File: can-execute.txt

package info (click to toggle)
llvm-toolchain-21 1%3A21.1.6-3
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 2,245,028 kB
  • sloc: cpp: 7,619,726; ansic: 1,434,018; asm: 1,058,748; python: 252,740; f90: 94,671; objc: 70,685; lisp: 42,813; pascal: 18,401; sh: 8,601; ml: 5,111; perl: 4,720; makefile: 3,675; awk: 3,523; javascript: 2,409; xml: 892; fortran: 770
file content (22 lines) | stat: -rw-r--r-- 823 bytes parent folder | download | duplicates (10)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
REQUIRES: can-execute

This tests that we abstract two peculiarities of unix in can_execute:

* Directories are executable, but we don't want to try to execute them.
* For shell scripts, we also need to be able to read them.

The PATH is constructed such that 'not' will first find a directory named
TestProg, then a file with executable bit but not readable and finally a
shell script which always returns false, which is what it actually tries to
execute.

If we want, it is probably OK to change the semantics of can_execute and this
test, but for now this test serves as a reminder to audit all the callers if
we do that.

RUN: rm -rf %t && mkdir -p %t
RUN: cp -f %S/Inputs/TestProg/TestProg %t/TestProg
RUN: chmod 111 %t/TestProg
RUN: export PATH=%S/Inputs:%t:%S/Inputs/TestProg:$PATH
RUN: not TestProg
RUN: rm -rf %t