urllibparse ~master

urllibparse

Direct D translation of Python's urllib.parse (beauty spots, warts and all).

Features

• Supports parsing, manipulation, relative URL joining, quoting, etc
• Supports URLs containing non-UTF-8 character sets
• Reuses Python's test suite based on RFCs and bug reports
• Single-file library with no non-std dependencies
• @safe

Usage example

Either add to an existing DUB project with dub add urllibparse, or just add the urllibparse.d file to somewhere in your D compiler's import path (such as the main project's source directory).

import urllibparse;

// Parsing
auto parsed = urlParse("https://dlang.org/blog/");
assert (parsed.scheme == "https");
assert (parsed.path == "/blog/");
assert (parsed.password is null);

parsed.path = "documentation.html";
assert (parsed.getUrl() == "https://dlang.org/documentation.html");

// Relative URL handling
assert (urlJoin("https://dlang.org/blog/", "../articles.html") == "https://dlang.org/articles.html");

// Low-level quoting/unquoting
assert (quote("URLs in D!") == "URLs%20in%20D%21");
assert (quotePlus("URLs in D!") == "URLs+in+D%21");
assert (unquote("URLs%20in%20D%21") == "URLs in D!");
assert (unquotePlus("URLs+in+D%21") == "URLs in D!");

// Query encoding
const args = [
	QueryArg("query", "d tools"),
	QueryArg("page", "1"),
	QueryArg("orderBy", "date"),
];
assert (urlEncode(args) == "query=d+tools&page=1&orderBy=date");
// can also use associative arrays, but can't guarantee order or handle repeated keys
assert (urlEncode([
	"query": "d tools",
	"page": "1",
	"orderBy": "date",
]));

// Query parsing
assert (parseQSL("query=d+tools&page=1&orderBy=date") == args);
// likewise, parseQS returns an associative array

Documentation

See the original Python documentation and the auto-generated D documentation.

Differences from Python:

• Function names have been changed to match D's camelCase functionNamingConvention
• ValueException is thrown where the Python throws ValueError, and EncodingException is thrown instead of UnicodeEncodeError
• The ParseResult class has been changed to a URLParseResult struct (and similarly for other *Result classes)
• Character encodings (and names) are those supported by std.encoding
• D isn't duck-typey like Python (but pull requests are welcome if you think an overload should be added)
• Some deprecated (trivial) functions are missing. (They're not in the Python docs now, anyway.)