c5bd0c37ce
- Refactored all builder (make_*) functions into separate Python modules along to the build tree - Introduced utility function to wrap all invocations on Windows, but does not change it elsewhere - Introduced stub to use the builders module as a stand alone script and invoke a selected function There is a problem with file handles related to writing generated content (*.gen.h and *.gen.cpp) on Windows, which randomly causes a SHARING VIOLATION error to the compiler resulting in flaky builds. Running all such content generators in a new subprocess instead of directly inside the build script works around the issue. Yes, I tried the multiprocessing module. It did not work due to conflict with SCons on cPickle. Suggested workaround did not fully work either. Using the run_in_subprocess wrapper on osx and x11 platforms as well for consistency. In case of running a cross-compilation on Windows they would still be used, but likely it will not happen in practice. What counts is that the build itself is running on which platform, not the target platform. Some generated files are written directly in an SConstruct or SCsub file, before the parallel build starts. They don't need to be written in a subprocess, apparently, so I left them as is.
412 lines
12 KiB
Python
412 lines
12 KiB
Python
"""Functions used to generate source files during build time
|
|
|
|
All such functions are invoked in a subprocess on Windows to prevent build flakiness.
|
|
|
|
"""
|
|
import os
|
|
import os.path
|
|
from platform_methods import subprocess_main
|
|
from compat import encode_utf8, byte_to_str, open_utf8, escape_string
|
|
|
|
|
|
def make_certs_header(target, source, env):
|
|
|
|
src = source[0]
|
|
dst = target[0]
|
|
f = open(src, "rb")
|
|
g = open_utf8(dst, "w")
|
|
buf = f.read()
|
|
decomp_size = len(buf)
|
|
import zlib
|
|
buf = zlib.compress(buf)
|
|
|
|
g.write("/* THIS FILE IS GENERATED DO NOT EDIT */\n")
|
|
g.write("#ifndef _CERTS_RAW_H\n")
|
|
g.write("#define _CERTS_RAW_H\n")
|
|
g.write("static const int _certs_compressed_size = " + str(len(buf)) + ";\n")
|
|
g.write("static const int _certs_uncompressed_size = " + str(decomp_size) + ";\n")
|
|
g.write("static const unsigned char _certs_compressed[] = {\n")
|
|
for i in range(len(buf)):
|
|
g.write("\t" + byte_to_str(buf[i]) + ",\n")
|
|
g.write("};\n")
|
|
g.write("#endif")
|
|
|
|
g.close()
|
|
f.close()
|
|
|
|
|
|
def make_doc_header(target, source, env):
|
|
|
|
dst = target[0]
|
|
g = open_utf8(dst, "w")
|
|
buf = ""
|
|
docbegin = ""
|
|
docend = ""
|
|
for src in source:
|
|
if not src.endswith(".xml"):
|
|
continue
|
|
with open_utf8(src, "r") as f:
|
|
content = f.read()
|
|
buf += content
|
|
|
|
buf = encode_utf8(docbegin + buf + docend)
|
|
decomp_size = len(buf)
|
|
import zlib
|
|
buf = zlib.compress(buf)
|
|
|
|
g.write("/* THIS FILE IS GENERATED DO NOT EDIT */\n")
|
|
g.write("#ifndef _DOC_DATA_RAW_H\n")
|
|
g.write("#define _DOC_DATA_RAW_H\n")
|
|
g.write("static const int _doc_data_compressed_size = " + str(len(buf)) + ";\n")
|
|
g.write("static const int _doc_data_uncompressed_size = " + str(decomp_size) + ";\n")
|
|
