Automatic generation produced by ISE Eiffel
note description: "[ Converter from/to UTF-8, UTF-16 and UTF-32 encodings. Handling of invalid encodings ============================= Whenever a UTF-8 or UTF-16 sequence is decoded, the decoding routines also check that the sequence is valid. If it is not, it will replace the invalid unit (e.g. a byte for UTF-8 and a 2-byte for UTF-16 by the replacement character U+FFFD as described by variant #3 of the recommended practice for replacement character in Unicode (see http://www.unicode.org/review/pr-121.html for more details). However it means that you cannot roundtrip incorrectly encoded sequence back and forth between the encoded version and the decoded STRING_32 version. To allow roundtrip, an escaped representation of a bad encoded sequence has been introduced. It is adding a a fourth variant (which is a slight modification of variant #3) to the recommended practice where the replacement character is followed by the printed hexadecimal value of the invalid byte or the invalid 2-byte sequence. To provide an example (assuming that the Unicode character U+FFFD is represented as ? textually): 1 - on UNIX, any invalid UTF-8 byte sequence such as 0x8F 0x8F is encoded as the following Unicode sequence: U+FFFD U+0038 U+0046 U+FFFF U+0038 U+0046, and textually it looks like "?8F?8F". 2 - on Windows, any invalid UTF-16 2-byte sequence such as 0xD800 0x0054 is encoded as the following Unicode sequence: U+FFFD U+0075 U+0044 U+0038 U+0030 U+0030 U+FFFD U+0035 U+0033, and textually it looks like "?uD800?54". The rule is that if the 2-byte sequence does not fit into 1 byte, it uses the letter u followed by the hexadecimal value of the 2-byte sequence, otherwise it simply uses the 1-byte hexadecimal representation. ]" date: "$Date: 2021-06-18 09:01:52 -0800 (Fri, 18 Jun 2021) $" revision: "$Revision: 105548 $" expanded class interface UTF_CONVERTER create default_create -- Process instances of classes with no creation clause. -- (Default: do nothing.) -- (from ANY) feature -- Access Escape_character: CHARACTER_32 = '�' -- Unicode replacement character to escape invalid UTF-8 or UTF-16 encoding. -- UTF-8 encoding: 0xEF 0xBF 0xBD -- Binary UTF-8 encoding: 11101111 10111111 10111101 -- UTF-16 encoding: 0xFFFD generating_type: TYPE [detachable UTF_CONVERTER] -- Type of current object -- (type of which it is a direct instance) -- (from ANY) ensure -- from ANY generating_type_not_void: Result /= Void generator: STRING_8 -- Name of current object's generating class -- (base class of the type of which it is a direct instance) -- (from ANY) ensure -- from ANY generator_not_void: Result /= Void generator_not_empty: not Result.is_empty feature -- Measurement utf_16_bytes_count (s: READABLE_STRING_GENERAL; start_pos, end_pos: INTEGER_32): INTEGER_32 -- Number of bytes necessary at the very least to encode in UTF-16 s.substring (start_pos, end_pos). -- Note that this feature can be used for both escaped and non-escaped string. -- In the case of escaped strings, the result will be possibly higher than really needed. -- It does not include the terminating null character. require start_position_big_enough: start_pos >= 1 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos <= s.count ensure instance_free: class utf_16_characters_count_form_pointer (m: MANAGED_POINTER; start_pos, end_pos: INTEGER_32): INTEGER_32 -- Number of characters of the UTF-16 encoded m starting at start_pos in m up to end_pos - 1. -- It does not include the terminating null character. require start_position_big_enough: start_pos >= 0 end_position: start_pos <= end_pos + 2 end_pos_small_enought: end_pos < m.count even_start_position: start_pos \\ 2 = 0 even_end_position: end_pos \\ 2 = 0 ensure instance_free: class utf_8_bytes_count (s: READABLE_STRING_GENERAL; start_pos, end_pos: INTEGER_32): INTEGER_32 -- Number of bytes necessary to encode in UTF-8 s.substring (start_pos, end_pos). -- Note that this feature can be used for both escaped and non-escaped string. -- In the case of escaped strings, the result will be possibly higher than really needed. -- It does not include the terminating null character. require start_position_big_enough: start_pos >= 1 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos <= s.count ensure instance_free: class utf_8_to_string_32_count (s: SPECIAL [CHARACTER_8]; start_pos, end_pos: INTEGER_32): INTEGER_32 -- Count of characters corresponding to UTF-8 sequence s. require start_position_big_enough: start_pos >= 0 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos < s.count ensure instance_free: class feature -- Comparison frozen deep_equal (a: detachable ANY; b: like arg #1): BOOLEAN -- Are a and b either both void -- or attached to isomorphic object structures? -- (from