bounds checking on subfiles

python/examples/play.py

@@ -14,99 +14,103 @@ base = Path('/mnt/sls_det_storage/matterhorn_data/aare_test_data/')
 
 f = File(base/'Moench03new/cu_half_speed_master_4.json')
 
+for i, frame in enumerate(f):
+    print(f'{i}', end='\r')
+print()
+
 
 from aare._aare import ClusterFinderMT, ClusterCollector, ClusterFileSink
 
 
-cf = ClusterFinderMT((400,400), (3,3), n_threads = 3)
+# cf = ClusterFinderMT((400,400), (3,3), n_threads = 3)
-# collector = ClusterCollector(cf)
+# # collector = ClusterCollector(cf)
-out_file = ClusterFileSink(cf, "test.clust")
+# out_file = ClusterFileSink(cf, "test.clust")
 
-for i in range(1000):
-    img = f.read_frame()
-    cf.push_pedestal_frame(img)
-print('Pedestal done')
-cf.sync()
-
-for i in range(100):
-    img = f.read_frame()
-    cf.find_clusters(img)
-
-
-# time.sleep(1)
-cf.stop()  
-time.sleep(1)
-print('Second run')
-cf.start()
-for i in range(100):
-    img = f.read_frame()
-    cf.find_clusters(img)
-
-cf.stop()
-print('Third run')
-cf.start()
-for i in range(129):
-    img = f.read_frame()
-    cf.find_clusters(img)
-
-cf.stop()
-out_file.stop()
-print('Done')
-
-
-cfile = ClusterFile("test.clust")
-i = 0
-while True:
-    try:
-        cv = cfile.read_frame()
-        i+=1
-    except RuntimeError:
-        break
-print(f'Read {i} frames') 
-
-
-
-
-# cf = ClusterFinder((400,400), (3,3))
 # for i in range(1000):
-#     cf.push_pedestal_frame(f.read_frame())
+#     img = f.read_frame()
+#     cf.push_pedestal_frame(img)
+# print('Pedestal done')
+# cf.sync()
 
-# fig, ax = plt.subplots()
+# for i in range(100):
-# im = ax.imshow(cf.pedestal())
+#     img = f.read_frame()
-# cf.pedestal()
+#     cf.find_clusters(img)
-# cf.noise()
+
+
+# # time.sleep(1)
+# cf.stop()  
+# time.sleep(1)
+# print('Second run')
+# cf.start()
+# for i in range(100):
+#     img = f.read_frame()
+#     cf.find_clusters(img)
+
+# cf.stop()
+# print('Third run')
+# cf.start()
+# for i in range(129):
+#     img = f.read_frame()
+#     cf.find_clusters(img)
+
+# cf.stop()
+# out_file.stop()
+# print('Done')
+
+
+# cfile = ClusterFile("test.clust")
+# i = 0
+# while True:
+#     try:
+#         cv = cfile.read_frame()
+#         i+=1
+#     except RuntimeError:
+#         break
+# print(f'Read {i} frames') 
 
 
 
-# N = 500
-# t0 = time.perf_counter()
-# hist1 = bh.Histogram(bh.axis.Regular(40, -2, 4000))
-# f.seek(0)
 
-# t0 = time.perf_counter()
+# # cf = ClusterFinder((400,400), (3,3))
-# data = f.read_n(N)
+# # for i in range(1000):
-# t_elapsed = time.perf_counter()-t0
+# #     cf.push_pedestal_frame(f.read_frame())
+
+# # fig, ax = plt.subplots()
+# # im = ax.imshow(cf.pedestal())
+# # cf.pedestal()
+# # cf.noise()
 
 
-# n_bytes = data.itemsize*data.size
 
-# print(f'Reading {N} frames took {t_elapsed:.3f}s {N/t_elapsed:.0f} FPS, {n_bytes/1024**2:.4f} GB/s')
+# # N = 500
+# # t0 = time.perf_counter()
+# # hist1 = bh.Histogram(bh.axis.Regular(40, -2, 4000))
+# # f.seek(0)
+
+# # t0 = time.perf_counter()
+# # data = f.read_n(N)
+# # t_elapsed = time.perf_counter()-t0
 
