defaultRebase
Function parameters
Parameters
- w:*Writer
- preserve:usize
- minimum_len:usize
Type definitions in this namespace
Types
- Allocating
- Maintains `Writer` state such that it writes to the unused capacity of an
Writes to `buffer` and returns `error.WriteFailed` when it is full.
Functions
- fixed
- Writes to `buffer` and returns `error.WriteFailed` when it is full.
- buffered
- Returns the contents not yet drained.
- writeVec
- If the total number of bytes of `data` fits inside `unusedCapacitySlice`,
- writeSplat
- If the number of bytes to write based on `data` and `splat` fits inside
- writeSplatHeader
- Returns how many bytes were consumed from `header` and `data`.
- writeSplatHeaderLimit
- Equivalent to `writeSplatHeader` but writes at most `limit` bytes.
- flush
- Drains all remaining buffered data.
- defaultFlush
- Repeatedly calls `VTable.drain` until `end` is zero.
- noopFlush
- Does nothing.
- writableArray
- Asserts the provided buffer has total capacity enough for `len`.
- writableSlice
- Asserts the provided buffer has total capacity enough for `len`.
- writableSliceGreedy
- Asserts the provided buffer has total capacity enough for `minimum_len`.
- writableSliceGreedyPreserve
- Asserts the provided buffer has total capacity enough for `minimum_len`
- writableSlicePreserve
- Asserts the provided buffer has total capacity enough for `len`.
- advance
- After calling `writableSliceGreedy`, this function tracks how many bytes
- writeVecAll
- The `data` parameter is mutable because this function needs to mutate the
- writeSplatAll
- The `data` parameter is mutable because this function needs to mutate the
- writeAll
- Calls `drain` as many times as necessary such that all of `bytes` are
- Renders fmt string with args, calling `writer` with slices of bytes.
- writeByte
- Calls `drain` as many times as necessary such that `byte` is transferred.
- writeBytePreserve
- When draining the buffer, ensures that at least `preserve` bytes
- splatByteAll
- Writes the same byte many times, performing the underlying write call as
- splatByte
- Writes the same byte many times, allowing short writes.
- splatBytesAll
- Writes the same slice many times, performing the underlying write call as
- splatBytes
- Writes the same slice many times, allowing short writes.
- writeInt
- Asserts the `buffer` was initialized with a capacity of at least `@sizeOf(T)` bytes.
- writeStruct
- The function is inline to avoid the dead code in case `endian` is
- sendFile
- Unlike `writeSplat` and `writeVec`, this function will call into `VTable`
- sendFileHeader
- Returns how many bytes from `header` and `file_reader` were consumed.
- sendFileReading
- Asserts nonzero buffer capacity and nonzero `limit`.
- sendFileAll
- Number of bytes logically written is returned.
- sendFileReadingAll
- Equivalent to `sendFileAll` but uses direct `pread` and `read` calls on
- printValue
- Asserts `buffer` capacity of at least 2 if `value` is a union.
- printIntAny
- In general, prefer `printInt` to avoid generic explosion.
- printFloat
- Uses a larger stack buffer; asserts mode is decimal or scientific.
- printFloatHexOptions
- Uses a smaller stack buffer; asserts mode is not decimal or scientific.
- printByteSize
- Format option `precision` is ignored when `value` is less than 1kB
- printDuration
- Writes number of nanoseconds according to its signed magnitude:
- writeUleb128
- Write a single unsigned integer as LEB128 to the given writer.
- writeSleb128
- Write a single signed integer as LEB128 to the given writer.
- writeLeb128
- Write a single integer as LEB128 to the given writer.
- consume
- Removes the first `n` bytes from `buffer` by shifting buffer contents,
- consumeAll
- Shortcut for setting `end` to zero and returning zero.
- unimplementedSendFile
- For use when the `Writer` implementation can cannot offer a more efficient
- fixedDrain
- When this function is called it usually means the buffer got full, so it's
- Hashed
- Provides a `Writer` implementation based on calling `Hasher.update`, sending
- Hashing
- Provides a `Writer` implementation based on calling `Hasher.update`,
Error sets in this namespace
Error Sets
= .{ .vtable = &.{ .drain = failingDrain, .sendFile = failingSendFile, .rebase = failingRebase, }, .buffer = &.{}, }
Values
- failing
- = .{ .vtable = &.{ .drain = failingDrain, .sendFile = failingSendFile, .rebase = failingRebase, }, .buffer = &.{}, }
Source
Implementation
pub fn defaultRebase(w: *Writer, preserve: usize, minimum_len: usize) Error!void {
while (w.buffer.len - w.end < minimum_len) {
{
// TODO: instead of this logic that "hides" data from
// the implementation, introduce a seek index to Writer
const preserved_head = w.end -| preserve;
const preserved_tail = w.end;
const preserved_len = preserved_tail - preserved_head;
w.end = preserved_head;
defer w.end += preserved_len;
assert(0 == try w.vtable.drain(w, &.{""}, 1));
assert(w.end <= preserved_head + preserved_len);
@memmove(w.buffer[w.end..][0..preserved_len], w.buffer[preserved_head..preserved_tail]);
}
// If the loop condition was false this assertion would have passed
// anyway. Otherwise, give the implementation a chance to grow the
// buffer before asserting on the buffer length.
assert(w.buffer.len - preserve >= minimum_len);
}
}