ANY) ensure -- from ANY instance_free: class shallow_implies_deep: standard_equal (a, b) implies Result both_or_none_void: (a = Void) implies (Result = (b = Void)) same_type: (Result and (a /= Void)) implies (b /= Void and then a.same_type (b)) symmetric: Result implies deep_equal (b, a) frozen equal (a: detachable ANY; b: like arg #1): BOOLEAN -- Are a and b either both void or attached -- to objects considered equal? -- (from ANY) ensure -- from ANY instance_free: class definition: Result = (a = Void and b = Void) or else ((a /= Void and b /= Void) and then a.is_equal (b)) frozen is_deep_equal alias "≡≡≡" (other: UTF_CONVERTER): BOOLEAN -- Are Current and other attached to isomorphic object structures? -- (from ANY) require -- from ANY other_not_void: other /= Void ensure -- from ANY shallow_implies_deep: standard_is_equal (other) implies Result same_type: Result implies same_type (other) symmetric: Result implies other.is_deep_equal (Current) is_equal (other: UTF_CONVERTER): BOOLEAN -- Is other attached to an object considered -- equal to current object? -- (from ANY) require -- from ANY other_not_void: other /= Void ensure -- from ANY symmetric: Result implies other ~ Current consistent: standard_is_equal (other) implies Result frozen standard_equal (a: detachable ANY; b: like arg #1): BOOLEAN -- Are a and b either both void or attached to -- field-by-field identical objects of the same type? -- Always uses default object comparison criterion. -- (from ANY) ensure -- from ANY instance_free: class definition: Result = (a = Void and b = Void) or else ((a /= Void and b /= Void) and then a.standard_is_equal (b)) frozen standard_is_equal alias "≜" (other: UTF_CONVERTER): BOOLEAN -- Is other attached to an object of the same type -- as current object, and field-by-field identical to it? -- (from ANY) require -- from ANY other_not_void: other /= Void ensure -- from ANY same_type: Result implies same_type (other) symmetric: Result implies other.standard_is_equal (Current) feature -- Status report conforms_to (other: ANY): BOOLEAN -- Does type of current object conform to type -- of other (as per Eiffel: The Language, chapter 13)? -- (from ANY) require -- from ANY other_not_void: other /= Void is_valid_utf_16 (s: SPECIAL [NATURAL_16]): BOOLEAN -- Is s a valid UTF-16 Unicode sequence? ensure instance_free: class is_valid_utf_16_subpointer (p: MANAGED_POINTER; start_pos, end_pos: INTEGER_32; a_stop_at_null: BOOLEAN): BOOLEAN -- Is p a valid UTF-16 Unicode sequence between code unit start_pos and end_pos? -- If a_stop_at_null we stop checking after finding a null character. ensure instance_free: class is_valid_utf_16le_string_8 (s: READABLE_STRING_8): BOOLEAN -- Is s a valid UTF-16LE Unicode sequence? ensure instance_free: class is_valid_utf_8_string_8 (s: READABLE_STRING_8): BOOLEAN -- Is s a valid UTF-8 Unicode sequence? ensure instance_free: class same_type (other: ANY): BOOLEAN -- Is type of current object identical to type of other? -- (from ANY) require -- from ANY other_not_void: other /= Void ensure -- from ANY definition: Result = (conforms_to (other) and other.conforms_to (Current)) feature -- Duplication copy (other: UTF_CONVERTER) -- Update current object using fields of object attached -- to other, so as to yield equal objects. -- (from ANY) require -- from ANY other_not_void: other /= Void type_identity: same_type (other) ensure -- from ANY is_equal: Current ~ other frozen deep_copy (other: UTF_CONVERTER) -- Effect equivalent to that of: -- `copy` (other . `deep_twin`) -- (from ANY) require -- from ANY other_not_void: other /= Void ensure -- from ANY deep_equal: deep_equal (Current, other) frozen deep_twin: UTF_CONVERTER -- New object structure recursively duplicated from Current. -- (from ANY) ensure -- from ANY deep_twin_not_void: Result /= Void deep_equal: deep_equal (Current, Result) frozen standard_copy (other: UTF_CONVERTER) -- Copy every field of other onto corresponding field -- of current object. -- (from ANY) require -- from ANY other_not_void: other /= Void type_identity: same_type (other) ensure -- from ANY is_standard_equal: standard_is_equal (other) frozen standard_twin: UTF_CONVERTER -- New object field-by-field identical to other. -- Always uses default copying semantics. -- (from ANY) ensure -- from ANY standard_twin_not_void: Result /= Void equal: standard_equal (Result, Current) frozen twin: UTF_CONVERTER -- New object equal to Current -- `twin` calls `copy`; to change copying/twinning semantics, redefine `copy`. -- (from ANY) ensure -- from ANY twin_not_void: Result /= Void is_equal: Result ~ Current feature -- Basic operations frozen default: detachable UTF_CONVERTER -- Default value of object's type -- (from ANY) frozen default_pointer: POINTER -- Default value of type POINTER -- (Avoid the need to write p.