#37484

pending completion

Build # #37484

Build Type

push

local

Committed by

web-flow

Commit Message

minor follow up on #48996 (#49129)

- removed unnecessary `Union`-signature
- use one-liner definition when applicable
- add type assertion to make it more robust against invalidation

Run Details

5 of 5 new or added lines in 2 files covered. (100.0%)

70963 of 82484 relevant lines covered (86.03%)

33340438.71 hits per line

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

96.71

/base/strings/string.jl

# This file is a part of Julia. License is MIT: https://julialang.org/license

"""
    StringIndexError(str, i)

An error occurred when trying to access `str` at index `i` that is not valid.
"""
struct StringIndexError <: Exception
    string::AbstractString
    index::Integer
end
@noinline string_index_err(s::AbstractString, i::Integer) =
    throw(StringIndexError(s, Int(i)))
function Base.showerror(io::IO, exc::StringIndexError)
    s = exc.string
    print(io, "StringIndexError: ", "invalid index [$(exc.index)]")
    if firstindex(s) <= exc.index <= ncodeunits(s)
        iprev = thisind(s, exc.index)
        inext = nextind(s, iprev)
        escprev = escape_string(s[iprev:iprev])
        if inext <= ncodeunits(s)
            escnext = escape_string(s[inext:inext])
            print(io, ", valid nearby indices [$iprev]=>'$escprev', [$inext]=>'$escnext'")
        else
            print(io, ", valid nearby index [$iprev]=>'$escprev'")
        end
    end
end

const ByteArray = Union{CodeUnits{UInt8,String}, Vector{UInt8},Vector{Int8}, FastContiguousSubArray{UInt8,1,CodeUnits{UInt8,String}}, FastContiguousSubArray{UInt8,1,Vector{UInt8}}, FastContiguousSubArray{Int8,1,Vector{Int8}}}

@inline between(b::T, lo::T, hi::T) where {T<:Integer} = (lo ≤ b) & (b ≤ hi)

"""
    String <: AbstractString

The default string type in Julia, used by e.g. string literals.

`String`s are immutable sequences of `Char`s. A `String` is stored internally as
a contiguous byte array, and while they are interpreted as being UTF-8 encoded,
they can be composed of any byte sequence. Use [`isvalid`](@ref) to validate
that the underlying byte sequence is valid as UTF-8.
"""
String

## constructors and conversions ##

# String constructor docstring from boot.jl, workaround for #16730
# and the unavailability of @doc in boot.jl context.
"""
    String(v::AbstractVector{UInt8})

Create a new `String` object using the data buffer from byte vector `v`.
If `v` is a `Vector{UInt8}` it will be truncated to zero length and future
modification of `v` cannot affect the contents of the resulting string.
To avoid truncation of `Vector{UInt8}` data, use `String(copy(v))`; for other
`AbstractVector` types, `String(v)` already makes a copy.

When possible, the memory of `v` will be used without copying when the `String`
object is created. This is guaranteed to be the case for byte vectors returned
by [`take!`](@ref) on a writable [`IOBuffer`](@ref) and by calls to
[`read(io, nb)`](@ref). This allows zero-copy conversion of I/O data to strings.
In other cases, `Vector{UInt8}` data may be copied, but `v` is truncated anyway
to guarantee consistent behavior.
"""
String(v::AbstractVector{UInt8}) = String(copyto!(StringVector(length(v)), v))
String(v::Vector{UInt8}) = ccall(:jl_array_to_string, Ref{String}, (Any,), v)

"""
    unsafe_string(p::Ptr{UInt8}, [length::Integer])

Copy a string from the address of a C-style (NUL-terminated) string encoded as UTF-8.
(The pointer can be safely freed afterwards.) If `length` is specified
(the length of the data in bytes), the string does not have to be NUL-terminated.

This function is labeled "unsafe" because it will crash if `p` is not
a valid memory address to data of the requested length.
"""
function unsafe_string(p::Union{Ptr{UInt8},Ptr{Int8}}, len::Integer)
    p == C_NULL && throw(ArgumentError("cannot convert NULL to string"))
    ccall(:jl_pchar_to_string, Ref{String}, (Ptr{UInt8}, Int), p, len)
end
function unsafe_string(p::Union{Ptr{UInt8},Ptr{Int8}})
    p == C_NULL && throw(ArgumentError("cannot convert NULL to string"))
    ccall(:jl_cstr_to_string, Ref{String}, (Ptr{UInt8},), p)
end

# This is @assume_effects :effect_free :nothrow :terminates_globally @ccall jl_alloc_string(n::Csize_t)::Ref{String},
# but the macro is not available at this time in bootstrap, so we write it manually.
@eval _string_n(n::Integer) = $(Expr(:foreigncall, QuoteNode(:jl_alloc_string), Ref{String}, Expr(:call, Expr(:core, :svec), :Csize_t), 1, QuoteNode((:ccall,0xe)), :(convert(Csize_t, n))))

