|
This channel is intended for people just starting with the Raku Programming Language (raku.org). Logs are available at irclogs.raku.org/raku-beginner/live.html Set by lizmat on 8 June 2022. |
|||
|
05:33
siavash joined
07:50
guest23 joined
|
|||
| guest23 | Hi. I'd like to titlecase some stuff. Is there a good way to do so in raku? I thought about using Inline::Perl5 and Lingua::EN::Titlecase or Lingua::EN::Titlecase::Simple, but don't seem to be able to alter the rules via “our” variables (e.g. @SMALL_WORDS in *::Simple). Is there a way do to this? Thanks | 07:53 | |
|
08:01
ab5tract joined
|
|||
| guest23 | I found this on the Inline::Perl5 github. I'm not sure whether it means it's possible, but needs examples, or isn't possible: github.com/niner/Inline-Perl5/issues/139 | 08:08 | |
| siavash | You can do it on words with `wordcase` method and on strings with `tclc` method, but it seems `wordcase` is what you want, unless you want something that comes with predefined rules docs.raku.org/type/Str#routine_wordcase | 08:21 | |
|
08:30
ab5tract left
|
|||
| guest23 | I'd prefer a module since there're lots of edge cases (e.g. handling first/last words, first word in parens, hyphenated words, etc.), but I'll have. look. Thanks | 08:34 | |
|
08:45
tonyo left,
tonyo joined
09:14
siavash left
09:15
siavash joined
09:16
guest23 left,
guest23 joined
09:24
siavash left,
CIAvash joined
09:28
CIAvash left
09:48
CIAvash joined
10:02
CIAvash left
10:11
guest23 left
12:40
ab5tract joined
12:43
lizmat_ joined
12:46
ab5tract_ left,
lizmat left
12:47
lizmat_ left,
lizmat joined
12:48
ab5tract left
13:03
guest23 joined
14:07
guest23 left
14:16
guest23 joined
|
|||
| guest23 | I'm having a problem with IO::Path. When I use .Str (or .path) on one with certain characters such as ÿ in them, I get weird symbols printed to my terminal. When I use uniname on the characters I get <private-use-10FFFD> printed. Any idea what's going on? I'm getting the paths by using the code from the bottom of docs.raku.org/routine/dir. | 14:20 | |
| I'm on macOS using Rakudo™ v2023.09 | |||
| librasteve | guest23: I just tried that on my mac (example program that lists all files and directories recursively) - works fine for me | 14:25 | |
| I am on 2023.08 | |||
| it's fine on 2023.09 also | 14:27 | ||
| can you maybe post your code to a gist and I can try that? | |||
| guest23 | hmm. It must be something with the filename. I just created a test directory and manually typed the same name and it was fine | 14:34 | |
| librasteve | i suspect that unicode in filenames is still very novel and maybe some apps do not handle well (Finder rename?) | 14:36 | |
| here is something from reddit (when getting files saved by others) ... In macOS, you can do this by going to System Preferences > Language & Region > Advanced, and selecting the encoding used by her coworkers under "Number, Currency, and Date Formatting". | 14:39 | ||
| maybe what you have is not utf-8? sis you read pawels blogs on unicode, there's likely a way to check what you have in raku | |||
| guest23 | Thank you. I'll have a look | 14:40 | |
| Making the directory using iTerm 2 works fine, but making it using the Finder causes the issue. | |||
| librasteve | dev.to/bbkr/series/23930 | ||
| guest23 | System Preferences > Language & Region > Advanced doesn't seem to exist anymore | 14:42 | |
| (System Settings, rather) | |||
| librasteve | what do you get with echo $LANG? | 14:45 | |
| I get en_GB.UTF-8 | |||
| guest23 | en_GB.UTF-8 | 14:47 | |
| I've created two dictories: A/ÿ using iTerm and B/ÿ using Finder. ls A | hexdump gives me bfc3 000a, ls B | hexdump gives me cc79 0a88. | 14:59 | ||
| Blob.new(0xC3, 0xBF).decode.say gives me ÿ | |||
| librasteve | hmmm looks like iTerm is round tripping correctly but is set to a differenet encoding | 15:00 | |
| My (vanilla) Terminal > Prefs > Encodings has 'Unicode (UTF-8)' and a bunch of other like Western (ASCII), Japanese (MacOS) and so on ... | 15:02 | ||
| guest23 | But $path.path.comb.map: *.uniname.say; outputs the weird private-use things | 15:07 | |
| librasteve | Can you maybe try with vanilla Terminal? | 15:08 | |
| guest23 | raku -e 'dir("A").map: *.basename.say' | 15:09 | |
| ÿ | |||
| raku -e 'dir("B").map: *.basename.say' | |||
| yxCCx88 | |||
| raku -e 'dir("A").map: *.basename.uninames.say' | |||
| (LATIN SMALL LETTER Y WITH DIAERESIS) | |||
| raku -e 'dir("B").map: *.basename.uninames.say' | |||
| (LATIN SMALL LETTER Y <private-use-10FFFD> LATIN SMALL LETTER X LATIN CAPITAL LETTER C LATIN CAPITAL LETTER C <private-use-10FFFD> LATIN SMALL LETTER X DIGIT EIGHT DIGIT EIGHT) | |||
| nemokosch | could it consist of combining code points? | 15:10 | |
| the second one | 15:11 | ||
| guest23 | hexdump says the filename created with Finder contains the bytes cc79 0a88, rather than the bytes bfc3 000a are used for the filename created with Terminal. Combining diaresis is CC88, so it looks like Finder is using a combining codepoint and Terminal is using a combined character (C3BF) | 15:18 | |
