5872154557

pending completion

Build # 5872154557

Build Type

push

github

Committed by

neon-sunset

Commit Message

feat: basic .Trim() implementation

Run Details

105 of 740 branches covered (14.19%)

Branch coverage included in aggregate %.

38 of 38 new or added lines in 5 files covered. (100.0%)

372 of 1597 relevant lines covered (23.29%)

445.62 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

0.0

/src/U8String.Manipulation.cs

using System.Buffers;
using System.Text;
using U8Primitives.InteropServices;

namespace U8Primitives;

#pragma warning disable IDE0046, IDE0057 // Why: range slicing and ternary expressions do not produce desired codegen
public readonly partial struct U8String
{
    // TODO: Optimize/deduplicate Concat variants
    // TODO: Investigate if it is possible fold validation for u8 literals
    public static U8String Concat(U8String left, U8String right)
    {
        if (!left.IsEmpty)
        {
            if (!right.IsEmpty)
            {
                var length = left.Length + right.Length;
                var value = new byte[length];

                left.UnsafeSpan.CopyTo(value);
                right.UnsafeSpan.CopyTo(value.AsSpan(left.Length));

                return new U8String(value, 0, length);
            }

            return left;
        }

        return right;
    }

    public static U8String Concat(U8String left, ReadOnlySpan<byte> right)
    {
        if (!right.IsEmpty)
        {
            Validate(right);
            if (!left.IsEmpty)
            {
                var length = left.Length + right.Length;
                var value = new byte[length];

                left.UnsafeSpan.CopyTo(value);
                right.CopyTo(value.AsSpan(left.Length));

                return new U8String(value, 0, length);
            }

            return new U8String(right, skipValidation: true);
        }

        return left;
    }

    public static U8String Concat(ReadOnlySpan<byte> left, U8String right)
    {
        if (!left.IsEmpty)
        {
            Validate(left);
            if (!right.IsEmpty)
            {
                var length = left.Length + right.Length;
                var value = new byte[length];

                left.CopyTo(value);
                right.UnsafeSpan.CopyTo(value.AsSpan(left.Length));

                return new U8String(value, 0, length);
            }

            return new U8String(left, skipValidation: true);
        }

        return right;
    }

    public static U8String Concat(ReadOnlySpan<byte> left, ReadOnlySpan<byte> right)
    {
        var length = left.Length + right.Length;
        if (length != 0)
        {
            var value = new byte[length];

            left.CopyTo(value);
            right.CopyTo(value.SliceUnsafe(left.Length, right.Length));

            Validate(value);
            return new U8String(value, 0, length);
        }

        return default;
    }

    /// <summary>
    /// Normalizes current <see cref="U8String"/> to the specified Unicode normalization form (default: <see cref="NormalizationForm.FormC"/>).
    /// </summary>
    /// <returns>A new <see cref="U8String"/> normalized to the specified form.</returns>
    public U8String Normalize(NormalizationForm form = NormalizationForm.FormC)
    {
        throw new NotImplementedException();
    }

    [MethodImpl(MethodImplOptions.AggressiveInlining)]
    public U8String Replace(byte oldValue, byte newValue)
    {
        return U8Manipulation.Replace(this, oldValue, newValue);
    }

    [MethodImpl(MethodImplOptions.AggressiveInlining)]
    public U8String Replace(char oldValue, char newValue)
    {
        return U8Manipulation.Replace(this, oldValue, newValue);
    }

    [MethodImpl(MethodImplOptions.AggressiveInlining)]
    public U8String Replace(Rune oldValue, Rune newValue)
    {
        return U8Manipulation.Replace(this, oldValue, newValue);
    }

    [MethodImpl(MethodImplOptions.AggressiveInlining)]
    public U8String Replace(ReadOnlySpan<byte> oldValue, ReadOnlySpan<byte> newValue)
    {
        return U8Manipulation.Replace(this, oldValue, newValue);
    }

