"""
Panflute filter to allow file includes
"""

import os
import json
import glob
import re
import itertools
from pathlib import Path

import panflute as pf
import lxml.etree as xml

from natsort import natsorted
from urllib.parse import urlparse

from .format_heuristics import formatFromPath
from .config import parseConfig, parseOptions, TEMP_FILE, Env


# Global variables
INCLUDE_INVALID  = 0
INCLUDE_FILE     = 1
INCLUDE_HEADER   = 2

# Regex patterns
RE_IS_INCLUDE_HEADER  = r"(\\?(!|\$))include-header"
RE_IS_INCLUDE_LINE    = r"^(\\?(!|\$))include(-header)?"
RE_INCLUDE_PATTERN    = r"^(\\?(!|\$))include(-header)?(\`(?P<args>[^\`]+(, ?[^\`]+)*)\`)? ((?P<fname>[^\`\'\"]+)|([\`\'\"])(?P<fnamealt>.+)\9)$"

# Inherited options
options = None

# parse env config
Env.parse()

def extract_info(rawString):
    global options

    includeType = INCLUDE_INVALID
    config = {}
    filename = None

    # wildcards '*' are escaped which needs to be undone because of path globing
    # convert_text has a tendency to produce multiline text which can not be matched correctly
    # Also here we should unescape underscores from markdown_strict.
    rawString = rawString.replace('\\*', '*').replace('\n', ' ').replace('\\_', '_')

    if re.match(RE_IS_INCLUDE_HEADER, rawString):
        includeType = INCLUDE_HEADER
    else:
        includeType = INCLUDE_FILE

    matches = re.match(RE_INCLUDE_PATTERN, rawString)
    if not matches:
        # Pattern was not able to extract args and file glob... Hence, abort
        raise ValueError(f"Unable to extract info from include line {rawString}")

    groups = matches.groupdict()

    # Get filename from Regex capture group
    filename = groups.get('fname', None)
    if not filename:
        filename = groups.get('fnamealt', None)

    # Get args from RegEx capture group
    if 'args' in groups and groups['args']:
        config = parseConfig(groups['args'])

    if not filename:
        raise ValueError(f"Unable to extract info from include line {rawString}")

    return includeType, filename, config

def is_include_line(elem, raw=False):
    # Revert to Markdown for regex matching
    if not raw:
        rawString = pf.convert_text(
            elem,
            input_format='panflute',
            output_format='markdown_strict',
            standalone=True,
            pandoc_path=Env.PandocBin
        )
    else:
        rawString = elem

    includeType = INCLUDE_INVALID
    config = {}
    name = None

    if re.match(RE_IS_INCLUDE_LINE, rawString):
        includeType, name, config = extract_info(rawString)

    return includeType, name, config


def is_code_include(elem):
    includeType, name, config = is_include_line(elem.text, raw=True)
    if includeType == INCLUDE_HEADER:
        pf.debug("[WARN] Invalid !include-header in code blocks")
        includeType = INCLUDE_INVALID

    return includeType, name, config


# Skip whitespaces until newline
def skipWhitespaces(content):
    whiteSpaceReg = re.compile(r"[^\s]|\n")
    m = whiteSpaceReg.search(content)
    if m == None:
        return None
    pos = m.span()[0]
    if content[pos] == "\n":
        pos += 1
    return pos

def removeLeadingWhitespaces(s, num):
    regex = re.compile(r"[^\s]")
    m = regex.search(s)
    if m == None:
        return
    pos = m.span()[0]
    if num < 0:
        return s[pos:]
    else:
        return s[min(pos, num):]

def dedent(content: str, num):
    lines = content.split("\n")
    return list(map(lambda s: removeLeadingWhitespaces(s, num), lines))


def findFile(filename: str):
    resource_paths = options['include-resources'].split(':')

