From a6a1294944d4ce02cbb429ca1aec07ca78d79a88 Mon Sep 17 00:00:00 2001
From: Carl Hetherington <cth@carlh.net>
Date: Thu, 27 May 2021 14:21:35 +0200
Subject: Tidy handling of _raw_xml.

Before this if we tried to get the XML of an encrypted asset we
would just get an empty string.  Now we get a boost::none which
means the verifier can avoid trying to check details of the XML
(and instead raise a warning that you are trying to verify data
that it cannot decrypt).
---
 src/verify.cc | 34 +++++++++++++++++++++++++---------
 1 file changed, 25 insertions(+), 9 deletions(-)

(limited to 'src/verify.cc')

diff --git a/src/verify.cc b/src/verify.cc
index 623794cc..40f7c0b6 100644
--- a/src/verify.cc
+++ b/src/verify.cc
@@ -728,7 +728,11 @@ verify_subtitle_asset (
 	/* Note: we must not use SubtitleAsset::xml_as_string() here as that will mean the data on disk
 	 * gets passed through libdcp which may clean up and therefore hide errors.
 	 */
-	validate_xml (asset->raw_xml(), xsd_dtd_directory, notes);
+	if (asset->raw_xml()) {
+		validate_xml (asset->raw_xml().get(), xsd_dtd_directory, notes);
+	} else {
+		notes.push_back ({VerificationNote::Type::WARNING, VerificationNote::Code::MISSED_CHECK_OF_ENCRYPTED});
+	}
 
 	auto smpte = dynamic_pointer_cast<const SMPTESubtitleAsset>(asset);
 	if (smpte) {
@@ -752,16 +756,20 @@ verify_closed_caption_asset (
 	/* Note: we must not use SubtitleAsset::xml_as_string() here as that will mean the data on disk
 	 * gets passed through libdcp which may clean up and therefore hide errors.
 	 */
-	validate_xml (asset->raw_xml(), xsd_dtd_directory, notes);
+	auto raw_xml = asset->raw_xml();
+	if (raw_xml) {
+		validate_xml (*raw_xml, xsd_dtd_directory, notes);
+		if (raw_xml->size() > 256 * 1024) {
+			notes.push_back ({VerificationNote::Type::BV21_ERROR, VerificationNote::Code::INVALID_CLOSED_CAPTION_XML_SIZE_IN_BYTES, raw_convert<string>(raw_xml->size()), *asset->file()});
+		}
+	} else {
+		notes.push_back ({VerificationNote::Type::WARNING, VerificationNote::Code::MISSED_CHECK_OF_ENCRYPTED});
+	}
 
 	auto smpte = dynamic_pointer_cast<const SMPTESubtitleAsset>(asset);
 	if (smpte) {
 		verify_smpte_timed_text_asset (smpte, reel_asset_duration, notes);
 	}
-
-	if (asset->raw_xml().size() > 256 * 1024) {
-		notes.push_back ({VerificationNote::Type::BV21_ERROR, VerificationNote::Code::INVALID_CLOSED_CAPTION_XML_SIZE_IN_BYTES, raw_convert<string>(asset->raw_xml().size()), *asset->file()});
-	}
 }
 
 
@@ -772,7 +780,7 @@ verify_text_timing (
 	int edit_rate,
 	vector<VerificationNote>& notes,
 	std::function<bool (shared_ptr<Reel>)> check,
-	std::function<string (shared_ptr<Reel>)> xml,
+	std::function<optional<string> (shared_ptr<Reel>)> xml,
 	std::function<int64_t (shared_ptr<Reel>)> duration
 	)
 {
@@ -823,6 +831,12 @@ verify_text_timing (
 			continue;
 		}
 
+		auto reel_xml = xml(reels[i]);
+		if (!reel_xml) {
+			notes.push_back ({VerificationNote::Type::WARNING, VerificationNote::Code::MISSED_CHECK_OF_ENCRYPTED});
+			continue;
+		}
+
 		/* We need to look at <Subtitle> instances in the XML being checked, so we can't use the subtitles
 		 * read in by libdcp's parser.
 		 */
@@ -832,7 +846,7 @@ verify_text_timing (
 		optional<Time> start_time;
 		try {
 			doc = make_shared<cxml::Document>("SubtitleReel");
-			doc->read_string (xml(reels[i]));
+			doc->read_string (*reel_xml);
 			tcr = doc->number_child<int>("TimeCodeRate");
 			auto start_time_string = doc->optional_string_child("StartTime");
 			if (start_time_string) {
@@ -840,7 +854,7 @@ verify_text_timing (
 			}
 		} catch (...) {
 			doc = make_shared<cxml::Document>("DCSubtitle");
-			doc->read_string (xml(reels[i]));
+			doc->read_string (*reel_xml);
 		}
 		parse (doc, tcr, start_time, edit_rate, i == 0);
 		auto end = reel_offset + duration(reels[i]);
@@ -1605,6 +1619,8 @@ dcp::note_to_string (VerificationNote note)
 		DCP_ASSERT (parts.size() == 2);
 		return String::compose("The reel duration of some timed text (%1) is not the same as the ContainerDuration of its MXF (%2).", parts[0], parts[1]);
 	}
+	case VerificationNote::Code::MISSED_CHECK_OF_ENCRYPTED:
+		return "Some aspect of this DCP could not be checked because it is encrypted.";
 	}
 
 	return "";
-- 
cgit v1.2.3