#!/usr/bin/env python
import argparse
import json
import sarif_cli.traverse as S
import sys

parser = argparse.ArgumentParser(description='list source files referenced by sarif file')
parser.add_argument('file', metavar='sarif-file', type=str,
                    help='input file, - for stdin')
args = parser.parse_args()

# Grab the file
with open(args.file, 'r') if args.file != '-' else sys.stdin as fp:
    sarif_struct = json.load(fp)

# File name collection
uris = set()

# TODO: list files from the relatedLocations property
# Traverse all runs
for runi in S.indices(sarif_struct, 'runs'):
    # Make sure there are some results
    num_results = len(S.get(sarif_struct, 'runs', runi, 'results'))
    if num_results == 0: continue

    # Collect the file names

    # Locations for @kind problem
    # e.g.,
    # sarif_struct['runs'][0]['results'][5]['locations'][0]['physicalLocation']['artifactLocation']
    for resi in S.indices(sarif_struct, 'runs', runi, 'results'):
        uri = S.get(sarif_struct, 'runs', runi, 'results', resi, 'locations', 0,
                    'physicalLocation', 'artifactLocation', 'uri')
        uris.add(uri)

    # Locations for @kind path-problem
    # e.g. sarif_struct['runs'][0]['results'][22]['codeFlows'][0]['threadFlows'][0]['locations'][1]['location']
    for resi in S.indices(sarif_struct, 'runs', runi, 'results'):
        if 'codeFlows' in S.get(sarif_struct, 'runs', runi, 'results', resi):
            locations = S.get(sarif_struct, 'runs', runi, 'results', resi, 'codeFlows', 0,
                              'threadFlows', 0, 'locations')
            for loci in range(0, len(locations)):
                uri = S.get(locations, loci, 'location', 'physicalLocation',
                            'artifactLocation', 'uri')
                uris.add(uri)
# Dump the file names
uris = list(uris)
uris.sort()
for u in uris:
    print(u)