"""
    String(s::AbstractString)

Create a new `String` from an existing `AbstractString`.
"""
String(s::AbstractString) = print_to_string(s)
@assume_effects :total String(s::Symbol) = unsafe_string(unsafe_convert(Ptr{UInt8}, s))

unsafe_wrap(::Type{Vector{UInt8}}, s::String) = ccall(:jl_string_to_array, Ref{Vector{UInt8}}, (Any,), s)

Vector{UInt8}(s::CodeUnits{UInt8,String}) = copyto!(Vector{UInt8}(undef, length(s)), s)
Vector{UInt8}(s::String) = Vector{UInt8}(codeunits(s))
Array{UInt8}(s::String)  = Vector{UInt8}(codeunits(s))

String(s::CodeUnits{UInt8,String}) = s.s

## low-level functions ##

pointer(s::String) = unsafe_convert(Ptr{UInt8}, s)
pointer(s::String, i::Integer) = pointer(s) + Int(i)::Int - 1

ncodeunits(s::String) = Core.sizeof(s)
codeunit(s::String) = UInt8

codeunit(s::String, i::Integer) = codeunit(s, Int(i))
@assume_effects :foldable @inline function codeunit(s::String, i::Int)
    @boundscheck checkbounds(s, i)
    b = GC.@preserve s unsafe_load(pointer(s, i))
    return b
end

## comparison ##

@assume_effects :total _memcmp(a::String, b::String) = @invoke _memcmp(a::Union{Ptr{UInt8},AbstractString},b::Union{Ptr{UInt8},AbstractString})

_memcmp(a::Union{Ptr{UInt8},AbstractString}, b::Union{Ptr{UInt8},AbstractString}) = _memcmp(a, b, min(sizeof(a), sizeof(b)))
function _memcmp(a::Union{Ptr{UInt8},AbstractString}, b::Union{Ptr{UInt8},AbstractString}, len::Int)
    ccall(:memcmp, Cint, (Ptr{UInt8}, Ptr{UInt8}, Csize_t), a, b, len % Csize_t) % Int
end

function cmp(a::String, b::String)
    al, bl = sizeof(a), sizeof(b)
    c = _memcmp(a, b)
    return c < 0 ? -1 : c > 0 ? +1 : cmp(al,bl)
end

==(a::String, b::String) = a===b

typemin(::Type{String}) = ""
typemin(::String) = typemin(String)

## thisind, nextind ##

@propagate_inbounds thisind(s::String, i::Int) = _thisind_str(s, i)

# s should be String or SubString{String}
@inline function _thisind_str(s, i::Int)
    i == 0 && return 0
    n = ncodeunits(s)
    i == n + 1 && return i
    @boundscheck between(i, 1, n) || throw(BoundsError(s, i))
    @inbounds b = codeunit(s, i)
    (b & 0xc0 == 0x80) & (i-1 > 0) || return i
    @inbounds b = codeunit(s, i-1)
    between(b, 0b11000000, 0b11110111) && return i-1
    (b & 0xc0 == 0x80) & (i-2 > 0) || return i
    @inbounds b = codeunit(s, i-2)
    between(b, 0b11100000, 0b11110111) && return i-2
    (b & 0xc0 == 0x80) & (i-3 > 0) || return i
    @inbounds b = codeunit(s, i-3)
    between(b, 0b11110000, 0b11110111) && return i-3
    return i
end

@propagate_inbounds nextind(s::String, i::Int) = _nextind_str(s, i)

# s should be String or SubString{String}
@inline function _nextind_str(s, i::Int)
    i == 0 && return 1
    n = ncodeunits(s)
    @boundscheck between(i, 1, n) || throw(BoundsError(s, i))
    @inbounds l = codeunit(s, i)
    (l < 0x80) | (0xf8 ≤ l) && return i+1
    if l < 0xc0
        i′ = @inbounds thisind(s, i)
        return i′ < i ? @inbounds(nextind(s, i′)) : i+1
    end
    # first continuation byte
    (i += 1) > n && return i
    @inbounds b = codeunit(s, i)
    b & 0xc0 ≠ 0x80 && return i
    ((i += 1) > n) | (l < 0xe0) && return i
    # second continuation byte
    @inbounds b = codeunit(s, i)
    b & 0xc0 ≠ 0x80 && return i
    ((i += 1) > n) | (l < 0xf0) && return i
    # third continuation byte
    @inbounds b = codeunit(s, i)
    ifelse(b & 0xc0 ≠ 0x80, i, i+1)
end

