michael@0: /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
michael@0: /* This Source Code Form is subject to the terms of the Mozilla Public
michael@0:  * License, v. 2.0. If a copy of the MPL was not distributed with this
michael@0:  * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
michael@0: // IWYU pragma: private, include "nsString.h"
michael@0: 
michael@0: #ifndef nsReadableUtils_h___
michael@0: #define nsReadableUtils_h___
michael@0: 
michael@0:   /**
michael@0:    * I guess all the routines in this file are all mis-named.
michael@0:    * According to our conventions, they should be |NS_xxx|.
michael@0:    */
michael@0: 
michael@0: #include "nsAString.h"
michael@0: 
michael@0: #include "nsTArrayForwardDeclare.h"
michael@0: 
michael@0: inline size_t Distance( const nsReadingIterator<char16_t>& start, const nsReadingIterator<char16_t>& end )
michael@0:   {
michael@0:     return end.get() - start.get();
michael@0:   }
michael@0: inline size_t Distance( const nsReadingIterator<char>& start, const nsReadingIterator<char>& end )
michael@0:   {
michael@0:     return end.get() - start.get();
michael@0:   }
michael@0: 
michael@0: void LossyCopyUTF16toASCII( const nsAString& aSource, nsACString& aDest );
michael@0: void CopyASCIItoUTF16( const nsACString& aSource, nsAString& aDest );
michael@0: 
michael@0: void LossyCopyUTF16toASCII( const char16_t* aSource, nsACString& aDest );
michael@0: void CopyASCIItoUTF16( const char* aSource, nsAString& aDest );
michael@0: 
michael@0: void CopyUTF16toUTF8( const nsAString& aSource, nsACString& aDest );
michael@0: void CopyUTF8toUTF16( const nsACString& aSource, nsAString& aDest );
michael@0: 
michael@0: void CopyUTF16toUTF8( const char16_t* aSource, nsACString& aDest );
michael@0: void CopyUTF8toUTF16( const char* aSource, nsAString& aDest );
michael@0: 
michael@0: void LossyAppendUTF16toASCII( const nsAString& aSource, nsACString& aDest );
michael@0: void AppendASCIItoUTF16( const nsACString& aSource, nsAString& aDest );
michael@0: bool AppendASCIItoUTF16( const nsACString& aSource, nsAString& aDest,
michael@0:                          const mozilla::fallible_t& ) NS_WARN_UNUSED_RESULT;
michael@0: 
michael@0: void LossyAppendUTF16toASCII( const char16_t* aSource, nsACString& aDest );
michael@0: void AppendASCIItoUTF16( const char* aSource, nsAString& aDest );
michael@0: 
michael@0: void AppendUTF16toUTF8( const nsAString& aSource, nsACString& aDest );
michael@0: bool AppendUTF16toUTF8( const nsAString& aSource, nsACString& aDest,
michael@0:                         const mozilla::fallible_t& ) NS_WARN_UNUSED_RESULT;
michael@0: void AppendUTF8toUTF16( const nsACString& aSource, nsAString& aDest );
michael@0: bool AppendUTF8toUTF16( const nsACString& aSource, nsAString& aDest,
michael@0:                         const mozilla::fallible_t& ) NS_WARN_UNUSED_RESULT;
michael@0: 
michael@0: void AppendUTF16toUTF8( const char16_t* aSource, nsACString& aDest );
michael@0: void AppendUTF8toUTF16( const char* aSource, nsAString& aDest );
michael@0: 
michael@0: #ifdef MOZ_USE_CHAR16_WRAPPER
michael@0: inline void AppendUTF16toUTF8( char16ptr_t aSource, nsACString& aDest )
michael@0:   {
michael@0:     return AppendUTF16toUTF8(static_cast<const char16_t*>(aSource), aDest);
michael@0:   }
michael@0: #endif
michael@0: 
michael@0:   /**
michael@0:    * Returns a new |char| buffer containing a zero-terminated copy of |aSource|.
michael@0:    *
michael@0:    * Allocates and returns a new |char| buffer which you must free with |nsMemory::Free|.
michael@0:    * Performs a lossy encoding conversion by chopping 16-bit wide characters down to 8-bits wide while copying |aSource| to your new buffer.
michael@0:    * This conversion is not well defined; but it reproduces legacy string behavior.
michael@0:    * The new buffer is zero-terminated, but that may not help you if |aSource| contains embedded nulls.
michael@0:    *
michael@0:    * @param aSource a 16-bit wide string
michael@0:    * @return a new |char| buffer you must free with |nsMemory::Free|.
