#37433

pending completion

Build # #37433

Build Type

push

local

Committed by

web-flow

Commit Message

Merge pull request #48513 from JuliaLang/jn/extend-once

ensure extension triggers are only run by the package that satified them

Run Details

60 of 60 new or added lines in 1 file covered. (100.0%)

72324 of 82360 relevant lines covered (87.81%)

31376331.4 hits per line

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

98.14

/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

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

## comparison ##

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

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

function ==(a::String, b::String)
    pointer_from_objref(a) == pointer_from_objref(b) && return true
    al = sizeof(a)
    return al == sizeof(b) && 0 == _memcmp(a, b, al)
end

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

length(s::String) = length_continued(s, 1, ncodeunits(s), ncodeunits(s))

@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)
    length_continued(s, i, j, c)
end

@inline function length_continued(s::String, i::Int, n::Int, c::Int)
    i < n || return c
    @inbounds b = codeunit(s, i)
    @inbounds 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

function isascii(s::String)
    @inbounds for i = 1:ncodeunits(s)
        codeunit(s, i) >= 0x80 && return false
    end
    return true
end

"""
    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"
```
"""
repeat(c::AbstractChar, r::Integer) = repeat(Char(c), r) # fallback
function repeat(c::Char, r::Integer)
    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	561✔
10	index::Integer
11	end
12	@noinline string_index_err(s::AbstractString, i::Integer) =	557✔
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)	149,437,510✔
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))	2,475,240✔
67	String(v::Vector{UInt8}) = ccall(:jl_array_to_string, Ref{String}, (Any,), v)	5,467,373✔
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)	1,201✔
80	p == C_NULL && throw(ArgumentError("cannot convert NULL to string"))	380,288✔
81	ccall(:jl_pchar_to_string, Ref{String}, (Ptr{UInt8}, Int), p, len)	380,287✔
82	end
83	function unsafe_string(p::Union{Ptr{UInt8},Ptr{Int8}})	21,312✔
84	p == C_NULL && throw(ArgumentError("cannot convert NULL to string"))	2,891,068✔
85	ccall(:jl_cstr_to_string, Ref{String}, (Ptr{UInt8},), p)	2,891,067✔
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))))	139,127,548✔
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))	530,268✔
99
100	unsafe_wrap(::Type{Vector{UInt8}}, s::String) = ccall(:jl_string_to_array, Ref{Vector{UInt8}}, (Any,), s)	6,407,055✔
101
102	Vector{UInt8}(s::CodeUnits{UInt8,String}) = copyto!(Vector{UInt8}(undef, length(s)), s)	731✔
103	Vector{UInt8}(s::String) = Vector{UInt8}(codeunits(s))	658✔
104	Array{UInt8}(s::String) = Vector{UInt8}(codeunits(s))	×
105
106	String(s::CodeUnits{UInt8,String}) = s.s	4✔
107
108	## low-level functions ##
109
110	pointer(s::String) = unsafe_convert(Ptr{UInt8}, s)	393,952,398✔
111	pointer(s::String, i::Integer) = pointer(s) + Int(i)::Int - 1	241,530,693✔
112
113	ncodeunits(s::String) = Core.sizeof(s)	253,847,872✔
114	codeunit(s::String) = UInt8	×
