Issue #5768: Change to Unicode output logic and test case for same.
[python.git] / Modules / stropmodule.c
blob2d884744aa7fe78e472a964c935c0dcbd6d64ef9
1 /* strop module */
3 #define PY_SSIZE_T_CLEAN
4 #include "Python.h"
5 #include <ctype.h>
7 PyDoc_STRVAR(strop_module__doc__,
8 "Common string manipulations, optimized for speed.\n"
9 "\n"
10 "Always use \"import string\" rather than referencing\n"
11 "this module directly.");
13 /* XXX This file assumes that the <ctype.h> is*() functions
14 XXX are defined for all 8-bit characters! */
16 #define WARN if (PyErr_Warn(PyExc_DeprecationWarning, \
17 "strop functions are obsolete; use string methods")) \
18 return NULL
20 /* The lstrip(), rstrip() and strip() functions are implemented
21 in do_strip(), which uses an additional parameter to indicate what
22 type of strip should occur. */
24 #define LEFTSTRIP 0
25 #define RIGHTSTRIP 1
26 #define BOTHSTRIP 2
29 static PyObject *
30 split_whitespace(char *s, Py_ssize_t len, Py_ssize_t maxsplit)
32 Py_ssize_t i = 0, j;
33 int err;
34 Py_ssize_t countsplit = 0;
35 PyObject* item;
36 PyObject *list = PyList_New(0);
38 if (list == NULL)
39 return NULL;
41 while (i < len) {
42 while (i < len && isspace(Py_CHARMASK(s[i]))) {
43 i = i+1;
45 j = i;
46 while (i < len && !isspace(Py_CHARMASK(s[i]))) {
47 i = i+1;
49 if (j < i) {
50 item = PyString_FromStringAndSize(s+j, i-j);
51 if (item == NULL)
52 goto finally;
54 err = PyList_Append(list, item);
55 Py_DECREF(item);
56 if (err < 0)
57 goto finally;
59 countsplit++;
60 while (i < len && isspace(Py_CHARMASK(s[i]))) {
61 i = i+1;
63 if (maxsplit && (countsplit >= maxsplit) && i < len) {
64 item = PyString_FromStringAndSize(
65 s+i, len - i);
66 if (item == NULL)
67 goto finally;
69 err = PyList_Append(list, item);
70 Py_DECREF(item);
71 if (err < 0)
72 goto finally;
74 i = len;
78 return list;
79 finally:
80 Py_DECREF(list);
81 return NULL;
85 PyDoc_STRVAR(splitfields__doc__,
86 "split(s [,sep [,maxsplit]]) -> list of strings\n"
87 "splitfields(s [,sep [,maxsplit]]) -> list of strings\n"
88 "\n"
89 "Return a list of the words in the string s, using sep as the\n"
90 "delimiter string. If maxsplit is nonzero, splits into at most\n"
91 "maxsplit words. If sep is not specified, any whitespace string\n"
92 "is a separator. Maxsplit defaults to 0.\n"
93 "\n"
94 "(split and splitfields are synonymous)");
96 static PyObject *
97 strop_splitfields(PyObject *self, PyObject *args)
99 Py_ssize_t len, n, i, j, err;
100 Py_ssize_t splitcount, maxsplit;
101 char *s, *sub;
102 PyObject *list, *item;
104 WARN;
105 sub = NULL;
106 n = 0;
107 splitcount = 0;
108 maxsplit = 0;
109 if (!PyArg_ParseTuple(args, "t#|z#n:split", &s, &len, &sub, &n, &maxsplit))
110 return NULL;
111 if (sub == NULL)
112 return split_whitespace(s, len, maxsplit);
113 if (n == 0) {
114 PyErr_SetString(PyExc_ValueError, "empty separator");
115 return NULL;
118 list = PyList_New(0);
119 if (list == NULL)
120 return NULL;
122 i = j = 0;
123 while (i+n <= len) {
124 if (s[i] == sub[0] && (n == 1 || memcmp(s+i, sub, n) == 0)) {
125 item = PyString_FromStringAndSize(s+j, i-j);
126 if (item == NULL)
127 goto fail;
128 err = PyList_Append(list, item);
129 Py_DECREF(item);
130 if (err < 0)
131 goto fail;
132 i = j = i + n;
133 splitcount++;
134 if (maxsplit && (splitcount >= maxsplit))
135 break;
137 else
138 i++;
140 item = PyString_FromStringAndSize(s+j, len-j);
141 if (item == NULL)
142 goto fail;
143 err = PyList_Append(list, item);
144 Py_DECREF(item);
145 if (err < 0)
146 goto fail;
148 return list;
150 fail:
151 Py_DECREF(list);
152 return NULL;
156 PyDoc_STRVAR(joinfields__doc__,
157 "join(list [,sep]) -> string\n"
158 "joinfields(list [,sep]) -> string\n"
159 "\n"
160 "Return a string composed of the words in list, with\n"
161 "intervening occurrences of sep. Sep defaults to a single\n"
162 "space.\n"
163 "\n"
164 "(join and joinfields are synonymous)");
166 static PyObject *
167 strop_joinfields(PyObject *self, PyObject *args)
169 PyObject *seq;
170 char *sep = NULL;
171 Py_ssize_t seqlen, seplen = 0;
172 Py_ssize_t i, reslen = 0, slen = 0, sz = 100;
173 PyObject *res = NULL;
174 char* p = NULL;
175 ssizeargfunc getitemfunc;
177 WARN;
178 if (!PyArg_ParseTuple(args, "O|t#:join", &seq, &sep, &seplen))
179 return NULL;
180 if (sep == NULL) {
181 sep = " ";
182 seplen = 1;
185 seqlen = PySequence_Size(seq);
186 if (seqlen < 0 && PyErr_Occurred())
187 return NULL;
189 if (seqlen == 1) {
190 /* Optimization if there's only one item */
191 PyObject *item = PySequence_GetItem(seq, 0);
192 if (item && !PyString_Check(item)) {
193 PyErr_SetString(PyExc_TypeError,
194 "first argument must be sequence of strings");
195 Py_DECREF(item);
196 return NULL;
198 return item;
201 if (!(res = PyString_FromStringAndSize((char*)NULL, sz)))
202 return NULL;
203 p = PyString_AsString(res);
205 /* optimize for lists, since it's the most common case. all others
206 * (tuples and arbitrary sequences) just use the sequence abstract
207 * interface.
