# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/main/LICENSE
# Copyright (c) https://github.com/PyCQA/pylint/blob/main/CONTRIBUTORS.txt

"""Check source code is ascii only or has an encoding declaration (PEP 263)."""

import re
import tokenize
from typing import TYPE_CHECKING, List, Optional

from astroid import nodes

from pylint.checkers import BaseChecker
from pylint.interfaces import IRawChecker, ITokenChecker
from pylint.typing import ManagedMessage
from pylint.utils.pragma_parser import OPTION_PO, PragmaParserError, parse_pragma

if TYPE_CHECKING:
    from pylint.lint import PyLinter


class ByIdManagedMessagesChecker(BaseChecker):

    """Checks for messages that are enabled or disabled by id instead of symbol."""

    __implements__ = IRawChecker
    name = "miscellaneous"
    msgs = {
        "I0023": (
            "%s",
            "use-symbolic-message-instead",
            "Used when a message is enabled or disabled by id.",
        )
    }
    options = ()

    def _clear_by_id_managed_msgs(self) -> None:
        self.linter._by_id_managed_msgs.clear()

    def _get_by_id_managed_msgs(self) -> List[ManagedMessage]:
        return self.linter._by_id_managed_msgs

    def process_module(self, node: nodes.Module) -> None:
        """Inspect the source file to find messages activated or deactivated by id."""
        managed_msgs = self._get_by_id_managed_msgs()
        for (mod_name, msgid, symbol, lineno, is_disabled) in managed_msgs:
            if mod_name == node.name:
                verb = "disable" if is_disabled else "enable"
                txt = f"'{msgid}' is cryptic: use '# pylint: {verb}={symbol}' instead"
                self.add_message("use-symbolic-message-instead", line=lineno, args=txt)
        self._clear_by_id_managed_msgs()


class EncodingChecker(BaseChecker):

    """BaseChecker for encoding issues.

    Checks for:
    * warning notes in the code like FIXME, XXX
    * encoding issues.
    """

    __implements__ = (IRawChecker, ITokenChecker)

    # configuration section name
    name = "miscellaneous"
    msgs = {
        "W0511": (
            "%s",
            "fixme",
            "Used when a warning note as FIXME or XXX is detected.",
        )
    }

    options = (
        (
            "notes",
            {
                "type": "csv",
                "metavar": "<comma separated values>",
                "default": ("FIXME", "XXX", "TODO"),
                "help": (
                    "List of note tags to take in consideration, "
                    "separated by a comma."
                ),
            },
        ),
        (
            "notes-rgx",
            {
                "type": "string",
                "metavar": "<regexp>",
                "help": "Regular expression of note tags to take in consideration.",
                "default": "",
            },
        ),
    )

    def open(self):
        super().open()

        notes = "|".join(re.escape(note) for note in self.linter.namespace.notes)
        if self.linter.namespace.notes_rgx:
            regex_string = (
                rf"#\s*({notes}|{self.linter.namespace.notes_rgx})(?=(:|\s|\Z))"
            )
        else:
            regex_string = rf"#\s*({notes})(?=(:|\s|\Z))"

        self._fixme_pattern = re.compile(regex_string, re.I)

    def _check_encoding(
        self, lineno: int, line: bytes, file_encoding: str
    ) -> Optional[str]:
        try:
            return line.decode(file_encoding)
        except UnicodeDecodeError:
            pass
        except LookupError:
            if (
                line.startswith(b"#")
                and "coding" in str(line)
                and file_encoding in str(line)
            ):
                msg = f"Cannot decode using encoding '{file_encoding}', bad encoding"
                self.add_message("syntax-error", line=lineno, args=msg)
        return None

    def process_module(self, node: nodes.Module) -> None:
        """Inspect the source file to find encoding problem."""
        encoding = node.file_encoding if node.file_encoding else "ascii"

        with node.stream() as stream:
            for lineno, line in enumerate(stream):
                self._check_encoding(lineno + 1, line, encoding)

    def process_tokens(self, tokens):
        """Inspect the source to find fixme problems."""
        if not self.linter.namespace.notes:
            return
        comments = (
            token_info for token_info in tokens if token_info.type == tokenize.COMMENT
        )
        for comment in comments:
            comment_text = comment.string[1:].lstrip()  # trim '#' and whitespaces

            # handle pylint disable clauses
            disable_option_match = OPTION_PO.search(comment_text)
            if disable_option_match:
                try:
                    values = []
                    try:
                        for pragma_repr in (
                            p_rep
                            for p_rep in parse_pragma(disable_option_match.group(2))
                            if p_rep.action == "disable"
                        ):
                            values.extend(pragma_repr.messages)
                    except PragmaParserError:
                        # Printing useful information dealing with this error is done in the lint package
                        pass
                    if set(values) & set(self.linter.namespace.notes):
                        continue
                except ValueError:
                    self.add_message(
                        "bad-inline-option",
                        args=disable_option_match.group(1).strip(),
                        line=comment.start[0],
                    )
                    continue

            # emit warnings if necessary
            match = self._fixme_pattern.search("#" + comment_text.lower())
            if match:
                self.add_message(
                    "fixme",
                    col_offset=comment.start[1] + 1,
                    args=comment_text,
                    line=comment.start[0],
                )


def register(linter: "PyLinter") -> None:
    linter.register_checker(EncodingChecker(linter))
    linter.register_checker(ByIdManagedMessagesChecker(linter))