115
116	@inline function codeunit(s::String, i::Integer)	1,052,316✔
117	@boundscheck checkbounds(s, i)	223,396,467✔
118	b = GC.@preserve s unsafe_load(pointer(s, i))	223,396,467✔
119	return b	223,396,467✔
120	end
121
122	## comparison ##
123
124	_memcmp(a::Union{Ptr{UInt8},AbstractString}, b::Union{Ptr{UInt8},AbstractString}, len) =	6,457,622✔
125	ccall(:memcmp, Cint, (Ptr{UInt8}, Ptr{UInt8}, Csize_t), a, b, len % Csize_t) % Int
126
127	function cmp(a::String, b::String)	1✔
128	al, bl = sizeof(a), sizeof(b)	99,457✔
129	c = _memcmp(a, b, min(al,bl))	99,457✔
130	return c < 0 ? -1 : c > 0 ? +1 : cmp(al,bl)	182,345✔
131	end
132
133	function ==(a::String, b::String)	9,849,233✔
134	pointer_from_objref(a) == pointer_from_objref(b) && return true	9,849,233✔
135	al = sizeof(a)	9,774,793✔
136	return al == sizeof(b) && 0 == _memcmp(a, b, al)	9,774,793✔
137	end
138
139	typemin(::Type{String}) = ""	2✔
140	typemin(::String) = typemin(String)	1✔
141
142	## thisind, nextind ##
143
144	@propagate_inbounds thisind(s::String, i::Int) = _thisind_str(s, i)	178,141,360✔
145
146	# s should be String or SubString{String}
147	@inline function _thisind_str(s, i::Int)	97✔
148	i == 0 && return 0	89,767,499✔
149	n = ncodeunits(s)	89,739,990✔
150	i == n + 1 && return i	89,739,990✔
151	@boundscheck between(i, 1, n) \|\| throw(BoundsError(s, i))	89,717,548✔
152	@inbounds b = codeunit(s, i)	89,717,524✔
153	(b & 0xc0 == 0x80) & (i-1 > 0) \|\| return i	159,367,572✔
154	@inbounds b = codeunit(s, i-1)	20,067,476✔
155	between(b, 0b11000000, 0b11110111) && return i-1	20,067,476✔
156	(b & 0xc0 == 0x80) & (i-2 > 0) \|\| return i	15,659,909✔
157	@inbounds b = codeunit(s, i-2)	7,170,721✔
158	between(b, 0b11100000, 0b11110111) && return i-2	7,170,721✔
159	(b & 0xc0 == 0x80) & (i-3 > 0) \|\| return i	5,906,601✔
160	@inbounds b = codeunit(s, i-3)	1,027,051✔
161	between(b, 0b11110000, 0b11110111) && return i-3	1,027,051✔
162	return i	241,004✔
163	end
164
165	@propagate_inbounds nextind(s::String, i::Int) = _nextind_str(s, i)	13,619,070✔
166
167	# s should be String or SubString{String}
168	@inline function _nextind_str(s, i::Int)	14,239,096✔
169	i == 0 && return 1	14,239,096✔
170	n = ncodeunits(s)	14,218,223✔
171	@boundscheck between(i, 1, n) \|\| throw(BoundsError(s, i))	14,218,235✔
172	@inbounds l = codeunit(s, i)	14,218,211✔
173	(l < 0x80) \| (0xf8 ≤ l) && return i+1	14,218,211✔
174	if l < 0xc0	131,410✔
175	i′ = @inbounds thisind(s, i)	50,983✔
176	return i′ < i ? @inbounds(nextind(s, i′)) : i+1	30,216✔
177	end
178	# first continuation byte
179	(i += 1) > n && return i	101,194✔
180	@inbounds b = codeunit(s, i)	100,584✔
181	b & 0xc0 ≠ 0x80 && return i	100,584✔
182	((i += 1) > n) \| (l < 0xe0) && return i	93,340✔
183	# second continuation byte
184	@inbounds b = codeunit(s, i)	75,972✔
185	b & 0xc0 ≠ 0x80 && return i	75,972✔
186	((i += 1) > n) \| (l < 0xf0) && return i	72,339✔
187	# third continuation byte
188	@inbounds b = codeunit(s, i)	20,796✔
189	ifelse(b & 0xc0 ≠ 0x80, i, i+1)	20,796✔
190	end
191
192	## checking UTF-8 & ACSII validity ##
193
194	byte_string_classify(s::Union{String,Vector{UInt8},FastContiguousSubArray{UInt8,1,Vector{UInt8}}}) =	20,311✔
195	ccall(:u8_isvalid, Int32, (Ptr{UInt8}, Int), s, sizeof(s))
