merged with pythonware's SRE 2.1.1 codebase

Modules/_sre.c

@@ -28,6 +28,8 @@
  * 2001-01-16 fl  fixed memory leak in pattern destructor
  * 2001-03-20 fl  lots of fixes for 2.1b2
  * 2001-04-15 fl  export copyright as Python attribute, not global
+ * 2001-04-28 fl  added __copy__ methods (work in progress)
+ * 2001-05-14 fl  fixes for 1.5.2
  *
  * Copyright (c) 1997-2001 by Secret Labs AB.  All rights reserved.
  *
@@ -43,9 +45,10 @@
 #ifndef SRE_RECURSIVE
 
 static char copyright[] =
-    " SRE 2.1b2 Copyright (c) 1997-2001 by Secret Labs AB ";
+    " SRE 2.1.1 Copyright (c) 1997-2001 by Secret Labs AB ";
 
 #include "Python.h"
+#include "structmember.h" /* offsetof */
 
 #include "sre.h"
 
@@ -86,6 +89,9 @@ static char copyright[] =
 /* enables aggressive inlining (always on for Visual C) */
 #undef USE_INLINE
 
+/* enables copy/deepcopy handling (work in progress) */
+#undef USE_BUILTIN_COPY
+
 #if PY_VERSION_HEX < 0x01060000
 #define PyObject_DEL(op) PyMem_DEL((op))
 #endif
@@ -444,6 +450,7 @@ SRE_AT(SRE_STATE* state, SRE_CHAR* ptr, SRE_CODE at)
             SRE_LOC_IS_WORD((int) ptr[0]) : 0;
         return this == that;
 
+#if defined(HAVE_UNICODE)
     case SRE_AT_UNI_BOUNDARY:
         if (state->beginning == state->end)
             return 0;
@@ -461,6 +468,8 @@ SRE_AT(SRE_STATE* state, SRE_CHAR* ptr, SRE_CODE at)
         this = ((void*) ptr < state->end) ?
             SRE_UNI_IS_WORD((int) ptr[0]) : 0;
         return this == that;
+#endif
+
     }
 
     return 0;
@@ -1286,6 +1295,8 @@ _compile(PyObject* self_, PyObject* args)
     if (!self)
         return NULL;
 
+    self->codesize = n;
+
     for (i = 0; i < n; i++) {
         PyObject *o = PyList_GET_ITEM(code, i);
         self->code[i] = (SRE_CODE) PyInt_AsLong(o);
@@ -1844,6 +1855,42 @@ error:
     
 }
 
+static PyObject*
+pattern_copy(PatternObject* self, PyObject* args)
+{
+#if USE_BUILTIN_COPY
+    PatternObject* copy;
+    int offset;
+
+    /* work in progress */
+    
+    copy = PyObject_NEW_VAR(PatternObject, &Pattern_Type, self->codesize);
+    if (!copy)
+        return NULL;
+
+    offset = offsetof(PatternObject, groups);
+
+    Py_XINCREF(self->groupindex);
+    Py_XINCREF(self->indexgroup);
+    Py_XINCREF(self->pattern);
+
+    memcpy((char*) copy + offset, (char*) self + offset,
+           sizeof(PatternObject) + self->codesize * sizeof(SRE_CODE) - offset);
+
+    return (PyObject*) copy;
+#else
+    PyErr_SetString(PyExc_TypeError, "cannot copy this pattern object");
+    return NULL;
+#endif
+}
+
+static PyObject*
+pattern_deepcopy(PatternObject* self, PyObject* args)
+{
+    PyErr_SetString(PyExc_TypeError, "cannot deepcopy this pattern object");
+    return NULL;
+}
+
 static PyMethodDef pattern_methods[] = {
     {"match", (PyCFunction) pattern_match, METH_VARARGS|METH_KEYWORDS},
     {"search", (PyCFunction) pattern_search, METH_VARARGS|METH_KEYWORDS},
@@ -1851,8 +1898,9 @@ static PyMethodDef pattern_methods[] = {
     {"subn", (PyCFunction) pattern_subn, METH_VARARGS|METH_KEYWORDS},
     {"split", (PyCFunction) pattern_split, METH_VARARGS|METH_KEYWORDS},
     {"findall", (PyCFunction) pattern_findall, METH_VARARGS|METH_KEYWORDS},
-    /* experimental */
     {"scanner", (PyCFunction) pattern_scanner, METH_VARARGS},
+    {"__copy__", (PyCFunction) pattern_copy, METH_VARARGS},
+    {"__deepcopy__", (PyCFunction) pattern_deepcopy, METH_VARARGS},
     {NULL, NULL}
 };
 
