[beast: 5/11] MISC: packres.py: add script to generate compressed C++ strings from binaries

[Tim Janik <timj src gnome org>]

commit 2108080760c0100432afebd4adaf4eb8890fd8bc
Author: Tim Janik <timj gnu org>
Date:   Mon Sep 18 16:43:58 2017 +0200

    MISC: packres.py: add script to generate compressed C++ strings from binaries
    
    Implementation is based on rapidres from Rapicorn commit id:
        bf228016cba3f6d252ee2cc38e1ed32607f37bf0
    
    Signed-off-by: Tim Janik <timj gnu org>

 misc/packres.py |   95 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 95 insertions(+), 0 deletions(-)
---
diff --git a/misc/packres.py b/misc/packres.py
new file mode 100755
index 0000000..8d609a9
--- /dev/null
+++ b/misc/packres.py
@@ -0,0 +1,95 @@
+#!/usr/bin/env python2
+# This Source Code Form is licensed MPL-2.0: http://mozilla.org/MPL/2.0
+"""
+Generate C source code from a binary file.
+"""
+import sys, re, zlib
+
+class Printer:
+  def print_data (self, data):
+    self.pad = False
+    self.lines = []
+    self.s = '"'
+    self.pos = -2
+    for c in data:
+      self.print_char (c)
+    self.lines.append (self.s + '"')
+    return '  ' + '\n  '.join (self.lines)
+  def print_char (self, c):
+    o = ord (c)
+    if self.pos >= 70-1:
+      self.lines.append (self.s + '"')
+      self.s = '"'
+      self.pos = 1
+      self.pad = False
+    need_pad = False
+    if c in '\\"':
+      self.s += '\\' + c
+      self.pos += 2
+    elif c == '\n':
+      self.s += '\\n'
+      self.pos += 2
+    elif o < 32 or o > 126 or c == '?':
+      self.s += '\\%o' % o
+      self.pos += 1 + (o > 63) + (o > 7) + 1
+      need_pad = not (o > 63)
+    elif self.pad and o >= ord ('0') and o <= ord ('9'):
+      self.s += '""' + c
+      self.pos += 3
+    else:
+      self.s += c
+      self.pos += 1
+    self.pad = need_pad
+
+prefix = 'PACKRES_'
+with_resource_entry = True
+
+def print_file (filename, strip_prefix = ''):
+  stripped_filename = filename
+  if strip_prefix:
+    stripped_filename = re.sub (strip_prefix, '', filename)
+  ident = re.sub ('[^a-z0-9_A-Z]', '_', stripped_filename)
+  idd, ide = prefix + ident, prefix + ident
+  if with_resource_entry:
+    idd += '_'
+  f = open (filename, 'rb')
+  raw = f.read()
+  f.close()
+  del f
+  l = len (raw)
+  data = zlib.compress (raw, 9)
+  # two things to consider for using the compressed data:
+  # 1) For the reader code, compressed size + 1 must be smaller than the original data size to identify 
compressed data.
+  # 2) Using compressed data requires runtime unpacking overhead and extra dynamic memory allocation.
+  # So it should provide a *significant* benefit if it's used.
+  j = len (data)
+  if j + 1 >= l or j > 0.9 * l:
+    data = raw  # skip compression
+  p = Printer()
+  out = p.print_data (data)
+  print 'static const char %s[] __attribute__ ((__aligned__ (16))) =' % idd
+  print out + '; // %u + 1' % len (data)
+  if with_resource_entry:
+    print 'static const LocalResourceEntry %s = {' % ide
+    print '  "%s", %u,' % (stripped_filename, l)
+    print '  %s, sizeof %s' % (idd, idd)
+    print '};'
+
+
+if __name__ == "__main__":
+  # parse args
+  strip_prefix = ''
+  files = []
+  i = 1
+  while i < len (sys.argv):
+    if sys.argv[i] == '-s':
+      i += 1
+      strip_prefix = sys.argv[i]
+    else:
+      files.append (sys.argv[i])
+    i += 1
+
+  # process files
+  for f in files:
+    print
+    print_file (f, strip_prefix)