196	# 0: neither valid ASCII nor UTF-8
197	# 1: valid ASCII
198	# 2: valid UTF-8
199
200	isvalid(::Type{String}, s::Union{Vector{UInt8},FastContiguousSubArray{UInt8,1,Vector{UInt8}},String}) = byte_string_classify(s) ≠ 0	20,311✔
201	isvalid(s::String) = isvalid(String, s)	28✔
202
203	is_valid_continuation(c) = c & 0xc0 == 0x80	10✔
204
205	## required core functionality ##
206
207	@inline function iterate(s::String, i::Int=firstindex(s))	54,791✔
208	(i % UInt) - 1 < ncodeunits(s) \|\| return nothing	24,180,071✔
209	b = @inbounds codeunit(s, i)	14,871,513✔
210	u = UInt32(b) << 24	14,871,513✔
211	between(b, 0x80, 0xf7) \|\| return reinterpret(Char, u), i+1	28,826,126✔
212	return iterate_continued(s, i, u)	916,900✔
213	end
214
215	function iterate_continued(s::String, i::Int, u::UInt32)	916,900✔
216	u < 0xc0000000 && (i += 1; @goto ret)	916,974✔
217	n = ncodeunits(s)	916,826✔
218	# first continuation byte
219	(i += 1) > n && @goto ret	916,826✔
220	@inbounds b = codeunit(s, i)	916,818✔
221	b & 0xc0 == 0x80 \|\| @goto ret	916,818✔
222	u \|= UInt32(b) << 16	916,757✔
223	# second continuation byte
224	((i += 1) > n) \| (u < 0xe0000000) && @goto ret	916,757✔
225	@inbounds b = codeunit(s, i)	852,909✔
226	b & 0xc0 == 0x80 \|\| @goto ret	852,911✔
227	u \|= UInt32(b) << 8	852,907✔
228	# third continuation byte
229	((i += 1) > n) \| (u < 0xf0000000) && @goto ret	852,907✔
230	@inbounds b = codeunit(s, i)	273,592✔
231	b & 0xc0 == 0x80 \|\| @goto ret	273,593✔
232	u \|= UInt32(b); i += 1	547,182✔
233	@label ret	×
234	return reinterpret(Char, u), i	916,900✔
235	end
236
237	@propagate_inbounds function getindex(s::String, i::Int)	151,346✔
238	b = codeunit(s, i)	2,572,587✔
239	u = UInt32(b) << 24	2,572,587✔
240	between(b, 0x80, 0xf7) \|\| return reinterpret(Char, u)	5,142,019✔
241	return getindex_continued(s, i, u)	3,155✔
242	end
243
244	function getindex_continued(s::String, i::Int, u::UInt32)	3,155✔
245	if u < 0xc0000000	3,155✔
246	# called from `getindex` which checks bounds
247	@inbounds isvalid(s, i) && @goto ret	4✔
248	string_index_err(s, i)	1✔
249	end
250	n = ncodeunits(s)	3,151✔
251
252	(i += 1) > n && @goto ret	3,151✔
253	@inbounds b = codeunit(s, i) # cont byte 1	3,150✔
254	b & 0xc0 == 0x80 \|\| @goto ret	3,150✔
255	u \|= UInt32(b) << 16	3,149✔
256
257	((i += 1) > n) \| (u < 0xe0000000) && @goto ret	3,149✔
258	@inbounds b = codeunit(s, i) # cont byte 2	2,080✔
259	b & 0xc0 == 0x80 \|\| @goto ret	2,080✔
260	u \|= UInt32(b) << 8	2,080✔
261
262	((i += 1) > n) \| (u < 0xf0000000) && @goto ret	2,080✔
263	@inbounds b = codeunit(s, i) # cont byte 3	468✔
264	b & 0xc0 == 0x80 \|\| @goto ret	468✔
265	u \|= UInt32(b)	468✔
266	@label ret	×
267	return reinterpret(Char, u)	3,154✔
268	end
269
270	getindex(s::String, r::AbstractUnitRange{<:Integer}) = s[Int(first(r)):Int(last(r))]	4✔
271
272	@inline function getindex(s::String, r::UnitRange{Int})	1,458,776✔
273	isempty(r) && return ""	941,283✔
274	i, j = first(r), last(r)	71,795✔
275	@boundscheck begin	918,724✔
276	checkbounds(s, r)	918,706✔
277	@inbounds isvalid(s, i) \|\| string_index_err(s, i)	918,694✔
278	@inbounds isvalid(s, j) \|\| string_index_err(s, j)	918,695✔
279	end
280	j = nextind(s, j) - 1	918,717✔
281	n = j - i + 1	918,717✔
282	ss = _string_n(n)	918,717✔
