Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-20 Thread Amirouche Boubekki
Thanks for the feedback! Also, if you are interested in r7rs in guile you might be interested in `r7rs-wip` branch of guile Le dim. 20 janv. 2019 à 14:50, Jéssica Milaré a écrit : > > > Em dom, 20 de jan de 2019 às 10:58, Amirouche Boubekki < > amirouche.boube...@gmail.com> escreveu: > >> Than

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-20 Thread Jéssica Milaré
Em dom, 20 de jan de 2019 às 10:58, Amirouche Boubekki < amirouche.boube...@gmail.com> escreveu: > Thanks! > > I have done mostly a cosmetic review of your code. Here is the summary: > Your comments are very good, I'll do most of the suggested changes. > - There is two big procedures that I thi

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-20 Thread Amirouche Boubekki
Thanks! I have done mostly a cosmetic review of your code. Here is the summary: - There is two big procedures that I think should be split ` get-hash-functions` and `hash-table` to help readability - I am not very fond of generic-hash-table macros, I would rather use apply / call-with-values, if

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-17 Thread Jéssica Milaré
*Note Em qui, 17 de jan de 2019 às 19:13, Jéssica Milaré escreveu: > Done. > > Not that there are many commits. The first commit only fixes bug 33827 of > SRFI 69, and SRFI 69 is reimplemented in the third commit, which pretty > much erases the changes of the first commit. > > https://github.com

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-17 Thread Jéssica Milaré
Done. Not that there are many commits. The first commit only fixes bug 33827 of SRFI 69, and SRFI 69 is reimplemented in the third commit, which pretty much erases the changes of the first commit. https://github.com/jessymilare/guile/pull/1 Em qua, 16 de jan de 2019 às 22:19, Amirouche Boubekki

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-16 Thread Arne Babenhauserheide
Jéssica Milaré writes: > Finally finished the libraries. > > SRFI-125 is implemented. Cool — thank you! I cannot test it right now, but I think having SRFI-125 will be very useful. Best wishes, Arne -- Unpolitisch sein heißt politisch sein ohne es zu merken signature.asc Description: PGP s

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-16 Thread Amirouche Boubekki
I will try to look at the commits. One last niptick, can you push guile master branch to you repository, and make a pull request against it so that I can directly comment in the pull requests. please. It's easier to way to see the overall changes in particular. Le dim. 13 janv. 2019 à 23:52, Jéssi

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-13 Thread Jéssica Milaré
Finally finished the libraries. SRFI-125 is implemented. I took the liberty to create a procedure scm_hash_n_items in 'libguile/hashtab.c' that works with both normal and weak hash tables, so the GENERIC-HASH-TABLES module don't need to keep track of the hash table size anymore. `make check` runs

Re: Proposal of a better hash table implementation based on SRFI 125

2019-01-12 Thread Jéssica Milaré
(It seems I mistakenly responded only to a personal e-mail, sorry, responded now to everyone in the list with updates). So here are the news: I've created a module called (ice-9 generic-hash-tables), which is similar to SRFI-125, and used it to implement SRFI-69, (RNRS HASHTABLES), SRFI-126 and S

Re: Proposal of a better hash table implementation based on SRFI 125

2018-12-30 Thread Amirouche Boubekki
Le 2018-12-28 17:11, Jéssica Milaré a écrit : Hello, As I said in a previous e-mail, currently SRFI-69 is broken for weak hash tables - and I've sent a patch to fix it. However, I think there are many other problems with current implementation of hash tables. There are guile standard hash tables