"""djLint add indentation to html."""

from functools import partial

import json5 as json
import regex as re

from ..helpers import (
    inside_ignored_block,
    is_ignored_block_closing,
    is_ignored_block_opening,
    is_safe_closing_tag,
    is_script_style_block_closing,
    is_script_style_block_opening,
)
from ..settings import Config
from .attributes import format_attributes


def indent_html(rawcode: str, config: Config) -> str:
    """Indent raw code."""
    if config.profile not in ["handlebars", "golang"]:
        # we can try to fix template tags. ignore handlebars
        # this should be done before indenting to line length
        # calc is preserved.

        def fix_tag_spacing(html: str, match: re.Match) -> str:
            if inside_ignored_block(config, html, match):
                return match.group()

            return f"{match.group(1)} {match.group(2)} {match.group(3)}"

        """
        We should have tags like this:
        {{ tag }}
        {%- tag atrib -%}
        """
        func = partial(fix_tag_spacing, rawcode)

        rawcode = re.sub(
            r"({%-?\+?)[ ]*?(\w(?:(?!%}).)*?)[ ]*?(\+?-?%})", func, rawcode
        )

        rawcode = re.sub(r"({{)[ ]*?(\w(?:(?!}}).)*?)[ ]*?(\+?-?}})", func, rawcode)

    elif config.profile == "handlebars":

        def fix_handlebars_template_tags(html: str, match: re.Match) -> str:
            if inside_ignored_block(config, html, match):
                return match.group()

            return f"{match.group(1)} {match.group(2)}"

        func = partial(fix_handlebars_template_tags, rawcode)
        # handlebars templates
        rawcode = re.sub(r"({{#(?:each|if).+?[^ ])(}})", func, rawcode)

    rawcode_flat_list = re.split("\n", rawcode)

    indent = config.indent

    beautified_code = ""
    indent_level = 0
    in_set_tag = False
    is_raw_first_line = False
    in_script_style_tag = False
    is_block_raw = False

    slt_html = config.indent_html_tags

    # here using all tags cause we allow empty tags on one line
    always_self_closing_html = config.always_self_closing_html_tags

    # here using all tags cause we allow empty tags on one line
    slt_template = config.optional_single_line_template_tags

    # nested ignored blocks..
    ignored_level = 0

    for item in rawcode_flat_list:
        # if a raw tag first line
        if not is_block_raw and is_ignored_block_opening(config, item):
            is_raw_first_line = True

        # if a raw tag then start ignoring
        if is_ignored_block_opening(config, item):
            is_block_raw = True
            ignored_level += 1

        if is_script_style_block_opening(config, item):
            in_script_style_tag = True

        if is_safe_closing_tag(config, item):
            ignored_level -= 1
            ignored_level = max(ignored_level, 0)
            if is_block_raw is True and ignored_level == 0:
                is_block_raw = False

        if (
            re.findall(
                rf"^\s*?(?:{config.ignored_inline_blocks})",
                item,
                flags=re.IGNORECASE | re.VERBOSE | re.MULTILINE,
            )
            and is_block_raw is False
        ):
            tmp = (indent * indent_level) + item + "\n"

        # if a one-line, inline tag, just process it, only if line starts w/ it
        # or if it is trailing text

        elif (
            (
                re.findall(
                    re.compile(
                        rf"""^(?:[^<\s].*?)? # start of a line, optionally with some text
                    (?:
                        (?:<({slt_html})>)(?:.*?)(?:</(?:\1)>) # <span>stuff</span> >>>> match 1
                       |(?:<({slt_html})\b[^>]+?>)(?:.*?)(?:</(?:\2)>) # <span stuff>stuff</span> >>> match 2
                       |(?:<(?:{always_self_closing_html})\b[^>]*?/?>) # <img stuff />
                       |(?:<(?:{slt_html})\b[^>]*?/>) # <img />
                       |(?:{{%[ ]*?({slt_template})[ ]+?.*?%}})(?:.*?)(?:{{%[ ]+?end(?:\3)[ ]+?.*?%}}) # >>> match 3
                       |{config.ignored_inline_blocks}
                    )[ \t]*?
                    (?:
                    .*? # anything
                    (?: # followed by another slt
                        (?:<({slt_html})>)(?:.*?)(?:</(?:\4)>) # <span>stuff</span> >>>> match 1
                       |(?:<({slt_html})\b[^>]+?>)(?:.*?)(?:</(?:\5)>) # <span stuff>stuff</span> >>> match 2
                       |(?:<(?:{always_self_closing_html})\b[^>]*?/?>) # <img stuff />
                       |(?:<(?:{slt_html})\b[^>]*?/>) # <img />
                       |(?:{{%[ ]*?({slt_template})[ ]+?.*?%}})(?:.*?)(?:{{%[ ]+?end(?:\6)[ ]+?.*?%}}) # >>> match 3
                       |{config.ignored_inline_blocks}
                    )[ \t]*?
                    )*? # optional of course
                    [^<]*?$ # with no other tags following until end of line
                """,
                        re.IGNORECASE | re.VERBOSE | re.MULTILINE,
                    ),
                    item,
                )
            )
            and is_block_raw is False
        ):
            tmp = (indent * indent_level) + item + "\n"