## checking UTF-8 & ACSII validity ##

byte_string_classify(s::Union{String,Vector{UInt8},FastContiguousSubArray{UInt8,1,Vector{UInt8}}}) =
    ccall(:u8_isvalid, Int32, (Ptr{UInt8}, Int), s, sizeof(s))
    # 0: neither valid ASCII nor UTF-8
    # 1: valid ASCII
    # 2: valid UTF-8

isvalid(::Type{String}, s::Union{Vector{UInt8},FastContiguousSubArray{UInt8,1,Vector{UInt8}},String}) = byte_string_classify(s) ≠ 0
isvalid(s::String) = isvalid(String, s)

is_valid_continuation(c) = c & 0xc0 == 0x80

## required core functionality ##

@inline function iterate(s::String, i::Int=firstindex(s))
    (i % UInt) - 1 < ncodeunits(s) || return nothing
    b = @inbounds codeunit(s, i)
    u = UInt32(b) << 24
    between(b, 0x80, 0xf7) || return reinterpret(Char, u), i+1
    return iterate_continued(s, i, u)
end

function iterate_continued(s::String, i::Int, u::UInt32)
    u < 0xc0000000 && (i += 1; @goto ret)
    n = ncodeunits(s)
    # first continuation byte
    (i += 1) > n && @goto ret
    @inbounds b = codeunit(s, i)
    b & 0xc0 == 0x80 || @goto ret
    u |= UInt32(b) << 16
    # second continuation byte
    ((i += 1) > n) | (u < 0xe0000000) && @goto ret
    @inbounds b = codeunit(s, i)
    b & 0xc0 == 0x80 || @goto ret
    u |= UInt32(b) << 8
    # third continuation byte
    ((i += 1) > n) | (u < 0xf0000000) && @goto ret
    @inbounds b = codeunit(s, i)
    b & 0xc0 == 0x80 || @goto ret
    u |= UInt32(b); i += 1
@label ret
    return reinterpret(Char, u), i
end

@propagate_inbounds function getindex(s::String, i::Int)
    b = codeunit(s, i)
    u = UInt32(b) << 24
    between(b, 0x80, 0xf7) || return reinterpret(Char, u)
    return getindex_continued(s, i, u)
end

function getindex_continued(s::String, i::Int, u::UInt32)
    if u < 0xc0000000
        # called from `getindex` which checks bounds
        @inbounds isvalid(s, i) && @goto ret
        string_index_err(s, i)
    end
    n = ncodeunits(s)

    (i += 1) > n && @goto ret
    @inbounds b = codeunit(s, i) # cont byte 1
    b & 0xc0 == 0x80 || @goto ret
    u |= UInt32(b) << 16

    ((i += 1) > n) | (u < 0xe0000000) && @goto ret
    @inbounds b = codeunit(s, i) # cont byte 2
    b & 0xc0 == 0x80 || @goto ret
    u |= UInt32(b) << 8

    ((i += 1) > n) | (u < 0xf0000000) && @goto ret
    @inbounds b = codeunit(s, i) # cont byte 3
    b & 0xc0 == 0x80 || @goto ret
    u |= UInt32(b)
@label ret
    return reinterpret(Char, u)
end

getindex(s::String, r::AbstractUnitRange{<:Integer}) = s[Int(first(r)):Int(last(r))]

@inline function getindex(s::String, r::UnitRange{Int})
    isempty(r) && return ""
    i, j = first(r), last(r)
    @boundscheck begin
        checkbounds(s, r)
        @inbounds isvalid(s, i) || string_index_err(s, i)
        @inbounds isvalid(s, j) || string_index_err(s, j)
    end
    j = nextind(s, j) - 1
    n = j - i + 1
    ss = _string_n(n)
    GC.@preserve s ss unsafe_copyto!(pointer(ss), pointer(s, i), n)
    return ss
end

# nothrow because we know the start and end indices are valid
@assume_effects :nothrow length(s::String) = length_continued(s, 1, ncodeunits(s), ncodeunits(s))

# effects needed because @inbounds
@assume_effects :consistent :effect_free @inline function length(s::String, i::Int, j::Int)
    @boundscheck begin
        0 < i ≤ ncodeunits(s)+1 || throw(BoundsError(s, i))
        0 ≤ j < ncodeunits(s)+1 || throw(BoundsError(s, j))
    end
    j < i && return 0
    @inbounds i, k = thisind(s, i), i
    c = j - i + (i == k)
    @inbounds length_continued(s, i, j, c)
end

@assume_effects :terminates_locally @inline @propagate_inbounds function length_continued(s::String, i::Int, n::Int, c::Int)
    i < n || return c
    b = codeunit(s, i)
    while true
        while true
            (i += 1) ≤ n || return c
            0xc0 ≤ b ≤ 0xf7 && break
            b = codeunit(s, i)
        end
        l = b
        b = codeunit(s, i) # cont byte 1
        c -= (x = b & 0xc0 == 0x80)
        x & (l ≥ 0xe0) || continue

