"""
* Version: 0.1.0.0.0

* Purposes:
    * This is a result-driven style English to Simplified Chinese translation tool which is better than the static dictionary only style, it blocks the wrong translation while still allows the translation flexibility for auto translator.
    * This is an intenral utilities specifically for processing English to Simplified Chinese translation for pingcap.pe.flash module.
"""

import re, csv, importlib_resources
import pingcap.pe.flash.util as util
import pingcap.pe.flash.util_translator as translator
import pingcap.pe.flash.ai_editor as editor

_linematch_replace_resultmatch_ci = [
    e[1:]
    for e in [
        *csv.reader(
            importlib_resources.open_text(
                "pingcap.pe.flash", "flash-en2zh-blocklist-dictionary-ci.csv"
            )
        )
    ]
]

_linematch_replace_resultmatch_cs = [
    e[1:]
    for e in [
        *csv.reader(
            importlib_resources.open_text(
                "pingcap.pe.flash", "flash-en2zh-blocklist-dictionary-cs.csv"
            )
        )
    ]
]

_errata_cs = [
    e[1:]
    for e in [
        *csv.reader(
            importlib_resources.open_text(
                "pingcap.pe.flash", "flash-en2zh-static-errata-cs.csv"
            )
        )
    ]
]


def _line_lower_result_replace(
    line, result, line_match: str, result_match: str, result_replacement: str
) -> str:
    if line_match in line.lower() and result_match in result:
        return result.replace(result_match, result_replacement)
    else:
        return result


def _line_lower_result_replace_multi(
    line, result, line_match_lower: str, results_match: list, result_replacement: str
) -> str:
    for r in results_match:
        result = _line_lower_result_replace(
            line, result, line_match_lower, r, result_replacement
        )
    additional_result_match = []
    additional_lower_result_match = line_match_lower.split(" ")
    # All lower case:
    additional_result_match += [" ".join(additional_lower_result_match)]
    # All title case:
    additional_result_match += [
        " ".join(e.title() for e in additional_lower_result_match)
    ]
    if " " in line_match_lower:
        # Dash connected case:
        additional_line_match_lower = line_match_lower.replace(" ", "-")
        result = _line_lower_result_replace_multi(
            line, result, additional_line_match_lower, results_match, result_replacement
        )
        # First word title case:
        additional_result_match += [
            additional_lower_result_match[0].title()
            + " "
            + " ".join(additional_lower_result_match[1:])
        ]
    for x in additional_result_match:
        result = _line_lower_result_replace(
            line, result, line_match_lower, x, result_replacement
        )
    return result


def _line_result_replace(
    line, result, line_match: str, result_match: str, result_replacement: str
) -> str:
    if line_match in line and result_match in result:
        return result.replace(result_match, result_replacement)
    else:
        return result


def _result_replace(result, result_match: str, result_replacement: str) -> str:
    if result_match in result:
        return result.replace(result_match, result_replacement)
    else:
        return result


def _translate_cache_en2zh(line: str) -> str:
    """Return the cached Markdown translated result for English Markdown to Simplified Chinese Markdown.

    Args:
        line (str): The English source Markdown line.

    Returns:
        str: The cached Markdown translation result.
    """
    if (
        line.strip("\n").strip() == "No speaker notes."
        or line.strip("\n").strip() == "[no notes]"
    ):
        return "[无内容]"
    if line.strip("\n").strip() == "## Overview":
        return "## 概述"
    if line.strip("\n").strip() == "## Notes":
        return "## 注意事项"
    if line.strip("\n").strip() == "## Duration":
        return "## 持续时间"
    if (
        line.strip("\n").strip() == "## Solution Steps"
        or line.strip("\n").strip() == "## Steps"
    ):
        return "## 步骤"
    if line.strip("\n").strip() == "## Outline":
        return "## 大纲"
    if line.strip("\n").strip() == "## Domain Details":
        return "## 知识领域"
    if line.strip("\n").strip() == "# Objectives":
        return "# 章节目标"
    return None


def pre_en2zh_auto_translation_md_prefix_aware(line_in: str) -> str:
    """
    This routine currently is NOT in-use.
    Args:
        line_in (str): The English line to be translated.

