Mini Shell

"""
This module represents the various types of requirement that can be specified for
a project. It is somewhat redundant to re-implement here as we could use
`pip.req.InstallRequirement`, but that would require depending on pip which is not
easy to do since it will usually be installed by the user at a specific version.
Additionally, the pip implementation has a lot of extra features that we don't need -
we don't expect relative file paths to exist, for example. Note that the parsing here
is also intentionally more lenient - it is not our job to validate the requirements
list.
"""
import os
import re
from pathlib import Path
from typing import Optional
from urllib import parse

from packaging.requirements import Requirement


def _is_filepath(req):
    # this is (probably) a file
    return os.path.sep in req or req.startswith(".")


def _parse_egg_name(url_fragment):
    """
    >>> _parse_egg_name('egg=fish&cake=lala')
    fish
    >>> _parse_egg_name('something_spurious')
    None
    """
    if "=" not in url_fragment:
        return None
    parts = parse.parse_qs(url_fragment)
    if "egg" not in parts:
        return None
    return parts["egg"][0]  # taking the first value mimics pip's behaviour


def _strip_fragment(urlparts):
    new_urlparts = (
        urlparts.scheme,
        urlparts.netloc,
        urlparts.path,
        urlparts.params,
        urlparts.query,
        None,
    )
    return parse.urlunparse(new_urlparts)


class DetectedRequirement:
    def __init__(
        self, name: str = None, url: str = None, requirement: Requirement = None, location_defined: Path = None
    ):
        if requirement is not None:
            self.name = requirement.name
            self.requirement = requirement
            self.version_specs = [(s.operator, s.version) for s in requirement.specifier]
            self.url = None
        else:
            self.name = name
            self.version_specs = []
            self.url = url
            self.requirement = None
        self.location_defined = location_defined

    def _format_specs(self) -> str:
        return ",".join(["%s%s" % (comp, version) for comp, version in self.version_specs])

    def pip_format(self) -> str:
        if self.url:
            if self.name:
                return "%s#egg=%s" % (self.url, self.name)
            return self.url
        if self.name:
            if self.version_specs:
                return "%s%s" % (self.name, self._format_specs())
            return self.name
        raise ValueError(f"Cannot convert {self} to pip format, no name or URL")

    def __str__(self):
        rep = self.name or "Unknown"
        if self.version_specs:
            specs = ",".join(["%s%s" % (comp, ver) for comp, ver in self.version_specs])
            rep = "%s%s" % (rep, specs)
        if self.url:
            rep = "%s (%s)" % (rep, self.url)
        return rep

    def __hash__(self):
        return hash(str(self.name) + str(self.url) + str(self.version_specs))

    def __repr__(self):
        return "<DetectedRequirement:%s>" % str(self)

    def __eq__(self, other):
        return self.name == other.name and self.url == other.url and self.version_specs == other.version_specs

    def __gt__(self, other):
        return (self.name or "") > (other.name or "")

    @staticmethod
    def parse(line, location_defined: Path = None) -> Optional["DetectedRequirement"]:
        # the options for a Pip requirements file are:
        #
        # 1) <dependency_name>
        # 2) <dependency_name><version_spec>
        # 3) <vcs_url>(#egg=<dependency_name>)?
        # 4) <url_to_archive>(#egg=<dependency_name>)?
        # 5) <path_to_dir>
        # 6) (-e|--editable) <path_to_dir>(#egg=<dependency_name)?
        # 7) (-e|--editable) <vcs_url>#egg=<dependency_name>
        line = line.strip()

        if line.startswith("--hash=sha256:"):
            # skip multi-line shas, produced by poetry export
            return None

        # We need to match whitespace + # because url based requirements specify
        # egg_name after a '#'
        comment_pos = re.search(r"\s#", line)
        if comment_pos:
            line = line[: comment_pos.start()]

        # strip the editable flag
        line = re.sub("^(-e|--editable) ", "", line)

        # remove the python version stuff from poetry files
        line = line.split(";")[0]

        url = parse.urlparse(line)

        # if it is a VCS URL, then we want to strip off the protocol as urlparse
        # might not handle it correctly
        vcs_scheme = None
        if "+" in url.scheme or url.scheme in ("git",):
            if url.scheme == "git":
                vcs_scheme = "git+git"
            else:
                vcs_scheme = url.scheme
            url = parse.urlparse(re.sub(r"^%s://" % re.escape(url.scheme), "", line))

        if vcs_scheme is None and url.scheme == "" and not _is_filepath(line):
            # if we are here, it is a simple dependency
            try:
                req = Requirement(line)
            except ValueError:
                # this happens if the line is invalid
                return None
            else:
                return DetectedRequirement(requirement=req, location_defined=location_defined)

        # otherwise, this is some kind of URL
        name = _parse_egg_name(url.fragment)
        url = _strip_fragment(url)

        if vcs_scheme:
            url = "%s://%s" % (vcs_scheme, url)

        return DetectedRequirement(name=name, url=url, location_defined=location_defined)