        (i += 1) ≤ n || return c
        b = codeunit(s, i) # cont byte 2
        c -= (x = b & 0xc0 == 0x80)
        x & (l ≥ 0xf0) || continue

        (i += 1) ≤ n || return c
        b = codeunit(s, i) # cont byte 3
        c -= (b & 0xc0 == 0x80)
    end
end

## overload methods for efficiency ##

isvalid(s::String, i::Int) = checkbounds(Bool, s, i) && thisind(s, i) == i

isascii(s::String) = isascii(codeunits(s))

# don't assume effects for general integers since we cannot know their implementation
@assume_effects :foldable repeat(c::Char, r::BitInteger) = @invoke repeat(c::Char, r::Integer)

"""
    repeat(c::AbstractChar, r::Integer) -> String

Repeat a character `r` times. This can equivalently be accomplished by calling
[`c^r`](@ref :^(::Union{AbstractString, AbstractChar}, ::Integer)).

# Examples
```jldoctest
julia> repeat('A', 3)
"AAA"
```
"""
function repeat(c::AbstractChar, r::Integer)
    c = Char(c)::Char
    r == 0 && return ""
    r < 0 && throw(ArgumentError("can't repeat a character $r times"))
    u = bswap(reinterpret(UInt32, c))
    n = 4 - (leading_zeros(u | 0xff) >> 3)
    s = _string_n(n*r)
    p = pointer(s)
    GC.@preserve s if n == 1
        ccall(:memset, Ptr{Cvoid}, (Ptr{UInt8}, Cint, Csize_t), p, u % UInt8, r)
    elseif n == 2
        p16 = reinterpret(Ptr{UInt16}, p)
        for i = 1:r
            unsafe_store!(p16, u % UInt16, i)
        end
    elseif n == 3
        b1 = (u >> 0) % UInt8
        b2 = (u >> 8) % UInt8
        b3 = (u >> 16) % UInt8
        for i = 0:r-1
            unsafe_store!(p, b1, 3i + 1)
            unsafe_store!(p, b2, 3i + 2)
            unsafe_store!(p, b3, 3i + 3)
        end
    elseif n == 4
        p32 = reinterpret(Ptr{UInt32}, p)
        for i = 1:r
            unsafe_store!(p32, u, i)
        end
    end
    return s
end