 
-# for frame in data:
+# # n_bytes = data.itemsize*data.size
-#     a = cf.find_clusters(frame)
 
-# clusters = cf.steal_clusters()
+# # print(f'Reading {N} frames took {t_elapsed:.3f}s {N/t_elapsed:.0f} FPS, {n_bytes/1024**2:.4f} GB/s')
-
-# t_elapsed = time.perf_counter()-t0
-# print(f'Clustering {N} frames took {t_elapsed:.2f}s  {N/t_elapsed:.0f} FPS')
 
 
-# t0 = time.perf_counter()
+# # for frame in data:
-# total_clusters = clusters.size
+# #     a = cf.find_clusters(frame)
 
-# hist1.fill(clusters.sum())
+# # clusters = cf.steal_clusters()
 
-# t_elapsed = time.perf_counter()-t0
+# # t_elapsed = time.perf_counter()-t0
-# print(f'Filling histogram with the sum of {total_clusters} clusters took: {t_elapsed:.3f}s, {total_clusters/t_elapsed:.3g} clust/s')
+# # print(f'Clustering {N} frames took {t_elapsed:.2f}s  {N/t_elapsed:.0f} FPS')
-# print(f'Average number of clusters per frame {total_clusters/N:.3f}')
+
+
+# # t0 = time.perf_counter()
+# # total_clusters = clusters.size
+
+# # hist1.fill(clusters.sum())
+
+# # t_elapsed = time.perf_counter()-t0
+# # print(f'Filling histogram with the sum of {total_clusters} clusters took: {t_elapsed:.3f}s, {total_clusters/t_elapsed:.3g} clust/s')
+# # print(f'Average number of clusters per frame {total_clusters/N:.3f}')

src/RawFile.cpp

@@ -278,6 +278,10 @@ void RawFile::get_frame_into(size_t frame_index, std::byte *frame_buffer, Detect
     if (n_subfile_parts != 1) {
         for (size_t part_idx = 0; part_idx != n_subfile_parts; ++part_idx) {
             auto subfile_id = frame_index / m_master.max_frames_per_file();
+            if (subfile_id >= subfiles.size()) {
+                throw std::runtime_error(LOCATION +
+                                         " Subfile out of range. Possible missing data.");
+            }
             frame_numbers[part_idx] =
                 subfiles[subfile_id][part_idx]->frame_number(
                     frame_index % m_master.max_frames_per_file());
@@ -311,6 +315,10 @@ void RawFile::get_frame_into(size_t frame_index, std::byte *frame_buffer, Detect
         for (size_t part_idx = 0; part_idx != n_subfile_parts; ++part_idx) {
             auto corrected_idx = frame_indices[part_idx];
             auto subfile_id = corrected_idx / m_master.max_frames_per_file();
+            if (subfile_id >= subfiles.size()) {
+                throw std::runtime_error(LOCATION +
+                                         " Subfile out of range. Possible missing data.");
+            }
 
             // This is where we start writing
             auto offset = (m_module_pixel_0[part_idx].y * m_cols +
@@ -343,6 +351,10 @@ void RawFile::get_frame_into(size_t frame_index, std::byte *frame_buffer, Detect
             auto pos = m_module_pixel_0[part_idx];
             auto corrected_idx = frame_indices[part_idx];
             auto subfile_id = corrected_idx / m_master.max_frames_per_file();
+            if (subfile_id >= subfiles.size()) {
+                throw std::runtime_error(LOCATION +
+                                         " Subfile out of range. Possible missing data.");
+            }
 
             subfiles[subfile_id][part_idx]->seek(corrected_idx % m_master.max_frames_per_file());
             subfiles[subfile_id][part_idx]->read_into(part_buffer, header);