    Returns:
        str: The English line refined for translating into Simplified Chinese.
    """
    processed_line = None
    processed_line = line_in
    return processed_line


def post_en2zh_auto_translation(
    line: str, auto_translated_line: str, mode="standard", md_file_path=None
) -> str:
    """
    This routine will be called in a context WITHOUT the Markdown prefix such as "#, +, -, *", etc.
    :param line: The English source line.
    :param auto_translated_line: The translated line by auto translation service.
    :returns: The post processed result.
    """
    result = auto_translated_line.strip("\n")
    print(util.speak_result("Begin Post Handling for Input: " + result))

    # 1. Placeholder.
    print(util.speak_result("Post Phase 1 Completed: " + result))

    # 2. Placeholder.
    print(util.speak_result("Post Phase 2 Completed: " + result))

    # 4. Correct case sensitive translation result by a static dictionary.
    # Load `line_match_cs, replacement, result_match_cs` from CSV.
    for e in _linematch_replace_resultmatch_cs[1:]:
        if not type(e[2]) == str:
            for i in e[2]:
                result = _line_result_replace(line, result, e[0], i, e[1])
        else:
            result = _line_result_replace(line, result, e[0], e[2], e[1])
    print(util.speak_result("Post Phase 4 Completed: " + result))

    # 5. Correct case insensitive translation result.
    # Load `line_match_ci, replacement, result_match_cs` from CSV.
    for e in _linematch_replace_resultmatch_ci[1:]:
        result = _line_lower_result_replace_multi(
            line, result, e[0], e[2].split(","), e[1]
        )
    print(util.speak_result("Post Phase 5 Completed: " + result))

    # 6. Correct complex translation result. Goal keeper solutions for those issues that can not be solved in step 5 completely.
    if "key-value pair" in line.lower():
        result = result.replace("键值对", "Key-Value pair")

    if "key-value" in line.lower():
        result = result.replace("键-值", "Key-Value")

    if (
        "key" in line.lower()
        and "keyword" not in line.lower()
        and "point" not in line.lower()
    ):
        if "关键词" in result:
            result = result.replace("关键词", "KEY")
        if "密钥" in result:
            result = result.replace("密钥", "KEY")
        if "关键字" in result:
            result = result.replace("关键字", "KEY")
        if "钥匙" in result:
            result = result.replace("钥匙", "KEY")
        if "键" in result:
            result = result.replace("键", "KEY")

    if "tidb/tikv/pd" in line.lower():
        result = re.sub("tidb/tikv/pd", "TiDB/TiKV/PD", result, flags=re.IGNORECASE)

    print(util.speak_result("Post Phase 6 Completed: " + result))

    # 7. REMOVERS:
    # Remove trailing "。" for bullets
    if result.strip().startswith("+"):
        result = result.strip("。")
    if result.strip().startswith("#"):
        result = result.strip("。")
    # Remove the generated **
    if "**" in result and "**" not in line:
        result = result.replace("**", "")
    # Remove the unwanted "（PD 一乘一）" at the end
    if re.match("^`[^`]+` *（PD 一乘一）[。]*$", result):
        result = result.replace("（PD 一乘一）", "")
    print(util.speak_result("Post Phase 7 Completed: " + result))
    return result


def post_en2zh_auto_translation_md_prefix_aware(in_line: str, source_line: str) -> str:
    """
    This routine will be called in a context with the Markdown prefix such as "#, +, -, *", etc.
    """
    result = in_line
    if result.strip().startswith("+"):
        result = result.strip("。")
    if result.strip().startswith("#"):
        result = result.strip("。")
    if result.strip().startswith("["):
        result = result.strip("。")
    if not "(" in source_line and not ")" in source_line:
        result = re.sub("\(.*\)", "", result)
    if not '"' in source_line:
        result = result.replace('"', "")
    return result


def translate_en2zh_md_line(
    line: str, en_1gram_list: list, mode="standard", md4ppt=True, md_file_path=None
) -> str:
    """
    Translate markdown line from EN to JA.
    Here goes the logic:
        0. Translation cache.
        1. Short-cut the line in case translation is not needed.
        2. Pack the prefixed whitepsaces from the line, for example: `_____## Hello world`
        3. Pack the prefixed `#`, `+`, `-`, `+`, `*`, `$numbering` (whitespace included) from the line.
        4. Replace the inline codeline - `CODE` with placeholders.

        Translate the processed line.

        5. Put the inline codeline back.
        6. Post translation handling.
        7-12. Unpack `#`, `+`, `-`, `+`, `*`, `$numbering` and prefixed whitspaces back to the line.
    """
    # Source
    print(util.speak_result("SOURCE: [ " + line + " ]"))
    # Pre-process
    line = pre_en2zh_auto_translation_md_prefix_aware(line)
    print(util.speak_result("AFTER PRE-PROCESS MD-AWARE: [ " + line + " ]"))

    print(util.speak_result("BEFORE TRANSLATION: [ " + line + " ]"))

    # 0. Translation cache.
    cache_result = _translate_cache_en2zh(line)
    if cache_result != None:
        print(util.speak_result("Translate from cache: " + cache_result))
        return cache_result

