import os, re, datetime
from xmlrpc.client import Boolean
from pingcap.pe.flash.markdown_processor import parse_markdown_folder
from pingcap.pe.flash.util import load_1gram_terms_en, pretty_format_title


# Keyword chapters chain
keyword_and_chapters = {}


def _get_table_header(lang_code: str) -> list:

    title_suffix = None
    column_header = []

    if lang_code[-2:] == "zh":
        title_suffix = "关键字快照"
        column_header = ["| 关键字 | 章节:幻灯片号.[幻灯片标题] | 关键字 | 章节:幻灯片号.[幻灯片标题] |\n"]
    elif lang_code[-2:] == "en":
        title_suffix = "Keywords Snapshot"
        column_header = [
            "| Keyword | Chapter:Slide_No.[Slide_Title] | Keyword | Chapter:Slide_No.[Slide_Title] |\n"
        ]

    output_lines = [
        "# "
        + pretty_format_title(
            os.path.basename(os.getcwd()).replace("-", " "), load_1gram_terms_en("pe")
        )
        + " "
        + title_suffix
        + "\n"
    ]

    s_now = str(datetime.datetime.now())
    output_lines += ["+ " + s_now + "\n\n"]

    output_lines += column_header

    output_lines += ["| :--   | :-- | :--   | :--  |\n"]

    return output_lines


def _seek_keyword_in_line(
    keyword: str, line: str, chapter_folder: str, view: tuple
) -> Boolean:
    if "\\" in keyword:
        keyword = keyword.replace("\\", "\\\\")
    m = re.findall("__\*" + keyword.strip().lower() + "\*__", line.strip().lower())
    if m != None and len(m) > 0:
        slide_lines = view[0]
        for slide_line in slide_lines:
            if slide_line[0] in ["title:slide"]:
                slide_title = slide_line[1]
                internal_slide_no = view[2].replace(".md", "")
        if "\\\\" in keyword:
            keyword = keyword.replace("\\\\", "\\")
        keyword_and_chapters[keyword] += [
            chapter_folder.strip().replace("chapter-0", "").replace("chapter-", "")
            + ":"
            + str(internal_slide_no)
            + ".["
            + slide_title.strip()
            + "]"
        ]
        return True
    else:
        return False


def dump_keyword_table(lang_code: str, keyword_list: list) -> None:
    """
    Dump a Markdown table for knowledge touched from a PE course structure.

    :param lang_code: en, en2zh, zh, zh2en.
    :param keyword_list: A list of keywords as the knowledge point.

    :returns: None.

    """

    # Table header
    course_folder_base = "student-guide/view/" + lang_code
    course_chapter_folders = sorted(os.listdir(course_folder_base))
    output_lines = _get_table_header(lang_code)

    # Driving keyword
    for keyword in keyword_list:
        keyword_and_chapters.update({keyword: []})
        # Driven chapters
        for chapter_folder in course_chapter_folders:
            found_in_chapter = False
            chapter_folder_pattern_matches = re.findall(
                "^chapter-[0-9][0-9]", chapter_folder
            )
            if (
                chapter_folder_pattern_matches != None
                and len(chapter_folder_pattern_matches) > 0
                and (not chapter_folder.endswith("-lms"))
            ):
                parsed_chapter = parse_markdown_folder(
                    course_folder_base + "/" + chapter_folder
                )
                for view in parsed_chapter:
                    # Found in slide lines, proceed to next chapter
                    # if found_in_chapter:
                    #  break
                    # Slide lines
                    for slide_lines in view[0]:
                        if (
                            slide_lines[1] != None
                            and len(slide_lines[1].strip()) > 0
                            and not "/" in slide_lines[1]
                        ):
                            found_in_chapter = _seek_keyword_in_line(
                                keyword, slide_lines[1], chapter_folder, view
                            )
                            if found_in_chapter:
                                print(keyword, "\t-> Found in slide of", chapter_folder)
                                break
                    if not found_in_chapter:
                        # Speaker notes if nothing found in slide lines
                        for note_line in view[1]:
                            if note_line != None and len(note_line.strip()) > 0:
                                found_in_chapter = _seek_keyword_in_line(
                                    keyword, note_line, chapter_folder, view
                                )
                                if found_in_chapter:
                                    print(
                                        keyword,
                                        "\t-> Found in speaker notes of",
                                        chapter_folder,
                                    )
                                    break

    # Write result to file
    s_now = str(datetime.datetime.now())
    dest_file = (
        "metadata/raw/keywords-"
        + os.path.basename(os.getcwd())
        + "-"
        + lang_code
        + "-"
        + s_now
        + ".md"
    )
    f = open(dest_file, "w")
    key_index = 0

    for k in keyword_and_chapters.keys():
        if keyword_and_chapters[k] != None and len(keyword_and_chapters[k]) > 0:
            key_index += 1
            if key_index % 2 == 0:
                output_lines += [
                    " | "
                    + k
                    + " | "
                    + "; ".join(sorted(keyword_and_chapters[k]))
                    + " |\n"
                ]
            else:
                output_lines += [
                    "| " + k + " | " + "; ".join(sorted(keyword_and_chapters[k]))
                ]

    f.write("".join(output_lines))
    f.close()

    print("\nPlease check", dest_file, "\n")