1	# This file is a part of Julia. License is MIT: https://julialang.org/license
2
3	"""
4	StringIndexError(str, i)
5
6	An error occurred when trying to access `str` at index `i` that is not valid.
7	"""
8	struct StringIndexError <: Exception
9	string::AbstractString	661✔
10	index::Integer
11	end
12	@noinline string_index_err(s::AbstractString, i::Integer) =	657✔
13	throw(StringIndexError(s, Int(i)))
14	function Base.showerror(io::IO, exc::StringIndexError)	4✔
15	s = exc.string	4✔
16	print(io, "StringIndexError: ", "invalid index [$(exc.index)]")	4✔
17	if firstindex(s) <= exc.index <= ncodeunits(s)	4✔
18	iprev = thisind(s, exc.index)	4✔
19	inext = nextind(s, iprev)	4✔
20	escprev = escape_string(s[iprev:iprev])	4✔
21	if inext <= ncodeunits(s)	4✔
22	escnext = escape_string(s[inext:inext])	3✔
23	print(io, ", valid nearby indices [$iprev]=>'$escprev', [$inext]=>'$escnext'")	3✔
24	else
25	print(io, ", valid nearby index [$iprev]=>'$escprev'")	1✔
26	end
27	end
28	end
29
30	const ByteArray = Union{CodeUnits{UInt8,String}, Vector{UInt8},Vector{Int8}, FastContiguousSubArray{UInt8,1,CodeUnits{UInt8,String}}, FastContiguousSubArray{UInt8,1,Vector{UInt8}}, FastContiguousSubArray{Int8,1,Vector{Int8}}}
31
32	@inline between(b::T, lo::T, hi::T) where {T<:Integer} = (lo ≤ b) & (b ≤ hi)	152,438,762✔
33
34	"""
35	String <: AbstractString
36
37	The default string type in Julia, used by e.g. string literals.
38
39	`String`s are immutable sequences of `Char`s. A `String` is stored internally as
40	a contiguous byte array, and while they are interpreted as being UTF-8 encoded,
41	they can be composed of any byte sequence. Use [`isvalid`](@ref) to validate
42	that the underlying byte sequence is valid as UTF-8.
43	"""
44	String
45
46	## constructors and conversions ##
47
48	# String constructor docstring from boot.jl, workaround for #16730
49	# and the unavailability of @doc in boot.jl context.
50	"""
51	String(v::AbstractVector{UInt8})
52
53	Create a new `String` object using the data buffer from byte vector `v`.
54	If `v` is a `Vector{UInt8}` it will be truncated to zero length and future
55	modification of `v` cannot affect the contents of the resulting string.
56	To avoid truncation of `Vector{UInt8}` data, use `String(copy(v))`; for other
57	`AbstractVector` types, `String(v)` already makes a copy.
58
59	When possible, the memory of `v` will be used without copying when the `String`
60	object is created. This is guaranteed to be the case for byte vectors returned
61	by [`take!`](@ref) on a writable [`IOBuffer`](@ref) and by calls to
62	[`read(io, nb)`](@ref). This allows zero-copy conversion of I/O data to strings.
63	In other cases, `Vector{UInt8}` data may be copied, but `v` is truncated anyway
64	to guarantee consistent behavior.
65	"""
66	String(v::AbstractVector{UInt8}) = String(copyto!(StringVector(length(v)), v))	2,612,856✔
67	String(v::Vector{UInt8}) = ccall(:jl_array_to_string, Ref{String}, (Any,), v)	6,369,215✔
68
69	"""
70	unsafe_string(p::Ptr{UInt8}, [length::Integer])
71
72	Copy a string from the address of a C-style (NUL-terminated) string encoded as UTF-8.
73	(The pointer can be safely freed afterwards.) If `length` is specified
74	(the length of the data in bytes), the string does not have to be NUL-terminated.
75
76	This function is labeled "unsafe" because it will crash if `p` is not
77	a valid memory address to data of the requested length.
78	"""
79	function unsafe_string(p::Union{Ptr{UInt8},Ptr{Int8}}, len::Integer)	1,188✔
80	p == C_NULL && throw(ArgumentError("cannot convert NULL to string"))	550,374✔
81	ccall(:jl_pchar_to_string, Ref{String}, (Ptr{UInt8}, Int), p, len)	550,373✔
82	end
83	function unsafe_string(p::Union{Ptr{UInt8},Ptr{Int8}})	10,190✔
84	p == C_NULL && throw(ArgumentError("cannot convert NULL to string"))	2,961,221✔
85	ccall(:jl_cstr_to_string, Ref{String}, (Ptr{UInt8},), p)	2,961,220✔
