package jstream

import (
	"unicode/utf8"
)

type scratch struct {
	data []byte
	fill int
}

// reset scratch buffer
func (s *scratch) reset() { s.fill = 0 }

// bytes returns the written contents of scratch buffer
func (s *scratch) bytes() []byte { return s.data[0:s.fill] }

// grow scratch buffer
func (s *scratch) grow() {
	ndata := make([]byte, cap(s.data)*2)
	copy(ndata, s.data)
	s.data = ndata
}

FROM scratch

COPY minio /minio

    ByteBuffer scratch = scratch();
    scratch.putLong(l);
    return update(scratch, Longs.BYTES);
  }

  @Override
  @CanIgnoreReturnValue
  public Hasher putChar(char c) {
    ByteBuffer scratch = scratch();
    scratch.putChar(c);
    return update(scratch, Chars.BYTES);
  }

  private ByteBuffer scratch() {
    if (scratch == null) {

			}
			c = d.next()
		}
	}

scan_esc:
	switch c {
	case '"', '\\', '/', '\'':
		d.scratch.add(c)
	case 'u':
		goto scan_u
	case 'b':
		d.scratch.add('\b')
	case 'f':
		d.scratch.add('\f')
	case 'n':
		d.scratch.add('\n')
	case 'r':
		d.scratch.add('\r')
	case 't':
		d.scratch.add('\t')
	default:
		return "", d.mkError(ErrSyntax, "in string escape code")
	}
	c = d.next()

### Target language

Translate to Turkish (Türkçe).

Language code: tr.

### Core principle

Don't translate word-by-word. Rewrite naturally in Turkish as if writing the doc from scratch. Preserve meaning, but prioritize fluency over literal accuracy.

### Grammar and tone

- Use instructional Turkish, consistent with existing Turkish docs.
- Use imperative/guide language (e.g. "açalım", "gidin", "kopyalayalım", "bir bakalım").

* loop (as in async loop): цикл событий
* Machine Learning: Машинное обучение
* Deep Learning: Глубокое обучение
* callback hell: callback hell (clarify as `ад обратных вызовов`)
* on the fly: на лету
* scratch the surface: поверхностно ознакомиться
* tip: совет (or `подсказка` depending on context)
* Pydantic model: Pydantic-модель (`модель Pydantic` and `Pydantic модель` are also fine)
* declare: объявить

	return p.ctxt.PosTable.XPos(src.MakePos(p.lex.Base(), uint(p.lineNum), 0))
}

func (p *Parser) Parse() (*obj.Prog, bool) {
	scratch := make([][]lex.Token, 0, 3)
	for {
		word, cond, operands, ok := p.line(scratch)
		if !ok {
			break
		}
		scratch = operands

		if p.pseudo(word, operands) {
			continue
		}
		i, present := p.arch.Instructions[word]
		if present {

                metadata = new Metadata(new MetadataStaxReader().read(input, false));
            }
        }

        boolean changed;

        // If file could not be found or was not valid, start from scratch
        if (metadata == null) {
            metadata = this.metadata;

            changed = true;
        } else {
            changed = metadata.merge(this.metadata);
        }

## Build a Docker Image for FastAPI { #build-a-docker-image-for-fastapi }

Okay, let's build something now! 🚀

I'll show you how to build a **Docker image** for FastAPI **from scratch**, based on the **official Python** image.

This is what you would want to do in **most cases**, for example:

* Using **Kubernetes** or similar tools
* When running on a **Raspberry Pi**

            + positionFrac * (dataset[positionFloor + 1] - dataset[positionFloor]);
      }
    }
  },

  /**
   * Uses quickselect. When calculating multiple quantiles, each quickselect starts from scratch.
   */
  QUICKSELECT {

    @Override
    double singleQuantile(int index, int scale, double[] dataset) {
      long numerator = (long) index * (dataset.length - 1);