209 if (PyList_Check(seq)) {
210 for (i = 0; i < seqlen; i++) {
211 PyObject *item = PyList_GET_ITEM(seq, i);
212 if (!PyString_Check(item)) {
213 PyErr_SetString(PyExc_TypeError,
214 "first argument must be sequence of strings");
215 Py_DECREF(res);
216 return NULL;
218 slen = PyString_GET_SIZE(item);
219 if (slen > PY_SSIZE_T_MAX - reslen ||
220 seplen > PY_SSIZE_T_MAX - reslen - seplen) {
221 PyErr_SetString(PyExc_OverflowError,
222 "input too long");
223 Py_DECREF(res);
224 return NULL;
226 while (reslen + slen + seplen >= sz) {
227 if (_PyString_Resize(&res, sz * 2) < 0)
228 return NULL;
229 sz *= 2;
230 p = PyString_AsString(res) + reslen;
232 if (i > 0) {
233 memcpy(p, sep, seplen);
234 p += seplen;
235 reslen += seplen;
237 memcpy(p, PyString_AS_STRING(item), slen);
238 p += slen;
239 reslen += slen;
241 _PyString_Resize(&res, reslen);
242 return res;
245 if (seq->ob_type->tp_as_sequence == NULL ||
246 (getitemfunc = seq->ob_type->tp_as_sequence->sq_item) == NULL)
248 PyErr_SetString(PyExc_TypeError,
249 "first argument must be a sequence");
250 return NULL;
252 /* This is now type safe */
253 for (i = 0; i < seqlen; i++) {
254 PyObject *item = getitemfunc(seq, i);
255 if (!item || !PyString_Check(item)) {
256 PyErr_SetString(PyExc_TypeError,
257 "first argument must be sequence of strings");
258 Py_DECREF(res);
259 Py_XDECREF(item);
260 return NULL;
262 slen = PyString_GET_SIZE(item);
263 if (slen > PY_SSIZE_T_MAX - reslen ||
264 seplen > PY_SSIZE_T_MAX - reslen - seplen) {
265 PyErr_SetString(PyExc_OverflowError,
266 "input too long");
267 Py_DECREF(res);
268 Py_XDECREF(item);
269 return NULL;
271 while (reslen + slen + seplen >= sz) {
272 if (_PyString_Resize(&res, sz * 2) < 0) {
273 Py_DECREF(item);
274 return NULL;
276 sz *= 2;
277 p = PyString_AsString(res) + reslen;
279 if (i > 0) {
280 memcpy(p, sep, seplen);
281 p += seplen;
282 reslen += seplen;
284 memcpy(p, PyString_AS_STRING(item), slen);
285 p += slen;
286 reslen += slen;
287 Py_DECREF(item);
289 _PyString_Resize(&res, reslen);
290 return res;
294 PyDoc_STRVAR(find__doc__,
295 "find(s, sub [,start [,end]]) -> in\n"
296 "\n"
297 "Return the lowest index in s where substring sub is found,\n"
298 "such that sub is contained within s[start,end]. Optional\n"
299 "arguments start and end are interpreted as in slice notation.\n"
300 "\n"
301 "Return -1 on failure.");
303 static PyObject *
304 strop_find(PyObject *self, PyObject *args)
306 char *s, *sub;
307 Py_ssize_t len, n, i = 0, last = PY_SSIZE_T_MAX;
309 WARN;
310 if (!PyArg_ParseTuple(args, "t#t#|nn:find", &s, &len, &sub, &n, &i, &last))
311 return NULL;
313 if (last > len)
314 last = len;
315 if (last < 0)
316 last += len;
317 if (last < 0)
318 last = 0;
319 if (i < 0)
320 i += len;
321 if (i < 0)
322 i = 0;
324 if (n == 0 && i <= last)
325 return PyInt_FromLong((long)i);
327 last -= n;
328 for (; i <= last; ++i)
329 if (s[i] == sub[0] &&
330 (n == 1 || memcmp(&s[i+1], &sub[1], n-1) == 0))
331 return PyInt_FromLong((long)i);
333 return PyInt_FromLong(-1L);
337 PyDoc_STRVAR(rfind__doc__,
338 "rfind(s, sub [,start [,end]]) -> int\n"
339 "\n"
340 "Return the highest index in s where substring sub is found,\n"
341 "such that sub is contained within s[start,end]. Optional\n"
342 "arguments start and end are interpreted as in slice notation.\n"
343 "\n"
344 "Return -1 on failure.");
346 static PyObject *
347 strop_rfind(PyObject *self, PyObject *args)
349 char *s, *sub;
350 Py_ssize_t len, n, j;
351 Py_ssize_t i = 0, last = PY_SSIZE_T_MAX;
353 WARN;
354 if (!PyArg_ParseTuple(args, "t#t#|nn:rfind", &s, &len, &sub, &n, &i, &last))
355 return NULL;
357 if (last > len)
358 last = len;
359 if (last < 0)
360 last += len;
361 if (last < 0)
362 last = 0;
363 if (i < 0)
364 i += len;