    # 1. Do nothing: Shortcut.
    if translator.is_skippable(line):
        for e in _errata_cs[1:]:
            line = _result_replace(line, e[0], e[1])
        return line

    # Assign `line` to `compress_space_raw` to keep the line intact.
    compress_space_raw = line.strip("\n")

    # 2. Pack the prefixed whitespace
    prefix_whitespace_pack = translator.pack_prefix_whitespace(compress_space_raw)
    compress_space_raw = prefix_whitespace_pack["processed_line"]

    # 2.1 Markdown table processing
    if translator.is_table_row(line.strip("\n")):
        compress_space_raw = translator.translate_table_row(
            compress_space_raw, translate_en2zh_md_line, en_1gram_list
        )
        print(util.speak_result("Table row translated: " + compress_space_raw))
    else:
        # 3.1 Pack the prefixed hash
        prefix_hash_pack = translator.pack_prefix_hash(compress_space_raw)
        compress_space_raw = prefix_hash_pack["processed_line"]

        # 3.2 Pack the prefixed plus
        prefix_plus_pack = translator.pack_prefix_plus(compress_space_raw)
        compress_space_raw = prefix_plus_pack["processed_line"]

        # 3.3 Pack the prefixed minus
        prefix_minus_pack = translator.pack_prefix_minus(compress_space_raw)
        compress_space_raw = prefix_minus_pack["processed_line"]

        # 3.4 Pack the prefixed star
        prefix_star_pack = translator.pack_prefix_star(compress_space_raw)
        compress_space_raw = prefix_star_pack["processed_line"]

        # 3.5 Pack the prefixed number
        prefix_number_pack = translator.pack_prefix_number(compress_space_raw)
        compress_space_raw = prefix_number_pack["processed_line"]

        # 3.6 Do nothing: Shortcut.
        if not translator.is_skippable(compress_space_raw):

            # 4. Pack inline codeline
            inline_codeline_pack = translator.pack_inline_codeline(compress_space_raw)
            compress_space_raw = inline_codeline_pack["processed_line"]

            # Ready for translation
            before_translation = compress_space_raw.strip()

            # Translate
            translated_line = editor.translate_en2zh_by_openai(before_translation)
            print(
                util.speak_result(
                    "BEFORE OPENAI TRANSLATION: [ " + before_translation + " ]"
                )
            )
            print(
                util.speak_result(
                    "AFTER OPENAI TRANSLATION: [ " + translated_line + " ]"
                )
            )
            compress_space_raw = translated_line.strip()

            # 5. Inline code handling
            compress_space_raw = translator.unpack_inline_codeline(
                compress_space_raw, inline_codeline_pack
            )

            # Cut-off the whitespaces if any
            compress_space_raw = compress_space_raw.strip()

            # 6. Post translate
            print(
                util.speak_result(
                    "BEFORE POST-TRANSLATION: [ " + compress_space_raw + " ]"
                )
            )
            compress_space_raw = post_en2zh_auto_translation(
                before_translation, compress_space_raw, mode, md_file_path
            )
            print(
                util.speak_result(
                    "AFTER POST-TRANSLATION: [ " + compress_space_raw + " ]"
                )
            )

        # 7. Adding Hash title prefix back
        compress_space_raw = translator.unpack_prefix_hash(
            compress_space_raw, prefix_hash_pack, en_1gram_list, md4ppt
        )
        # 8. Adding plus prefix back
        compress_space_raw = translator.unpack_prefix_common(
            compress_space_raw, prefix_plus_pack
        )
        # 9. Adding minus prefix back
        compress_space_raw = translator.unpack_prefix_common(
            compress_space_raw, prefix_minus_pack
        )
        # 10. Adding star prefix back
        compress_space_raw = translator.unpack_prefix_common(
            compress_space_raw, prefix_star_pack
        )
        # 11. Adding number prefix back
        compress_space_raw = translator.unpack_prefix_common(
            compress_space_raw, prefix_number_pack
        )
    # 12. Give back the prefix whitespace
    compress_space_raw = translator.unpack_prefix_common(
        compress_space_raw, prefix_whitespace_pack
    )
    # 13. Correct the general results in all cases.
    for e in _errata_cs[1:]:
        compress_space_raw = _result_replace(compress_space_raw, e[0], e[1])
    # Finally
    print(util.speak_result("AFTER TRANSLATION: [ " + compress_space_raw + " ]"))
    if md4ppt:
        compress_space_raw = post_en2zh_auto_translation_md_prefix_aware(
            compress_space_raw, line
        )
        print(
            util.speak_result(
                "AFTER TRANSLATION MD-AWARE: [ " + compress_space_raw + " ]"
            )
        )
    return compress_space_raw