86	end
87
88	# This is @assume_effects :effect_free :nothrow :terminates_globally @ccall jl_alloc_string(n::Csize_t)::Ref{String},
89	# but the macro is not available at this time in bootstrap, so we write it manually.
90	@eval _string_n(n::Integer) = $(Expr(:foreigncall, QuoteNode(:jl_alloc_string), Ref{String}, Expr(:call, Expr(:core, :svec), :Csize_t), 1, QuoteNode((:ccall,0xe)), :(convert(Csize_t, n))))	108,677,923✔
91
92	"""
93	String(s::AbstractString)
94
95	Create a new `String` from an existing `AbstractString`.
96	"""
97	String(s::AbstractString) = print_to_string(s)	397✔
98	@assume_effects :total String(s::Symbol) = unsafe_string(unsafe_convert(Ptr{UInt8}, s))	588,068✔
99
100	unsafe_wrap(::Type{Vector{UInt8}}, s::String) = ccall(:jl_string_to_array, Ref{Vector{UInt8}}, (Any,), s)	6,132,963✔
101
102	Vector{UInt8}(s::CodeUnits{UInt8,String}) = copyto!(Vector{UInt8}(undef, length(s)), s)	20,937✔
103	Vector{UInt8}(s::String) = Vector{UInt8}(codeunits(s))	20,860✔
104	Array{UInt8}(s::String) = Vector{UInt8}(codeunits(s))	×
105
106	String(s::CodeUnits{UInt8,String}) = s.s	3✔
107
108	## low-level functions ##
109
110	pointer(s::String) = unsafe_convert(Ptr{UInt8}, s)	406,482,074✔
111	pointer(s::String, i::Integer) = pointer(s) + Int(i)::Int - 1	283,147,007✔
112
113	ncodeunits(s::String) = Core.sizeof(s)	344,787,842✔
114	codeunit(s::String) = UInt8	×
115
116	codeunit(s::String, i::Integer) = codeunit(s, Int(i))	×
117	@assume_effects :foldable @inline function codeunit(s::String, i::Int)	658,517✔
118	@boundscheck checkbounds(s, i)	264,035,233✔
119	b = GC.@preserve s unsafe_load(pointer(s, i))	264,035,233✔
120	return b	264,035,233✔
121	end
122
123	## comparison ##
124
125	@assume_effects :total _memcmp(a::String, b::String) = @invoke _memcmp(a::Union{Ptr{UInt8},AbstractString},b::Union{Ptr{UInt8},AbstractString})	370,280✔
126
127	_memcmp(a::Union{Ptr{UInt8},AbstractString}, b::Union{Ptr{UInt8},AbstractString}) = _memcmp(a, b, min(sizeof(a), sizeof(b)))	715,233✔
128	function _memcmp(a::Union{Ptr{UInt8},AbstractString}, b::Union{Ptr{UInt8},AbstractString}, len::Int)	297✔
129	ccall(:memcmp, Cint, (Ptr{UInt8}, Ptr{UInt8}, Csize_t), a, b, len % Csize_t) % Int	1,347,491✔
130	end
131
132	function cmp(a::String, b::String)	1✔
133	al, bl = sizeof(a), sizeof(b)	370,280✔
134	c = _memcmp(a, b)	370,280✔
135	return c < 0 ? -1 : c > 0 ? +1 : cmp(al,bl)	588,147✔
136	end
137
138	==(a::String, b::String) = a===b	10,363,626✔
139
140	typemin(::Type{String}) = ""	2✔
141	typemin(::String) = typemin(String)	1✔
142
143	## thisind, nextind ##
144
145	@propagate_inbounds thisind(s::String, i::Int) = _thisind_str(s, i)	177,371,204✔
146
147	# s should be String or SubString{String}
148	@inline function _thisind_str(s, i::Int)	×
149	i == 0 && return 0	89,254,034✔
150	n = ncodeunits(s)	89,235,486✔
151	i == n + 1 && return i	89,235,486✔
152	@boundscheck between(i, 1, n) \|\| throw(BoundsError(s, i))	89,213,044✔
153	@inbounds b = codeunit(s, i)	89,213,020✔
154	(b & 0xc0 == 0x80) & (i-1 > 0) \|\| return i	158,358,481✔
155	@inbounds b = codeunit(s, i-1)	20,067,559✔
156	between(b, 0b11000000, 0b11110111) && return i-1	20,067,559✔
157	(b & 0xc0 == 0x80) & (i-2 > 0) \|\| return i	15,659,928✔
158	@inbounds b = codeunit(s, i-2)	7,170,732✔
159	between(b, 0b11100000, 0b11110111) && return i-2	7,170,732✔
160	(b & 0xc0 == 0x80) & (i-3 > 0) \|\| return i	5,906,547✔
161	@inbounds b = codeunit(s, i-3)	1,026,997✔
162	between(b, 0b11110000, 0b11110111) && return i-3	1,026,997✔
163	return i	241,004✔
164	end
165
166	@propagate_inbounds nextind(s::String, i::Int) = _nextind_str(s, i)	13,404,867✔
167
168	# s should be String or SubString{String}
169	@inline function _nextind_str(s, i::Int)	13,865,932✔
170	i == 0 && return 1	13,865,932✔
171	n = ncodeunits(s)	13,845,352✔
172	@boundscheck between(i, 1, n) \|\| throw(BoundsError(s, i))	13,845,364✔
