Files
java-core-examples/bufferedreader
..

bufferedreader

Companion code for the ankurm.com post "Java BufferedReader: Read a File Line by Line (vs Scanner, Files.lines and readAllLines)." Module bufferedreader in java-core-examples. The explanation lives in the post; this module holds the runnable evidence and the captured output.

Versions

Component Version
JDK 25.0.5+8 (Temurin, LTS)
JUnit Jupiter 5.11.0
OS Linux 6.18, 2 vCPU x86-64 VM (timings are indicative, not a leaderboard)
Big file 1,000,000,064 bytes, 9,623,347 ASCII log lines, generated then deleted by the script

Quickstart

export JDK25_HOME=/path/to/jdk-25
./scripts/run-all.sh                      # needs ~1.3 GB free in /tmp/bufferedreader-data and ~6 GB RAM; regenerates output/
BIG_BYTES=200000000 ./scripts/run-all.sh  # smaller "big" file if disk or RAM is tight
mvn -q -pl bufferedreader -am test        # 10 assertions, tiny files only

What is in here

File Shows Output
TestData deterministic generator: ~100-byte log lines, or one giant line with no newline 02
LineCount BufferedReader, Scanner, Files.lines, Files.readAllLines on the same file: which survive -Xmx64m, what readAllLines needs, time and GC 03, 04
ReadCalls read calls that reach the underlying stream, and per-character vs readLine() time 05
LongLineDemo readLine() on one 200 MB line at -Xmx128m vs chunked read(char[]) 06
CharsetDemo invalid UTF-8 through five readers, the JDK 18+ default charset, -Dfile.encoding=COMPAT, and a UTF-8 BOM 07, 08
LineEndingsDemo \n, \r\n, \r, final newline, empty file: what counts as a line 09
ScannerSwallowsDemo a reader that fails mid-file: BufferedReader throws, Scanner looks like end of file 10
ReadLineContractTest 10 assertions behind the claims above 11

01 records the environment. Timings move with whatever else the machine is doing; counts, exceptions, line totals and heap outcomes do not.

License

MIT - see LICENSE.