scripts/library: add human_to_size

This is a bash wrapper around an awk function that parses human readable
sizes and returns their representative values in bytes, as a string. A
small test harness is added to validate the functionality.

Signed-off-by: Dave Reisner <dreisner@archlinux.org>

3 files changed, 61 insertions, 1 deletions

diff --git a/scripts/Makefile.am b/scripts/Makefile.am
index 6bb72198..75699b2e 100644
--- a/scripts/Makefile.am
+++ b/scripts/Makefile.am
@@ -27,7 +27,8 @@ EXTRA_DIST = \
 
 LIBRARY = \
 	library/output_format.sh \
-	library/parseopts.sh
+	library/parseopts.sh \
+	library/human_to_size.sh
 
 # Files that should be removed, but which Automake does not know.
 MOSTLYCLEANFILES = $(bin_SCRIPTS)
diff --git a/scripts/library/README b/scripts/library/README
index c71c0714..44748ee2 100644
--- a/scripts/library/README
+++ b/scripts/library/README
@@ -27,3 +27,11 @@ Reccommended Usage:
 Returns:
   0: parse success
   1: parse failure (error message supplied)
+
+human_to_size.sh:
+A function to convert human readable sizes (such as "5.3 GiB") to raw byte
+equivalents. base10 and base2 suffixes are supported, case sensitively. If
+successful, the converted byte value is written to stdout and the function
+returns 0. If an error occurs, nothing in written and the function returns 1.
+Results may be inaccurate when using a broken implementation of awk, such
+as mawk or busybox awk.
diff --git a/scripts/library/human_to_size.sh b/scripts/library/human_to_size.sh
new file mode 100644
index 00000000..11613207
--- /dev/null
+++ b/scripts/library/human_to_size.sh
@@ -0,0 +1,51 @@
+human_to_size() {
+  awk -v human="$1" '
+  function trim(s) {
+    gsub(/^[[:space:]]+|[[:space:]]+$/, "", s)
+    return s
+  }
+
+  function parse_units(units) {
+    if (!units || units == "B")
+      return 1
+    if (match(units, /^.iB$/))
+      return 1024
+    if (match(units, /^.B$/))
+      return 1000
+    if (length(units) == 1)
+      return 1024
+
+    # parse failure: invalid base
+    return -1
+  }
+
+  function parse_scale(s) {
+    return index("BKMGTPE", s) - 1
+  }
+
+  function isnumeric(string) {
+    return match(string, /^[-+]?[[:digit:]]*(\.[[:digit:]]*)?/)
+  }
+
+  BEGIN {
+    # peel off the leading number as the size, fail on invalid number
+    human = trim(human)
+    if (isnumeric(human))
+      size = substr(human, RSTART, RLENGTH)
+    else
+      exit 1
+
+    # the trimmed remainder is assumed to be the units
+    units = trim(substr(human, RLENGTH + 1))
+
+    base = parse_units(units)
+    if (base < 0)
+      exit 1
+
+    scale = parse_scale(substr(units, 1, 1))
+    if (scale < 0)
+      exit 1
+
+    printf "%d\n", size * base^scale + (size + 0 > 0 ? 0.5 : -0.5)
+  }'
+}