Translate guides/models/relationships.md

[Jasonlgrd]

Closes #87

[Jasonlgrd]

Some snippets were updated with the code formatter, I pushed it but idk if it's a good idea 😅

[MrChocolatine]

Thanks a lot for your contribution, we greatly appreciate.

Please, don't be "scared" by the number of comments I made, most of them are the same (putting English terms in italic , typing errors, ...) while others are "real suggestions" with more content.

Besides, this high number is also due to how the original text is formatted, with sentences split over several lines... ¯_(ツ)_/¯

[MrChocolatine]

I am sharing my thoughts to see how everyone feels about this.

We could "translate" these relationship names if we base our approach on the Merise methodology, or the UML notation, at least to provide somehow familiar names so that people (French readers, let's recall it) can eventually recognise these with something they might know.

This would give us:

• One-to-one => association de type (1,1)
• One-to-many => association de type (1,N)
• Many-to-many => association de type (1,N)

I do not pretend "we should be using these" nor this would be the exact text to use, but at least you get the idea so we can eventually discuss it.

[MrChocolatine]

@BlueCutOfficial what do you think?

[BlueCutOfficial]

I like the idea, we could give this translation at least once. Your suggestion matches the little I remember about SQL relationships, works for me :)

I don't have a strong opinion about using French everywhere or giving a translation once with the One-to-one (association de type (1,1)) syntax.

[MrChocolatine]

I don't have a strong opinion about using French everywhere or giving a translation once with the One-to-one (association de type (1,1)) syntax.

Yes, not necessarily every time but at least once, like here in this (main) chapter.

[MrChocolatine]

Wrong click with my previous review.

[Jasonlgrd]

I have fix all suggestions, you can review me before I rebase for 1 commit ? :)

[BlueCutOfficial]

@Jasonlgrd thanks a lot for the PR. Could you please add !guides/models/relationships.md to the file .remarkignore then run npm run lint to do the spellcheck :)

(I added this step in the new contributing guide I am working on, you've been faster than me :D)

[Jasonlgrd]

@Jasonlgrd thanks a lot for the PR. Could you please add !guides/models/relationships.md to the file .remarkignore then run npm run lint to do the spellcheck :)

(I added this step in the new contributing guide I am working on, you've been faster than me :D)

Yes I can ! I have run npm run lint and he say me I have written associtation instead of association ^^

Just, how I can ignore when the word is specific ?

[BlueCutOfficial]

You need to add words like one-to-one, many-to-many, and Paypal to the file called .local.dic. I don't know why they have both lowercase and uppercase versions in the English file but if it's the same context I would go for simplifying and go only with lowercase everywhere.

[MrChocolatine]

High number of comments BUT 97% of them are for the exact same typo. I tried to merge the suggestions as much as possible but again, due to how the text is formatted... We end up with a high number of micro suggestions.

I think I have caught all the typo, but please triple-check on your side as well, it is needed.

The rest of the 3% is just commas, plural, ...

---

Next time, I highly recommend that, once you think a document is fully translated, copy and paste its content into a Google Document (or whatever you prefer) with the spellchecker enabled, this will highlight what is (or potentially) wrong, and other stuff.

@BlueCutOfficial , we could also recommend that in the general guide, which would help the contributors but also, and especially, the reviewers.

[BlueCutOfficial]

@MrChocolatine I think we have 2 different topics here. One is partially solved and the other not at all:

• The non-existing words is a partially-solved issue: the new contributing guide we merged explains that the file must be added to the .remarkignore and the contributor has to run npm run lint to spellcheck. Our linter relies only on dictionaries, so it knows whether a word exists. For instance, in the associtation case, last week I explained to @Jasonlgrd how to get the file spellchecked because the new contributing guide was still in progress and @Jasonlgrd replied that indeed the linter detected associtation, but I think he didn't have time to repush yet? It's just painful that the file we modify is not added automatically to .remarkignore, because we can forget this step: Improve the shared hook to lint the French files  #179

• The second problem is about grammar/conjugation: if I remember correctly, our French dictionary relies on Grammalecte dictionary, but it's not the "Grammalecte" tool as it can be used in OpenOffice, Texmaker, and whatnot. Using only the dictionary part is not good enough to spot all the typos. Copy-pasting the text in a different editor with a proper spellcheck tool is a fine workaround, but it's very inconvenient. I am ok to recommend this workaround in the contributing guide as long as we don't have a better solution, but I think we should open a dedicated issue about this. I can open this by the end of the week.

[MrChocolatine]

ember-fr-guides-source/CONTRIBUTING.md

Line 46 in 90aa6f1

If there are errors because you kept an English term commonly used in French, add this word to the `.local.dic` file (in alphabetical order, lowercase words are below uppercase acronyms).

This is not a deal-breaker, just a reminder.