bpo-47152: Automatically regenerate sre_constants.h (GH-91439)

* Move the code for generating Modules/_sre/sre_constants.h from
  Lib/re/_constants.py into a separate script
  Tools/scripts/generate_sre_constants.py.
* Add target `regen-sre` in the makefile.
* Make target `regen-all` depending on `regen-sre`.

Lib/re/_constants.py

@@ -215,46 +215,3 @@ SRE_FLAG_ASCII = 256 # use ascii "locale"
 SRE_INFO_PREFIX = 1 # has prefix
 SRE_INFO_LITERAL = 2 # entire pattern is literal (given by prefix)
 SRE_INFO_CHARSET = 4 # pattern starts with character from given set
-
-if __name__ == "__main__":
-    def dump(f, d, prefix):
-        items = sorted(d)
-        for item in items:
-            f.write("#define %s_%s %d\n" % (prefix, item, item))
-    with open("sre_constants.h", "w") as f:
-        f.write("""\
-/*
- * Secret Labs' Regular Expression Engine
- *
- * regular expression matching engine
- *
- * NOTE: This file is generated by Lib/re/_constants.py.  If you need
- * to change anything in here, edit Lib/re/_constants.py and run it.
- *
- * Copyright (c) 1997-2001 by Secret Labs AB.  All rights reserved.
- *
- * See the sre.c file for information on usage and redistribution.
- */
-
-""")
-
-        f.write("#define SRE_MAGIC %d\n" % MAGIC)
-
-        dump(f, OPCODES, "SRE_OP")
-        dump(f, ATCODES, "SRE")
-        dump(f, CHCODES, "SRE")
-
-        f.write("#define SRE_FLAG_IGNORECASE %d\n" % SRE_FLAG_IGNORECASE)
-        f.write("#define SRE_FLAG_LOCALE %d\n" % SRE_FLAG_LOCALE)
-        f.write("#define SRE_FLAG_MULTILINE %d\n" % SRE_FLAG_MULTILINE)
-        f.write("#define SRE_FLAG_DOTALL %d\n" % SRE_FLAG_DOTALL)
-        f.write("#define SRE_FLAG_UNICODE %d\n" % SRE_FLAG_UNICODE)
-        f.write("#define SRE_FLAG_VERBOSE %d\n" % SRE_FLAG_VERBOSE)
-        f.write("#define SRE_FLAG_DEBUG %d\n" % SRE_FLAG_DEBUG)
-        f.write("#define SRE_FLAG_ASCII %d\n" % SRE_FLAG_ASCII)
-
-        f.write("#define SRE_INFO_PREFIX %d\n" % SRE_INFO_PREFIX)
-        f.write("#define SRE_INFO_LITERAL %d\n" % SRE_INFO_LITERAL)
-        f.write("#define SRE_INFO_CHARSET %d\n" % SRE_INFO_CHARSET)
-
-    print("done")

Makefile.pre.in

@@ -1197,7 +1197,7 @@ regen-limited-abi: all
 # Regenerate all generated files
 
 regen-all: regen-opcode regen-opcode-targets regen-typeslots \
-	regen-token regen-ast regen-keyword regen-frozen clinic \
+	regen-token regen-ast regen-keyword regen-sre regen-frozen clinic \
 	regen-pegen-metaparser regen-pegen regen-test-frozenmain \
 	regen-global-objects
 	@echo
@@ -1350,6 +1350,13 @@ regen-stdlib-module-names: build_all Programs/_testembed
 		> $(srcdir)/Python/stdlib_module_names.h.new
 	$(UPDATE_FILE) $(srcdir)/Python/stdlib_module_names.h $(srcdir)/Python/stdlib_module_names.h.new
 
+regen-sre:
+	# Regenerate Modules/_sre/sre_constants.h from Lib/re/_constants.py
+	# using Tools/scripts/generate_sre_constants.py
+	$(PYTHON_FOR_REGEN) $(srcdir)/Tools/scripts/generate_sre_constants.py \
+		$(srcdir)/Lib/re/_constants.py \
+		$(srcdir)/Modules/_sre/sre_constants.h
+
 Python/compile.o Python/symtable.o Python/ast_unparse.o Python/ast.o Python/future.o: $(srcdir)/Include/internal/pycore_ast.h
 
 Python/getplatform.o: $(srcdir)/Python/getplatform.c

Misc/NEWS.d/next/Build/2022-04-10-16-33-31.bpo-47152.TLkxKm.rst

@@ -0,0 +1 @@
+Add script and make target for generating ``sre_constants.h``.

Modules/_sre/sre_constants.h

@@ -3,8 +3,8 @@
  *
  * regular expression matching engine
  *
- * NOTE: This file is generated by Lib/re/_constants.py.  If you need
+ * Auto-generated by Tools/scripts/generate_sre_constants.py from
- * to change anything in here, edit Lib/re/_constants.py and run it.
+ * Lib/re/_constants.py.
  *
  * Copyright (c) 1997-2001 by Secret Labs AB.  All rights reserved.
  *

Tools/scripts/generate_sre_constants.py

@@ -0,0 +1,62 @@
+#! /usr/bin/env python3
+# This script generates Modules/_sre/sre_constants.h from Lib/re/_constants.py.
+
+
+def update_file(file, content):
+    try:
+        with open(file, 'r') as fobj:
+            if fobj.read() == content:
+                return False
+    except (OSError, ValueError):
+        pass
+    with open(file, 'w') as fobj:
+        fobj.write(content)
+    return True
+
+sre_constants_header = """\
+/*
+ * Secret Labs' Regular Expression Engine
+ *
+ * regular expression matching engine
+ *
+ * Auto-generated by Tools/scripts/generate_sre_constants.py from
+ * Lib/re/_constants.py.
+ *
+ * Copyright (c) 1997-2001 by Secret Labs AB.  All rights reserved.
+ *
+ * See the sre.c file for information on usage and redistribution.
+ */
+
+"""
+
+def main(infile='Lib/re/_constants.py', outfile='Modules/_sre/sre_constants.h'):
+    ns = {}
+    with open(infile) as fp:
+        code = fp.read()
+    exec(code, ns)
+
+    def dump(d, prefix):
+        items = sorted(d)
+        for item in items:
+            yield "#define %s_%s %d\n" % (prefix, item, item)
+
+    def dump2(d, prefix):
+        items = [(value, name) for name, value in d.items()
+                 if name.startswith(prefix)]
+        for value, name in sorted(items):
+            yield "#define %s %d\n" % (name, value)
+
+    content = [sre_constants_header]
+    content.append("#define SRE_MAGIC %d\n" % ns["MAGIC"])
+    content.extend(dump(ns["OPCODES"], "SRE_OP"))
+    content.extend(dump(ns["ATCODES"], "SRE"))
+    content.extend(dump(ns["CHCODES"], "SRE"))
+    content.extend(dump2(ns, "SRE_FLAG_"))
+    content.extend(dump2(ns, "SRE_INFO_"))
+
+    update_file(outfile, ''.join(content))
+
+
+if __name__ == '__main__':
+    import sys
+    main(*sys.argv[1:])