Make list compression ratio more pessimistic

With the recently added h2_huge.lzh, I noticed that the compressed file
was listed as being 0.0% of the original size, since the compression
ratio is so high. This happens because we were using the regular %f
printf formatting, which presumably rounds to the nearest 0.1%. However,
it nonetheless seems rather misleading.

This instead changes the output to always round upwards to the next
0.1%. This is more pessimistic than the old rounding but seems more
honest. With this change, we never show any file as having a 0.0% ratio,
0.1% is the best that can ever be achieved. A bunch of the expected
outputs from the list command had to be updated to match the new logic.

Author: fragglet

NEWS.md

@@ -15,6 +15,10 @@
  * Support was added for the 64-bit file sizes header, allowing huge
    files (>=4GiB) generated by the MorphOS port of lha to be
    extracted correctly.
+ * The compression ratio shown in list output now always rounds up to
+   the next 0.1%, which is a more pessimistic but more honest. For
+   example, very highly compressed files might be shown as "0.1%" of
+   their original size, but never "0.0%".
  * The manual page now includes more detailed information about the
    different list subcommands.
  * The liblhasa headers are now installed into a directory with a name

doc/lha.1

@@ -124,12 +124,13 @@ like \fB-lh0-\fR is used, this will be equal to SIZE.
 Size of the file in bytes before it was compressed.
 .TP
 \fBFile compression ratio (RATIO)\fR
-Compressed size divided by uncompressed size. This measures the effectiveness
-of the compression format; the smaller the percentage, the better. If an
+Compressed size as a percentage of uncompressed size. This measures the
+effectiveness of the compression format; the smaller the percentage, the
+better. The value is (pessimistically) rounded up to the next 0.1%. If an
 uncompressed format like \fB-lh0-\fR is used, this will be equal to 100%. For
-directories, "******" is shown.
-Note that the value here is backwards relative to \fBunzip\fR(1), which instead
-shows the percentage of the original file size reduced.
+directories, "******" is shown. Note that the value here is backwards relative
+to \fBunzip\fR(1), which instead shows the percentage of the original file size
+reduced.
 .TP
 \fBCompression format (METHOD)\fR
 The compression format used to store this file. See the \fBCOMPRESSION

src/list.c

@@ -242,22 +242,31 @@ static const ListColumn size_column = {
 
 // Compression ratio
 
-static float compression_percent(size_t compressed, size_t uncompressed)
+static const char *compression_percent(size_t compressed, size_t uncompressed)
 {
+	static char buf[10];
+	int permille;
+
+	// We pessimistically round the compression ratio up to the next 0.1%,
+	// so that even if eg. a 10,000:1 ratio was achieved, it will be shown
+	// as "0.1%", not "0.0%". This is marginally more honest.
 	if (uncompressed > 0) {
-		return ((float) compressed * 100.0f) / (float) uncompressed;
+		permille = (compressed * 1000 + uncompressed - 1) / uncompressed;
 	} else {
-		return 100.0f;
+		permille = 1000;
 	}
+
+	snprintf(buf, sizeof(buf), "%3d.%1d%%", permille / 10, permille % 10);
+	return buf;
 }
 
 static void ratio_column_print(LHAFileHeader *header)
 {
 	if (!strcmp(header->compress_method, "-lhd-")) {
 		printf("******");
 	} else {
-		printf("%5.1f%%", compression_percent(header->compressed_length,
-		                                      header->length));
+		printf("%s", compression_percent(header->compressed_length,
+		                                 header->length));
 	}
 }
 
@@ -266,8 +275,8 @@ static void ratio_column_footer(FileStatistics *stats)
 	if (stats->length == 0) {
 		printf("******");
 	} else {
-		printf("%5.1f%%", compression_percent(stats->compressed_length,
-		                                      stats->length));
+		printf("%s", compression_percent(stats->compressed_length,
+		                                 stats->length));
 	}
 }
 

test/output/generated/lzs/lzs.lzs-l.txt

@@ -1,5 +1,5 @@
  PERMSSN    UID  GID      SIZE  RATIO     STAMP           NAME
 ---------- ----------- ------- ------ ------------ --------------------
-[generic]  *****/*****   18092  70.0% May  6  2010 gpl-2
+[generic]  *****/*****   18092  70.1% May  6  2010 gpl-2
 ---------- ----------- ------- ------ ------------ --------------------
- Total         1 file    18092  70.0% Jan  1  2000
+ Total         1 file    18092  70.1% Jan  1  2000

