v8  3.25.30(node0.11.13)
V8 is Google's open source JavaScript engine
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Pages
bignum-dtoa.h
Go to the documentation of this file.
1 // Copyright 2010 the V8 project authors. All rights reserved.
2 // Redistribution and use in source and binary forms, with or without
3 // modification, are permitted provided that the following conditions are
4 // met:
5 //
6 // * Redistributions of source code must retain the above copyright
7 // notice, this list of conditions and the following disclaimer.
8 // * Redistributions in binary form must reproduce the above
9 // copyright notice, this list of conditions and the following
10 // disclaimer in the documentation and/or other materials provided
11 // with the distribution.
12 // * Neither the name of Google Inc. nor the names of its
13 // contributors may be used to endorse or promote products derived
14 // from this software without specific prior written permission.
15 //
16 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
17 // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
18 // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
19 // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
20 // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
21 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
22 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
26 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27 
28 #ifndef V8_BIGNUM_DTOA_H_
29 #define V8_BIGNUM_DTOA_H_
30 
31 namespace v8 {
32 namespace internal {
33 
34 enum BignumDtoaMode {
35  // Return the shortest correct representation.
36  // For example the output of 0.299999999999999988897 is (the less accurate but
37  // correct) 0.3.
38  BIGNUM_DTOA_SHORTEST,
39  // Return a fixed number of digits after the decimal point.
40  // For instance fixed(0.1, 4) becomes 0.1000
41  // If the input number is big, the output will be big.
42  BIGNUM_DTOA_FIXED,
43  // Return a fixed number of digits, no matter what the exponent is.
44  BIGNUM_DTOA_PRECISION
45 };
46 
47 // Converts the given double 'v' to ASCII.
48 // The result should be interpreted as buffer * 10^(point-length).
49 // The buffer will be null-terminated.
50 //
51 // The input v must be > 0 and different from NaN, and Infinity.
52 //
53 // The output depends on the given mode:
54 // - SHORTEST: produce the least amount of digits for which the internal
55 // identity requirement is still satisfied. If the digits are printed
56 // (together with the correct exponent) then reading this number will give
57 // 'v' again. The buffer will choose the representation that is closest to
58 // 'v'. If there are two at the same distance, than the number is round up.
59 // In this mode the 'requested_digits' parameter is ignored.
60 // - FIXED: produces digits necessary to print a given number with
61 // 'requested_digits' digits after the decimal point. The produced digits
62 // might be too short in which case the caller has to fill the gaps with '0's.
63 // Example: toFixed(0.001, 5) is allowed to return buffer="1", point=-2.
64 // Halfway cases are rounded up. The call toFixed(0.15, 2) thus returns
65 // buffer="2", point=0.
66 // Note: the length of the returned buffer has no meaning wrt the significance
67 // of its digits. That is, just because it contains '0's does not mean that
68 // any other digit would not satisfy the internal identity requirement.
69 // - PRECISION: produces 'requested_digits' where the first digit is not '0'.
70 // Even though the length of produced digits usually equals
71 // 'requested_digits', the function is allowed to return fewer digits, in
72 // which case the caller has to fill the missing digits with '0's.
73 // Halfway cases are again rounded up.
74 // 'BignumDtoa' expects the given buffer to be big enough to hold all digits
75 // and a terminating null-character.
76 void BignumDtoa(double v, BignumDtoaMode mode, int requested_digits,
77  Vector<char> buffer, int* length, int* point);
78 
79 } } // namespace v8::internal
80 
81 #endif // V8_BIGNUM_DTOA_H_
v8::internal::mode
enable upcoming ES6 features enable harmony block scoping enable harmony enable harmony proxies enable harmony generators enable harmony numeric enable harmony string enable harmony math functions harmony_scoping harmony_symbols harmony_collections harmony_iteration harmony_strings harmony_scoping harmony_maths tracks arrays with only smi values Optimize object Array DOM strings and string pretenure call new trace pretenuring decisions of HAllocate instructions track fields with only smi values track fields with heap values track_fields track_fields Enables optimizations which favor memory size over execution speed use string slices optimization filter maximum number of GVN fix point iterations use function inlining use allocation folding eliminate write barriers targeting allocations in optimized code maximum source size in bytes considered for a single inlining maximum cumulative number of AST nodes considered for inlining crankshaft harvests type feedback from stub cache trace check elimination phase hydrogen tracing filter trace hydrogen to given file name trace inlining decisions trace store elimination trace all use positions trace global value numbering trace hydrogen escape analysis trace the tracking of allocation sites trace map generalization environment for every instruction deoptimize every n garbage collections put a break point before deoptimizing deoptimize uncommon cases use on stack replacement trace array bounds check elimination perform array index dehoisting use load elimination use store elimination use constant folding eliminate unreachable code number of stress runs when picking a function to watch for shared function not JSFunction itself flushes the cache of optimized code for closures on every GC functions with arguments object maximum number of escape analysis fix point iterations allow uint32 values on optimize frames if they are used only in safe operations track concurrent recompilation artificial compilation delay in ms concurrent on stack replacement do not emit check maps for constant values that have a leaf deoptimize the optimized code if the layout of the maps changes number of stack frames inspected by the profiler percentage of ICs that must have type info to allow optimization extra verbose compilation tracing generate extra emit comments in code disassembly enable use of SSE3 instructions if available enable use of CMOV instruction if available enable use of VFP3 instructions if available enable use of NEON instructions if enable use of SDIV and UDIV instructions if enable loading bit constant by means of movw movt instruction enable unaligned accesses for enable use of d16 d31 registers on ARM this requires VFP3 force all emitted branches to be in long mode(MIPS only)") DEFINE_string(expose_natives_as
v8::internal::BignumDtoa
void BignumDtoa(double v, BignumDtoaMode mode, int requested_digits, Vector< char > buffer, int *length, int *decimal_point)
Definition: bignum-dtoa.cc:92