Add some nice statistics output to tar2sqfs and gensquashfs

Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>

2 files changed, 74 insertions, 1 deletions

diff --git a/lib/Makemodule.am b/lib/Makemodule.am
index 5f0cdae..8258136 100644
--- a/lib/Makemodule.am
+++ b/lib/Makemodule.am
@@ -37,7 +37,7 @@ libsquashfs_a_SOURCES += lib/sqfs/data_writer.c lib/sqfs/write_xattr.c
 libsquashfs_a_SOURCES += include/data_writer.h
 libsquashfs_a_SOURCES += include/data_reader.h lib/sqfs/data_reader.c
 libsquashfs_a_SOURCES += lib/sqfs/write_export_table.c
-libsquashfs_a_SOURCES += lib/sqfs/read_table.c
+libsquashfs_a_SOURCES += lib/sqfs/read_table.c lib/sqfs/statistics.c
 
 libutil_a_SOURCES = lib/util/canonicalize_name.c lib/util/write_data.c
 libutil_a_SOURCES += lib/util/read_data.c include/util.h
diff --git a/lib/sqfs/statistics.c b/lib/sqfs/statistics.c
new file mode 100644
index 0000000..2bd4acf
--- /dev/null
+++ b/lib/sqfs/statistics.c
@@ -0,0 +1,73 @@
+/* SPDX-License-Identifier: GPL-3.0-or-later */
+#include "config.h"
+#include "highlevel.h"
+
+#include <stdio.h>
+
+void sqfs_print_statistics(fstree_t *fs, sqfs_super_t *super)
+{
+	size_t blocks_written = 0, duplicate_blocks = 0, sparse_blocks = 0;
+	size_t ratio, file_count = 0, file_dup_count = 0;
+	size_t frag_count = 0, frag_dup = 0;
+	size_t i, num_blocks, sparse;
+	uint64_t output_bytes = 0;
+	uint64_t input_bytes = 0;
+	file_info_t *fi;
+
+	for (fi = fs->files; fi != NULL; fi = fi->next) {
+		num_blocks = fi->size / fs->block_size;
+
+		if ((fi->size % fs->block_size) &&
+		    !(fi->flags & FILE_FLAG_HAS_FRAGMENT)) {
+			++num_blocks;
+		}
+
+		for (sparse = 0, i = 0; i < num_blocks; ++i) {
+			if (fi->blocks[i].size == 0)
+				sparse += 1;
+		}
+
+		if (fi->flags & FILE_FLAG_BLOCKS_ARE_DUPLICATE) {
+			duplicate_blocks += num_blocks - sparse;
+		} else {
+			blocks_written += num_blocks - sparse;
+		}
+
+		if ((fi->flags & FILE_FLAG_FRAGMENT_IS_DUPLICATE) &&
+		    (fi->flags & FILE_FLAG_BLOCKS_ARE_DUPLICATE)) {
+			file_dup_count += 1;
+		}
+
+		if (fi->flags & FILE_FLAG_HAS_FRAGMENT) {
+			if (fi->flags & FILE_FLAG_FRAGMENT_IS_DUPLICATE) {
+				frag_dup += 1;
+			} else {
+				frag_count += 1;
+			}
+		}
+
+		sparse_blocks += sparse;
+		file_count += 1;
+		input_bytes += fi->size;
+	}
+
+	if (input_bytes > 0) {
+		output_bytes = super->inode_table_start - sizeof(*super);
+		ratio = (100 * output_bytes) / input_bytes;
+	} else {
+		ratio = 100;
+	}
+
+	fputs("---------------------------------------------------\n", stdout);
+	printf("Input files processed: %zu\n", file_count);
+	printf("Files that were complete duplicates: %zu\n", file_dup_count);
+	printf("Data blocks actually written: %zu\n", blocks_written);
+	printf("Fragment blocks written: %u\n", super->fragment_entry_count);
+	printf("Duplicate data blocks omitted: %zu\n", duplicate_blocks);
+	printf("Sparse blocks omitted: %zu\n", sparse_blocks);
+	printf("Fragments actually written: %zu\n", frag_count);
+	printf("Duplicated fragments omitted: %zu\n", frag_dup);
+	printf("Total number of inodes: %u\n", super->inode_count);
+	printf("Number of unique group/user IDs: %u\n", super->id_count);
+	printf("Data compression ratio: %zu%%\n", ratio);
+}