    [MethodImpl(MethodImplOptions.AggressiveInlining)]
    public U8String Replace(U8String oldValue, U8String newValue)
    {
        return U8Manipulation.ReplaceUnchecked(this, oldValue, newValue);
    }

    /// <inheritdoc />
    public void CopyTo(byte[] destination, int index)
    {
        var src = this;
        var dst = destination.AsSpan()[index..];
        if (src.Length > dst.Length)
        {
            ThrowHelpers.ArgumentOutOfRange(nameof(index));
        }

        src.UnsafeSpan.CopyTo(dst);
    }

    /// <summary>
    /// Retrieves a substring from this instance. The substring starts at a specified
    /// character position and continues to the end of the string.
    /// </summary>
    /// <param name="start">The zero-based starting character position of a substring in this instance.</param>
    /// <returns>A substring view that begins at <paramref name="start"/>.</returns>
    /// <exception cref="ArgumentOutOfRangeException">
    /// <paramref name="start"/> is less than zero or greater than the length of this instance.
    /// </exception>
    /// <exception cref="ArgumentException">
    /// The resulting substring splits at a UTF-8 code point boundary and would result in an invalid UTF-8 string.
    /// </exception>
    public U8String Slice(int start)
    {
        var source = this;
        // From ReadOnly/Span<T> Slice(int) implementation
        if ((ulong)(uint)start > (ulong)(uint)source.Length)
        {
            ThrowHelpers.ArgumentOutOfRange();
        }

        var length = source.Length - start;
        if (length > 0)
        {
            if (U8Info.IsContinuationByte(in source.UnsafeRefAdd(start)))
            {
                ThrowHelpers.InvalidSplit();
            }

            return new(source._value, source.Offset + start, length);
        }

        return default;
    }

    /// <summary>
    /// Retrieves a substring from this instance. The substring starts at a specified
    /// character position and has a specified length.
    /// </summary>
    /// <param name="start">The zero-based starting character position of a substring in this instance.</param>
    /// <param name="length">The number of bytes in the substring.</param>
    /// <returns>A substring view that begins at <paramref name="start"/> and has <paramref name="length"/> bytes.</returns>
    /// <exception cref="ArgumentOutOfRangeException">
    /// <paramref name="start"/> or <paramref name="length"/> is less than zero, or the sum of <paramref name="start"/> and <paramref name="length"/> is greater than the length of the current instance.
    /// </exception>
    /// <exception cref="ArgumentException">
    /// The resulting substring splits at a UTF-8 code point boundary and would result in an invalid UTF-8 string.
    /// </exception>
    public U8String Slice(int start, int length)
    {
        var source = this;
        // From ReadOnly/Span<T> Slice(int, int) implementation
        if ((ulong)(uint)start + (ulong)(uint)length > (ulong)(uint)source.Length)
        {
            ThrowHelpers.ArgumentOutOfRange();
        }

        var result = default(U8String);
        if (length > 0)
        {
            // TODO: Is there really no way to get rid of length < source.Length when checking the last+1 byte?
            if ((start > 0 && U8Info.IsContinuationByte(source.UnsafeRefAdd(start))) || (
                length < source.Length && U8Info.IsContinuationByte(source.UnsafeRefAdd(start + length))))
            {
                // TODO: Exception message UX
                ThrowHelpers.InvalidSplit();
            }

            result = new(source._value, source.Offset + start, length);
        }

        return result;
    }

    public U8String Trim()
    {
        // TODO: Optimize fast path on no whitespace
        // TODO 2: Do not convert to runes and have proper
        // whitespace LUT to evaluate code points in a branchless way
        var source = this;
        ref var ptr = ref source.DangerousRef;

        var start = 0;
        for (; start < source.Length; start++)
        {
            var b = ptr.Add(start);
            if (!U8Info.IsContinuationByte(b) && !(
                U8Info.IsAsciiByte(b)
                    ? U8Info.IsAsciiWhitespace(b)
                    : U8Info.IsNonAsciiWhitespace(ref ptr.Add(start))))
            {
                break;
            }
        }

