Add uuid feature flag, based on uuid crate.

I originally opened, wrongly, a feature request #524 there but then realized it should be part of this crate instead.

I tried to implement it by mimicking how the other supported data are implemented,

but I would like additional review because some inner mechanisms I didn't really grasp (at least it seemed, when I had the test passes something doesn't seem right, does it ?)

I'm receiving the following error when trying to use Dummy trait on a struct that contains an emum.

proc-macro derive panicked

help: message: called `Result::unwrap()` on an `Err` value: Error { kind: UnsupportedShape("enum"), locations: [], span: None }

From reviewing the crate, it seems like enums should be "fakeable", but minor tweaking doesn't help. If this is not the case, a use case on implementing the Dummy Trait would be great to have in the docs or examples.

Thank you for providing this excellent tool!

This PR introduces Decimal faking. I chose not to localize it and go with an enum for configuation. I will most likely supply more configuration options in the future :)

I think, filtering on stable/unstable version is an important feature.

I added a choice but i'm not really satisfied (breaking changed) : Ideally i wish i could do :

Semver(filter: crate::faker::SemverFilter);

and

Semver();

I didn't find an elegant solution.

This should implement issue #11 . This is implemented using an optional Cargo feature, so it's disabled by default to minimize unnecessary dependencies and compile time. That can be changed.

@svenstaro , can you take a look over this and let me know if this is what you had in mind? The tests in src/lib.rs are instructive, and the implementation is all in src/faker/time.rs

Hi team,

From scanning the repo, I detected below unsafe code, and it can be convert to safe version: unsafe { String::from_utf8_unchecked(bytes) } https://github.com/cksac/fake-rs/blob/5a26c9c3b7f81f5dcc5cbec999e8025e7399676d/fake/src/faker/impls/creditcard.rs#L43

Safe version: String::from_utf8(bytes).unwrap()

I just realized that there are crate examples, not quite discoverable.

Usually I need to do stuff offline so I think having offline documentation is good, since it is not easy to browse the README or the examples as compared to each specific items generated by rustdoc, having some explanation for the traits can roughly describes how the trait works.

Fix #46

I have fixed some typos in the ADDRESS_COUNTRY constant:

• "Democratic People's Republic of Korea Republic of Korea Kuwait Kyrgyz Republic Lao People's Democratic Republic" was in one word.
• "'Virgin Islands, British'" and "'Virgin Islands, U.S.'" had additional quotes.

@cksac Thanks for merging my other PRs.

As I was using the 'dummy" enhancement PR locally I found a few other niceties and expanded the test.

• this adds a "default" field option as well that will use Default:::default()

#[derive(Dummy)]
struct Obj(
    #[dummy(default)]
    String,
);

#[derive(Dummy)]
#[dummy(default_impl)]
struct Obj {
    pub value: i32,
}

let a : Obj = Default::default();

Hi there! We are currently using faker (via pyo3) in synth and want to move to use fake-rs. However, to do that, we'd want fake-rs to match more of the functionality faker provides.

So we'd like to give this project a boost by trying to speed up that implementation work on that (I did the creditcardnumber stuff to check how hard that is). But before we go that route (and risk overwhelming you with PRs) I wanted to openly address you to learn your needs as a maintainer of this project and how we can best work together to our mutual benefit.

I am using Dummy to create an object for testing so I've used [cfg_attr(test, derive(Dummy)] for the struct I want to fake. I have a "domain" crate where I have my models and then an "http_api" crate which depends on "domain". When I try to get a fake instance of my model struct, I get an error saying that Dummy is not implemented. Is there a way to "expose" that I've implemented the trait here that I'm missing or is this not possible when using cfg(test)? It works perfectly fine if I remove the #[cfg_attr(test, derive(Dummy)] from the struct and just derive it normally.

error[E0277]: the trait bound `domain::models::api_requests::PutRequest: Dummy<Faker>` is not satisfied
   --> http_api/src/local/filters.rs:154:48
    |
154 |         let request: PutRequest = Faker.fake();
    |                                                ^^^^ the trait `Dummy<Faker>` is not implemented for `domain::models::api_requests::PutRequest`

Hi @cksac, I've been using your library thanks to LukeMathWalker/zero-to-production

Since I've had the need to generate a bit of german fake data I took a closer look into the project.