365 if (i < 0)
366 i = 0;
368 if (n == 0 && i <= last)
369 return PyInt_FromLong((long)last);
371 for (j = last-n; j >= i; --j)
372 if (s[j] == sub[0] &&
373 (n == 1 || memcmp(&s[j+1], &sub[1], n-1) == 0))
374 return PyInt_FromLong((long)j);
376 return PyInt_FromLong(-1L);
380 static PyObject *
381 do_strip(PyObject *args, int striptype)
383 char *s;
384 Py_ssize_t len, i, j;
387 if (PyString_AsStringAndSize(args, &s, &len))
388 return NULL;
390 i = 0;
391 if (striptype != RIGHTSTRIP) {
392 while (i < len && isspace(Py_CHARMASK(s[i]))) {
393 i++;
397 j = len;
398 if (striptype != LEFTSTRIP) {
399 do {
400 j--;
401 } while (j >= i && isspace(Py_CHARMASK(s[j])));
402 j++;
405 if (i == 0 && j == len) {
406 Py_INCREF(args);
407 return args;
409 else
410 return PyString_FromStringAndSize(s+i, j-i);
414 PyDoc_STRVAR(strip__doc__,
415 "strip(s) -> string\n"
416 "\n"
417 "Return a copy of the string s with leading and trailing\n"
418 "whitespace removed.");
420 static PyObject *
421 strop_strip(PyObject *self, PyObject *args)
423 WARN;
424 return do_strip(args, BOTHSTRIP);
428 PyDoc_STRVAR(lstrip__doc__,
429 "lstrip(s) -> string\n"
430 "\n"
431 "Return a copy of the string s with leading whitespace removed.");
433 static PyObject *
434 strop_lstrip(PyObject *self, PyObject *args)
436 WARN;
437 return do_strip(args, LEFTSTRIP);
441 PyDoc_STRVAR(rstrip__doc__,
442 "rstrip(s) -> string\n"
443 "\n"
444 "Return a copy of the string s with trailing whitespace removed.");
446 static PyObject *
447 strop_rstrip(PyObject *self, PyObject *args)
449 WARN;
450 return do_strip(args, RIGHTSTRIP);
454 PyDoc_STRVAR(lower__doc__,
455 "lower(s) -> string\n"
456 "\n"
457 "Return a copy of the string s converted to lowercase.");
459 static PyObject *
460 strop_lower(PyObject *self, PyObject *args)
462 char *s, *s_new;
463 Py_ssize_t i, n;
464 PyObject *newstr;
465 int changed;
467 WARN;
468 if (PyString_AsStringAndSize(args, &s, &n))
469 return NULL;
470 newstr = PyString_FromStringAndSize(NULL, n);
471 if (newstr == NULL)
472 return NULL;
473 s_new = PyString_AsString(newstr);
474 changed = 0;
475 for (i = 0; i < n; i++) {
476 int c = Py_CHARMASK(*s++);
477 if (isupper(c)) {
478 changed = 1;
479 *s_new = tolower(c);
480 } else
481 *s_new = c;
482 s_new++;
484 if (!changed) {
485 Py_DECREF(newstr);
486 Py_INCREF(args);
487 return args;
489 return newstr;
493 PyDoc_STRVAR(upper__doc__,
494 "upper(s) -> string\n"
495 "\n"
496 "Return a copy of the string s converted to uppercase.");
498 static PyObject *
499 strop_upper(PyObject *self, PyObject *args)
501 char *s, *s_new;
502 Py_ssize_t i, n;
503 PyObject *newstr;
504 int changed;
506 WARN;
507 if (PyString_AsStringAndSize(args, &s, &n))
508 return NULL;
509 newstr = PyString_FromStringAndSize(NULL, n);
510 if (newstr == NULL)
511 return NULL;
512 s_new = PyString_AsString(newstr);
513 changed = 0;
514 for (i = 0; i < n; i++) {
515 int c = Py_CHARMASK(*s++);
516 if (islower(c)) {
517 changed = 1;
518 *s_new = toupper(c);
519 } else
520 *s_new = c;
521 s_new++;
523 if (!changed) {
524 Py_DECREF(newstr);
525 Py_INCREF(args);
526 return args;
528 return newstr;
532 PyDoc_STRVAR(capitalize__doc__,
533 "capitalize(s) -> string\n"
534 "\n"
535 "Return a copy of the string s with only its first character\n"
536 "capitalized.");
538 static PyObject *
539 strop_capitalize(PyObject *self, PyObject *args)
541 char *s, *s_new;
542 Py_ssize_t i, n;
543 PyObject *newstr;
544 int changed;
546 WARN;
547 if (PyString_AsStringAndSize(args, &s, &n))
548 return NULL;
549 newstr = PyString_FromStringAndSize(NULL, n);
550 if (newstr == NULL)
551 return NULL;
552 s_new = PyString_AsString(newstr);
553 changed = 0;
554 if (0 < n) {
555 int c = Py_CHARMASK(*s++);
556 if (islower(c)) {