`default` for -- some p of type POINTER.) -- (from ANY) ensure -- from ANY instance_free: class default_rescue -- Process exception for routines with no Rescue clause. -- (Default: do nothing.) -- (from ANY) frozen do_nothing -- Execute a null action. -- (from ANY) ensure -- from ANY instance_free: class feature -- Byte Order Mark (BOM) Utf_16be_bom_to_string_8: STRING_8 = "þÿ" -- UTF-16BE BOM sequence. Utf_16le_bom_to_string_8: STRING_8 = "ÿþ" -- UTF-16LE BOM sequence. Utf_32be_bom_to_string_8: STRING_8 = "%U%Uþÿ" -- UTF-32BE BOM sequence. Utf_32le_bom_to_string_8: STRING_8 = "ÿþ%U%U" -- UTF-32LE BOM sequence. Utf_8_bom_to_string_8: STRING_8 = "" -- UTF-8 BOM sequence. feature -- Output Io: STD_FILES -- Handle to standard file setup -- (from ANY) ensure -- from ANY instance_free: class io_not_void: Result /= Void out: STRING_8 -- New string containing terse printable representation -- of current object -- (from ANY) ensure -- from ANY out_not_void: Result /= Void print (o: detachable ANY) -- Write terse external representation of o -- on standard output. -- (from ANY) ensure -- from ANY instance_free: class frozen tagged_out: STRING_8 -- New string containing terse printable representation -- of current object -- (from ANY) ensure -- from ANY tagged_out_not_void: Result /= Void feature -- Platform Operating_environment: OPERATING_ENVIRONMENT -- Objects available from the operating system -- (from ANY) ensure -- from ANY instance_free: class operating_environment_not_void: Result /= Void feature -- UTF-16 to UTF-32 utf_16_0_pointer_into_escaped_string_32 (p: MANAGED_POINTER; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-16 sequence p which is zero-terminated, -- where invalid UTF-16LE sequences are escaped, appended into a_result. require minimum_size: p.count >= 2 valid_count: p.count \\ 2 = 0 ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_16le_string_8 (a_result.substring (old a_result.count + 1, a_result.count)) as l_utf and then ∀ c: l_utf.new_cursor.incremented (1) ¦ (c.natural_32_code | (l_utf.code (@c.target_index + 1) |<< 8)) = -- Unable to show all postconditions! utf_16_0_pointer_into_string_32 (p: MANAGED_POINTER; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-16 sequence p which is zero-terminated -- appended into a_result. require minimum_size: p.count >= 2 valid_count: p.count \\ 2 = 0 ensure instance_free: class roundtrip: is_valid_utf_16_subpointer (p, 0, p.count // 2, True) implies ∀ n: string_32_to_utf_16 (a_result.substring (old a_result.count + 1, a_result.count)) ¦ n = p.read_natural_16 (@n.target_index * 2) utf_16_0_pointer_to_escaped_string_32 (p: MANAGED_POINTER): STRING_32 -- {STRING_32} object corresponding to UTF-16 sequence p which is zero-terminated, -- where invalid UTF-16LE sequences are escaped. require minimum_size: p.count >= 2 valid_count: p.count \\ 2 = 0 ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_16le_string_8 (Result) as l_utf and then ∀ c: l_utf.new_cursor.incremented (1) ¦ (c.natural_32_code | (l_utf.code (@c.target_index + 1) |<< 8)) = -- Unable to show all postconditions! utf_16_0_pointer_to_string_32 (p: MANAGED_POINTER): STRING_32 -- {STRING_32} object corresponding to UTF-16 sequence p which is zero-terminated. require minimum_size: p.count >= 2 valid_count: p.count \\ 2 = 0 ensure instance_free: class roundtrip: is_valid_utf_16_subpointer (p, 0, p.count // 2, True) implies ∀ n: string_32_to_utf_16 (Result) ¦ n = p.read_natural_16 ((@n.target_index + 1) * 2) utf_16_0_subpointer_into_escaped_string_32 (p: MANAGED_POINTER; start_pos, end_pos: INTEGER_32; a_stop_at_null: BOOLEAN; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-16 sequence p between code units start_pos and -- end_pos or the first null character encountered if a_stop_at_null, where invalid -- UTF-16LE sequences are escaped, appended into a_result. require minimum_size: p.count >= 2 start_position_big_enough: start_pos >= 0 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos < p.count // 2 ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_16le_string_8 (a_result.substring (old a_result.count + 1, a_result.count)) as l_utf and then ∀ c: l_utf.new_cursor.incremented (1) ¦ (c.natural_32_code | (l_utf.code (@c.target_index + 1) |<< 8)) = -- Unable to show all postconditions! utf_16_0_subpointer_into_string_32 (p: MANAGED_POINTER; start_pos, end_pos: INTEGER_32; a_stop_at_null: BOOLEAN; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-16 sequence p between code units start_pos and -- end_pos or the first null character encountered if a_stop_at_null appended into a_result. require minimum_size: p.count >= 2 start_position_big_enough: start_pos >= 0 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos < p.count // 2 ensure instance_free: class roundtrip: is_valid_utf_16_subpointer (p, start_pos, end_pos, a_stop_at_null) implies ∀ x: string_32_to_utf_16 (a_result.substring (old a_result.count + 1, a_result.count)) ¦ x = p.read_natural_16 (@x.target_index * 2) utf_16_0_subpointer_to_escaped_string_32 (p: MANAGED_POINTER; start_pos, end_pos: INTEGER_32; a_stop_at_null: BOOLEAN): STRING_32 -- {STRING_32} object corresponding to UTF-16 sequence p between code units start_pos and -- end_pos or the first null character encountered if a_stop_at_null, where invalid -- UTF-16LE sequences are escaped. require minimum_size: p.count >= 2 start_position_big_enough: start_pos >= 0 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos < p.count // 2 ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_16le_string_8 (Result) as l_utf and then ∀ c: l_utf.new_cursor.incremented (1) ¦ (c.natural_32_code | (l_utf.code (@c.target_index + 1) |<< 8)) = -- Unable to show all postconditions! utf_16_0_subpointer_to_string_32 (p: MANAGED_POINTER; start_pos, end_pos: INTEGER_32; a_stop_at_null: BOOLEAN): STRING_32 -- {STRING_32} object corresponding to UTF-16 sequence p between code units start_pos and -- end_pos or the first null character encountered if a_stop_at_null. require minimum_size: p.count >= 2 start_position_big_enough: start_pos >= 0 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos < p.count // 2 ensure instance_free: class roundtrip: is_valid_utf_16_subpointer (p, start_pos, end_pos, a_stop_at_null) implies ∀ n: string_32_to_utf_16 (Result) ¦ n = p.read_natural_16 (@n.target_index * 2) utf_16_into_string_32 (s: SPECIAL [NATURAL_16]; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-16 sequence s -- appended into a_result. ensure instance_free: class roundtrip: is_valid_utf_16 (s) implies string_32_to_utf_16 (a_result.substring (old a_result.count + 1, a_result.count)).is_equal (s) utf_16_to_string_32 (s: SPECIAL [NATURAL_16]): STRING_32 -- {STRING_32} object corresponding to UTF-16 sequence s. ensure instance_free: class roundtrip: is_valid_utf_16 (s) implies string_32_to_utf_16 (Result).is_equal (s) utf_16le_string_8_into_escaped_string_32 (s: READABLE_STRING_8; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-16LE sequence s, where invalid UTF-16LE -- sequences are escaped, appended into a_result. ensure instance_free: class roundtrip: escaped_utf_32_string_to_utf_16le_string_8 (a_result.substring (old a_result.count + 1, a_result.count)).same_string (s) utf_16le_string_8_into_string_32 (s: READABLE_STRING_8; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-16LE sequence s appended into a_result. ensure instance_free: class roundtrip: is_valid_utf_16le_string_8 (s) implies escaped_utf_32_string_to_utf_16le_string_8 (a_result.substring (old a_result.count + 1, a_result.count)).same_string (s) utf_16le_string_8_to_escaped_string_32 (s: READABLE_STRING_8): STRING_32 -- {STRING_32} object corresponding to UTF-16LE sequence s, where invalid UTF-16LE -- sequences are escaped. ensure instance_free: class roundtrip: escaped_utf_32_string_to_utf_16le_string_8 (Result).same_string (s) utf_16le_string_8_to_string_32 (s: READABLE_STRING_8): STRING_32 -- {STRING_32} object corresponding to UTF-16LE sequence s. ensure instance_free: class roundtrip: is_valid_utf_16le_string_8 (s) implies escaped_utf_32_string_to_utf_16le_string_8 (Result).same_string (s) feature -- UTF-16 to UTF-8 utf_16_into_utf_8_string_8 (s: SPECIAL [NATURAL_16]; a_result: STRING_8) -- Copy UTF-8 sequence corresponding to UTF-16 sequence s appended into a_result. ensure instance_free: class roundtrip: is_valid_utf_16 (s) implies string_32_to_utf_16 (utf_8_string_8_to_string_32 (a_result.substring (old a_result.count + 1, a_result.count))).is_equal (s) utf_16_to_utf_8_string_8 (s: SPECIAL [NATURAL_16]): STRING_8 -- UTF-8 sequence corresponding to UTF-16 sequence s. ensure instance_free: class roundtrip: is_valid_utf_16 (s) implies string_32_to_utf_16 (utf_8_string_8_to_string_32 (Result)).is_equal (s) utf_16le_string_8_into_utf_8_string_8 (s: READABLE_STRING_8; a_result: STRING_8) -- Copy UTF-8 sequence corresponding to UTF-16LE sequence s appended into a_result. require even_count: (s.count & 1) = 0 ensure instance_free: class roundtrip: is_valid_utf_16le_string_8 (s) implies utf_32_string_to_utf_16le_string_8 (utf_8_string_8_to_string_32 (a_result.substring (old a_result.count + 1, a_result.count))).same_string (s) utf_16le_string_8_to_utf_8_string_8 (s: READABLE_STRING_8): STRING_8 -- UTF-8 sequence corresponding to UTF-16LE sequence s. ensure instance_free: class roundtrip: is_valid_utf_16le_string_8 (s) implies utf_32_string_to_utf_16le_string_8 (utf_8_string_8_to_string_32 (Result)).same_string (s) feature -- UTF-32 