I added a de_de.rs module to provide a few basic attributes like State,Company Type,Names and so on from the official government ID test card data

https://persosim.secunet.com/fileadmin/user_upload/PersoSim_Personalisierungsdaten_eID_UB.pdf

It's working with the check_determinism test as far as I can tell.

To verify the data that gets generated I added a de_manual_test.rs just so see some generic company struct on the terminal.

Run via cargo test de_manual_test -- --nocapture it produces some results like

Company { name: "Schuster e.V.", street: "Hillebrandtgang", zipcode: "29346", city: "Musterfrau land", state: "Berlin" }

So far so good. :smile:

But that is the reason I opened the PR.

"Musterfrau land" is not a valid city name in german.

Most city names would consist of two nouns or names and a suffix.

Germany and to further extend the "DACH" Region (Germany, Austria, Switzerland) all have a pretty different layout for thinks like street names and addresses.

And I would like to discuss how I could implement it in a way it suites the library and their users.

For example a street name might be separated by dashes if it contains one or multiple names e.g. Karl-Marx-Allee https://en.wikipedia.org/wiki/Karl-Marx-Allee or can contain an apostrophe as seen here in https://de.wikipedia.org/wiki/Stra%C3%9Fenname in Laehr'scher Jagdweg

There are already some impl with random branching for data generation https://github.com/cksac/fake-rs/blob/32646496eba2918f133c31e594e46ee907577c50/fake/src/faker/impls/address.rs#L108

And I don't know if I should start adding impls special for the DE_DE locale or how it would fit the overall project structure.

Any feedback or guidance is appreciated and I would gladly implement more of the german locale.

Cheers xoryouyou

All the examples I can find fix the locale to a specific one:

let val: String = Name(EN).fake();
let val: String = Name(ZH_TW).fake();

I'd like to generate data based on user-supplied argument, e.g

enum Locale {
    English,
    Taiwan,
    ...etc..
}

fn enthusiastic_fake_data(lo: Locale) -> String {
    format!("{}!!!", ...)
}

I couldn't find a neat way to do this - I can of course just do

let val: String = match {
    English => Name(EN).fake(),
    Taiwan => Name(ZH_TW).fake(),
}

..but this gets very repetitive if you also have a similar match for type of data (e.g Name, Country, etc)

I was hoping to do something like

let lo = match locale_arg {
    Locale::English => fake::locales::EN,
    Locale::Taiwan => fake::Local::ZH_TW,
};
let val: String = Name(lo).fake();

However this doesn't work as each local is it's own struct. I tried let lo: Box<dyn fake::locale::Data> = match ... and started running into more complex errors, so I started wondering if I was missing a simpler way to do this

Thanks!

Option<T> is something like Boolean (it has Some(T) value or None). So I'd like to use ratio for this. But now we have a 50:50 ratio (look at percentage on 1 screenshot).
I have a code something like this:

    let mut rng = rand::thread_rng();
    let file = File::create("data.csv")?;
    let mut wtr = csv::Writer::from_writer(file);
    let mut i = 0;
    let rows = rng.gen_range(10000..100000);
    loop {
        wtr.serialize(Faker.fake::<Foo>())?;
        if i == rows {
            break;
        }
        i += 1;
    }
    wtr.flush()?;

Current behavior - we have 50:50 ration:

After this commit:

P.S. Of course to start it working we should add #[dummy(faker = 'Opt((), 98)') to our struct's members like this

    #[dummy(faker = "Opt(0..2000, 93)")]
    order_id: Option<u64>,
    #[dummy(faker = "Opt(0..2000, 81)")]
    make_id: Option<u64>,
    #[dummy(faker = "Opt(0..2000, 98)")]
    time_id: Option<u64>,
    #[dummy(faker = "Opt(0..2000, 90)")]
    this_id: Option<u64>,
    #[dummy(faker = "Opt(0.0..2000.0, 78)")]
    my_float: Option<f64>,