#!/usr/bin/env python

# Scrubs the taxonpaths in VegBIEN using TNRS.

# Runs continuously until no new rows are added after max_pause.

import os.path

import StringIO

import sys

import time

sys.path.append(os.path.dirname(__file__)+"/../lib")

import csvs

import opts

import sql

import sql_gen

import sql_io

import streams

import strings

import tnrs

# Config

pause = 60 # sec

max_pause = 2*60*60 # sec; = 2 hr; must be >= max import time of one partition

assert pause <= max_pause

tnrs_data = sql_gen.Table('tnrs')

def main():

    # Input

    env_names = []

    db_config = opts.get_env_vars(sql.db_config_names, None, env_names)

    verbosity = float(opts.get_env_var('verbosity', 3, env_names))

    if not 'engine' in db_config: raise SystemExit('Usage: '

        +opts.env_usage(env_names)+' '+sys.argv[0]+' 2>>log')

    def log(msg, level=1):

        '''Higher level -> more verbose'''

        if level <= verbosity:

            sys.stderr.write(strings.to_raw_str(msg.rstrip('\n')+'\n'))

    # Connect to DB

    db = sql.connect(db_config, log_debug=log)

    # Iterate over unscrubbed verbatim taxonpaths

    start = 0

    total_pause = 0

    while True:

        # Fetch next set

        cur = sql.select(db, 'taxonpath', ['taxonomicnamewithauthor'],

            [('canon_taxonpath_id', None)], limit=tnrs.max_taxons, start=start,

            cacheable=False)

        this_ct = cur.rowcount

        start += this_ct # advance start to fetch next set

        if this_ct == 0:

            total_pause += pause

            if total_pause > max_pause: break

            log('Waited '+str(total_pause)+' sec. Waiting...')

            time.sleep(pause) # wait for more rows

            continue # try again

        # otherwise, rows found

        total_pause = 0

        # Run TNRS

        stream = tnrs.tnrs_request(list(sql.values(cur)))

        stream_info = csvs.stream_info(stream, parse_header=True)

        stream = streams.ProgressInputStream(stream, sys.stderr, n=1000)

        sql_io.append_csv(db, tnrs_data, stream_info, stream)

main()