v/vlib/compiler/tests/repl
joe-conigliaro 59378dce46 compiler/vlib: replace substr/left/right with `[start..end]` everywhere 2019-10-27 10:03:15 +03:00
..
runner compiler/vlib: replace substr/left/right with `[start..end]` everywhere 2019-10-27 10:03:15 +03:00
.gitattributes
.gitignore
README.md
arr_decl.repl
chained_fields.correct.repl
chained_fields.repl
conditional_blocks.repl
default_printing.repl
empty_struct.repl
error.repl parser: undefined error 2019-10-20 20:47:27 +03:00
error_nosave.repl parser: undefined error 2019-10-20 20:47:27 +03:00
function.repl
immutable_len_fields.repl
interpolation.repl
multiple_decl.repl
multiple_println.repl
naked_strings.repl
newlines.repl
nomain.repl
nothing.repl
println.repl
repl_test.v escape os.exec('"$vexec"') to support directories with spaces 2019-10-16 02:49:51 +03:00
run.v tools: new tool to extracts function names declared in V files 2019-10-21 14:14:28 +03:00
var_decl.repl

README.md

V REPL Tests Script

How to write a new test

  • Create a new file named *.repl
  • Write the input to be given to REPL
  • Add ===output===
  • Write the output expected

Notes

Keep in mind, that the way V repl works for now, every non empty line would cause a new recompilation of the entire repl content that was collected so far.

Longer REPL files would cause measurably longer recompilation/testing times.

Also, longer repl files would be slower to debug when they fail, It is better to have several smaller files vs one huge REPL file.

Example :

a := 1
println(a)
===output===
1