g.write("static const unsigned char _doc_data_compressed[] = {\n")
|
|
for i in range(len(buf)):
|
|
g.write("\t" + byte_to_str(buf[i]) + ",\n")
|
|
g.write("};\n")
|
|
|
|
g.write("#endif")
|
|
|
|
g.close()
|
|
|
|
|
|
def make_fonts_header(target, source, env):
|
|
|
|
dst = target[0]
|
|
|
|
g = open_utf8(dst, "w")
|
|
|
|
g.write("/* THIS FILE IS GENERATED DO NOT EDIT */\n")
|
|
g.write("#ifndef _EDITOR_FONTS_H\n")
|
|
g.write("#define _EDITOR_FONTS_H\n")
|
|
|
|
# saving uncompressed, since freetype will reference from memory pointer
|
|
xl_names = []
|
|
for i in range(len(source)):
|
|
with open(source[i], "rb")as f:
|
|
buf = f.read()
|
|
|
|
name = os.path.splitext(os.path.basename(source[i]))[0]
|
|
|
|
g.write("static const int _font_" + name + "_size = " + str(len(buf)) + ";\n")
|
|
g.write("static const unsigned char _font_" + name + "[] = {\n")
|
|
for i in range(len(buf)):
|
|
g.write("\t" + byte_to_str(buf[i]) + ",\n")
|
|
|
|
g.write("};\n")
|
|
|
|
g.write("#endif")
|
|
|
|
g.close()
|
|
|
|
|
|
def make_translations_header(target, source, env):
|
|
|
|
dst = target[0]
|
|
|
|
g = open_utf8(dst, "w")
|
|
|
|
g.write("/* THIS FILE IS GENERATED DO NOT EDIT */\n")
|
|
g.write("#ifndef _EDITOR_TRANSLATIONS_H\n")
|
|
g.write("#define _EDITOR_TRANSLATIONS_H\n")
|
|
|
|
import zlib
|
|
import os.path
|
|
|
|
sorted_paths = sorted(source, key=lambda path: os.path.splitext(os.path.basename(path))[0])
|
|
|
|
xl_names = []
|
|
for i in range(len(sorted_paths)):
|
|
with open(sorted_paths[i], "rb") as f:
|
|
buf = f.read()
|
|
decomp_size = len(buf)
|
|
buf = zlib.compress(buf)
|
|
name = os.path.splitext(os.path.basename(sorted_paths[i]))[0]
|
|
|
|
g.write("static const unsigned char _translation_" + name + "_compressed[] = {\n")
|
|
for i in range(len(buf)):
|
|
g.write("\t" + byte_to_str(buf[i]) + ",\n")
|
|
|
|
g.write("};\n")
|
|
|
|
xl_names.append([name, len(buf), str(decomp_size)])
|
|
|
|
g.write("struct EditorTranslationList {\n")
|
|
g.write("\tconst char* lang;\n")
|
|
g.write("\tint comp_size;\n")
|
|
g.write("\tint uncomp_size;\n")
|
|
g.write("\tconst unsigned char* data;\n")
|
|
g.write("};\n\n")
|
|
g.write("static EditorTranslationList _editor_translations[] = {\n")
|
|
for x in xl_names:
|
|
g.write("\t{ \"" + x[0] + "\", " + str(x[1]) + ", " + str(x[2]) + ", _translation_" + x[0] + "_compressed},\n")
|
|
g.write("\t{NULL, 0, 0, NULL}\n")
|
|
g.write("};\n")
|
|
|
|
g.write("#endif")
|
|
|
|
g.close()
|
|
|
|
|
|
def make_authors_header(target, source, env):
|
|
|
|
sections = ["Project Founders", "Lead Developer", "Project Manager", "Developers"]
|
|
sections_id = ["dev_founders", "dev_lead", "dev_manager", "dev_names"]
|
|
|
|
src = source[0]
|
|
dst = target[0]
|
|
f = open_utf8(src, "r")
|
|
g = open_utf8(dst, "w")
|
|
|
|
g.write("/* THIS FILE IS GENERATED DO NOT EDIT */\n")
|
|
g.write("#ifndef _EDITOR_AUTHORS_H\n")
|
|
g.write("#define _EDITOR_AUTHORS_H\n")
|
|
|
|
current_section = ""
|
|
reading = False
|
|
|
|
def close_section():
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
|
|
for line in f:
|
|
if reading:
|
|
if line.startswith(" "):
|
|