| nemokosch | dang | 15:19 | |
| what are you trying to do again? | |||
| guest23 | Should be able to treat them the same in raku though, right? | ||
| Just process some albums. Broke on Queensrÿche. I can just use mv to fix it, but I don't think I should need to, should I? | 15:20 | ||
| librasteve | I just tried this the way you describe (ie use finder to rename to 'Broke on Queensrÿche' and then use the raku -e you show) - I now get the same private-use issue as you | 15:44 | |
| it was working ok for me before since I was just making then reading the filename in Terminal | 15:45 | ||
| my guess is that macOS file system and raku have different ideas on how to manage combined unicode chars (as nemo suggests) ... it's probably worth making an issue for this on rakudo since afaik macOS is a gold standard for this | 15:47 | ||
| guest23 | Gotcha. Yeah, I was discovering as I went so wasn't clear. Glad you can reproduce | ||
| librasteve | anyway, back to titlecase, this should work for you:use Inline::Perl5; my $p5 = Inline::Perl5.new; $p5.use('Lingua::EN::Titlecase'); my $tc = $p5.invoke('Lingua::EN::Titlecase', 'new', 'word_punctuation => "[^]"'); $p5.run( 'print join(" ", sort(keys(%Lingua::EN::Titlecase::LC)), "\n")' ); $p5.run( '$Lingua::EN::Titlecase::LC{"yep"} = 1' ); $p5.run( 'print join(" ", | 15:48 | |
| sort(keys(%Lingua::EN::Titlecase::LC)), "\n")' ); print $tc.title("and again and again yep, cow^catcher not cow£catcher but differently"); | |||
| And Again and Again yep, Cow^catcher Not Cow£Catcher but Differently% | |||
| I ended up going with the non Simple cpan module since that has an explicit constructor where you can control the ATTRS such as word_punctuation => 1, original => 1, title => 1, uc_threshold => 1, mixed_threshold => 1, | 15:49 | ||
|
15:50
habere-et-disper joined
|
|||
| metacpan.org/release/ASHLEY/Lingua...tlecase.pm | 15:50 | ||
| guest23 | Ah, thanks. That's a lot more complex than I'd tried. I was just doing `use Lingua::EN::Titlecase::Simple 'titlecase'; @Lingua::EN::Titlecase::Simple::SHORD_WORD.push: ...` and hoping it'd work. | 15:51 | |
| I think I'll just use Lingua::EN::Titlecase::Simple and fix up a few bits instead. | |||
| Thanks for that for future Inline::Perl5 bits | |||
| Do you want me to create the issue about IO::Path or do you have a better idea of what to write? | 15:52 | ||
| habere-et-disper | m: say sqrt: 100 | ||
| camelia | ===SORRY!=== Error while compiling <tmp> Unsupported use of bare "sqrt". In Raku please use: .sqrt if you meant to call it as a method on $_, or use an explicit invocant or argument, or use &sqrt to refer to the function as a noun. at <… |
||
| guest23 | (Made a typo in my code above) | ||
| m: say sqrt 100 | 15:54 | ||
| camelia | 10 | ||
| habere-et-disper | In the REPL I get `100` for `sqrt: 100` | 15:57 | |
|
16:02
guifa left
|
|||
| guest23 | It is treating sqrt: as a label there or something? (I'm not sure. I should leave this to the professionals.) | 16:02 | |
| *Is it | |||
| habere-et-disper | That sounds plausible, thanks. :) | 16:03 | |
| nemokosch | I think so, actually | 16:11 | |
| librasteve | you can make an issue at github search rakudo | 16:12 | |
|
16:20
habere-et-disper left
|
|||
| m: say sqrt 100 | 16:46 | ||
| Raku eval | 10 | ||
| librasteve | m: say 100.sqrt | ||
| Raku eval | 10 | ||
| librasteve | the colon is only needed when there are additional arguments to a method call | 16:47 | |
| m: say [1,2].push: 3 | 16:48 | ||
| Raku eval | [1 2 3] | ||
|
18:17
ab5tract_ joined
19:20
Nemokosch joined
20:04
Nemokosch left
22:11
guest23 left
|
|||
| nhail | sub altsum(Array @x) { return sum @x } say altsum [0,1]; I don't understand what's wrong here? I get the error Type check failed in binding to parameter '@x'; expected Positional[Array] but got Array ([0, 1]) | 23:02 | |
|
23:12
teatime joined
|
|||
| nemokosch | Array @x is @x of Array, not @x is Array | 23:16 | |
| nhail | So what should I put in the type signature? | 23:17 | |
|
23:42
teatime left,
teatime joined
|
|||
| nemokosch | m: sub altsum(Array $x) { return sum $x } say altsum [0,1]; | 23:47 | |
| Raku eval | 1 | ||
| nemokosch | One way, kind of a Gordian knot solution | 23:48 | |
| m: sub altsum(@x where Array) { return sum @x } say altsum [0,1]; | |||
| Raku eval | 1 | ||
|
23:48
teatwo joined
|
|||
| nemokosch | With a custom type check | 23:48 | |
| I wonder know | 23:49 | ||
| m: sub altsum(@x of Int) { return sum @x } my Int @x = 0, 1; say altsum @x; | 23:50 | ||
| Raku eval | Exit code: 1 ===SORRY!=== Error while compiling /home/glot/main.raku Cannot resolve caller trait_mod:<of>(Parameter:D, Int:U); none of these signatures matches: (Mu:U $target, Mu:U $type) (Routine:D $target, Mu:U $type) at /home/glot/main.raku:1 | ||
| nemokosch | Yeah lest traits do anything type-related in a signature :\ | 23:51 | |
|
23:52
teatime left
23:55
teatwo left,
teatwo joined
|
|||