557 changed = 1;
558 *s_new = toupper(c);
559 } else
560 *s_new = c;
561 s_new++;
563 for (i = 1; i < n; i++) {
564 int c = Py_CHARMASK(*s++);
565 if (isupper(c)) {
566 changed = 1;
567 *s_new = tolower(c);
568 } else
569 *s_new = c;
570 s_new++;
572 if (!changed) {
573 Py_DECREF(newstr);
574 Py_INCREF(args);
575 return args;
577 return newstr;
581 PyDoc_STRVAR(expandtabs__doc__,
582 "expandtabs(string, [tabsize]) -> string\n"
583 "\n"
584 "Expand tabs in a string, i.e. replace them by one or more spaces,\n"
585 "depending on the current column and the given tab size (default 8).\n"
586 "The column number is reset to zero after each newline occurring in the\n"
587 "string. This doesn't understand other non-printing characters.");
589 static PyObject *
590 strop_expandtabs(PyObject *self, PyObject *args)
592 /* Original by Fredrik Lundh */
593 char* e;
594 char* p;
595 char* q;
596 Py_ssize_t i, j, old_j;
597 PyObject* out;
598 char* string;
599 Py_ssize_t stringlen;
600 int tabsize = 8;
602 WARN;
603 /* Get arguments */
604 if (!PyArg_ParseTuple(args, "s#|i:expandtabs", &string, &stringlen, &tabsize))
605 return NULL;
606 if (tabsize < 1) {
607 PyErr_SetString(PyExc_ValueError,
608 "tabsize must be at least 1");
609 return NULL;
612 /* First pass: determine size of output string */
613 i = j = old_j = 0; /* j: current column; i: total of previous lines */
614 e = string + stringlen;
615 for (p = string; p < e; p++) {
616 if (*p == '\t') {
617 j += tabsize - (j%tabsize);
618 if (old_j > j) {
619 PyErr_SetString(PyExc_OverflowError,
620 "new string is too long");
621 return NULL;
623 old_j = j;
624 } else {
625 j++;
626 if (*p == '\n') {
627 i += j;
628 j = 0;
633 if ((i + j) < 0) {
634 PyErr_SetString(PyExc_OverflowError, "new string is too long");
635 return NULL;
638 /* Second pass: create output string and fill it */
639 out = PyString_FromStringAndSize(NULL, i+j);
640 if (out == NULL)
641 return NULL;
643 i = 0;
644 q = PyString_AS_STRING(out);
646 for (p = string; p < e; p++) {
647 if (*p == '\t') {
648 j = tabsize - (i%tabsize);
649 i += j;
650 while (j-- > 0)
651 *q++ = ' ';
652 } else {
653 *q++ = *p;
654 i++;
655 if (*p == '\n')
656 i = 0;
660 return out;
664 PyDoc_STRVAR(count__doc__,
665 "count(s, sub[, start[, end]]) -> int\n"
666 "\n"
667 "Return the number of occurrences of substring sub in string\n"
668 "s[start:end]. Optional arguments start and end are\n"
669 "interpreted as in slice notation.");
671 static PyObject *
672 strop_count(PyObject *self, PyObject *args)
674 char *s, *sub;
675 Py_ssize_t len, n;
676 Py_ssize_t i = 0, last = PY_SSIZE_T_MAX;
677 Py_ssize_t m, r;
679 WARN;
680 if (!PyArg_ParseTuple(args, "t#t#|nn:count", &s, &len, &sub, &n, &i, &last))
681 return NULL;
682 if (last > len)
683 last = len;
684 if (last < 0)
685 last += len;
686 if (last < 0)
687 last = 0;
688 if (i < 0)
689 i += len;
690 if (i < 0)
691 i = 0;
692 m = last + 1 - n;
693 if (n == 0)
694 return PyInt_FromLong((long) (m-i));
696 r = 0;
697 while (i < m) {
698 if (!memcmp(s+i, sub, n)) {
699 r++;
700 i += n;
701 } else {
702 i++;
705 return PyInt_FromLong((long) r);
709 PyDoc_STRVAR(swapcase__doc__,
710 "swapcase(s) -> string\n"
711 "\n"
712 "Return a copy of the string s with upper case characters\n"
713 "converted to lowercase and vice versa.");
715 static PyObject *
716 strop_swapcase(PyObject *self, PyObject *args)
718 char *s, *s_new;
719 Py_ssize_t i, n;
720 PyObject *newstr;
721 int changed;
723 WARN;
724 if (PyString_AsStringAndSize(args, &s, &n))
725 return NULL;
726 newstr = PyString_FromStringAndSize(NULL, n);
727 if (newstr == NULL)
728 return NULL;
729 s_new = PyString_AsString(newstr);
730 changed = 0;
731 for (i = 0; i < n; i++) {
732 int c = Py_CHARMASK(*s++);
733 if (islower(c)) {