g.write("\t\"" + escape_string(line.strip()) + "\",\n")
|
|
continue
|
|
if line.startswith("## "):
|
|
if reading:
|
|
close_section()
|
|
reading = False
|
|
for i in range(len(sections)):
|
|
if line.strip().endswith(sections[i]):
|
|
current_section = escape_string(sections_id[i])
|
|
reading = True
|
|
g.write("static const char *" + current_section + "[] = {\n")
|
|
break
|
|
|
|
if reading:
|
|
close_section()
|
|
|
|
g.write("#endif\n")
|
|
|
|
g.close()
|
|
f.close()
|
|
|
|
def make_donors_header(target, source, env):
|
|
|
|
sections = ["Platinum sponsors", "Gold sponsors", "Mini sponsors", "Gold donors", "Silver donors", "Bronze donors"]
|
|
sections_id = ["donor_s_plat", "donor_s_gold", "donor_s_mini", "donor_gold", "donor_silver", "donor_bronze"]
|
|
|
|
src = source[0]
|
|
dst = target[0]
|
|
f = open_utf8(src, "r")
|
|
g = open_utf8(dst, "w")
|
|
|
|
g.write("/* THIS FILE IS GENERATED DO NOT EDIT */\n")
|
|
g.write("#ifndef _EDITOR_DONORS_H\n")
|
|
g.write("#define _EDITOR_DONORS_H\n")
|
|
|
|
current_section = ""
|
|
reading = False
|
|
|
|
def close_section():
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
|
|
for line in f:
|
|
if reading >= 0:
|
|
if line.startswith(" "):
|
|
g.write("\t\"" + escape_string(line.strip()) + "\",\n")
|
|
continue
|
|
if line.startswith("## "):
|
|
if reading:
|
|
close_section()
|
|
reading = False
|
|
for i in range(len(sections)):
|
|
if line.strip().endswith(sections[i]):
|
|
current_section = escape_string(sections_id[i])
|
|
reading = True
|
|
g.write("static const char *" + current_section + "[] = {\n")
|
|
break
|
|
|
|
if reading:
|
|
close_section()
|
|
|
|
g.write("#endif\n")
|
|
|
|
g.close()
|
|
f.close()
|
|
|
|
|
|
def make_license_header(target, source, env):
|
|
|
|
src_copyright = source[0]
|
|
src_license = source[1]
|
|
dst = target[0]
|
|
f = open_utf8(src_license, "r")
|
|
fc = open_utf8(src_copyright, "r")
|
|
g = open_utf8(dst, "w")
|
|
|
|
g.write("/* THIS FILE IS GENERATED DO NOT EDIT */\n")
|
|
g.write("#ifndef _EDITOR_LICENSE_H\n")
|
|
g.write("#define _EDITOR_LICENSE_H\n")
|
|
g.write("static const char *about_license =")
|
|
|
|
for line in f:
|
|
escaped_string = escape_string(line.strip())
|
|
g.write("\n\t\"" + escaped_string + "\\n\"")
|
|
|
|
g.write(";\n")
|
|
|
|
tp_current = 0
|
|
tp_file = ""
|
|
tp_comment = ""
|
|
tp_copyright = ""
|
|
tp_license = ""
|
|
|
|
tp_licensename = ""
|
|
tp_licensebody = ""
|
|
|
|
tp = []
|
|
tp_licensetext = []
|
|
for line in fc:
|
|
if line.startswith("#"):
|
|
continue
|
|
|
|
if line.startswith("Files:"):
|
|
tp_file = line[6:].strip()
|
|
tp_current = 1
|
|
elif line.startswith("Comment:"):
|
|
tp_comment = line[8:].strip()
|
|
tp_current = 2
|
|
elif line.startswith("Copyright:"):
|
|
tp_copyright = line[10:].strip()
|
|
tp_current = 3
|
|
elif line.startswith("License:"):
|
|
if tp_current != 0:
|
|
tp_license = line[8:].strip()
|
|
tp_current = 4
|
|
else:
|
|
tp_licensename = line[8:].strip()
|
|
tp_current = 5
|
|
elif line.startswith(" "):
|
|
if tp_current == 1:
|
|
tp_file += "\n" + line.strip()
|
|
elif tp_current == 3:
|
|
tp_copyright += "\n" + line.strip()
|
|