        # closing set tag
        elif (
            config.no_set_formatting is False
            and re.search(
                re.compile(
                    r"^(?!.*\{\%).*%\}.*$",
                    re.IGNORECASE | re.MULTILINE | re.VERBOSE,
                ),
                item,
            )
            and is_block_raw is False
            and in_set_tag is True
        ):
            indent_level = max(indent_level - 1, 0)
            in_set_tag = False
            tmp = (indent * indent_level) + item + "\n"

        # closing curly brace inside a set tag
        elif (
            config.no_set_formatting is False
            and re.search(
                re.compile(
                    r"^[ ]*}|^[ ]*]",
                    re.IGNORECASE | re.MULTILINE | re.VERBOSE,
                ),
                item,
            )
            and is_block_raw is False
            and in_set_tag is True
        ):
            indent_level = max(indent_level - 1, 0)
            tmp = (indent * indent_level) + item + "\n"

        # if unindent, move left
        elif (
            re.search(
                config.tag_unindent,
                item,
                re.IGNORECASE | re.MULTILINE | re.VERBOSE,
            )
            and is_block_raw is False
            and not is_safe_closing_tag(config, item)
            # and not ending in a slt like <span><strong></strong>.
            and not re.findall(
                rf"(<({slt_html})>)(.*?)(</(\2)>[^<]*?$)",
                item,
                re.IGNORECASE | re.VERBOSE | re.MULTILINE,
            )
            and not re.findall(
                rf"(<({slt_html})\\b[^>]+?>)(.*?)(</(\2)>[^<]*?$)",
                item,
                re.IGNORECASE | re.VERBOSE | re.MULTILINE,
            )
        ):
            # block to catch inline block followed by a non-break tag
            if (
                len(
                    re.findall(
                        rf"(^<({slt_html})>)(.*?)(</(\2)>)",
                        item,
                        re.IGNORECASE | re.VERBOSE | re.MULTILINE,
                    )
                    or re.findall(
                        re.compile(
                            rf"(^<({slt_html})\b[^>]+?>)(.*?)(</(\2)>)",
                            re.IGNORECASE | re.VERBOSE | re.MULTILINE,
                        ),
                        item,
                    )
                )
                > 0
            ):
                # unindent after instead of before
                tmp = (indent * indent_level) + item + "\n"
                indent_level = max(indent_level - 1, 0)
            else:
                indent_level = max(indent_level - 1, 0)
                tmp = (indent * indent_level) + item + "\n"

        elif (
            re.search(
                r"^" + str(config.tag_unindent_line),
                item,
                re.IGNORECASE | re.MULTILINE | re.VERBOSE,
            )
            and is_block_raw is False
        ):
            tmp = (indent * (indent_level - 1)) + item + "\n"

        # if indent, move right

        # opening set tag
        elif (
            config.no_set_formatting is False
            and re.search(
                re.compile(
                    r"^([ ]*{%[ ]*?set)(?!.*%}).*$",
                    re.IGNORECASE | re.MULTILINE | re.VERBOSE,
                ),
                item,
            )
            and is_block_raw is False
            and in_set_tag is False
        ):
            tmp = (indent * indent_level) + item + "\n"
            indent_level = indent_level + 1
            in_set_tag = True

        # opening curly brace inside a set tag
        elif (
            config.no_set_formatting is False
            and re.search(
                re.compile(
                    r"(\{(?![^{}]*%[}\s])(?=[^{}]*$)|\[(?=[^\]]*$))",
                    re.IGNORECASE | re.MULTILINE | re.VERBOSE,
                ),
                item,
            )
            and is_block_raw is False
            and in_set_tag is True
        ):
            tmp = (indent * indent_level) + item + "\n"
            indent_level = indent_level + 1

        elif (
            re.search(
                re.compile(
                    r"^(?:" + str(config.tag_indent) + r")",
                    re.IGNORECASE | re.MULTILINE | re.VERBOSE,
                ),
                item,
            )
            and is_block_raw is False
        ):
            tmp = (indent * indent_level) + item + "\n"
            indent_level = indent_level + 1

        elif is_raw_first_line is True or (
            is_safe_closing_tag(config, item) and is_block_raw is False
        ):
            tmp = (indent * indent_level) + item + "\n"

        elif is_block_raw is True or not item.strip():
            tmp = item + "\n"

