# Copyright (C) 2018 Robin Krahl <robin.krahl@ireas.org>
# SPDX-License-Identifier: MIT

import mimetypes
import re
import urllib.request

import bibtexparser
import pdftotext
import PyPDF2.pdf


PDF_INFO_DOI = '/doi'


def _extract_doi(filename):
    with open(filename, 'rb') as f:
        reader = PyPDF2.pdf.PdfFileReader(f)
        if PDF_INFO_DOI in reader.documentInfo:
            return reader.documentInfo[PDF_INFO_DOI]

    with open(filename, 'rb') as f:
        pdf = pdftotext.PDF(f)

    if len(pdf):
        text = pdf[0]

        pattern = re.compile(r'\b(10\.\d{4,}(?:\.\d+)*/(?:(?!["&\'<>])\S)+)\b')
        m = pattern.search(text)
        if m:
            return m.group(1)

    return None


def _get_data_for_doi(doi):
    url = 'http://dx.doi.org/{}'.format(doi)
    headers = {'Accept': 'text/bibliography; style=bibtex'}
    request = urllib.request.Request(url, headers=headers)
    with urllib.request.urlopen(request) as response:
        bibtex = response.read().decode('utf-8')
    return bibtexparser.loads(bibtex)


def _handle_pdf(filename):
    doi = _extract_doi(filename)

    if not doi:
        raise Exception('Could not extract a DOI from the PDF file')

    return _get_data_for_doi(doi)


def get_bibtex_data(filename):
    mtype, _ = mimetypes.guess_type(filename)
    if mtype == 'application/pdf':
        return _handle_pdf(filename)
    raise Exception('Unsupported file type ' + str(mtype))