|
| From: | Taylan Kammer |
| Subject: | bug#72365: srfi-64: test-on-bad-end-name-simple is not allowed to raise an exception |
| Date: | Wed, 2 Oct 2024 23:07:01 +0200 |
| User-agent: | Mozilla Thunderbird |
Taylan Kammer <taylan.kammer@gmail.com> writes: Hi, sorry for taking so long to respond to your comments, work has been bit busy lately.
I took two months whereas you took two days, so I'm not going to complain. :D
On 30.07.2024 21:51, Tomas Volf wrote:Hello, I think I found a bug in (srfi srfi-64) module shipped with GNU Guile.Hi Tomas, Thanks for stress-testing the SRFI 64 spec & implementation and reporting all these discrepancies. :-)No problem, it was necessary during implementing my own version of SRFI-64. The full test suite is available as part of my library[2], the runner[3] is quite simple, so using it to test your implementation should be fairly easy. I think there are few more bugs than I reported, I lost the willpower to do the rest somewhere on the way. :/ 2: https://git.wolfsden.cz/guile-wolfsden/tree/tests/srfi-64 3: https://git.wolfsden.cz/guile-wolfsden/tree/build-aux/srfi64test-driver.scm
Do I understand correctly that this is an additional test suite
for testing SRFI-64 itself? Like the "meta test suite" shipped
with SRFI-64?
Is there a brief description somewhere on how to run it with Guile? Would be really neat if I can use it to further test my implementation.
Firstly, to reiterate some things I've already mentioned in the thread on bug 71300, just so it goes on record here as well: I have a SRFI 64 implementation of my own. I hope Guile will switch to it eventually because I find the upstream reference implementation to be somewhat unpleasant to work with. (It's monolithic, and not the cleanest code.)I plan to attempt to upstream my version as well, so I guess it is a race :)Until then, my implementation can be used by following these steps: 1. Cloning this repo: https://codeberg.org/taylan/scheme-srfis/ 2. Running Guile like so: GUILE_LOAD_PATH=/path/to/scheme-srfis/guile-srfi-64 guile (Replacing /path/to/scheme-srfis with the actual path to wherein the repo was cloned, of course.) Then, loading SRFI-64 the regular way should load my implementation rather than the one that ships with Guile (which is the reference implementation from the SRFI author).You can find my version here[0]. If you do not use Guix, building from tarball[1] might be easier. Contrary to your version, mine is available as (wolfsden srfi srfi-64). 0: https://git.wolfsden.cz/guile-wolfsden/ 1: https://wolfsden.cz/project/guile-wolfsden.html
Your implementation seems written specifically with Guile in mind, which is a big plus I guess.
Mine was written with R7RS in mind, and has compatibility shims
for Guile as well as other Scheme implementations like Kawa (shims
adopted from the original implementation). This makes it hacky in
some parts, like when it comes to how it gathers source code
information for file name & line number reporting. It also
lacks docstrings, doesn't make use of Guile features like
`define*` and so on...
If the quality of the implementations is the same or higher, in terms of observable behavior, then it should be preferred for Guile, I think. If I find the time, I'll see if I can use your implementation to run some of my test suites, like the bytestructures test suite, and report if I notice any issues.
Unfortunately, I'm not motivated to work on the implementation that's in Guile, because I find it too cumbersome to navigate its code and the unclean coding practices too distracting.While in principle I agree, let us not be too harsh, the implementation is really old. I assume coding practices available at the time to achieve portability were bit different. The implementation even considers SRFI-9 optional, these days I think that would be considered bit absurd.
That's true. I don't mean to insult the author or anything. It is what it is.
The spec is very sparse on what the simple test runner does, so I'm not sure if the intention is to imply that it does nothing other than what's stated.I am not sure how to read the spec regarding this. But in my readingCreates a new simple test-runner, that prints errors and a summary on the standard output port.is clear enough. Same way I would think (for example) reporting a telemetry (e.g. on number of tests executed) would violate the spec.In one case, the reference implementation clearly violates the specification: The simple test runner uses the `aux` field which the spec claims it doesn't use. (My implementation fixes this.) However, in this case it's not that clear-cut. In this case, I think raising an error is good default behavior, since the mismatched end name indicates a problem with the test suite itself rather than the code being tested. If it poses a problem to the user, one can override that callback with the `test-runner-on-bad-end-name!` setter. What do you think?I agree that raising an error is good behavior. However I do not think that on-bad-end-name-function is a place where to do it. In my opinion the name mismatch is a hard error, in my implementation subclass of &programming-error[4]. If I am writing new test runner, the specification does not mention that raising the error is *my* responsibility, just that test-end will signal an error. To rephrase that: test-end is mandated to signal error, but custom test runner has no provision requiring it to do it in on-bad-end-name-function. Hence I believe test-end needs to be the one to signal the error.
Makes sense I guess. I've generally tried to imitate the
reference implementation's behavior as closely as possible in such
matters, worrying that there might be code out there that relies
on its various quirks, but maybe I'm being too paranoid.
I don't have a strong opinion either way. The number of people, who want to write a test runner that does something special on bad-end-name (something other than raise an error), is probably very small.
- Making `test-end` itself raise an error would probably be most convenient, so test runner authors don't have to take care of it.
- But if `test-end` doesn't do it, it's not a big deal either IMO, because all they would need to do is to call `(test-runner-on-bad-end-name! my-runner test-on-bad-end-name-simple)` to make their custom runner raise an error as well. (And, if they want to do something before, they can use a procedure that ends with the call `(test-on-bad-end-name-simple ...)`.)
The latter is my preference, because enabling the behavior via a single line of code is easy, whereas disabling it would be difficult / impossible if `test-end` were to be hardcoded to raise an error. But if a SRFI-64 implementation made its `test-end` always raise an error, it probably wouldn't anyone in practice, so I wouldn't see it as a real problem.
However! That does not make on-bad-end-name-function useless. The specification does not mandate *how* the error signaled by test-end should look like, hence there is no *portable* way to detect it. Custom runner, if it needs to report name mismatch specially, can just produce specific log line in the callback (or even signal its own exception first before test-end does). 4: https://git.wolfsden.cz/guile-wolfsden/tree/wolfsden/srfi/srfi-64.scm#n960 Let me know what you think.
Right, it doesn't make it useless per se. It could still be called before `test-end` raises an error.
Have a nice day, Tomas
Likewise!
- Taylan
| [Prev in Thread] | Current Thread | [Next in Thread] |