Question: syntax for abs paths passed as file:// URLs in Windows?

[jaimergp]

What's the actual way? Given this absolute path in Windows D:\a_\temp\popen-gw0\test_croot_with_spaces0\space path:

>>> from libmambapy.specs import CondaURL
>>> CondaURL.parse("file:///D:/a/_temp/popen-gw0/test_croot_with_spaces0/space path")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
libmambapy.bindings.specs.ParseError: Failed to parse URL "file:///D:/a/_temp/popen-gw0/test_croot_with_spaces0/space path": Malformed input to a URL function
>>> CondaURL.parse("file:///D:/a/_temp/popen-gw0/test_croot_with_spaces0/space%20path
")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
libmambapy.bindings.specs.ParseError: Failed to parse URL "file:///D:/a/_temp/popen-gw0/test_croot_with_spaces0/space%20path": Bad file:// URL
>>> CondaURL.parse("file:///D/a/_temp/popen-gw0/test_croot_with_spaces0/space path")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
libmambapy.bindings.specs.ParseError: Failed to parse URL "file:///D/a/_temp/popen-gw0/test_croot_with_spaces0/space path": Malformed input to a URL function
>>> CondaURL.parse("file:///D/a/_temp/popen-gw0/test_croot_with_spaces0/space%20path"
)
file:///D/a/_temp/popen-gw0/test_croot_with_spaces0/space%20path

Only the last one (without the : in the drive letter) is accepted, but not sure if that's understood as a Windows path internally? Should I always remove the : in the drive?

[jaimergp]

Seeing comments like this one:

mamba/libmamba/include/mamba/util/url.hpp

Lines 186 to 192 in 31b2347

	/**
	* Set the path from a not encoded value.
	*
	* All '/' are not encoded but interpreted as separators.
	* On windows with a file scheme, the colon after the drive letter is not encoded.
	* A leading '/' is added if abscent.
	*/

makes me think it should be accepted?

[kenodegard]

Per this blogpost from the IE team back in 2006, the proper syntax is as follows:

[kenodegard]

Reading the file_uri_unc2_to_unc4 implementation:

mamba/libmamba/src/util/url_manip.cpp

Lines 88 to 133 in 0abd7d5

	auto file_uri_unc2_to_unc4(std::string_view uri) -> std::string
	{
	static constexpr std::string_view file_scheme = "file:";
	// Not "file:" scheme
	if (!util::starts_with(uri, file_scheme))
	{
	return std::string(uri);
	}
	// No hostname set in "file://hostname/path/to/data.xml"
	auto [slashes, rest] = util::lstrip_parts(util::remove_prefix(uri, file_scheme), '/');
	if (slashes.size() != 2)
	{
	return std::string(uri);
	}
	const auto s_idx = rest.find('/');
	const auto c_idx = rest.find(':');
	// ':' found before '/', a Windows drive is specified in "file://C:/path/to/data.xml" (not
	// really URI compliant, they should have "file:///" or "file:/"). Otherwise no path in
	// "file://hostname", also not URI compliant.
	if (c_idx < s_idx)
	{
	return std::string(uri);
	}
	const auto hostname = rest.substr(0, s_idx);
	// '\' are used as path separator in "file://\\hostname\path\to\data.xml" (also not RFC
	// compliant)
	if (util::starts_with(hostname, R"(\\)"))
	{
	return std::string(uri);
	}
	// Things that means localhost are kept for some reason in ``url_to_path``
	// in ``conda.common.path``
	if ((hostname == "localhost") || (hostname == "127.0.0.1") || (hostname == "::1"))
	{
	return std::string(uri);
	}
	return util::concat("file:////", rest);
	}

And testing some more, it seems CondaURL wants 4 leading / and %-encoded spaces:

>>> from libmambapy.specs import CondaURL
>>> CondaURL.parse("file:////D:/a_/temp/popen-gw0/test_croot_with_spaces0/space%20path")
file:////D:/a_/temp/popen-gw0/test_croot_with_spaces0/space%20path

[jaimergp]

It works with 4+ leading /, so I don't think it was intended. file:///C:/path/to/file should be legal imo.