173	@inbounds l = codeunit(s, i)	13,845,340✔
174	(l < 0x80) \| (0xf8 ≤ l) && return i+1	13,845,340✔
175	if l < 0xc0	131,779✔
176	i′ = @inbounds thisind(s, i)	51,357✔
177	return i′ < i ? @inbounds(nextind(s, i′)) : i+1	30,403✔
178	end
179	# first continuation byte
180	(i += 1) > n && return i	101,376✔
181	@inbounds b = codeunit(s, i)	100,738✔
182	b & 0xc0 ≠ 0x80 && return i	100,738✔
183	((i += 1) > n) \| (l < 0xe0) && return i	93,494✔
184	# second continuation byte
185	@inbounds b = codeunit(s, i)	76,070✔
186	b & 0xc0 ≠ 0x80 && return i	76,070✔
187	((i += 1) > n) \| (l < 0xf0) && return i	72,437✔
188	# third continuation byte
189	@inbounds b = codeunit(s, i)	20,743✔
190	ifelse(b & 0xc0 ≠ 0x80, i, i+1)	20,743✔
191	end
192
193	## checking UTF-8 & ACSII validity ##
194
195	byte_string_classify(s::Union{String,Vector{UInt8},FastContiguousSubArray{UInt8,1,Vector{UInt8}}}) =	20,322✔
196	ccall(:u8_isvalid, Int32, (Ptr{UInt8}, Int), s, sizeof(s))
197	# 0: neither valid ASCII nor UTF-8
198	# 1: valid ASCII
199	# 2: valid UTF-8
200
201	isvalid(::Type{String}, s::Union{Vector{UInt8},FastContiguousSubArray{UInt8,1,Vector{UInt8}},String}) = byte_string_classify(s) ≠ 0	20,322✔
202	isvalid(s::String) = isvalid(String, s)	39✔
203
204	is_valid_continuation(c) = c & 0xc0 == 0x80	4✔
205
206	## required core functionality ##
207
208	@inline function iterate(s::String, i::Int=firstindex(s))	62,239✔
209	(i % UInt) - 1 < ncodeunits(s) \|\| return nothing	27,607,902✔
210	b = @inbounds codeunit(s, i)	18,742,805✔
211	u = UInt32(b) << 24	18,742,805✔
212	between(b, 0x80, 0xf7) \|\| return reinterpret(Char, u), i+1	36,566,535✔
213	return iterate_continued(s, i, u)	919,075✔
214	end
215
216	function iterate_continued(s::String, i::Int, u::UInt32)	919,075✔
217	u < 0xc0000000 && (i += 1; @goto ret)	919,176✔
218	n = ncodeunits(s)	918,974✔
219	# first continuation byte
220	(i += 1) > n && @goto ret	918,974✔
221	@inbounds b = codeunit(s, i)	918,954✔
222	b & 0xc0 == 0x80 \|\| @goto ret	918,954✔
223	u \|= UInt32(b) << 16	918,893✔
224	# second continuation byte
225	((i += 1) > n) \| (u < 0xe0000000) && @goto ret	918,893✔
226	@inbounds b = codeunit(s, i)	854,119✔
227	b & 0xc0 == 0x80 \|\| @goto ret	854,121✔
228	u \|= UInt32(b) << 8	854,117✔
229	# third continuation byte
230	((i += 1) > n) \| (u < 0xf0000000) && @goto ret	854,117✔
231	@inbounds b = codeunit(s, i)	273,658✔
232	b & 0xc0 == 0x80 \|\| @goto ret	273,659✔
233	u \|= UInt32(b); i += 1	547,314✔
234	@label ret	×
235	return reinterpret(Char, u), i	919,075✔
236	end
237
238	@propagate_inbounds function getindex(s::String, i::Int)	156,418✔
239	b = codeunit(s, i)	2,522,002✔
240	u = UInt32(b) << 24	2,522,002✔
241	between(b, 0x80, 0xf7) \|\| return reinterpret(Char, u)	5,040,362✔
242	return getindex_continued(s, i, u)	3,642✔
243	end
244
245	function getindex_continued(s::String, i::Int, u::UInt32)	3,642✔
246	if u < 0xc0000000	3,642✔
247	# called from `getindex` which checks bounds
248	@inbounds isvalid(s, i) && @goto ret	16✔
249	string_index_err(s, i)	1✔
250	end
251	n = ncodeunits(s)	3,626✔
252
253	(i += 1) > n && @goto ret	3,626✔
254	@inbounds b = codeunit(s, i) # cont byte 1	3,625✔
255	b & 0xc0 == 0x80 \|\| @goto ret	3,625✔
256	u \|= UInt32(b) << 16	3,624✔
257
258	((i += 1) > n) \| (u < 0xe0000000) && @goto ret	3,624✔
259	@inbounds b = codeunit(s, i) # cont byte 2	2,398✔
260	b & 0xc0 == 0x80 \|\| @goto ret	2,398✔
261	u \|= UInt32(b) << 8	2,398✔
262
263	((i += 1) > n) \| (u < 0xf0000000) && @goto ret	2,398✔
264	@inbounds b = codeunit(s, i) # cont byte 3	461✔
265	b & 0xc0 == 0x80 \|\| @goto ret	461✔
266	u \|= UInt32(b)	461✔
267	@label ret	×
268	return reinterpret(Char, u)	3,641✔
269	end
270
271	getindex(s::String, r::AbstractUnitRange{<:Integer}) = s[Int(first(r)):Int(last(r))]	×
272
273	@inline function getindex(s::String, r::UnitRange{Int})	1,485,637✔
