Hi Ghansyam,
I just wanted to share two comments based on what I observed from the
recent changes.
On 13/08/24 22:54, Ghanshyam Thakkar wrote:
>
index 0000000000..e0dd50dc11
--- /dev/null
+++ b/t/unit-tests/t-urlmatch-normalization.c
@@ -0,0 +1,294 @@
+
+static void check_normalized_url_from_file(const char *file, const char *expect)
+{
+ struct strbuf content = STRBUF_INIT, path = STRBUF_INIT;
+ char *cwd_basename;
+
+ if (!check_int(strbuf_getcwd(&path), ==, 0))
+ return;
+
+ cwd_basename = basename(path.buf);
+ if (!check(!strcmp(cwd_basename, "t") || !strcmp(cwd_basename, "bin"))) {
I think comparing blindly for "bin" would not be an ideal thing to do as
this would let other locations which have the "bin" basename to be slip
through. For instance, running the test from "perl/blib/bin" would
result in this check letting it through. I suppose we could need a bit
more specific check that ensures that the test is indeed running from
"t/unit-tests/bin".
... snip ...
>
+static void t_url_high_bit(void)
+{
+ check_normalized_url_from_file("url-1",
+ "x://q/%01%02%03%04%05%06%07%08%0E%0F%10%11%12");
+ check_normalized_url_from_file("url-2",
+ "x://q/%13%14%15%16%17%18%19%1B%1C%1D%1E%1F%7F");
When we run the unit-test binary from other directories, the error
message is thrown appropriately. But it seems to be printed for every
test case which seems a bit spammy. I suppose it might be helpful to do
it once before actually running the test cases and skip the cases when
we realize the binary is being run from a different directory.
--
Sivaraam