to UTF-16 escaped_utf_32_string_into_utf_16le_string_8 (s: READABLE_STRING_GENERAL; a_result: STRING_8) -- Copy UTF-16LE sequence corresponding to s interpreted as a UTF-32 sequence that could be -- escaped appended into a_result. -- If s contains the `escape_character` followed by either "HH" or "uHHHH" where H stands -- for an hexadecimal digit, then s has been escaped and will be converted to what is -- expected by the current platform. -- Otherwise it will be ignored and it will be left as is. -- See the note clause for the class for more details on the encoding. ensure instance_free: class roundtrip: utf_16le_string_8_to_escaped_string_32 (a_result.substring (old a_result.count + 1, a_result.count)).same_string_general (s) escaped_utf_32_string_to_utf_16le_string_8 (s: READABLE_STRING_GENERAL): STRING_8 -- UTF-16LE sequence corresponding to s interpreted as a UTF-32 sequence that could be escaped. -- If s contains the `escape_character` followed by either "HH" or "uHHHH" where H stands -- for an hexadecimal digit, then s has been escaped and will be converted to what is -- expected by the current platform. -- Otherwise it will be ignored and it will be left as is. -- See the note clause for the class for more details on the encoding. ensure instance_free: class roundtrip: utf_16le_string_8_to_escaped_string_32 (Result).same_string_general (s) escaped_utf_32_substring_into_utf_16_0_pointer (s: READABLE_STRING_GENERAL; start_pos, end_pos: like {READABLE_STRING_32}.count; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-16 sequence corresponding to the substring of s, -- interpreted as a UTF-32 sequence, starting at index start_pos -- and ending at index end_pos to address p + p_offset and update the -- size of p to the number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is not zero-terminated. require start_position_big_enough: start_pos >= 1 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos <= s.count even_p_offset: (p_offset \\ 2) = 0 p_offset_non_negative: p_offset >= 0 ensure instance_free: class p_count_may_increase: p.count >= old p.count roundtrip: a_new_upper /= Void implies utf_16_0_subpointer_to_escaped_string_32 (p, p_offset // 2, (a_new_upper.item // 2) - 1, False).same_string_general (s.substring (start_pos, end_pos)) roundtrip: (a_new_upper = Void and then not s.substring (start_pos, end_pos).has ('%U'.to_character_32)) implies utf_16_0_subpointer_to_escaped_string_32 (p, p_offset // 2, (p.count // 2) - 1, True).same_string_general (s.substring (start_pos, end_pos)) string_32_into_utf_16_0_pointer (s: READABLE_STRING_32; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-16 sequence corresponding to s with terminating zero -- to address p + p_offset and update the size of p to the number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is zero-terminated. require even_p_offset: (p_offset \\ 2) = 0 p_offset_non_negative: p_offset >= 0 ensure instance_free: class roundtrip: a_new_upper /= Void implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (a_new_upper.item // 2) - 1, False).same_string (s) roundtrip: (a_new_upper = Void and then not s.has ('%U'.to_character_32)) implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (p.count // 2) - 1, True).same_string (s) string_32_into_utf_16_pointer (s: READABLE_STRING_32; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-16 sequence corresponding to s to address p + p_offset -- and update the size of p to the number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is not zero-terminated. require even_p_offset: (p_offset \\ 2) = 0 p_offset_non_negative: p_offset >= 0 ensure instance_free: class roundtrip: a_new_upper /= Void implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (a_new_upper.item // 2) - 1, False).same_string (s) roundtrip: (a_new_upper = Void and then not s.has ('%U'.to_character_32)) implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (p.count // 2) - 1, True).same_string (s) string_32_to_utf_16 (s: READABLE_STRING_32): SPECIAL [NATURAL_16] -- UTF-16 sequence corresponding to s. -- The sequence is not zero-terminated. ensure instance_free: class roundtrip: attached utf_32_string_to_utf_16le_string_8 (s) as l_ref and then ∀ n: Result ¦ n.to_natural_32 = (l_ref.code (@n.target_index * 2 + 1) | (l_ref.code ((.target_index + 1) * 2) |<< 16)) string_32_to_utf_16_0 (s: READABLE_STRING_32): SPECIAL [NATURAL_16] -- UTF-16 sequence corresponding to s with terminating zero. ensure instance_free: class roundtrip: attached utf_32_string_to_utf_16le_string_8 (s) as l_ref and then ∀ n: Result.resized_area_with_default (0, Result.count - 1) ¦ n.to_natural_32 = (l_ref.code (@n.target_index * 2 + 1) | ((l_ref.code ((.target_index + 1) * 2)) |<< 8)) utf_32_string_into_utf_16le_string_8 (s: READABLE_STRING_GENERAL; a_result: STRING_8) -- Copy UTF-16LE sequence corresponding to s interpreted as a UTF-32 sequence -- appended into a_result. ensure instance_free: class roundtrip: utf_16le_string_8_to_string_32 (a_result.substring (old a_result.count + 1, a_result.count)).same_string_general (s) utf_32_string_to_utf_16 (s: READABLE_STRING_GENERAL): SPECIAL [NATURAL_16] -- UTF-16 sequence corresponding to s interpreted as a UTF-32 sequence. -- The sequence is not zero-terminated. ensure instance_free: class roundtrip: attached utf_32_string_to_utf_16le_string_8 (s) as l_ref and then ∀ n: Result ¦ n.to_natural_32 = (l_ref.code (@n.target_index * 2 + 1) | (l_ref.code ((.target_index + 1) * 2) |<< 8)) utf_32_string_to_utf_16_0 (s: READABLE_STRING_GENERAL): SPECIAL [NATURAL_16] -- UTF-16 sequence corresponding to s, interpreted as a UTF-32 sequence, -- with terminating zero. ensure instance_free: class roundtrip: attached utf_32_string_to_utf_16le_string_8 (s) as l_ref and then ∀ x: Result.resized_area_with_default (0, Result.count - 1) ¦ x.to_natural_32 = (l_ref.code (@x.target_index * 2 + 1) | ((l_ref.code ((.target_index + 1) * 2)) |<< 8)) utf_32_string_to_utf_16le_string_8 (s: READABLE_STRING_GENERAL): STRING_8 -- UTF-16LE sequence corresponding to s interpreted as a UTF-32 sequence ensure instance_free: class roundtrip: utf_16le_string_8_to_string_32 (Result).same_string_general (s) utf_32_substring_into_utf_16_0_pointer (s: READABLE_STRING_GENERAL; start_pos, end_pos: like {READABLE_STRING_32}.count; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-16 sequence corresponding to the substring of s, -- interpreted as a UTF-32 sequence, starting at index start_pos -- and ending at index end_pos to address p + p_offset and update the -- size of p to the number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is zero-terminated. require start_position_big_enough: start_pos >= 1 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos <= s.count even_p_offset: (p_offset \\ 2) = 0 p_offset_non_negative: p_offset >= 0 ensure instance_free: class p_count_may_increase: p.count >= old p.count roundtrip: a_new_upper /= Void implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (a_new_upper.item // 2) - 1, False).same_string_general (s) roundtrip: (a_new_upper = Void and then not s.has ('%U'.to_character_32)) implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (p.count // 2) - 1, True).same_string_general (s) utf_32_substring_into_utf_16_pointer (s: READABLE_STRING_GENERAL; start_pos, end_pos: like {READABLE_STRING_32}.count; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-16 sequence corresponding to the substring of s, -- interpreted as a UTF-32 sequence, starting at index start_pos -- and ending at index end_pos to address p + p_offset and update the -- size of p to the number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is not zero-terminated. require start_position_big_enough: start_pos >= 1 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos <= s.count even_p_offset: (p_offset \\ 2) = 0 p_offset_non_negative: p_offset >= 0 ensure instance_free: class p_count_may_increase: p.count >= old p.count roundtrip: a_new_upper /= Void implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (a_new_upper.item // 2) - 1, False).same_string_general (s) roundtrip: (a_new_upper = Void and then not s.has ('%U'.to_character_32)) implies utf_16_0_subpointer_to_string_32 (p, p_offset // 2, (p.count // 2) - 1, True).same_string_general (s) feature -- UTF-32 to UTF-8 escaped_utf_32_string_into_utf_8_string_8 (s: READABLE_STRING_GENERAL; a_result: STRING_8) -- Copy the UTF-8 sequence corresponding to s interpreted as a UTF-32 sequence that could -- be escaped appended into a_result. -- If s contains the `escape_character` followed by either "HH" or "uHHHH" where H stands -- for an hexadecimal digit, then s has been escaped and will be converted to what is -- expected by the current platform. -- Otherwise it will be ignored and it will be left as is. -- See the note clause for the class for more details on the encoding. ensure instance_free: class roundtrip: utf_8_string_8_to_escaped_string_32 (a_result.substring (old a_result.count + 1, a_result.count)).same_string_general (s) escaped_utf_32_string_to_utf_8_string_8 (s: READABLE_STRING_GENERAL): STRING_8 -- UTF-8 sequence corresponding to s interpreted as a UTF-32 sequence that could be escaped. -- If s contains the `escape_character` followed by either "HH" or "uHHHH" where H stands -- for an hexadecimal digit, then s has been escaped and will be converted to what is -- expected by the current platform. -- Otherwise it will be ignored and it will be left as is. -- See the note clause for the class for more details on the encoding. ensure