        var end = source.Length - 1;
        for (; end >= start; end--)
        {
            var b = ptr.Add(end);
            if (!U8Info.IsContinuationByte(b) && !(
                U8Info.IsAsciiByte(b)
                    ? U8Info.IsAsciiWhitespace(b)
                    : U8Info.IsNonAsciiWhitespace(ref ptr.Add(end))))
            {
                break;
            }
        }

        return U8Marshal.Slice(source, start, end - start + 1);
    }

    /// <summary>
    /// Removes all leading and trailing ASCII white-space characters from the current string.
    /// </summary>
    /// <returns>
    /// A substring that remains after all ASCII white-space characters
    /// are removed from the start and end of the current string.
    /// </returns>
    public U8String TrimAscii()
    {
        var source = this;
        var range = Ascii.Trim(source);

        return !range.IsEmpty()
            ? U8Marshal.Slice(source, range)
            : default;
    }

    /// <summary>
    /// Removes all the leading ASCII white-space characters from the current string.
    /// </summary>
    /// <returns>
    /// A substring that remains after all white-space characters
    /// are removed from the start of the current string.
    /// </returns>
    [MethodImpl(MethodImplOptions.AggressiveInlining)]
    public U8String TrimStartAscii()
    {
        var source = this;
        var range = Ascii.TrimStart(source);

        return !range.IsEmpty()
            ? U8Marshal.Slice(source, range)
            : default;
    }

    /// <summary>
    /// Removes all the trailing ASCII white-space characters from the current string.
    /// </summary>
    /// <returns>
    /// A substring that remains after all white-space characters
    /// are removed from the end of the current string.
    /// </returns>
    [MethodImpl(MethodImplOptions.AggressiveInlining)]
    public U8String TrimEndAscii()
    {
        var source = this;
        var range = Ascii.TrimEnd(source);

        return !range.IsEmpty()
            ? U8Marshal.Slice(source, range)
            : default;
    }

    /// <summary>
    /// Returns a copy of this ASCII string converted to lower case.
    /// </summary>
    /// <returns>A lowercase equivalent of the current ASCII string.</returns>
    /// <exception cref="ArgumentException">
    /// The current string is not a valid ASCII sequence.
    /// </exception>
    public U8String ToLowerAscii()
    {
        var source = this;
        if (!source.IsEmpty)
        {
            var span = source.UnsafeSpan;
            var destination = new byte[span.Length];
            var result = Ascii.ToLower(span, destination, out _);
            if (result is OperationStatus.InvalidData)
            {
                ThrowHelpers.InvalidAscii();
            }

            return new U8String(destination, 0, span.Length);
        }

        return default;
    }

    /// <summary>
    /// Returns a copy of this ASCII string converted to upper case.
    /// </summary>
    /// <returns>The uppercase equivalent of the current ASCII string.</returns>
    /// <exception cref="ArgumentException">
    /// The current string is not a valid ASCII sequence.
    /// </exception>
    public U8String ToUpperAscii()
    {
        var source = this;
        if (!source.IsEmpty)
        {
            var span = source.UnsafeSpan;
            var destination = new byte[span.Length];
            var result = Ascii.ToUpper(span, destination, out _);
            if (result is OperationStatus.InvalidData)
            {
                ThrowHelpers.InvalidAscii();
            }

