Re: MySQL/MariaDB Boolean Type

[Jaft]

I'd been using commit e50772d19322a27425f85485c38306c8b093d72d, as I thought it should be current-enough, but, to be certain, I've just changed to the latest commit (a6a6e79b19533e1de31a31ef680ae1a1b969826a). When trying to run the migration

(migrate-create (create-table 'REVISED_WORDS
                              '(REVISED_WORD_ID  auto       (#:primary-key))
                              '(REVISED_WORD     char-field (#:not-null    #:unique #:maxlen  200))
                              '(NEW_SPELLING     char-field (#:not-null    #:unique #:maxlen  200))
                              '(TAKES_PRECEDENCE boolean    (#:not-null             #:default #t))
                              '(NOTES            text       (#:not-null             #:default ""))))

Artanis throws, when I run art migrate create REVISED_WORDS, the error

Creating table `REVISED_WORDS'......Backtrace:
In ice-9/boot-9.scm:
  1752:10 12 (with-exception-handler _ _ #:unwind? _ # _)
In unknown file:
          11 (apply-smob/0 #<thunk 7f4dd30af540>)
In ice-9/boot-9.scm:
    724:2 10 (call-with-prompt _ _ #<procedure default-prompt-handle…>)
In ice-9/eval.scm:
    619:8  9 (_ #(#(#<directory (guile-user) 7f4dd30b2c80>)))
In /gnu/store/6vzx5d827di2h7mfcvmc3wwk8j4iia1s-artanis-latest-0.5.1-3.a6a6e79/bin/.art-real:
    42:12  8 (_ _ _)
In artanis/mvc/migration.scm:
     95:4  7 (create-table REVISED_WORDS . _)
In artanis/fprm.scm:
   411:18  6 (_ REVISED_WORDS ((REVISED_WORD_ID serial (# # #)) # …) …)
In ice-9/boot-9.scm:
   222:29  5 (map1 ((REVISED_WORD_ID serial (#:no-edit # #)) (# …) …))
   222:29  4 (map1 ((REVISED_WORD varchar 200 (#:not-null #)) (# …) …))
   222:29  3 (map1 ((NEW_SPELLING varchar 200 (#:not-null #)) (# …) …))
   222:17  2 (map1 ((TAKES_PRECEDENCE boolean (#:not-null # #t)) (…)))
In artanis/fprm.scm:
   401:49  1 (->types _)
   153:10  0 (->mysql-type _ . _)

artanis/fprm.scm:153:10: In procedure ->mysql-type:
Throw to key `artanis-err' with args `(500 #<procedure ->mysql-type (name . args)> "Invalid type name `~a'" boolean)'.

I figured the error would still be present in the most recent work as, when I checked the ->mysql-type function earlier, I'd just looked at the GitLab master branch. Am I assuming the wrong location for vanilla? I figured you meant "latest master" but I could be mistaken.

Hopefully, I've explained the steps I'm taking clearly but I don't always so let me know, please, if anything I've said so far doesn't make sense and I'll try my best to clarify.

On Sunday, March 12, 2023 at 12:23:27 PM CDT, Nala Ginrut <mulei@gnu.org> wrote:

Hi Jaft!
I actually fixed several bugs in models and migrations in recent
months. Maybe you should try vanilla now.

Best regards.

Jaft writes:

>  Mmm; then maybe there's an issue just with migrations? When I tried using "boolean" for creating a table in a migration, I got an error in "->mysql-type"; checking that function out, I did notice that there's no case for "boolean".
>    On Friday, March 10, 2023 at 01:07:48 PM CST, Nala Ginrut <mulei@gnu.org> wrote:
>
>
> Sorry, I misunderstood you.
> You were asking why I provide it.
> Actually, for mysql/mariadb, Artanis won't do anything for boolean, it's
> only for sqlite3. So it won't cause any problem.
>
> Best regards.
>
>
> Nala Ginrut writes:
>
>> Hi Jaft!
>> Aratnis had abstracted boolean type in mvc/model.scm
>> --------------------code----------------
>> (define *db-specific-types*
>> '((mysql
>> (text . longtext)
>> (boolean . boolean))
>> (postgresql
>> (text . text)
>> (boolean . boolean))
>> (sqlite3
>> (text . text)
>> (boolean . integer))))
>> ---------------------end-----------------
>>
>>
>>
>> You may use `boolean' to define a table in a model, for example:
>>
>> -----------------code---------------
>> (create-table
>> 'user
>> '(id auto (#:primary-key))
>> '(name char-field (#:not-null #:maxlen 128))
>> '(is_admin boolean (#:not-null))
>> '(last_login bigint (#:unsigned #:not-null)))
>> ------------------end------------------
>>
>> I'm sorry I haven't put it in the document, I'll introduce it in the
>> recent release.
>>
>> Best regards.
>>
>>
>>
>> Jaft writes:
>>
>>> I think I always meant to ask but keep forgetting to – what wouldyou think, Nala, of providing a boolean type?
>>> Both MySQL and MariaDB abstract this type for the user, already, so doingthe same at the Artanis layer wouldn't be terribly different than the DB converting boolean values to 1 and 0, right?
>>> And it'd make working with those values easier to read when reading the code of a project, for the first time.


--
GNU Powered it
GPL Protected it
GOD Blessed it
HFG - NalaGinrut
Fingerprint F53B 4C56 95B5 E4D5 6093 4324 8469 6772 846A 0058