michael@0:    */
michael@0: char* ToNewCString( const nsAString& aSource );
michael@0: 
michael@0: 
michael@0:   /**
michael@0:    * Returns a new |char| buffer containing a zero-terminated copy of |aSource|.
michael@0:    *
michael@0:    * Allocates and returns a new |char| buffer which you must free with |nsMemory::Free|.
michael@0:    * The new buffer is zero-terminated, but that may not help you if |aSource| contains embedded nulls.
michael@0:    *
michael@0:    * @param aSource an 8-bit wide string
michael@0:    * @return a new |char| buffer you must free with |nsMemory::Free|.
michael@0:    */
michael@0: char* ToNewCString( const nsACString& aSource );
michael@0: 
michael@0:   /**
michael@0:    * Returns a new |char| buffer containing a zero-terminated copy of |aSource|.
michael@0:    *
michael@0:    * Allocates and returns a new |char| buffer which you must free with 
michael@0:    * |nsMemory::Free|.
michael@0:    * Performs an encoding conversion from a UTF-16 string to a UTF-8 string
michael@0:    * copying |aSource| to your new buffer.
michael@0:    * The new buffer is zero-terminated, but that may not help you if |aSource| 
michael@0:    * contains embedded nulls.
michael@0:    *
michael@0:    * @param aSource a UTF-16 string (made of char16_t's)
michael@0:    * @param aUTF8Count the number of 8-bit units that was returned
michael@0:    * @return a new |char| buffer you must free with |nsMemory::Free|.
michael@0:    */
michael@0: 
michael@0: char* ToNewUTF8String( const nsAString& aSource, uint32_t *aUTF8Count = nullptr );
michael@0: 
michael@0: 
michael@0:   /**
michael@0:    * Returns a new |char16_t| buffer containing a zero-terminated copy of 
michael@0:    * |aSource|.
michael@0:    *
michael@0:    * Allocates and returns a new |char16_t| buffer which you must free with 
michael@0:    * |nsMemory::Free|.
michael@0:    * The new buffer is zero-terminated, but that may not help you if |aSource| 
michael@0:    * contains embedded nulls.
michael@0:    *
michael@0:    * @param aSource a UTF-16 string
michael@0:    * @return a new |char16_t| buffer you must free with |nsMemory::Free|.
michael@0:    */
michael@0: char16_t* ToNewUnicode( const nsAString& aSource );
michael@0: 
michael@0: 
michael@0:   /**
michael@0:    * Returns a new |char16_t| buffer containing a zero-terminated copy of |aSource|.
michael@0:    *
michael@0:    * Allocates and returns a new |char16_t| buffer which you must free with |nsMemory::Free|.
michael@0:    * Performs an encoding conversion by 0-padding 8-bit wide characters up to 16-bits wide while copying |aSource| to your new buffer.
michael@0:    * This conversion is not well defined; but it reproduces legacy string behavior.
michael@0:    * The new buffer is zero-terminated, but that may not help you if |aSource| contains embedded nulls.
michael@0:    *
michael@0:    * @param aSource an 8-bit wide string (a C-string, NOT UTF-8)
michael@0:    * @return a new |char16_t| buffer you must free with |nsMemory::Free|.
michael@0:    */
michael@0: char16_t* ToNewUnicode( const nsACString& aSource );
michael@0: 
michael@0:   /**
michael@0:    * Returns the required length for a char16_t buffer holding
michael@0:    * a copy of aSource, using UTF-8 to UTF-16 conversion.
michael@0:    * The length does NOT include any space for zero-termination.
michael@0:    *
michael@0:    * @param aSource an 8-bit wide string, UTF-8 encoded
michael@0:    * @return length of UTF-16 encoded string copy, not zero-terminated
michael@0:    */
michael@0: uint32_t CalcUTF8ToUnicodeLength( const nsACString& aSource );
michael@0: 
michael@0:   /**
michael@0:    * Copies the source string into the specified buffer, converting UTF-8 to
michael@0:    * UTF-16 in the process. The conversion is well defined for valid UTF-8
michael@0:    * strings.
michael@0:    * The copied string will be zero-terminated! Any embedded nulls will be
michael@0:    * copied nonetheless. It is the caller's responsiblity to ensure the buffer
michael@0:    * is large enough to hold the string copy plus one char16_t for
michael@0:    * zero-termination!