            return new U8String(destination, 0, span.Length);
        }

        return default;
    }
}

1	using System.Buffers;
2	using System.Text;
3	using U8Primitives.InteropServices;
4
5	namespace U8Primitives;
6
7	#pragma warning disable IDE0046, IDE0057 // Why: range slicing and ternary expressions do not produce desired codegen
8	public readonly partial struct U8String
9	{
10	// TODO: Optimize/deduplicate Concat variants
11	// TODO: Investigate if it is possible fold validation for u8 literals
12	public static U8String Concat(U8String left, U8String right)
13	{
14	if (!left.IsEmpty)	×
15	{
16	if (!right.IsEmpty)	×
17	{
18	var length = left.Length + right.Length;	×
19	var value = new byte[length];	×
20
21	left.UnsafeSpan.CopyTo(value);	×
22	right.UnsafeSpan.CopyTo(value.AsSpan(left.Length));	×
23
24	return new U8String(value, 0, length);	×
25	}
26
27	return left;	×
28	}
29
30	return right;	×
31	}
32
33	public static U8String Concat(U8String left, ReadOnlySpan<byte> right)
34	{
35	if (!right.IsEmpty)	×
36	{
37	Validate(right);	×
38	if (!left.IsEmpty)	×
39	{
40	var length = left.Length + right.Length;	×
41	var value = new byte[length];	×
42
43	left.UnsafeSpan.CopyTo(value);	×
44	right.CopyTo(value.AsSpan(left.Length));	×
45
46	return new U8String(value, 0, length);	×
47	}
48
49	return new U8String(right, skipValidation: true);	×
50	}
51
52	return left;	×
53	}
54
55	public static U8String Concat(ReadOnlySpan<byte> left, U8String right)
56	{
57	if (!left.IsEmpty)	×
58	{
59	Validate(left);	×
60	if (!right.IsEmpty)	×
61	{
62	var length = left.Length + right.Length;	×
63	var value = new byte[length];	×
64
65	left.CopyTo(value);	×
66	right.UnsafeSpan.CopyTo(value.AsSpan(left.Length));	×
67
68	return new U8String(value, 0, length);	×
69	}
70
71	return new U8String(left, skipValidation: true);	×
72	}
73
74	return right;	×
75	}
76
77	public static U8String Concat(ReadOnlySpan<byte> left, ReadOnlySpan<byte> right)
78	{
79	var length = left.Length + right.Length;	×
80	if (length != 0)	×
81	{
82	var value = new byte[length];	×
83
84	left.CopyTo(value);	×
85	right.CopyTo(value.SliceUnsafe(left.Length, right.Length));	×
86
87	Validate(value);	×
88	return new U8String(value, 0, length);	×
89	}
90
91	return default;	×
92	}
93
94	/// <summary>
95	/// Normalizes current <see cref="U8String"/> to the specified Unicode normalization form (default: <see cref="NormalizationForm.FormC"/>).
96	/// </summary>
97	/// <returns>A new <see cref="U8String"/> normalized to the specified form.</returns>
98	public U8String Normalize(NormalizationForm form = NormalizationForm.FormC)
99	{
100	throw new NotImplementedException();	×
101	}
102
103	[MethodImpl(MethodImplOptions.AggressiveInlining)]
104	public U8String Replace(byte oldValue, byte newValue)
105	{
106	return U8Manipulation.Replace(this, oldValue, newValue);	×
107	}
108
109	[MethodImpl(MethodImplOptions.AggressiveInlining)]
110	public U8String Replace(char oldValue, char newValue)
111	{
112	return U8Manipulation.Replace(this, oldValue, newValue);	×
113	}
114
115	[MethodImpl(MethodImplOptions.AggressiveInlining)]
116	public U8String Replace(Rune oldValue, Rune newValue)
117	{
118	return U8Manipulation.Replace(this, oldValue, newValue);	×
119	}
120
121	[MethodImpl(MethodImplOptions.AggressiveInlining)]
122	public U8String Replace(ReadOnlySpan<byte> oldValue, ReadOnlySpan<byte> newValue)
123	{
124	return U8Manipulation.Replace(this, oldValue, newValue);	×
125	}
126
127	[MethodImpl(MethodImplOptions.AggressiveInlining)]
128	public U8String Replace(U8String oldValue, U8String newValue)
129	{
130	return U8Manipulation.ReplaceUnchecked(this, oldValue, newValue);	×
131	}
132
133	/// <inheritdoc />
134	public void CopyTo(byte[] destination, int index)
