URI.merge/2

[andrewtimberlake]

Merge a relative path/URI to a base URI as per RFC3986 §5.2

Examples:

iex> URI.merge("http://example.com/foo", "/bar") |> to_string
"http://example.com/bar"

iex> URI.parse("http://example.com/foo") |> URI.merge("bar") |> to_string
"http://example.com/bar"

…and more

[josevalim]

Error messages start with lower case.

[josevalim]

Thank you @andrewtimberlake. I have added some comments. I also believe we should call this function URI.join. I would expect a function named merge to merge all fields, so I think join reflects it a bit better.

[andrewtimberlake]

@josevalim I believe it is merging all fields as per the RFC which refers to this as a merge.
“The pseudocode above refers to a "merge" routine for merging a
relative-path reference with the path of the base URI”

[josevalim]

Right. We need to consider though the meaning "merge" has in Elixir and the meaning it has in the UFC. If we want to call it "merge", then I would recommend us to call it merge_path.

[andrewtimberlake]

I understand. I don’t want to belabour the point but we are doing more than just merging the path.

URI.merge("http://example.com", "http://google.com") |> to_string == "http://google.com"
URI.merge("http://example.com/foo?a=b", "?c=d") |> to_string == "http://example.com/foo?c=d"
URI.merge("http://example.com/foo?a=b", "") |> to_string == "http://example.com/foo?a=b"

From that info, you tell me what you’d like me to name it.

[josevalim]

Merge it is. :)

[lexmag]

We always have spaces around =.

[andrewtimberlake]

Thanks, I’ll update

[lexmag]

I think scheme != nil (instead of not is_nil(scheme)) reads more easily.

[josevalim]

@andrewtimberlake please ping us when this is ready to go. :)

[lexmag]

I think we can ditch this clause and simply call remove_dot_segments with [] on L#410.

[andrewtimberlake]

@josevalim It’s ready to go if you’re happy.