283	GC.@preserve s ss unsafe_copyto!(pointer(ss), pointer(s, i), n)	918,717✔
284	return ss	918,717✔
285	end
286
287	length(s::String) = length_continued(s, 1, ncodeunits(s), ncodeunits(s))	813,763✔
288
289	@inline function length(s::String, i::Int, j::Int)	68,251✔
290	@boundscheck begin	241,906✔
291	0 < i ≤ ncodeunits(s)+1 \|\| throw(BoundsError(s, i))	241,906✔
292	0 ≤ j < ncodeunits(s)+1 \|\| throw(BoundsError(s, j))	241,910✔
293	end
294	j < i && return 0	241,902✔
295	@inbounds i, k = thisind(s, i), i	377,422✔
296	c = j - i + (i == k)	188,711✔
297	length_continued(s, i, j, c)	188,711✔
298	end
299
300	@inline function length_continued(s::String, i::Int, n::Int, c::Int)	896,956✔
301	i < n \|\| return c	1,237,931✔
302	@inbounds b = codeunit(s, i)	767,017✔
303	@inbounds while true	4,260,009✔
304	while true	47,850,715✔
305	(i += 1) ≤ n \|\| return c	48,616,961✔
306	0xc0 ≤ b ≤ 0xf7 && break	47,084,469✔
307	b = codeunit(s, i)	43,590,706✔
308	end	43,590,706✔
309	l = b	3,493,763✔
310	b = codeunit(s, i) # cont byte 1	3,493,763✔
311	c -= (x = b & 0xc0 == 0x80)	3,493,763✔
312	x & (l ≥ 0xe0) \|\| continue	3,493,763✔
313
314	(i += 1) ≤ n \|\| return c	3,094,983✔
315	b = codeunit(s, i) # cont byte 2	3,093,521✔
316	c -= (x = b & 0xc0 == 0x80)	3,093,521✔
317	x & (l ≥ 0xf0) \|\| continue	5,290,143✔
318
319	(i += 1) ≤ n \|\| return c	896,939✔
320	b = codeunit(s, i) # cont byte 3	896,859✔
321	c -= (b & 0xc0 == 0x80)	896,859✔
322	end	3,492,992✔
323	end
324
325	## overload methods for efficiency ##
326
327	isvalid(s::String, i::Int) = checkbounds(Bool, s, i) && thisind(s, i) == i	82,074,970✔
328
329	function isascii(s::String)	25,625✔
330	@inbounds for i = 1:ncodeunits(s)	51,243✔
331	codeunit(s, i) >= 0x80 && return false	375,363✔
332	end	704,784✔
333	return true	5,301✔
334	end
335
336	"""
337	repeat(c::AbstractChar, r::Integer) -> String
338
339	Repeat a character `r` times. This can equivalently be accomplished by calling
340	[`c^r`](@ref :^(::Union{AbstractString, AbstractChar}, ::Integer)).
341
342	# Examples
343	```jldoctest
344	julia> repeat('A', 3)
345	"AAA"
346	```
347	"""
348	repeat(c::AbstractChar, r::Integer) = repeat(Char(c), r) # fallback	1✔
349	function repeat(c::Char, r::Integer)	24,955✔
350	r == 0 && return ""	24,955✔
351	r < 0 && throw(ArgumentError("can't repeat a character $r times"))	11,988✔
352	u = bswap(reinterpret(UInt32, c))	11,984✔
353	n = 4 - (leading_zeros(u \| 0xff) >> 3)	11,984✔
354	s = _string_n(n*r)	12,090✔
355	p = pointer(s)	12,089✔
356	GC.@preserve s if n == 1	12,089✔
357	ccall(:memset, Ptr{Cvoid}, (Ptr{UInt8}, Cint, Csize_t), p, u % UInt8, r)	12,064✔
358	elseif n == 2	25✔
359	p16 = reinterpret(Ptr{UInt16}, p)	5✔
360	for i = 1:r	9✔
361	unsafe_store!(p16, u % UInt16, i)	14✔
362	end	19✔
363	elseif n == 3	20✔
364	b1 = (u >> 0) % UInt8	16✔
365	b2 = (u >> 8) % UInt8	16✔
366	b3 = (u >> 16) % UInt8	16✔
367	for i = 0:r-1	32✔
368	unsafe_store!(p, b1, 3i + 1)	45✔
369	unsafe_store!(p, b2, 3i + 2)	45✔
370	unsafe_store!(p, b3, 3i + 3)	45✔
371	end	45✔
372	elseif n == 4	4✔
373	p32 = reinterpret(Ptr{UInt32}, p)	4✔
374	for i = 1:r	8✔
375	unsafe_store!(p32, u, i)	8✔
376	end	12,097✔
377	end
378	return s	11,983✔
379	end

JuliaLang / julia / #37433

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