test/output/generated/lzs/lzs.lzs-lv.txt

@@ -1,6 +1,6 @@
  PERMSSN    UID  GID      SIZE  RATIO     STAMP     LV
 ---------- ----------- ------- ------ ------------ ---
 gpl-2
-[generic]  *****/*****   18092  70.0% May  6  2010 [0]
+[generic]  *****/*****   18092  70.1% May  6  2010 [0]
 ---------- ----------- ------- ------ ------------ ---
- Total         1 file    18092  70.0% Jan  1  2000
+ Total         1 file    18092  70.1% Jan  1  2000

test/output/generated/lzs/lzs.lzs-v.txt

@@ -1,5 +1,5 @@
  PERMSSN    UID  GID    PACKED    SIZE  RATIO METHOD CRC     STAMP          NAME
 ---------- ----------- ------- ------- ------ ---------- ------------ -------------
-[generic]  *****/*****   12667   18092  70.0% -lzs- a33a May  6  2010 gpl-2
+[generic]  *****/*****   12667   18092  70.1% -lzs- a33a May  6  2010 gpl-2
 ---------- ----------- ------- ------- ------ ---------- ------------ -------------
- Total         1 file    12667   18092  70.0%            Jan  1  2000
+ Total         1 file    12667   18092  70.1%            Jan  1  2000

test/output/generated/lzs/lzs.lzs-vv.txt

@@ -1,6 +1,6 @@
  PERMSSN    UID  GID    PACKED    SIZE  RATIO METHOD CRC     STAMP            LV
 ---------- ----------- ------- ------- ------ ---------- ------------------- ---
 gpl-2
-[generic]  *****/*****   12667   18092  70.0% -lzs- a33a 2010-05-06 23:17:54 [0]
+[generic]  *****/*****   12667   18092  70.1% -lzs- a33a 2010-05-06 23:17:54 [0]
 ---------- ----------- ------- ------- ------ ---------- ------------------- ---
- Total         1 file    12667   18092  70.0%            2000-01-01 00:00:00
+ Total         1 file    12667   18092  70.1%            2000-01-01 00:00:00

test/output/larc333/long.lzs-l.txt

@@ -1,5 +1,5 @@
  PERMSSN    UID  GID      SIZE  RATIO     STAMP           NAME
 ---------- ----------- ------- ------ ------------ --------------------
-[generic]  *****/***** 1241658  18.2% Jun  9  2011 long.txt
+[generic]  *****/***** 1241658  18.3% Jun  9  2011 long.txt
 ---------- ----------- ------- ------ ------------ --------------------
- Total         1 file  1241658  18.2% Jan  1  2000
+ Total         1 file  1241658  18.3% Jan  1  2000

test/output/larc333/long.lzs-lv.txt

@@ -1,6 +1,6 @@
  PERMSSN    UID  GID      SIZE  RATIO     STAMP     LV
 ---------- ----------- ------- ------ ------------ ---
 long.txt
-[generic]  *****/***** 1241658  18.2% Jun  9  2011 [0]
+[generic]  *****/***** 1241658  18.3% Jun  9  2011 [0]
 ---------- ----------- ------- ------ ------------ ---
- Total         1 file  1241658  18.2% Jan  1  2000
+ Total         1 file  1241658  18.3% Jan  1  2000

test/output/larc333/long.lzs-v.txt

@@ -1,5 +1,5 @@
  PERMSSN    UID  GID    PACKED    SIZE  RATIO METHOD CRC     STAMP          NAME
 ---------- ----------- ------- ------- ------ ---------- ------------ -------------
-[generic]  *****/*****  226557 1241658  18.2% -lz5- 6a7c Jun  9  2011 long.txt
+[generic]  *****/*****  226557 1241658  18.3% -lz5- 6a7c Jun  9  2011 long.txt
 ---------- ----------- ------- ------- ------ ---------- ------------ -------------
- Total         1 file   226557 1241658  18.2%            Jan  1  2000
+ Total         1 file   226557 1241658  18.3%            Jan  1  2000