sarif-cli/bin/sarif-extract-multi

#!/usr/bin/env python
""" Extract data from multiple sarif files in table form.
"""
from dataclasses import dataclass
from sarif_cli import signature, signature_multi
from sarif_cli import typegraph
from sarif_cli import snowflake_id
import argparse
import csv
import dataclasses as dc
import json
import pandas as pd
import pathlib
import sarif_cli.table_joins as tj
import sys

#
# Start processing 
#
parser = argparse.ArgumentParser(description='Read a collection of sarif files and produce tabular output.')
parser.add_argument('file', metavar='sarif-files.json', type=str,
                    help="json file containing the metadata array.  Use - for stdin. ")
parser.add_argument('outdir', metavar='output-dir', type=str, help='output directory')
parser.add_argument('-c', '--combine-only', action="store_true",
                    help='Read the referenced input file(s) and write the combined structure to stdout')
args = parser.parse_args()

# Load meta info
with open(args.file, 'r') if args.file != '-' else sys.stdin as fp:
    meta_struct = json.load(fp)

# Attach referenced files
def load(fname):
    with open(fname, 'rb') as fp: 
        content = json.load(fp)
    return content

for sarif_meta in meta_struct:
    sarif_meta['sarif_content'] = load(sarif_meta['sarif_file_name'])

# Only output composite?
if args.combine_only:
    json.dump(meta_struct, sys.stdout, indent=4)
    sys.exit(0)
#
# Preprocess raw SARIF to get smaller signature
#
context = signature.Context(
    {
        "string" : "String",
        "int" : "Int",
        "bool" : "Bool"
    }
) 
meta_struct = signature.fillsig(args, meta_struct, context)
#
# Use reference type graph (signature) to traverse sarif and attach values to tables
#
tgraph = typegraph.Typegraph(signature_multi.struct_graph_2022_03_08)
typegraph.destructure(tgraph, signature_multi.start_node_2022_03_08, meta_struct)
#
# Form output tables
# 
typegraph.attach_tables(tgraph)

#
# Dataframe / table collection
# 
@dataclass
class BaseTables:
    artifacts : pd.DataFrame
    codeflows : pd.DataFrame
    kind_pathproblem : pd.DataFrame
    kind_problem : pd.DataFrame
    project : pd.DataFrame
    relatedLocations : pd.DataFrame
    rules : pd.DataFrame
    def __init__(self): pass

bt = BaseTables()
# 
# Add dataframes
# 
sf_2683 = tj.joins_for_location_info(tgraph)
af_0350_location = tj.joins_for_af_0350_location(tgraph)
bt.artifacts = tj.joins_for_artifacts(tgraph)
bt.codeflows = tj.joins_for_codeflows(tgraph, sf_2683)
bt.kind_pathproblem = tj.joins_for_path_problem(tgraph, af_0350_location)
bt.kind_problem = tj.joins_for_problem(tgraph, af_0350_location)
bt.project = tj.joins_for_project(tgraph) # multi-sarif only
bt.relatedLocations = tj.joins_for_relatedLocations(tgraph, sf_2683)
bt.rules = tj.joins_for_rules(tgraph)

#
# Replace the remaining internal ids with snowflake ids
# 
flakegen = snowflake_id.Snowflake(0)

columns_to_reindex = {
    # template from {field.name : [''] for field in dc.fields(bt)}
    'artifacts': ['artifacts_id'],
    'codeflows': ['codeflow_id'],
    'kind_pathproblem': ['results_array_id', 'codeFlows_id'],
    'kind_problem': ['results_array_id'],
    'project': ['artifacts', 'results', 'rules'],
    'relatedLocations': ['struct_id'],
    'rules': ['rules_array_id']}

_id_to_flake = {}
def _get_flake(id):
    flake = _id_to_flake.get(id, -1)
    if flake == -1:
        flake = flakegen.next()
        _id_to_flake[id] = flake
    return flake

# 
# Cleaner, but makes far too many copies; keep the loop below
# 
# def _reindex(table, colname):
#     newtable = table.astype({ colname : 'uint64'}).reset_index(drop=True)
#     for i in range(0, len(newtable)):
#         newtable.loc[i, colname] = _get_flake(newtable.loc[i, colname])
#     return newtable
# 
# for field in dc.fields(bt):
#     table_name = field.name   
#     for colname in columns_to_reindex[table_name]:
#         setattr(bt, field.name, _reindex(getattr(bt, field.name), colname))
# 
for field in dc.fields(bt):
    table_name = field.name   
    table = getattr(bt, field.name)
    # Turn all snowflake columns into uint64 and reset indexing to 0..len(table)
    newtable = table.astype(
        { colname : 'uint64'
          for colname in columns_to_reindex[table_name]}
    ).reset_index(drop=True)
    # Swap ids for flakes
    for colname in columns_to_reindex[table_name]:
        for i in range(0, len(newtable)):
            newtable.loc[i, colname] = _get_flake(newtable.loc[i, colname])
    # Replace the table
    setattr(bt, field.name, newtable)
#
# Write output
#
p = pathlib.Path(args.outdir)
p.mkdir(exist_ok=True)
def write(path, frame):
    with p.joinpath(path + ".csv").open(mode='wb') as fh:
        frame.to_csv(fh, index=False, quoting=csv.QUOTE_NONNUMERIC)
for field in dc.fields(bt):
    table = getattr(bt, field.name)
    write(field.name, table)