From 92d904ee100594df758b369bba65739357c83398 Mon Sep 17 00:00:00 2001
From: Michael Hohn <hohn@github.com>
Date: Mon, 6 Dec 2021 14:24:08 -0800
Subject: [PATCH] Add quick check to verify that input is serif

An occasional output from LGTM is
    {"code":404,"error":"The specified analysis could not be found"}

With this patch, the csv output is now
    "ERROR","invalid json contents %s","some-file.json"

and the plain text output becomes
    ERROR: invalid json contents in some-file.json
---
 bin/sarif-results-summary | 8 ++++++++
 sarif_cli/__init__.py     | 6 ++++++
 2 files changed, 14 insertions(+)

diff --git a/bin/sarif-results-summary b/bin/sarif-results-summary
index 5a4e361..127e751 100755
--- a/bin/sarif-results-summary
+++ b/bin/sarif-results-summary
@@ -23,6 +23,14 @@ with open(args.file, 'r') if args.file != '-' else sys.stdin as fp:
 if args.csv:
     cw = S.get_csv_writer()
 
+if not S.is_sarif_struct(sarif_struct):
+    if args.csv:
+        S.write_csv(cw, "ERROR", "invalid json contents %s", args.file)
+    else:
+        S.msg("ERROR: invalid json contents in %s\n" % (args.file))
+    S.dbg("invalid json contents in %s\n" % (args.file))
+    sys.exit(0)                 # No failure, just a warning
+
 for runi in S.indices(sarif_struct, 'runs'):
     num_results = len(S.get(sarif_struct, 'runs', runi, 'results'))
     if num_results == 0: continue
diff --git a/sarif_cli/__init__.py b/sarif_cli/__init__.py
index 94270e4..70cd3b3 100644
--- a/sarif_cli/__init__.py
+++ b/sarif_cli/__init__.py
@@ -18,6 +18,12 @@ class NoFile:
     whole file is to be used.
     """
 
+def is_sarif_struct(struct):
+    """A quick check to verify that `struct` is in fact a SARIF tree.
+    """
+    return type(struct) == dict and "$schema" in struct and \
+        "sarif" in struct["$schema"] and "version" in struct
+
 def get_csv_writer():
     """ Set up and return the default csv writer on stdout.
     """