    files = glob.glob(filename, recursive=True)
    if len(files) == 0 and resource_paths:
        for resource_path in resource_paths:
            if os.path.isabs(resource_path):
                files += glob.glob(os.path.normpath(os.path.join(resource_path, filename)), recursive=True)
            else:
                files += glob.glob(os.path.normpath(os.path.join(options['process-path'], resource_path, filename)), recursive=True)

    return files


def read_file(filename, config: dict):
    with open(filename, encoding="utf-8") as f:
        content = f.read()

    if "xslt" in config:
        xsltParam = config.get("xslt", None)

        if not xsltParam:
            raise ValueError(f"Invalid value for xsl file: '{xsltParam}'")

        xslTransformerFile = findFile(xsltParam)
        if len(xslTransformerFile) == 0:
            raise ValueError(f"xsl transformer file not found: '{xsltParam}'")
        elif len(xslTransformerFile) > 1:
            raise ValueError(f"Ambiguous xsl transformer file: '{xsltParam}'")
        else:
            xslTransformerFile = xslTransformerFile[0]

        pf.debug(f"[INFO] xslt transform {filename} with {xslTransformerFile}")

        dom = xml.parse(filename, xml.XMLParser(recover=True))

        xslt = xml.parse(xslTransformerFile)
        if not xslt:
            raise IOError("Unable to read XSLT file '{xslt}'")
        transform = xml.XSLT(xslt)
        transformedDom = transform(dom)

        content = str(transformedDom)

    if "startLine" in config or "endLine" in config:
        lines = content.split("\n")
        startLine = config.get("startLine", 1) - 1
        endLine = config.get("endLine", len(lines))
        # count from the end of file
        if startLine < 0:
            startLine += len(lines)
        if endLine < 0:
            endLine += len(lines) + 1
        result = lines[startLine:endLine]
        content = "\n".join(result)

    if "snippetStart" in config or "snippetEnd" in config:
        start = 0
        length = len(content)
        snippets = []
        includeSnippetDelimiters = config.get("includeSnippetDelimiters", False)

        while start < length:
            if "snippetStart" in config:
                pos = content.find(config["snippetStart"], start)
            else:
                pos = -1
            if pos != -1:
                start = pos
            else:
                # If not found for the first time, start from the beginning
                if start != 0:
                    break

            if not includeSnippetDelimiters:
                start += len(config.get("snippetStart", ""))
                # Skip whitespaces until newline
                pos = skipWhitespaces(content[start:])
                if pos == None:
                    break
                start += pos

            if "snippetEnd" in config:
                end = content.find(config["snippetEnd"], start)
            else:
                end = -1
            # no snippetEnd means the end of file
            if end == -1:
                snippets.append(content[start:])
                break

            if includeSnippetDelimiters:
                end += len(config.get("snippetEnd", ""))
                subEnd = end
            else:
                # Skip whitespaces until newline
                pos = skipWhitespaces(content[start:end][::-1])
                if pos == None:
                    subEnd = end
                else:
                    subEnd = end - pos

            snippets.append(content[start:subEnd])
            start = end
        content = "\n".join(snippets)

    if "dedent" in config:
        content = "\n".join(dedent(content, config["dedent"]))

    return content


def action(elem, doc):
    global options

    # Try to read inherited options from temp file
    if options is None:
        options = parseOptions(doc)

    # Change dir to entry file
    entry = options["include-entry"]
    if not entry["entered"]:
        os.chdir(entry["path"])
        entry["entered"] = True

    # --- Include statement ---
    if isinstance(elem, pf.Para):
        includeType, name, config = is_include_line(elem)

        if includeType == INCLUDE_INVALID:
            return

        files = findFile(name)
        if len(files) == 0:
            msg = f"Included file not found: {name}"
            if Env.NotFoundError:
                raise IOError(msg)
            else:
                pf.debug(f"[WARNING] {msg}")
                return

        # order
        include_order = options['include-order']
        if include_order == 'natural':
            files = natsorted(files)
        elif include_order == 'alphabetical':
            files = sorted(files)
        elif include_order == 'default':
            pass
        else:
            raise ValueError('Invalid file order: ' + include_order)

        elements = []
        for fn in files:
            pf.debug(f"[INFO] including file '{fn}'", end="", flush=True)
            if not os.path.isfile(fn):
                raise IOError(f"Included file not found: {fn}")
            pf.debug(f"... ok")

            raw = read_file(fn, config)

            # Save current path
            cur_path = os.getcwd()

            # Change to included file's path so that sub-include's path is correct
            target = os.path.dirname(fn)
            # Empty means relative to current dir
            if not target:
                target = '.'

            currentPath = options["current-path"]
            options["current-path"] = os.path.normpath(os.path.join(currentPath, target))
            os.chdir(target)

            # pass options by temp files
            with open(TEMP_FILE, 'w+') as f:
                json.dump(options, f)

            # Add recursive include support
            new_elems = None
            new_metadata = None
            if includeType == 1:
                # Set file format
                if "format" in config:
                    fmt = config["format"]
                else:
                    fmt = formatFromPath(fn)
                # default use markdown
                if fmt is None:
                    fmt = "markdown"

                # copy since pf will modify this argument
                pandoc_options = list(options["pandoc-options"])

                if "raw" in config:
                    rawFmt = config.get("raw")
                    # raw block
                    new_elems = [pf.RawBlock(raw, format=rawFmt)]
                else:
                    new_doc = pf.convert_text(
                        raw,
                        input_format=fmt,
                        standalone=True,
                        extra_args=pandoc_options,
                        pandoc_path=Env.PandocBin
                    )

                    new_metadata = new_doc.get_metadata(builtin=False)
                    new_elems = new_doc.content.list

            else:
                # Read header from yaml
                # Use pf to preserve all info
                new_metadata = pf.convert_text(
                    f"---\n{raw}\n---",
                    standalone=True,
                    pandoc_path=Env.PandocBin
                ).get_metadata(builtin=False)

            # Merge metadata
            if new_metadata is not None:
                for key in new_metadata.content:
                    if not key in doc.metadata.content:
                        doc.metadata[key] = new_metadata[key]

            # delete temp file (the file might have been deleted in subsequent executions)
            if os.path.exists(TEMP_FILE):
                os.remove(TEMP_FILE)
            # Restore to current path
            os.chdir(cur_path)
            options["current-path"] = currentPath

            # incremement headings
            increment = config.get('incrementSection', 0)

            if increment:
                for new_elem in new_elems:
                    if isinstance(new_elem, pf.Header):
                        new_elem.level += increment

            if new_elems != None:
                elements += new_elems

        return elements

    # --- Code Blocks ---
    elif isinstance(elem, pf.CodeBlock):
        includeType, name, config = is_code_include(elem)
        if includeType == 0:
            return

        # Enable shell-style wildcards
        files = findFile(name)
        if len(files) == 0:
            msg = f"Included file not found: {name}"
            if Env.NotFoundError:
                raise IOError(msg)
            else:
                pf.debug(f"[WARNING] {msg}")
                return

        codes = []
        for fn in files:
            codes.append(read_file(fn, config))

        elem.text = "\n".join(codes)

    # --- Images ---
    elif isinstance(elem, pf.Image):
        rewritePath = options.get("rewrite-path", True)
        if not rewritePath:
            return

        url = elem.url
        # try to parse the url first
        result = urlparse(url)
        # url
        if result.scheme != "":
            return
        # absolute path
        if os.path.isabs(url):
            return

        # rewrite relative path
        elem.url = str(
            Path(options["include-entry"]["path"])
                .joinpath(options["current-path"])
                .joinpath(url)
        )


def main(doc=None):
    return pf.run_filter(action, doc=doc)


if __name__ == '__main__':
    main()