instance_free: class roundtrip: utf_8_string_8_to_escaped_string_32 (Result).same_string_general (s) escaped_utf_32_substring_into_utf_8_0_pointer (s: READABLE_STRING_GENERAL; start_pos, end_pos: INTEGER_32; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-8 sequence corresponding to s, interpreted as a UTF-32 sequence that could -- be escaped, with terminating zero to address p + p_offset and update the size of p to the -- number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is zero-terminated. -- If s contains the `escape_character` followed by either "HH" or "uHHHH" where H stands -- for an hexadecimal digit, then s has been escaped and will be converted to what is -- expected by the current platform. -- Otherwise it will be ignored and it will be left as is. -- See the note clause for the class for more details on the encoding. require start_position_big_enough: start_pos >= 1 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos <= s.count p_offset_non_negative: p_offset >= 0 ensure instance_free: class roundtrip: a_new_upper /= Void implies utf_8_0_subpointer_to_escaped_string_32 (p, p_offset, a_new_upper.item - 1, False).same_string_general (s.substring (start_pos, end_pos)) roundtrip: (a_new_upper = Void and then not s.substring (start_pos, end_pos).has ('%U'.to_character_32)) implies utf_8_0_subpointer_to_escaped_string_32 (p, p_offset, p.count, True).same_string_general (s.substring (start_pos, end_pos)) string_32_into_utf_8_0_pointer (s: READABLE_STRING_32; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-8 sequence corresponding to s with terminating zero -- to address p + p_offset and update the size of p to the number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is zero-terminated. require p_offset_non_negative: p_offset >= 0 ensure instance_free: class roundtrip: a_new_upper /= Void implies utf_8_0_subpointer_to_escaped_string_32 (p, p_offset, a_new_upper.item - 1, False).same_string (s) roundtrip: (a_new_upper = Void and then not s.has ('%U'.to_character_32)) implies utf_8_0_subpointer_to_escaped_string_32 (p, p_offset, p.count, True).same_string_general (s) string_32_into_utf_8_string_8 (s: READABLE_STRING_32; a_result: STRING_8) -- Copy the UTF-8 sequence corresponding to s appended into a_result. ensure instance_free: class roundtrip: utf_8_string_8_to_string_32 (a_result.substring (old a_result.count + 1, a_result.count)).same_string (s) string_32_to_utf_8_string_8 (s: READABLE_STRING_32): STRING_8 -- UTF-8 sequence corresponding to s. ensure instance_free: class roundtrip: utf_8_string_8_to_string_32 (Result).same_string (s) utf_32_code_into_utf_8_string_8 (c: NATURAL_32; a_result: STRING_8) -- Copy the UTF-8 sequence corresponding to code c appended into a_result. ensure instance_free: class utf_32_string_into_utf_8_0_pointer (s: READABLE_STRING_GENERAL; p: MANAGED_POINTER; p_offset: INTEGER_32; a_new_upper: detachable CELL [INTEGER_32]) -- Write UTF-8 sequence corresponding to s, interpreted as a UTF-32 sequence, -- with terminating zero to address p + p_offset and update the size of p to the -- number of written bytes. -- If a_new_upper is provided, the upper index of p containing the zero-termination -- is written to a_new_upper. -- The sequence is zero-terminated. require p_offset_non_negative: p_offset >= 0 ensure instance_free: class roundtrip: a_new_upper /= Void implies utf_8_0_subpointer_to_escaped_string_32 (p, p_offset, a_new_upper.item - 1, False).same_string_general (s) roundtrip: (a_new_upper = Void and then not s.has ('%U'.to_character_32)) implies utf_8_0_subpointer_to_escaped_string_32 (p, p_offset, p.count, True).same_string_general (s) utf_32_string_into_utf_8_string_8 (s: READABLE_STRING_GENERAL; a_result: STRING_8) -- Copy the UTF-8 sequence corresponding to s interpreted as a UTF-32 sequence -- appended into a_result. ensure instance_free: class roundtrip: utf_8_string_8_to_string_32 (a_result.substring (old a_result.count + 1, a_result.count)).same_string_general (s) utf_32_string_to_utf_8 (s: READABLE_STRING_GENERAL): SPECIAL [NATURAL_8] -- UTF-8 sequence corresponding to s, interpreted as a UTF-32 sequence. -- The sequence is not zero-terminated. ensure instance_free: class roundtrip: attached utf_32_string_to_utf_8_string_8 (s) as l_ref and then ∀ n: Result ¦ n.to_natural_32 = l_ref.code (@n.target_index + 1) utf_32_string_to_utf_8_0 (s: READABLE_STRING_GENERAL): SPECIAL [NATURAL_8] -- UTF-8 sequence corresponding to s, interpreted as a UTF-32 sequence. -- The sequence is zero-terminated. ensure instance_free: class attached_utf_8_string: attached utf_32_string_to_utf_8_string_8 (s) as l_ref count: Result.count = l_ref.count + 1 roundtrip: ∀ x: l_ref ¦ x = Result [@x.target_index - 1].to_character_8 zero_terminated: Result [Result.upper] = 0 utf_32_string_to_utf_8_string_8 (s: READABLE_STRING_GENERAL): STRING_8 -- UTF-8 sequence corresponding to s interpreted as a UTF-32 sequence. ensure instance_free: class roundtrip: utf_8_string_8_to_string_32 (Result).same_string_general (s) feature -- UTF-8 to UTF-16 utf_8_string_8_to_utf_16 (s: READABLE_STRING_8): SPECIAL [NATURAL_16] -- UTF-16 sequence corresponding to UTF-8 sequence s. ensure instance_free: class roundtrip: is_valid_utf_8_string_8 (s) implies utf_16_to_utf_8_string_8 (Result).same_string (s) utf_8_string_8_to_utf_16_0 (s: READABLE_STRING_8): SPECIAL [NATURAL_16] -- UTF-16 sequence corresponding to UTF-8 sequence s with terminating zero. ensure instance_free: class roundtrip: is_valid_utf_8_string_8 (s) implies utf_16_to_utf_8_string_8 (Result).same_string (s) feature -- UTF-8 to UTF-32 utf_8_0_pointer_into_escaped_string_32 (p: MANAGED_POINTER; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-8 sequence p which is zero-terminated, -- where invalid UTF-8 sequences are escaped, appended into a_result. ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_8_string_8 (a_result.substring (old a_result.count + 1, a_result.count)) as s and then ∀ c: s ¦ c = p.read_natural_8 (@c.target_index - 1).to_character_8 utf_8_0_pointer_to_escaped_string_32 (p: MANAGED_POINTER): STRING_32 -- {STRING_32} object corresponding to UTF-8 sequence p which is zero-terminated, -- where invalid UTF-8 sequences are escaped. ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_8_string_8 (Result) as s and then ∀ c: s ¦ c = p.read_natural_8 (@c.target_index - 1).to_character_8 utf_8_0_subpointer_into_escaped_string_32 (p: MANAGED_POINTER; start_pos, end_pos: INTEGER_32; a_stop_at_null: BOOLEAN; a_result: STRING_32) -- Copy {STRING_32} object corresponding to UTF-8 sequence p between indexes start_pos and -- end_pos or the first null character encountered if a_stop_at_null, where invalid -- UTF-8 sequences are escaped, appended into a_result. require start_position_big_enough: start_pos >= 0 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos < p.count ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_8_string_8 (a_result.substring (old a_result.count + 1, a_result.count)) as s and then ∀ c: s ¦ c = p.read_natural_8 (start_pos + .target_index - 1).to_character_8 utf_8_0_subpointer_to_escaped_string_32 (p: MANAGED_POINTER; start_pos, end_pos: INTEGER_32; a_stop_at_null: BOOLEAN): STRING_32 -- {STRING_32} object corresponding to UTF-8 sequence p between indexes start_pos and -- end_pos or the first null character encountered if a_stop_at_null, where invalid -- UTF-8 sequences are escaped. require start_position_big_enough: start_pos >= 0 end_position_big_enough: start_pos <= end_pos + 1 end_pos_small_enough: end_pos < p.count ensure instance_free: class roundtrip: attached escaped_utf_32_string_to_utf_8_string_8 (Result) as s and then ∀ c: s ¦ c = p.read_natural_8 (start_pos + .target_index - 1).to_character_8 utf_8_string_8_into_escaped_string_32 (s: READABLE_STRING_8; a_result: STRING_32) -- Copy STRING_32 corresponding to UTF-8 sequence s, where invalid UTF-8 sequences are escaped, -- appended into a_result. ensure instance_free: class roundtrip: escaped_utf_32_string_to_utf_8_string_8 (a_result.substring (old a_result.count + 1, a_result.count)).same_string (s) utf_8_string_8_into_string_32 (s: READABLE_STRING_8; a_result: STRING_32) -- Copy STRING_32 corresponding to UTF-8 sequence s appended into a_result. ensure instance_free: class roundtrip: is_valid_utf_8_string_8 (s) implies utf_32_string_to_utf_8_string_8 (a_result.substring (old a_result.count + 1, a_result.count)).same_string (s) utf_8_string_8_to_escaped_string_32 (s: READABLE_STRING_8): STRING_32 -- STRING_32 corresponding to UTF-8 sequence s, where invalid UTF-8 sequences are escaped. ensure instance_free: class roundtrip: escaped_utf_32_string_to_utf_8_string_8 (Result).same_string (s) utf_8_string_8_to_string_32 (s: READABLE_STRING_8): STRING_32 -- STRING_32 corresponding to UTF-8 sequence s. ensure instance_free: class roundtrip: is_valid_utf_8_string_8 (s) implies utf_32_string_to_utf_8_string_8 (Result).same_string (s) invariant -- from ANY reflexive_equality: standard_is_equal (Current) reflexive_conformance: conforms_to (Current) note ca_ignore: "CA011", "CA011: too many arguments" copyright: "Copyright (c) 1984-2021, Eiffel Software and others" license: "Eiffel Forum License v2 (see http://www.eiffel.com/licensing/forum.txt)" source: "[ Eiffel Software 5949 Hollister Ave., Goleta, CA 93117 USA Telephone 805-685-1006, Fax 805-685-6869 Website http://www.eiffel.com Customer support http://support.eiffel.com ]" end -- class UTF_CONVERTER -- Generated by Eiffel Studio --
For more details: eiffel.org