734 changed = 1;
735 *s_new = toupper(c);
737 else if (isupper(c)) {
738 changed = 1;
739 *s_new = tolower(c);
741 else
742 *s_new = c;
743 s_new++;
745 if (!changed) {
746 Py_DECREF(newstr);
747 Py_INCREF(args);
748 return args;
750 return newstr;
754 PyDoc_STRVAR(atoi__doc__,
755 "atoi(s [,base]) -> int\n"
756 "\n"
757 "Return the integer represented by the string s in the given\n"
758 "base, which defaults to 10. The string s must consist of one\n"
759 "or more digits, possibly preceded by a sign. If base is 0, it\n"
760 "is chosen from the leading characters of s, 0 for octal, 0x or\n"
761 "0X for hexadecimal. If base is 16, a preceding 0x or 0X is\n"
762 "accepted.");
764 static PyObject *
765 strop_atoi(PyObject *self, PyObject *args)
767 char *s, *end;
768 int base = 10;
769 long x;
770 char buffer[256]; /* For errors */
772 WARN;
773 if (!PyArg_ParseTuple(args, "s|i:atoi", &s, &base))
774 return NULL;
776 if ((base != 0 && base < 2) || base > 36) {
777 PyErr_SetString(PyExc_ValueError, "invalid base for atoi()");
778 return NULL;
781 while (*s && isspace(Py_CHARMASK(*s)))
782 s++;
783 errno = 0;
784 if (base == 0 && s[0] == '0')
785 x = (long) PyOS_strtoul(s, &end, base);
786 else
787 x = PyOS_strtol(s, &end, base);
788 if (end == s || !isalnum(Py_CHARMASK(end[-1])))
789 goto bad;
790 while (*end && isspace(Py_CHARMASK(*end)))
791 end++;
792 if (*end != '\0') {
793 bad:
794 PyOS_snprintf(buffer, sizeof(buffer),
795 "invalid literal for atoi(): %.200s", s);
796 PyErr_SetString(PyExc_ValueError, buffer);
797 return NULL;
799 else if (errno != 0) {
800 PyOS_snprintf(buffer, sizeof(buffer),
801 "atoi() literal too large: %.200s", s);
802 PyErr_SetString(PyExc_ValueError, buffer);
803 return NULL;
805 return PyInt_FromLong(x);
809 PyDoc_STRVAR(atol__doc__,
810 "atol(s [,base]) -> long\n"
811 "\n"
812 "Return the long integer represented by the string s in the\n"
813 "given base, which defaults to 10. The string s must consist\n"
814 "of one or more digits, possibly preceded by a sign. If base\n"
815 "is 0, it is chosen from the leading characters of s, 0 for\n"
816 "octal, 0x or 0X for hexadecimal. If base is 16, a preceding\n"
817 "0x or 0X is accepted. A trailing L or l is not accepted,\n"
818 "unless base is 0.");
820 static PyObject *
821 strop_atol(PyObject *self, PyObject *args)
823 char *s, *end;
824 int base = 10;
825 PyObject *x;
826 char buffer[256]; /* For errors */
828 WARN;
829 if (!PyArg_ParseTuple(args, "s|i:atol", &s, &base))
830 return NULL;
832 if ((base != 0 && base < 2) || base > 36) {
833 PyErr_SetString(PyExc_ValueError, "invalid base for atol()");
834 return NULL;
837 while (*s && isspace(Py_CHARMASK(*s)))
838 s++;
839 if (s[0] == '\0') {
840 PyErr_SetString(PyExc_ValueError, "empty string for atol()");
841 return NULL;
843 x = PyLong_FromString(s, &end, base);
844 if (x == NULL)
845 return NULL;
846 if (base == 0 && (*end == 'l' || *end == 'L'))
847 end++;
848 while (*end && isspace(Py_CHARMASK(*end)))
849 end++;
850 if (*end != '\0') {
851 PyOS_snprintf(buffer, sizeof(buffer),
852 "invalid literal for atol(): %.200s", s);
853 PyErr_SetString(PyExc_ValueError, buffer);
854 Py_DECREF(x);
855 return NULL;
857 return x;
861 PyDoc_STRVAR(atof__doc__,
862 "atof(s) -> float\n"
863 "\n"
864 "Return the floating point number represented by the string s.");
866 static PyObject *
867 strop_atof(PyObject *self, PyObject *args)
869 char *s, *end;
870 double x;
871 char buffer[256]; /* For errors */
873 WARN;
874 if (!PyArg_ParseTuple(args, "s:atof", &s))
875 return NULL;
876 while (*s && isspace(Py_CHARMASK(*s)))
877 s++;
878 if (s[0] == '\0') {
879 PyErr_SetString(PyExc_ValueError, "empty string for atof()");
880 return NULL;
882 errno = 0;
883 PyFPE_START_PROTECT("strop_atof", return 0)