135	{
136	var src = this;	×
137	var dst = destination.AsSpan()[index..];	×
138	if (src.Length > dst.Length)	×
139	{
140	ThrowHelpers.ArgumentOutOfRange(nameof(index));	×
141	}
142
143	src.UnsafeSpan.CopyTo(dst);	×
144	}	×
145
146	/// <summary>
147	/// Retrieves a substring from this instance. The substring starts at a specified
148	/// character position and continues to the end of the string.
149	/// </summary>
150	/// <param name="start">The zero-based starting character position of a substring in this instance.</param>
151	/// <returns>A substring view that begins at <paramref name="start"/>.</returns>
152	/// <exception cref="ArgumentOutOfRangeException">
153	/// <paramref name="start"/> is less than zero or greater than the length of this instance.
154	/// </exception>
155	/// <exception cref="ArgumentException">
156	/// The resulting substring splits at a UTF-8 code point boundary and would result in an invalid UTF-8 string.
157	/// </exception>
158	public U8String Slice(int start)
159	{
160	var source = this;	×
161	// From ReadOnly/Span<T> Slice(int) implementation
162	if ((ulong)(uint)start > (ulong)(uint)source.Length)	×
163	{
164	ThrowHelpers.ArgumentOutOfRange();	×
165	}
166
167	var length = source.Length - start;	×
168	if (length > 0)	×
169	{
170	if (U8Info.IsContinuationByte(in source.UnsafeRefAdd(start)))	×
171	{
172	ThrowHelpers.InvalidSplit();	×
173	}
174
175	return new(source._value, source.Offset + start, length);	×
176	}
177
178	return default;	×
179	}
180
181	/// <summary>
182	/// Retrieves a substring from this instance. The substring starts at a specified
183	/// character position and has a specified length.
184	/// </summary>
185	/// <param name="start">The zero-based starting character position of a substring in this instance.</param>
186	/// <param name="length">The number of bytes in the substring.</param>
187	/// <returns>A substring view that begins at <paramref name="start"/> and has <paramref name="length"/> bytes.</returns>
188	/// <exception cref="ArgumentOutOfRangeException">
189	/// <paramref name="start"/> or <paramref name="length"/> is less than zero, or the sum of <paramref name="start"/> and <paramref name="length"/> is greater than the length of the current instance.
190	/// </exception>
191	/// <exception cref="ArgumentException">
192	/// The resulting substring splits at a UTF-8 code point boundary and would result in an invalid UTF-8 string.
193	/// </exception>
194	public U8String Slice(int start, int length)
195	{
196	var source = this;	×
197	// From ReadOnly/Span<T> Slice(int, int) implementation
198	if ((ulong)(uint)start + (ulong)(uint)length > (ulong)(uint)source.Length)	×
199	{
200	ThrowHelpers.ArgumentOutOfRange();	×
201	}
202
203	var result = default(U8String);	×
204	if (length > 0)	×
205	{
206	// TODO: Is there really no way to get rid of length < source.Length when checking the last+1 byte?
207	if ((start > 0 && U8Info.IsContinuationByte(source.UnsafeRefAdd(start))) \|\| (	×
208	length < source.Length && U8Info.IsContinuationByte(source.UnsafeRefAdd(start + length))))	×
209	{
210	// TODO: Exception message UX
211	ThrowHelpers.InvalidSplit();	×
212	}
213
214	result = new(source._value, source.Offset + start, length);	×
215	}
216
217	return result;	×
218	}
219
220	public U8String Trim()
221	{
222	// TODO: Optimize fast path on no whitespace
223	// TODO 2: Do not convert to runes and have proper
224	// whitespace LUT to evaluate code points in a branchless way
225	var source = this;	×
226	ref var ptr = ref source.DangerousRef;	×
227
228	var start = 0;	×
229	for (; start < source.Length; start++)	×
230	{
231	var b = ptr.Add(start);	×
232	if (!U8Info.IsContinuationByte(b) && !(	×
