \u2028  \u2029";
        enum expected6 = "\r\n\r\n\u2028\u2029";
        assert(testStr6.outdent() == expected6);
        static assert(testStr6.outdent() == expected6);
    }
    });
}
debug(string)void main(){}