michael@0:    *
michael@0:    * @see CalcUTF8ToUnicodeLength( const nsACString& )
michael@0:    * @see UTF8ToNewUnicode( const nsACString&, uint32_t* )
michael@0:    *
michael@0:    * @param aSource an 8-bit wide string, UTF-8 encoded
michael@0:    * @param aBuffer the buffer holding the converted string copy
michael@0:    * @param aUTF16Count receiving optionally the number of 16-bit units that
michael@0:    *                    were copied
michael@0:    * @return aBuffer pointer, for convenience 
michael@0:    */
michael@0: char16_t* UTF8ToUnicodeBuffer( const nsACString& aSource,
michael@0:                                 char16_t *aBuffer,
michael@0:                                 uint32_t *aUTF16Count = nullptr );
michael@0: 
michael@0:   /**
michael@0:    * Returns a new |char16_t| buffer containing a zero-terminated copy
michael@0:    * of |aSource|.
michael@0:    *
michael@0:    * Allocates and returns a new |char| buffer which you must free with
michael@0:    * |nsMemory::Free|.  Performs an encoding conversion from UTF-8 to UTF-16 
michael@0:    * while copying |aSource| to your new buffer.  This conversion is well defined
michael@0:    * for a valid UTF-8 string.  The new buffer is zero-terminated, but that 
michael@0:    * may not help you if |aSource| contains embedded nulls.
michael@0:    *
michael@0:    * @param aSource an 8-bit wide string, UTF-8 encoded
michael@0:    * @param aUTF16Count the number of 16-bit units that was returned
michael@0:    * @return a new |char16_t| buffer you must free with |nsMemory::Free|.
michael@0:    *         (UTF-16 encoded)
michael@0:    */
michael@0: char16_t* UTF8ToNewUnicode( const nsACString& aSource, uint32_t *aUTF16Count = nullptr );
michael@0: 
michael@0:   /**
michael@0:    * Copies |aLength| 16-bit code units from the start of |aSource| to the
michael@0:    * |char16_t| buffer |aDest|.
michael@0:    *
michael@0:    * After this operation |aDest| is not null terminated.
michael@0:    *
michael@0:    * @param aSource a UTF-16 string
michael@0:    * @param aSrcOffset start offset in the source string
michael@0:    * @param aDest a |char16_t| buffer
michael@0:    * @param aLength the number of 16-bit code units to copy
michael@0:    * @return pointer to destination buffer - identical to |aDest|
michael@0:    */
michael@0: char16_t* CopyUnicodeTo( const nsAString& aSource,
michael@0:                                  uint32_t aSrcOffset,
michael@0:                                  char16_t* aDest,
michael@0:                                  uint32_t aLength );
michael@0: 
michael@0: 
michael@0:   /**
michael@0:    * Copies 16-bit characters between iterators |aSrcStart| and
michael@0:    * |aSrcEnd| to the writable string |aDest|. Similar to the
michael@0:    * |nsString::Mid| method.
michael@0:    *
michael@0:    * After this operation |aDest| is not null terminated.
michael@0:    *
michael@0:    * @param aSrcStart start source iterator
michael@0:    * @param aSrcEnd end source iterator
michael@0:    * @param aDest destination for the copy
michael@0:    */
michael@0: void CopyUnicodeTo( const nsAString::const_iterator& aSrcStart,
michael@0:                            const nsAString::const_iterator& aSrcEnd,
michael@0:                            nsAString& aDest );
michael@0: 
michael@0:   /**
michael@0:    * Appends 16-bit characters between iterators |aSrcStart| and
michael@0:    * |aSrcEnd| to the writable string |aDest|. 
michael@0:    *
michael@0:    * After this operation |aDest| is not null terminated.
michael@0:    *
michael@0:    * @param aSrcStart start source iterator
michael@0:    * @param aSrcEnd end source iterator
michael@0:    * @param aDest destination for the copy
michael@0:    */
michael@0: void AppendUnicodeTo( const nsAString::const_iterator& aSrcStart,
michael@0:                              const nsAString::const_iterator& aSrcEnd,
michael@0:                              nsAString& aDest );
michael@0: 
michael@0:   /**
michael@0:    * Returns |true| if |aString| contains only ASCII characters, that is, characters in the range (0x00, 0x7F).
michael@0:    *
michael@0:    * @param aString a 16-bit wide string to scan
michael@0:    */
michael@0: bool IsASCII( const nsAString& aString );
michael@0: 
michael@0:   /**
michael@0:    * Returns |true| if |aString| contains only ASCII characters, that is, characters in the range (0x00, 0x7F).
michael@0:    *
michael@0:    * @param aString a 8-bit wide string to scan
michael@0:    */
michael@0: bool IsASCII( const nsACString& aString );
michael@0: 
michael@0:   /**
michael@0:    * Returns |true| if |aString| is a valid UTF-8 string.