884 x = PyOS_ascii_strtod(s, &end);
885 PyFPE_END_PROTECT(x)
886 while (*end && isspace(Py_CHARMASK(*end)))
887 end++;
888 if (*end != '\0') {
889 PyOS_snprintf(buffer, sizeof(buffer),
890 "invalid literal for atof(): %.200s", s);
891 PyErr_SetString(PyExc_ValueError, buffer);
892 return NULL;
894 else if (errno != 0) {
895 PyOS_snprintf(buffer, sizeof(buffer),
896 "atof() literal too large: %.200s", s);
897 PyErr_SetString(PyExc_ValueError, buffer);
898 return NULL;
900 return PyFloat_FromDouble(x);
904 PyDoc_STRVAR(maketrans__doc__,
905 "maketrans(frm, to) -> string\n"
906 "\n"
907 "Return a translation table (a string of 256 bytes long)\n"
908 "suitable for use in string.translate. The strings frm and to\n"
909 "must be of the same length.");
911 static PyObject *
912 strop_maketrans(PyObject *self, PyObject *args)
914 unsigned char *c, *from=NULL, *to=NULL;
915 Py_ssize_t i, fromlen=0, tolen=0;
916 PyObject *result;
918 if (!PyArg_ParseTuple(args, "t#t#:maketrans", &from, &fromlen, &to, &tolen))
919 return NULL;
921 if (fromlen != tolen) {
922 PyErr_SetString(PyExc_ValueError,
923 "maketrans arguments must have same length");
924 return NULL;
927 result = PyString_FromStringAndSize((char *)NULL, 256);
928 if (result == NULL)
929 return NULL;
930 c = (unsigned char *) PyString_AS_STRING((PyStringObject *)result);
931 for (i = 0; i < 256; i++)
932 c[i]=(unsigned char)i;
933 for (i = 0; i < fromlen; i++)
934 c[from[i]]=to[i];
936 return result;
940 PyDoc_STRVAR(translate__doc__,
941 "translate(s,table [,deletechars]) -> string\n"
942 "\n"
943 "Return a copy of the string s, where all characters occurring\n"
944 "in the optional argument deletechars are removed, and the\n"
945 "remaining characters have been mapped through the given\n"
946 "translation table, which must be a string of length 256.");
948 static PyObject *
949 strop_translate(PyObject *self, PyObject *args)
951 register char *input, *table, *output;
952 Py_ssize_t i;
953 int c, changed = 0;
954 PyObject *input_obj;
955 char *table1, *output_start, *del_table=NULL;
956 Py_ssize_t inlen, tablen, dellen = 0;
957 PyObject *result;
958 int trans_table[256];
960 WARN;
961 if (!PyArg_ParseTuple(args, "St#|t#:translate", &input_obj,
962 &table1, &tablen, &del_table, &dellen))
963 return NULL;
964 if (tablen != 256) {
965 PyErr_SetString(PyExc_ValueError,
966 "translation table must be 256 characters long");
967 return NULL;
970 table = table1;
971 inlen = PyString_GET_SIZE(input_obj);
972 result = PyString_FromStringAndSize((char *)NULL, inlen);
973 if (result == NULL)
974 return NULL;
975 output_start = output = PyString_AsString(result);
976 input = PyString_AsString(input_obj);
978 if (dellen == 0) {
979 /* If no deletions are required, use faster code */
980 for (i = inlen; --i >= 0; ) {
981 c = Py_CHARMASK(*input++);
982 if (Py_CHARMASK((*output++ = table[c])) != c)
983 changed = 1;
985 if (changed)
986 return result;
987 Py_DECREF(result);
988 Py_INCREF(input_obj);
989 return input_obj;
992 for (i = 0; i < 256; i++)
993 trans_table[i] = Py_CHARMASK(table[i]);
995 for (i = 0; i < dellen; i++)
996 trans_table[(int) Py_CHARMASK(del_table[i])] = -1;
998 for (i = inlen; --i >= 0; ) {
999 c = Py_CHARMASK(*input++);
1000 if (trans_table[c] != -1)
1001 if (Py_CHARMASK(*output++ = (char)trans_table[c]) == c)
1002 continue;
1003 changed = 1;
1005 if (!changed) {
1006 Py_DECREF(result);
1007 Py_INCREF(input_obj);
1008 return input_obj;
1010 /* Fix the size of the resulting string */
1011 if (inlen > 0)
1012 _PyString_Resize(&result, output - output_start);
1013 return result;
1017 /* What follows is used for implementing replace(). Perry Stoll. */
1020 mymemfind
1022 strstr replacement for arbitrary blocks of memory.