        # otherwise, just leave same level
        elif not config.preserve_leading_space:
            # if we are not trying to preserve indenting
            # on text, the add it now.
            tmp = (indent * indent_level) + item + "\n"
        else:
            tmp = item + "\n"

        # if a opening raw tag then start ignoring.. only if there is no closing tag
        # on the same line
        if is_ignored_block_opening(config, item):
            is_block_raw = True
            is_raw_first_line = False

        # if a normal tag, we can try to expand attributes
        elif is_block_raw is False:
            # get leading space, and attributes

            func = partial(format_attributes, config, item)

            tmp = re.sub(
                re.compile(
                    rf"(\s*?)(<(?:{config.indent_html_tags})\b)((?:\"[^\"]*\"|'[^']*'|{{[^}}]*}}|[^'\">{{}}\/])+?)(\s?/?>)",
                    re.VERBOSE | re.IGNORECASE,
                ),
                func,
                tmp,
            )

        # turn off raw block if we hit end - for one line raw blocks, but not an inline raw
        if is_ignored_block_closing(config, item) and (
            in_script_style_tag is False
            or (in_script_style_tag and is_script_style_block_closing(config, item))
        ):
            in_script_style_tag = False
            if not is_safe_closing_tag(config, item):
                ignored_level -= 1
                ignored_level = max(ignored_level, 0)
            if ignored_level == 0:
                is_block_raw = False

        beautified_code = beautified_code + tmp

    # try to fix internal formatting of set tag
    def format_data(config: Config, contents: str, tag_size: int, leading_space) -> str:
        try:
            # try to format the contents as json
            data = json.loads(contents)
            contents = json.dumps(
                data, trailing_commas=False, ensure_ascii=False, quote_keys=True
            )

            if tag_size + len(contents) >= config.max_line_length:
                # if the line is too long we can indent the json
                contents = json.dumps(
                    data,
                    indent=config.indent_size,
                    trailing_commas=False,
                    ensure_ascii=False,
                    quote_keys=True,
                )

        except:
            # was not json.. try to eval as set
            try:
                # if contents is a python keyword, do not evaluate it.
                evaluated = (
                    str(eval(contents)) if contents not in ["object"] else contents
                )
                # need to unwrap the eval
                contents = (
                    evaluated[1:-1]
                    if contents[:1] != "(" and evaluated[:1] == "("
                    else evaluated
                )
            except:
                contents = contents.strip()

        return (f"\n{leading_space}").join(contents.splitlines())

    def format_set(config: Config, html: str, match: re.Match) -> str:
        if inside_ignored_block(config, html, match):
            return match.group()

        leading_space = match.group(1)
        open_bracket = match.group(2)
        tag = match.group(3)
        close_bracket = match.group(5)
        contents = match.group(4).strip()
        contents_split = contents.split("=", 1)

        if len(contents_split) > 1:
            contents = (
                contents_split[0].strip()
                + " = "
                + format_data(
                    config,
                    contents_split[-1],
                    len(f"{open_bracket} {tag}  {close_bracket}"),
                    leading_space,
                )
            )

        return f"{leading_space}{open_bracket} {tag} {contents} {close_bracket}"

    def format_function(config: Config, html: str, match: re.Match) -> str:
        if inside_ignored_block(config, html, match):
            return match.group()

        leading_space = match.group(1)
        open_bracket = match.group(2)
        tag = match.group(3).strip()
        index = (match.group(5) or "").strip()
        close_bracket = match.group(6)
        contents = format_data(
            config,
            match.group(4).strip()[1:-1],
            len(f"{open_bracket} {tag}() {close_bracket}"),
            leading_space,
        )

        return f"{leading_space}{open_bracket} {tag}({contents}){index} {close_bracket}"

    if config.no_set_formatting is False:
        func = partial(format_set, config, beautified_code)
        # format set contents
        beautified_code = re.sub(
            re.compile(
                r"([ ]*)({%-?)[ ]*(set)[ ]+?((?:(?!%}).)*?)(-?%})",
                flags=re.IGNORECASE | re.MULTILINE | re.VERBOSE | re.DOTALL,
            ),
            func,
            beautified_code,
        )

    if config.no_function_formatting is False:
        func = partial(format_function, config, beautified_code)
        # format function contents
        beautified_code = re.sub(
            re.compile(
                r"([ ]*)({{-?\+?)[ ]*?((?:(?!}}).)*?\w)(\((?:\"[^\"]*\"|'[^']*'|[^\)])*?\)[ ]*)((?:\[[^\]]*?\]|\.[^\s]+)[ ]*)?((?:(?!}}).)*?-?\+?}})",
                flags=re.IGNORECASE | re.MULTILINE | re.VERBOSE | re.DOTALL,
            ),
            func,
            beautified_code,
        )

    if not config.preserve_blank_lines:
        beautified_code = beautified_code.lstrip()

    return beautified_code.rstrip() + "\n"