274	isempty(r) && return ""	975,347✔
275	i, j = first(r), last(r)	69,383✔
276	@boundscheck begin	950,546✔
277	checkbounds(s, r)	950,528✔
278	@inbounds isvalid(s, i) \|\| string_index_err(s, i)	950,516✔
279	@inbounds isvalid(s, j) \|\| string_index_err(s, j)	950,517✔
280	end
281	j = nextind(s, j) - 1	950,539✔
282	n = j - i + 1	950,539✔
283	ss = _string_n(n)	950,539✔
284	GC.@preserve s ss unsafe_copyto!(pointer(ss), pointer(s, i), n)	950,539✔
285	return ss	950,539✔
286	end
287
288	# nothrow because we know the start and end indices are valid
289	@assume_effects :nothrow length(s::String) = length_continued(s, 1, ncodeunits(s), ncodeunits(s))	744,761✔
290
291	# effects needed because @inbounds
292	@assume_effects :consistent :effect_free @inline function length(s::String, i::Int, j::Int)	68,251✔
293	@boundscheck begin	180,273✔
294	0 < i ≤ ncodeunits(s)+1 \|\| throw(BoundsError(s, i))	180,273✔
295	0 ≤ j < ncodeunits(s)+1 \|\| throw(BoundsError(s, j))	180,277✔
296	end
297	j < i && return 0	180,269✔
298	@inbounds i, k = thisind(s, i), i	261,932✔
299	c = j - i + (i == k)	130,966✔
300	@inbounds length_continued(s, i, j, c)	130,966✔
301	end
302
303	@assume_effects :terminates_locally @inline @propagate_inbounds function length_continued(s::String, i::Int, n::Int, c::Int)	896,893✔
304	i < n \|\| return c	1,041,892✔
305	b = codeunit(s, i)	709,562✔
306	while true	4,132,221✔
307	while true	48,082,961✔
308	(i += 1) ≤ n \|\| return c	48,791,752✔
309	0xc0 ≤ b ≤ 0xf7 && break	47,374,170✔
310	b = codeunit(s, i)	43,950,740✔
311	end	43,950,740✔
312	l = b	3,423,430✔
313	b = codeunit(s, i) # cont byte 1	3,423,430✔
314	c -= (x = b & 0xc0 == 0x80)	3,423,430✔
315	x & (l ≥ 0xe0) \|\| continue	3,423,430✔
316
317	(i += 1) ≤ n \|\| return c	3,024,583✔
318	b = codeunit(s, i) # cont byte 2	3,023,121✔
319	c -= (x = b & 0xc0 == 0x80)	3,023,121✔
320	x & (l ≥ 0xf0) \|\| continue	5,149,309✔
321
322	(i += 1) ≤ n \|\| return c	896,973✔
323	b = codeunit(s, i) # cont byte 3	896,893✔
324	c -= (b & 0xc0 == 0x80)	896,893✔
325	end	3,422,659✔
326	end
327
328	## overload methods for efficiency ##
329
330	isvalid(s::String, i::Int) = checkbounds(Bool, s, i) && thisind(s, i) == i	81,986,766✔
331
332	isascii(s::String) = isascii(codeunits(s))	25,782✔
333
334	# don't assume effects for general integers since we cannot know their implementation
335	@assume_effects :foldable repeat(c::Char, r::BitInteger) = @invoke repeat(c::Char, r::Integer)	14,173✔
336
337	"""
338	repeat(c::AbstractChar, r::Integer) -> String
339
340	Repeat a character `r` times. This can equivalently be accomplished by calling
341	[`c^r`](@ref :^(::Union{AbstractString, AbstractChar}, ::Integer)).
342
343	# Examples
344	```jldoctest
345	julia> repeat('A', 3)
346	"AAA"
347	```
348	"""
349	function repeat(c::AbstractChar, r::Integer)	14,174✔
350	c = Char(c)::Char	14,174✔
351	r == 0 && return ""	14,174✔
352	r < 0 && throw(ArgumentError("can't repeat a character $r times"))	10,019✔
353	u = bswap(reinterpret(UInt32, c))	10,015✔
354	n = 4 - (leading_zeros(u \| 0xff) >> 3)	10,015✔
355	s = _string_n(n*r)	10,015✔
356	p = pointer(s)	10,013✔
357	GC.@preserve s if n == 1	10,013✔
358	ccall(:memset, Ptr{Cvoid}, (Ptr{UInt8}, Cint, Csize_t), p, u % UInt8, r)	9,987✔
359	elseif n == 2	26✔
360	p16 = reinterpret(Ptr{UInt16}, p)	6✔
361	for i = 1:r	12✔
362	unsafe_store!(p16, u % UInt16, i)	20✔
363	end	20✔
364	elseif n == 3	20✔
365	b1 = (u >> 0) % UInt8	16✔
366	b2 = (u >> 8) % UInt8	16✔
367	b3 = (u >> 16) % UInt8	16✔
368	for i = 0:r-1	32✔
369	unsafe_store!(p, b1, 3i + 1)	45✔
370	unsafe_store!(p, b2, 3i + 2)	45✔
371	unsafe_store!(p, b3, 3i + 3)	45✔
372	end	45✔
373	elseif n == 4	4✔
374	p32 = reinterpret(Ptr{UInt32}, p)	4✔
375	for i = 1:r	8✔
376	unsafe_store!(p32, u, i)	8✔
377	end	10,021✔
378	end
379	return s	10,013✔
380	end

JuliaLang / julia / #37484

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