michael@0:    * XXX This is not bullet-proof and nor an all-purpose UTF-8 validator. 
michael@0:    * It is mainly written to replace and roughly equivalent to
michael@0:    *
michael@0:    *    str.Equals(NS_ConvertUTF16toUTF8(NS_ConvertUTF8toUTF16(str)))
michael@0:    *
michael@0:    * (see bug 191541)
michael@0:    * As such,  it does not check for non-UTF-8 7bit encodings such as 
michael@0:    * ISO-2022-JP and HZ. 
michael@0:    *
michael@0:    * It rejects sequences with the following errors:
michael@0:    *
michael@0:    * byte sequences that cannot be decoded into characters according to
michael@0:    *   UTF-8's rules (including cases where the input is part of a valid
michael@0:    *   UTF-8 sequence but starts or ends mid-character)
michael@0:    * overlong sequences (i.e., cases where a character was encoded
michael@0:    *   non-canonically by using more bytes than necessary)
michael@0:    * surrogate codepoints (i.e., the codepoints reserved for
michael@0:        representing astral characters in UTF-16)
michael@0:    * codepoints above the unicode range (i.e., outside the first 17
michael@0:    *   planes; higher than U+10FFFF), in accordance with
michael@0:    *   http://tools.ietf.org/html/rfc3629
michael@0:    * when aRejectNonChar is true (the default), any codepoint whose low
michael@0:    *   16 bits are 0xFFFE or 0xFFFF
michael@0: 
michael@0:    *
michael@0:    * @param aString an 8-bit wide string to scan
michael@0:    * @param aRejectNonChar a boolean to control the rejection of utf-8
michael@0:    *        non characters
michael@0:    */
michael@0: bool IsUTF8( const nsACString& aString, bool aRejectNonChar = true );
michael@0: 
michael@0: bool ParseString(const nsACString& aAstring, char aDelimiter, 
michael@0:                           nsTArray<nsCString>& aArray);
michael@0: 
michael@0:   /**
michael@0:    * Converts case in place in the argument string.
michael@0:    */
michael@0: void ToUpperCase( nsACString& );
michael@0: 
michael@0: void ToLowerCase( nsACString& );
michael@0: 
michael@0: void ToUpperCase( nsCSubstring& );
michael@0: 
michael@0: void ToLowerCase( nsCSubstring& );
michael@0: 
michael@0:   /**
michael@0:    * Converts case from string aSource to aDest.
michael@0:    */
michael@0: void ToUpperCase( const nsACString& aSource, nsACString& aDest );
michael@0: 
michael@0: void ToLowerCase( const nsACString& aSource, nsACString& aDest );
michael@0: 
michael@0:   /**
michael@0:    * Finds the leftmost occurrence of |aPattern|, if any in the range |aSearchStart|..|aSearchEnd|.
michael@0:    *
michael@0:    * Returns |true| if a match was found, and adjusts |aSearchStart| and |aSearchEnd| to
michael@0:    * point to the match.  If no match was found, returns |false| and makes |aSearchStart == aSearchEnd|.
michael@0:    *
michael@0:    * Currently, this is equivalent to the O(m*n) implementation previously on |ns[C]String|.
michael@0:    * If we need something faster, then we can implement that later.
michael@0:    */
michael@0: 
michael@0: bool FindInReadable( const nsAString& aPattern, nsAString::const_iterator&, nsAString::const_iterator&, const nsStringComparator& = nsDefaultStringComparator() );
michael@0: bool FindInReadable( const nsACString& aPattern, nsACString::const_iterator&, nsACString::const_iterator&, const nsCStringComparator& = nsDefaultCStringComparator() );
michael@0: 
michael@0: /* sometimes we don't care about where the string was, just that we
michael@0:  * found it or not */
michael@0: inline bool FindInReadable( const nsAString& aPattern, const nsAString& aSource, const nsStringComparator& compare = nsDefaultStringComparator() )
michael@0: {
michael@0:   nsAString::const_iterator start, end;
michael@0:   aSource.BeginReading(start);
michael@0:   aSource.EndReading(end);
michael@0:   return FindInReadable(aPattern, start, end, compare);
michael@0: }
michael@0: 
michael@0: inline bool FindInReadable( const nsACString& aPattern, const nsACString& aSource, const nsCStringComparator& compare = nsDefaultCStringComparator() )
michael@0: {
michael@0:   nsACString::const_iterator start, end;
michael@0:   aSource.BeginReading(start);
michael@0:   aSource.EndReading(end);
michael@0:   return FindInReadable(aPattern, start, end, compare);
michael@0: }
michael@0: 
michael@0: 
michael@0: bool CaseInsensitiveFindInReadable( const nsACString& aPattern, nsACString::const_iterator&, nsACString::const_iterator& );
michael@0: 
michael@0:   /**
michael@0:    * Finds the rightmost occurrence of |aPattern| 
michael@0:    * Returns |true| if a match was found, and adjusts |aSearchStart| and |aSearchEnd| to
michael@0:    * point to the match.  If no match was found, returns |false| and makes |aSearchStart == aSearchEnd|.
michael@0:    *
michael@0:    */
michael@0: bool RFindInReadable( const nsAString& aPattern, nsAString::const_iterator&, nsAString::const_iterator&, const nsStringComparator& = nsDefaultStringComparator() );
michael@0: bool RFindInReadable( const nsACString& aPattern, nsACString::const_iterator&, nsACString::const_iterator&, const nsCStringComparator& = nsDefaultCStringComparator() );
michael@0: 
michael@0:    /**
michael@0:    * Finds the leftmost occurrence of |aChar|, if any in the range 
michael@0:    * |aSearchStart|..|aSearchEnd|.
michael@0:    *
michael@0:    * Returns |true| if a match was found, and adjusts |aSearchStart| to
michael@0:    * point to the match.  If no match was found, returns |false| and 
michael@0:    * makes |aSearchStart == aSearchEnd|.
michael@0:    */
michael@0: bool FindCharInReadable( char16_t aChar, nsAString::const_iterator& aSearchStart, const nsAString::const_iterator& aSearchEnd );
michael@0: bool FindCharInReadable( char aChar, nsACString::const_iterator& aSearchStart, const nsACString::const_iterator& aSearchEnd );
michael@0: 
michael@0:     /**
michael@0:     * Finds the number of occurences of |aChar| in the string |aStr|
michael@0:     */
michael@0: uint32_t CountCharInReadable( const nsAString& aStr,
michael@0:                                      char16_t aChar );
michael@0: uint32_t CountCharInReadable( const nsACString& aStr,
michael@0:                                      char aChar );
michael@0: 
michael@0: bool
michael@0: StringBeginsWith( const nsAString& aSource, const nsAString& aSubstring,
michael@0:                   const nsStringComparator& aComparator =
michael@0:                                               nsDefaultStringComparator() );
michael@0: bool
michael@0: StringBeginsWith( const nsACString& aSource, const nsACString& aSubstring,
michael@0:                   const nsCStringComparator& aComparator =
michael@0:                                                nsDefaultCStringComparator() );
michael@0: bool
michael@0: StringEndsWith( const nsAString& aSource, const nsAString& aSubstring,
michael@0:                 const nsStringComparator& aComparator =
michael@0:                                             nsDefaultStringComparator() );
michael@0: bool
michael@0: StringEndsWith( const nsACString& aSource, const nsACString& aSubstring,
michael@0:                 const nsCStringComparator& aComparator =
michael@0:                                              nsDefaultCStringComparator() );
michael@0: 
michael@0: const nsAFlatString& EmptyString();
michael@0: const nsAFlatCString& EmptyCString();
michael@0: 
michael@0: const nsAFlatString& NullString();
michael@0: const nsAFlatCString& NullCString();
michael@0: 
michael@0:    /**
michael@0:    * Compare a UTF-8 string to an UTF-16 string.
michael@0:    *
michael@0:    * Returns 0 if the strings are equal, -1 if aUTF8String is less
michael@0:    * than aUTF16Count, and 1 in the reverse case.  In case of fatal
michael@0:    * error (eg the strings are not valid UTF8 and UTF16 respectively),
michael@0:    * this method will return INT32_MIN.
michael@0:    */
michael@0: int32_t
michael@0: CompareUTF8toUTF16(const nsASingleFragmentCString& aUTF8String,
michael@0:                    const nsASingleFragmentString& aUTF16String);
michael@0: 
michael@0: void
michael@0: AppendUCS4ToUTF16(const uint32_t aSource, nsAString& aDest);
michael@0: 
michael@0: template<class T>
michael@0: inline bool EnsureStringLength(T& aStr, uint32_t aLen)
michael@0: {
michael@0:     aStr.SetLength(aLen);
michael@0:     return (aStr.Length() == aLen);
michael@0: }
michael@0: 
michael@0: #endif // !defined(nsReadableUtils_h___)