1024 Locates the first occurrence in the memory pointed to by MEM of the
1025 contents of memory pointed to by PAT. Returns the index into MEM if
1026 found, or -1 if not found. If len of PAT is greater than length of
1027 MEM, the function returns -1.
1029 static Py_ssize_t
1030 mymemfind(const char *mem, Py_ssize_t len, const char *pat, Py_ssize_t pat_len)
1032 register Py_ssize_t ii;
1034 /* pattern can not occur in the last pat_len-1 chars */
1035 len -= pat_len;
1037 for (ii = 0; ii <= len; ii++) {
1038 if (mem[ii] == pat[0] &&
1039 (pat_len == 1 ||
1040 memcmp(&mem[ii+1], &pat[1], pat_len-1) == 0)) {
1041 return ii;
1044 return -1;
1048 mymemcnt
1050 Return the number of distinct times PAT is found in MEM.
1051 meaning mem=1111 and pat==11 returns 2.
1052 mem=11111 and pat==11 also return 2.
1054 static Py_ssize_t
1055 mymemcnt(const char *mem, Py_ssize_t len, const char *pat, Py_ssize_t pat_len)
1057 register Py_ssize_t offset = 0;
1058 Py_ssize_t nfound = 0;
1060 while (len >= 0) {
1061 offset = mymemfind(mem, len, pat, pat_len);
1062 if (offset == -1)
1063 break;
1064 mem += offset + pat_len;
1065 len -= offset + pat_len;
1066 nfound++;
1068 return nfound;
1072 mymemreplace
1074 Return a string in which all occurrences of PAT in memory STR are
1075 replaced with SUB.
1077 If length of PAT is less than length of STR or there are no occurrences
1078 of PAT in STR, then the original string is returned. Otherwise, a new
1079 string is allocated here and returned.
1081 on return, out_len is:
1082 the length of output string, or
1083 -1 if the input string is returned, or
1084 unchanged if an error occurs (no memory).
1086 return value is:
1087 the new string allocated locally, or
1088 NULL if an error occurred.
1090 static char *
1091 mymemreplace(const char *str, Py_ssize_t len, /* input string */
1092 const char *pat, Py_ssize_t pat_len, /* pattern string to find */
1093 const char *sub, Py_ssize_t sub_len, /* substitution string */
1094 Py_ssize_t count, /* number of replacements */
1095 Py_ssize_t *out_len)
1097 char *out_s;
1098 char *new_s;
1099 Py_ssize_t nfound, offset, new_len;
1101 if (len == 0 || pat_len > len)
1102 goto return_same;
1104 /* find length of output string */
1105 nfound = mymemcnt(str, len, pat, pat_len);
1106 if (count < 0)
1107 count = PY_SSIZE_T_MAX;
1108 else if (nfound > count)
1109 nfound = count;
1110 if (nfound == 0)
1111 goto return_same;
1113 new_len = len + nfound*(sub_len - pat_len);
1114 if (new_len == 0) {
1115 /* Have to allocate something for the caller to free(). */
1116 out_s = (char *)PyMem_MALLOC(1);
1117 if (out_s == NULL)
1118 return NULL;
1119 out_s[0] = '\0';
1121 else {
1122 assert(new_len > 0);
1123 new_s = (char *)PyMem_MALLOC(new_len);
1124 if (new_s == NULL)
1125 return NULL;
1126 out_s = new_s;
1128 for (; count > 0 && len > 0; --count) {
1129 /* find index of next instance of pattern */
1130 offset = mymemfind(str, len, pat, pat_len);
1131 if (offset == -1)
1132 break;
1134 /* copy non matching part of input string */
1135 memcpy(new_s, str, offset);
1136 str += offset + pat_len;
1137 len -= offset + pat_len;
1139 /* copy substitute into the output string */
1140 new_s += offset;
1141 memcpy(new_s, sub, sub_len);
1142 new_s += sub_len;
1144 /* copy any remaining values into output string */
1145 if (len > 0)
1146 memcpy(new_s, str, len);
1148 *out_len = new_len;
1149 return out_s;
1151 return_same:
1152 *out_len = -1;
1153 return (char *)str; /* cast away const */
1157 PyDoc_STRVAR(replace__doc__,
1158 "replace (str, old, new[, maxsplit]) -> string\n"
1159 "\n"
1160 "Return a copy of string str with all occurrences of substring\n"
1161 "old replaced by new. If the optional argument maxsplit is\n"
1162 "given, only the first maxsplit occurrences are replaced.");
1164 static PyObject *
1165 strop_replace(PyObject *self, PyObject *args)
1167 char *str, *pat,*sub,*new_s;
1168 Py_ssize_t len,pat_len,sub_len,out_len;
1169 Py_ssize_t count = -1;
1170 PyObject *newstr;
1172 WARN;
1173 if (!PyArg_ParseTuple(args, "t#t#t#|n:replace",
1174 &str, &len, &pat, &pat_len, &sub, &sub_len,
1175 &count))
1176 return NULL;
1177 if (pat_len <= 0) {
1178 PyErr_SetString(PyExc_ValueError, "empty pattern string");
1179 return NULL;
1181 /* CAUTION: strop treats a replace count of 0 as infinity, unlke
1182 * current (2.1) string.py and string methods. Preserve this for
1183 * ... well, hard to say for what <wink>.