233	U8Info.IsAsciiByte(b)	×
234	? U8Info.IsAsciiWhitespace(b)	×
235	: U8Info.IsNonAsciiWhitespace(ref ptr.Add(start))))	×
236	{
237	break;
238	}
239	}
240
241	var end = source.Length - 1;	×
242	for (; end >= start; end--)	×
243	{
244	var b = ptr.Add(end);	×
245	if (!U8Info.IsContinuationByte(b) && !(	×
246	U8Info.IsAsciiByte(b)	×
247	? U8Info.IsAsciiWhitespace(b)	×
248	: U8Info.IsNonAsciiWhitespace(ref ptr.Add(end))))	×
249	{
250	break;
251	}
252	}
253
254	return U8Marshal.Slice(source, start, end - start + 1);	×
255	}
256
257	/// <summary>
258	/// Removes all leading and trailing ASCII white-space characters from the current string.
259	/// </summary>
260	/// <returns>
261	/// A substring that remains after all ASCII white-space characters
262	/// are removed from the start and end of the current string.
263	/// </returns>
264	public U8String TrimAscii()
265	{
266	var source = this;	×
267	var range = Ascii.Trim(source);	×
268
269	return !range.IsEmpty()	×
270	? U8Marshal.Slice(source, range)	×
271	: default;	×
272	}
273
274	/// <summary>
275	/// Removes all the leading ASCII white-space characters from the current string.
276	/// </summary>
277	/// <returns>
278	/// A substring that remains after all white-space characters
279	/// are removed from the start of the current string.
280	/// </returns>
281	[MethodImpl(MethodImplOptions.AggressiveInlining)]
282	public U8String TrimStartAscii()
283	{
284	var source = this;	×
285	var range = Ascii.TrimStart(source);	×
286
287	return !range.IsEmpty()	×
288	? U8Marshal.Slice(source, range)	×
289	: default;	×
290	}
291
292	/// <summary>
293	/// Removes all the trailing ASCII white-space characters from the current string.
294	/// </summary>
295	/// <returns>
296	/// A substring that remains after all white-space characters
297	/// are removed from the end of the current string.
298	/// </returns>
299	[MethodImpl(MethodImplOptions.AggressiveInlining)]
300	public U8String TrimEndAscii()
301	{
302	var source = this;	×
303	var range = Ascii.TrimEnd(source);	×
304
305	return !range.IsEmpty()	×
306	? U8Marshal.Slice(source, range)	×
307	: default;	×
308	}
309
310	/// <summary>
311	/// Returns a copy of this ASCII string converted to lower case.
312	/// </summary>
313	/// <returns>A lowercase equivalent of the current ASCII string.</returns>
314	/// <exception cref="ArgumentException">
315	/// The current string is not a valid ASCII sequence.
316	/// </exception>
317	public U8String ToLowerAscii()
318	{
319	var source = this;	×
320	if (!source.IsEmpty)	×
321	{
322	var span = source.UnsafeSpan;	×
323	var destination = new byte[span.Length];	×
324	var result = Ascii.ToLower(span, destination, out _);	×
325	if (result is OperationStatus.InvalidData)	×
326	{
327	ThrowHelpers.InvalidAscii();	×
328	}
329
330	return new U8String(destination, 0, span.Length);	×
331	}
332
333	return default;	×
334	}
335
336	/// <summary>
337	/// Returns a copy of this ASCII string converted to upper case.
338	/// </summary>
339	/// <returns>The uppercase equivalent of the current ASCII string.</returns>
340	/// <exception cref="ArgumentException">
341	/// The current string is not a valid ASCII sequence.
342	/// </exception>
343	public U8String ToUpperAscii()
344	{
345	var source = this;	×
346	if (!source.IsEmpty)	×
347	{
348	var span = source.UnsafeSpan;	×
349	var destination = new byte[span.Length];	×
350	var result = Ascii.ToUpper(span, destination, out _);	×
351	if (result is OperationStatus.InvalidData)	×
352	{
353	ThrowHelpers.InvalidAscii();	×
354	}
355
356	return new U8String(destination, 0, span.Length);	×
357	}
358
359	return default;	×
360	}
361	}

neon-sunset / U8String / 5872154557

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous