#!c:\Program Files\Python39\python.exe
# -*- coding: UTF-8 -*-
import os, sys, json
from os.path import join, exists, getsize, splitext
from datetime import datetime
from urllib.parse import quote

# ***** include the biocase.lib directory in the python sys path for importing *****
exec(open(os.path.abspath(os.path.join(os.path.dirname(__file__), os.path.pardir, os.path.pardir, 'lib', 'biocase', 'adjustpath.py'))).read())

from biocase.wrapper.protocol.inventory.inventory_protocol import ProtocolClass as InventoryProtocolClass
from biocase.archive.general import read_inventory, list_archives_for_schema, drop_archives, ns_dataset_aware, sizeof_fmt, fmt
from biocase.configtool.general import dsaObj, form, dsa, schema, action, log
import biocase.configuration
cfg = biocase.configuration.Cfg()

archivePath = join(cfg.archiveWorkLocator, dsa) + os.sep
dldPath = join(cfg.archiveDownloadLocator, dsa) + os.sep
proc_dwca = archivePath + "DwCA.proc"
log_dwca = archivePath + "DwCA.log"
if schema is not None:
    config = archivePath + schema + ".config.xml"
    proc_xml = archivePath + schema + ".proc"
    log = archivePath + schema + ".log"
    faf = dldPath + schema + '.zip'


def pimp_export_fname(f):
    if len(f) == 19:
        return '%s-%s-%s %s:%s:%s' % (f[0:4], f[4:6], f[6:8], f[9:11], f[11:13], f[13:15])
    else:
        return f


print('Content-Type: text/html\n')
# Set encoding to UTF8 and encoding error handling to replace
# print(sys.stdout.encoding, hasattr(sys.stdout, 'buffer'), sys.stdout.errors)
sys.stdout.reconfigure(encoding='utf8', errors='replace')

# Print 30 last lines of log files
if action == "showlog":
    # First, print log
    if exists(log):
        size = getsize(log)
        f = open(log, encoding='utf8')
        # Only read 4KB at the end
        f.seek(max(size - 4096, 0))
        lines = f.readlines()
        f.close()
        # Restrict output to 25 lines
        if len(lines) > 30:
            sys.stdout.write("(continued)<br>")
        for line in lines[-30:]:
            line = line[:-1]
            if line != '' and line[:4] != "INFO":
                line = '<font color="%s">%s</font>' % ('#FF0000' if line.startswith('ERROR') else '#FF6600', line)
            # line += '&nbsp;<br>'
            # Use sys.out here in order to allow printing without newline or space
            print(line)
        # Add line that indicates finishing
        if not exists(proc_xml):
            if exists(proc_dwca):
                print("XML generation ended, conversion will be started...")
            else:
                print("... processing ended.")
    else:
        print(("Log file not found: " + log))

if action == "archiveexists":
    if list_archives_for_schema(dsa, schema):
        print("true")
    else:
        print("false")

if action == "list":
    url_rel = cfg.archiveDownloadLocator[len(cfg.wwwLocator) + 1:].replace('\\', '/')
    baseURL = cfg.getWebappURL() + url_rel + '/' + dsa + '/'

    # Downloads
    dld = [f for f in os.listdir(dldPath) if f.endswith('.zip')]

    # Then, list archives
    archives = []
    for aname, (schema, NS, dataset, records, dropped, size, modified, _dummy) in read_inventory(dsa).items():
        # URL-encode the name for the download link
        # This is a workaround for a bug in Python 2.7 urllib.quote (http://mail.python.org/pipermail/python-dev/2006-July/067248.html)
        if isinstance(aname, str):
            aname_encoded = aname.encode('utf-8')
        else:
            aname_encoded = aname
        # if NS is not ABCD2.06/HISPID5, remove it
        # This way, the interface can detect which archives an be converted to DwC-A
        srecs, sdropped = fmt(records, dropped if dropped else '')
        archives += [(dataset, schema, NS if NS in ns_dataset_aware else None, aname, baseURL + quote(aname_encoded),
                      datetime.fromtimestamp(modified).strftime("%y-%m-%d %H:%M"), sizeof_fmt(size), srecs, sdropped)]
        # Remove archives from the list of zip files (that will be the download list eventually)
        dld.remove(aname)

    # Proc and Log files of aborted runs
    log = []
    proc = None
    for fname in os.listdir(archivePath):
        fname_head = splitext(fname)[0]
        if fname.endswith('.proc'):
            proc = fname_head
        elif fname.endswith('.log'):
            addit = False
            if getsize(join(archivePath, fname)) > 0:  # Ignore empty files
                if exists(join(dldPath, fname)):
                    # Same file exists in download folder: Only list if it's not the same file
                    a = os.stat(join(archivePath, fname))
                    d = os.stat(join(dldPath, fname))
                    if (a.st_size != d.st_size) or (d.st_mtime - a.st_mtime > 5):
                        addit = True
                else:
                    # Doesn't exist in download folder: add to list
                    addit = True
                # Exception: If it's the log file of a download, don't add it
                if '<script>window.location=' in open(join(archivePath, fname), encoding='utf8').read():
                    addit = False
            if addit:
                log += [fname_head]

    # Make up result dictionary; only add incomplete log file if it is not the running process
    # Important archives need to be sorted by schema (and dataset title subsequently)
    res = {'archive': sorted(archives, key=lambda arch: arch[1] + arch[3]), 'proc': proc, 'log': [name for name in log if name != proc],
           'dld': [(pimp_export_fname(f), f, baseURL + f) for f in dld]}

    # Return dictionary as JSON
    #print res
    print(json.dumps(res))

if action == "list_datasets":
    try:
        invObj = InventoryProtocolClass(dsaObj, None)
        invObj.parseRequest(schema)
        invObj.processRequest()
        datasets = invObj.datasets
        if datasets == []:
            print("Dataset list could not be retrieved or is empty. \nPlease make sure that the DB connection is OK and that this mapping returns any results!")
        else:
            html = ""
            for dataset in datasets:
                html += "<option>%s</option>" % dataset.replace(" ", "&nbsp;")  # replace spaces with non-breakable spaces
            print(json.dumps(html))
    except Exception as err:
        print("Error trying to load dataset list: %s" % err.msg)

    #    print "<option>AlgaTerra</option> <option>AlgaTerraMovies</option> <option>Biosphärenpark Wienerwald - Wiener Steinhofgründe</option> <option selected>BoBO - Botanic Garden Berlin BDI Observations</option> <option>Colección de Monocotiledóneas Mexicanas (UAM-I)</option> <option>Herbarium Berolinense</option>"

if action == "remove":
    try:
        logfile = form.getvalue('logfile')
        filenames = form.getlist('filename')

        # If failed log for the schema is to be removed: easy
        if logfile == 'true':
            os.remove(archivePath + filenames[0] + '.log')
            print("Log file was removed.")

        else:
            if filenames:
                cnt = drop_archives(dsa, True, filenames)
            else:
                cnt = drop_archives(dsa, True, schema)
            print("%i file%s been removed." % (cnt, 's have' if cnt > 1 else ' has'))

    except Exception as err:
        print(err)