1185 if (count == 0)
1186 count = -1;
1187 new_s = mymemreplace(str,len,pat,pat_len,sub,sub_len,count,&out_len);
1188 if (new_s == NULL) {
1189 PyErr_NoMemory();
1190 return NULL;
1192 if (out_len == -1) {
1193 /* we're returning another reference to the input string */
1194 newstr = PyTuple_GetItem(args, 0);
1195 Py_XINCREF(newstr);
1197 else {
1198 newstr = PyString_FromStringAndSize(new_s, out_len);
1199 PyMem_FREE(new_s);
1201 return newstr;
1205 /* List of functions defined in the module */
1207 static PyMethodDef
1208 strop_methods[] = {
1209 {"atof", strop_atof, METH_VARARGS, atof__doc__},
1210 {"atoi", strop_atoi, METH_VARARGS, atoi__doc__},
1211 {"atol", strop_atol, METH_VARARGS, atol__doc__},
1212 {"capitalize", strop_capitalize, METH_O, capitalize__doc__},
1213 {"count", strop_count, METH_VARARGS, count__doc__},
1214 {"expandtabs", strop_expandtabs, METH_VARARGS, expandtabs__doc__},
1215 {"find", strop_find, METH_VARARGS, find__doc__},
1216 {"join", strop_joinfields, METH_VARARGS, joinfields__doc__},
1217 {"joinfields", strop_joinfields, METH_VARARGS, joinfields__doc__},
1218 {"lstrip", strop_lstrip, METH_O, lstrip__doc__},
1219 {"lower", strop_lower, METH_O, lower__doc__},
1220 {"maketrans", strop_maketrans, METH_VARARGS, maketrans__doc__},
1221 {"replace", strop_replace, METH_VARARGS, replace__doc__},
1222 {"rfind", strop_rfind, METH_VARARGS, rfind__doc__},
1223 {"rstrip", strop_rstrip, METH_O, rstrip__doc__},
1224 {"split", strop_splitfields, METH_VARARGS, splitfields__doc__},
1225 {"splitfields", strop_splitfields, METH_VARARGS, splitfields__doc__},
1226 {"strip", strop_strip, METH_O, strip__doc__},
1227 {"swapcase", strop_swapcase, METH_O, swapcase__doc__},
1228 {"translate", strop_translate, METH_VARARGS, translate__doc__},
1229 {"upper", strop_upper, METH_O, upper__doc__},
1230 {NULL, NULL} /* sentinel */
1234 PyMODINIT_FUNC
1235 initstrop(void)
1237 PyObject *m, *s;
1238 char buf[256];
1239 int c, n;
1240 m = Py_InitModule4("strop", strop_methods, strop_module__doc__,
1241 (PyObject*)NULL, PYTHON_API_VERSION);
1242 if (m == NULL)
1243 return;
1245 /* Create 'whitespace' object */
1246 n = 0;
1247 for (c = 0; c < 256; c++) {
1248 if (isspace(c))
1249 buf[n++] = c;
1251 s = PyString_FromStringAndSize(buf, n);
1252 if (s)
1253 PyModule_AddObject(m, "whitespace", s);
1255 /* Create 'lowercase' object */
1256 n = 0;
1257 for (c = 0; c < 256; c++) {
1258 if (islower(c))
1259 buf[n++] = c;
1261 s = PyString_FromStringAndSize(buf, n);
1262 if (s)
1263 PyModule_AddObject(m, "lowercase", s);
1265 /* Create 'uppercase' object */
1266 n = 0;
1267 for (c = 0; c < 256; c++) {
1268 if (isupper(c))
1269 buf[n++] = c;
1271 s = PyString_FromStringAndSize(buf, n);
1272 if (s)
1273 PyModule_AddObject(m, "uppercase", s);