[feature] output in ZIP format if extension is '.zip'

[woshilapin]

First commit is only about setting up some tests.
Second commit is about the new feature.

The new feature is to detect if --output contains .zip in the name. If no, the previous behavior is done: writing files to a folder. If yes, then we write the output files to a temporary folder, then ZIP all these files into the --output path.

🔍 Help for reviewing:

1. 8ec61ef: add tests on the binary for gtfs2ntfs
2. 7540763: create a ZIP file in gtfs2ntfs if output path has extension .zip
3. a6ef370: (1) add write() and write_to_zip() functions on the API of transit_model::netex_france and (2) create a ZIP file in gtfs2netexfr and ntfs2netexfr if output path has extension .zip
4. e8743b1: create a ZIP file in ntfs2ntfs if output path has extension .zip
5. f79827b: create a ZIP file in ntfs2gtfs if output path has extension .zip and add a write_to_zip() function to transit_model::gtfs

🏗️ All the steps that have to be implemented

• gtfs2ntfs supports ZIP export (7540763)
• gtfs2netexfr supports ZIP export (a6ef370)
• ntfs2netexfr supports ZIP export (a6ef370)
• ntfs2ntfs supports ZIP export (e8743b1)
• ntfs2gtfs supports ZIP export (f79827b)
• Bump versions of Cargo.toml everywhere (1e57620)

❓ Unresolved questions

• Tests added in gtfs2netexfr and ntfs2netexfr are covering more than what is in tests/write_netex_france.rs. Should I remove them from tests/write_netex_france.rs? -> On the binary tests, only test that output files are produced (not the entire output fixture) and extend to all the other similar tests (f2af969, 0ee7806, 0ee7806, b7e7f58)
• Should we remove transit_model::netex_france::Exporter from the public API? -> Make Exporter private and possibly document the migration to the new write() function (1e57620)
• Should we homogenize the ntfs::write(model: &Model, ...) and gtfs::write(model: Model, ...) (some with reference and some other with ownership)? -> To be done in another PR
• Add a test for --output with an extension that is not .zip (585b37e)

[antoine-de]

🎉
it's a pity that the zip file is not done using directly a zip archive writer but it's fine for me to do this in a separate PR 😁

[woshilapin]

tada
it's a pity that the zip file is not done using directly a zip archive writer but it's fine for me to do this in a separate PR grin

Compared to writing to a temporary folder, then into an archive, it sure would be better. But not sure that it'd be as simple to do because it would force us to call ZipWriter::start_file() in code that doesn't know (and doesn't have to know) that it is a ZIP or not. We might need an abstraction here... in a separate PR for sure 😛

[datanel]

approved but -1 for the "pouet" in the fixtures.