Lectures

A lecture about a “mindset shift” in I/O: a file is first and foremost bytes, and text is only an interpretation. We break down when readText() is not suitable and why binary data needs ByteArray, how the Byte type works (range - 128 .. 127), why a buffer is needed, and why arrays are compared via contentEquals rather than the == operator. Plus, a bridge between the String and byte worlds via encodeToByteArray()/ decodeToString().

We go through the shortest path “read file → write file” for binary data in Kotlin/JVM: the extension functions File.readBytes() and File.writeBytes(bytes), working with the ByteArray type, validating the result by size ( length()) and byte-by-byte comparison ( contentEquals()), as well as the limitations of the “read all at once” approach and typical mistakes (for example, comparing arrays via == and reading a binary file as text).

We break down low-level file work on the JVM via byte streams: how to read and write data using FileInputStream/ FileOutputStream, why InputStream.read() returns Int and what -1 (EOF) means, how to correctly connect the result of read(buffer) with write(buffer, 0, n), and why use {} is a must-have habit for closing resources.

We break down why copying “byte by byte” via read()/ write() is usually slow, how “our” buffer ByteArray differs from wrappers like BufferedInputStream/ BufferedOutputStream, what the contract of read(buffer) is (the n value and -1 on EOF), and what the canonical chunk-copy loop with output.write(buffer, 0, n) looks like. At the end: choosing a buffer size, a data flow diagram (mermaid), and common mistakes.

We look at ZIP on the JVM as a streaming “container” of entries: what ZipEntry stores (primarily the name and the directory flag), how to correctly write an archive via ZipOutputStream ( putNextEntry() → write bytes → closeEntry()), how to read via ZipInputStream (iterate nextEntry) and unpack safely, protecting against paths like ../../secret.txt. At the end — a mini CLI and typical mistakes.