@@ -2211,6 +2259,46 @@ match_regs(MatchObject* self)
     return regs;
 }
 
+static PyObject*
+match_copy(MatchObject* self, PyObject* args)
+{
+#if USE_BUILTIN_COPY
+    MatchObject* copy;
+    int slots, offset;
+    
+    /* works in progress */
+
+    slots = 2 * (self->pattern->groups+1);
+
+    copy = PyObject_NEW_VAR(MatchObject, &Match_Type, slots);
+    if (!copy)
+        return NULL;
+
+    /* this value a constant, but any compiler should be able to
+       figure that out all by itself */
+    offset = offsetof(MatchObject, string);
+
+    Py_XINCREF(self->pattern);
+    Py_XINCREF(self->string);
+    Py_XINCREF(self->regs);
+
+    memcpy((char*) copy + offset, (char*) self + offset,
+           sizeof(MatchObject) + slots * sizeof(int) - offset);
+
+    return (PyObject*) copy;
+#else
+    PyErr_SetString(PyExc_TypeError, "cannot deepcopy this match object");
+    return NULL;
+#endif
+}
+
+static PyObject*
+match_deepcopy(MatchObject* self, PyObject* args)
+{
+    PyErr_SetString(PyExc_TypeError, "cannot deepcopy this match object");
+    return NULL;
+}
+
 static PyMethodDef match_methods[] = {
     {"group", (PyCFunction) match_group, METH_VARARGS},
     {"start", (PyCFunction) match_start, METH_VARARGS},
@@ -2219,6 +2307,8 @@ static PyMethodDef match_methods[] = {
     {"groups", (PyCFunction) match_groups, METH_VARARGS|METH_KEYWORDS},
     {"groupdict", (PyCFunction) match_groupdict, METH_VARARGS|METH_KEYWORDS},
     {"expand", (PyCFunction) match_expand, METH_VARARGS},
+    {"__copy__", (PyCFunction) match_copy, METH_VARARGS},
+    {"__deepcopy__", (PyCFunction) match_deepcopy, METH_VARARGS},
     {NULL, NULL}
 };
 

Modules/sre.h

@@ -3,7 +3,7 @@
  *
  * regular expression matching engine
  *
- * Copyright (c) 1997-2000 by Secret Labs AB.  All rights reserved.
+ * Copyright (c) 1997-2001 by Secret Labs AB.  All rights reserved.
  *
  * See the _sre.c file for information on usage and redistribution.
  */
@@ -21,15 +21,18 @@
 #define SRE_CODE unsigned short
 #endif
 
+#define SRE_CODE unsigned short
+
 typedef struct {
     PyObject_VAR_HEAD
-    int groups;
+    int groups; /* must be first! */
     PyObject* groupindex;
     PyObject* indexgroup;
     /* compatibility */
     PyObject* pattern; /* pattern source (or None) */
     int flags; /* flags used when compiling pattern source */
     /* pattern code */
+    int codesize;
     SRE_CODE code[1];
 } PatternObject;
 
@@ -37,7 +40,7 @@ typedef struct {
 
 typedef struct {
     PyObject_VAR_HEAD
-    PyObject* string; /* link to the target string */
+    PyObject* string; /* link to the target string (must be first) */
     PyObject* regs; /* cached list of matching spans */
     PatternObject* pattern; /* link to the regex (pattern) object */
     int pos, endpos; /* current target slice */