Re: [GSoC][PATCH] t: migrate helper/test-urlmatch-normalization to unit tests
- From
Ghanshyam Thakkar <shyamthakkar001@gmail.com>
- Date
- Jul 22, 2024, 12:54 UTC
- Message-ID
- <D2W30F19QN7J.146CV16OMTO6M@gmail.com>
- In-Reply-To
- <CAOLa=ZQ8fuCksk+CGP2thMqKnT_KME4goLwEt6hKYvt+Gbn5tw@mail.gmail.com>
Karthik Nayak <karthik.188@gmail.com> wrote:
Show 31 quoted lines
> Ghanshyam Thakkar <shyamthakkar001@gmail.com> writes:
>
> [snip]
>
> > +static void compare_normalized_urls(const char *url1, const char *url2,
> > + size_t equal)
> [ 5 more citation lines. Click/Enter to show. ]
> > +{
> > + char *url1_norm = url_normalize(url1, NULL);
> > + char *url2_norm = url_normalize(url2, NULL);
> > +
> > + if (equal) {
> > + if (!check_str(url1_norm, url2_norm))
> > + test_msg("input url1: %s\n input url2: %s", url1,
> > + url2);
>
> check_str() checks and prints the values if they don't match, so here
> since the normalized urls will be printed by check_str(), we print the
> input urls. Makes sense.
>
> > + } else if (!check_int(strcmp(url1_norm, url2_norm), !=, 0))
> > + test_msg(" url1_norm: %s\n url2_norm: %s\n"
> > + " input url1: %s\n input url2: %s",
> > + url1_norm, url2_norm, url1, url2);
>
> Here we use strcmp and hence, it won't print the normalized urls, so we
> also print them. This is because we want to make sure they are not
> equal.
>
> I don't understand why there is inconsistent spacing in this message
> though.That is for alignment purposes, so the ':' matches vertically between them. I.e.
# url1_norm: https://@x.y/%5E url2_norm: https://x.y/%5E input url1: https://@x.y/^ input url2: httpS://x.y:443/^ Thanks.
> Apart from this small question, the patch looks great! > > Thanks