elif tp_current == 5:
|
|
if line.strip() == ".":
|
|
tp_licensebody += "\n"
|
|
else:
|
|
tp_licensebody += line[1:]
|
|
else:
|
|
if tp_current != 0:
|
|
if tp_current == 5:
|
|
tp_licensetext.append([tp_licensename, tp_licensebody])
|
|
|
|
tp_licensename = ""
|
|
tp_licensebody = ""
|
|
else:
|
|
added = False
|
|
for i in tp:
|
|
if i[0] == tp_comment:
|
|
i[1].append([tp_file, tp_copyright, tp_license])
|
|
added = True
|
|
break
|
|
if not added:
|
|
tp.append([tp_comment,[[tp_file, tp_copyright, tp_license]]])
|
|
|
|
tp_file = []
|
|
tp_comment = ""
|
|
tp_copyright = []
|
|
tp_license = ""
|
|
tp_current = 0
|
|
|
|
tp_licensetext.append([tp_licensename, tp_licensebody])
|
|
|
|
about_thirdparty = ""
|
|
about_tp_copyright_count = ""
|
|
about_tp_license = ""
|
|
about_tp_copyright = ""
|
|
about_tp_file = ""
|
|
|
|
for i in tp:
|
|
about_thirdparty += "\t\"" + i[0] + "\",\n"
|
|
about_tp_copyright_count += str(len(i[1])) + ", "
|
|
for j in i[1]:
|
|
file_body = ""
|
|
copyright_body = ""
|
|
for k in j[0].split("\n"):
|
|
if file_body != "":
|
|
file_body += "\\n\"\n"
|
|
escaped_string = escape_string(k.strip())
|
|
file_body += "\t\"" + escaped_string
|
|
for k in j[1].split("\n"):
|
|
if copyright_body != "":
|
|
copyright_body += "\\n\"\n"
|
|
escaped_string = escape_string(k.strip())
|
|
copyright_body += "\t\"" + escaped_string
|
|
|
|
about_tp_file += "\t" + file_body + "\",\n"
|
|
about_tp_copyright += "\t" + copyright_body + "\",\n"
|
|
about_tp_license += "\t\"" + j[2] + "\",\n"
|
|
|
|
about_license_name = ""
|
|
about_license_body = ""
|
|
|
|
for i in tp_licensetext:
|
|
body = ""
|
|
for j in i[1].split("\n"):
|
|
if body != "":
|
|
body += "\\n\"\n"
|
|
escaped_string = escape_string(j.strip())
|
|
body += "\t\"" + escaped_string
|
|
|
|
about_license_name += "\t\"" + i[0] + "\",\n"
|
|
about_license_body += "\t" + body + "\",\n"
|
|
|
|
g.write("static const char *about_thirdparty[] = {\n")
|
|
g.write(about_thirdparty)
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
g.write("#define THIRDPARTY_COUNT " + str(len(tp)) + "\n")
|
|
|
|
g.write("static const int about_tp_copyright_count[] = {\n\t")
|
|
g.write(about_tp_copyright_count)
|
|
g.write("0\n};\n")
|
|
|
|
g.write("static const char *about_tp_file[] = {\n")
|
|
g.write(about_tp_file)
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
|
|
g.write("static const char *about_tp_copyright[] = {\n")
|
|
g.write(about_tp_copyright)
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
|
|
g.write("static const char *about_tp_license[] = {\n")
|
|
g.write(about_tp_license)
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
|
|
g.write("static const char *about_license_name[] = {\n")
|
|
g.write(about_license_name)
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
g.write("#define LICENSE_COUNT " + str(len(tp_licensetext)) + "\n")
|
|
|
|
g.write("static const char *about_license_body[] = {\n")
|
|
g.write(about_license_body)
|
|
g.write("\t0\n")
|
|
g.write("};\n")
|
|
|
|
g.write("#endif\n")
|
|
|
|
g.close()
|
|
fc.close()
|
|
f.close()
|
|
|
|
|
|
if __name__ == '__main__':
|
|
subprocess_main(globals())
|