import os
import re
import fiona
from tqdm import tqdm
import numpy as np
import pandas as pd
import geopandas as gpd
from shapely.geometry import MultiLineString
from survey_tools import call_lastools

mga56 = fiona.crs.from_epsg(28356)

input_dir = 'output/01-las-classified'
output_dir = 'output/11-shp-shorelines'

contour = 0.7
smoothing = 2
clean = 5  # any lines shorted that this are automatically removed

las_names = os.listdir(input_dir)
for las_name in tqdm(las_names):
    shp_name = las_name.replace('.las', '_0_7_contour.shp')
    call_lastools(
        'blast2iso',
        input=os.path.join(input_dir, las_name),
        output=os.path.join(output_dir, shp_name),
        args=['-iso', contour, '-smooth', smoothing, '-clean', clean])

# Consolidate all surveys for each beach
shp_names = [f for f in os.listdir(output_dir) if f.endswith('contour.shp')]

master = gpd.GeoDataFrame({'shp_name': shp_names})
master.geometry = [[] for s in shp_names]

# Get date information
master['beach'] = master['shp_name'].str.extract('(.*)_\d{8}')
yyyy = master['shp_name'].str.extract('(\d{4})')
mm = master['shp_name'].str.extract('\d{4}(\d{2})')
dd = master['shp_name'].str.extract('\d{6}(\d{2})')

# Use US date format (required by DSAS)
master['date'] = mm + '/' + dd + '/' + yyyy

# Read shapefiles
for i, row in master.iterrows():
    geom = gpd.read_file(os.path.join(output_dir, row['shp_name'])).geometry
    # Combine all lines into mulit line string
    row.geometry = MultiLineString(list(geom))

    # Update fields
    master.loc[i, 'uncertainty'] = 0

# Export summary
master.crs = {'init': 'epsg:28356'}
master.to_file(os.path.join(output_dir, 'all_beaches.shp'),
               driver='ESRI Shapefile')