Annotation of OpenXM_contrib/gc/include/cord.h, Revision 1.1.1.2
1.1 maekawa 1: /*
2: * Copyright (c) 1993-1994 by Xerox Corporation. All rights reserved.
3: *
4: * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
5: * OR IMPLIED. ANY USE IS AT YOUR OWN RISK.
6: *
7: * Permission is hereby granted to use or copy this program
8: * for any purpose, provided the above notices are retained on all copies.
9: * Permission to modify the code and to distribute modified code is granted,
10: * provided the above notices are retained, and a notice that the code was
11: * modified is included with the above copyright notice.
12: *
13: * Author: Hans-J. Boehm (boehm@parc.xerox.com)
14: */
15: /* Boehm, October 5, 1995 4:20 pm PDT */
16:
17: /*
18: * Cords are immutable character strings. A number of operations
19: * on long cords are much more efficient than their strings.h counterpart.
20: * In particular, concatenation takes constant time independent of the length
21: * of the arguments. (Cords are represented as trees, with internal
22: * nodes representing concatenation and leaves consisting of either C
23: * strings or a functional description of the string.)
24: *
25: * The following are reasonable applications of cords. They would perform
26: * unacceptably if C strings were used:
27: * - A compiler that produces assembly language output by repeatedly
28: * concatenating instructions onto a cord representing the output file.
29: * - A text editor that converts the input file to a cord, and then
30: * performs editing operations by producing a new cord representing
31: * the file after echa character change (and keeping the old ones in an
32: * edit history)
33: *
34: * For optimal performance, cords should be built by
35: * concatenating short sections.
36: * This interface is designed for maximum compatibility with C strings.
37: * ASCII NUL characters may be embedded in cords using CORD_from_fn.
38: * This is handled correctly, but CORD_to_char_star will produce a string
39: * with embedded NULs when given such a cord.
40: *
41: * This interface is fairly big, largely for performance reasons.
42: * The most basic constants and functions:
43: *
1.1.1.2 ! maekawa 44: * CORD - the type of a cord;
1.1 maekawa 45: * CORD_EMPTY - empty cord;
46: * CORD_len(cord) - length of a cord;
47: * CORD_cat(cord1,cord2) - concatenation of two cords;
48: * CORD_substr(cord, start, len) - substring (or subcord);
49: * CORD_pos i; CORD_FOR(i, cord) { ... CORD_pos_fetch(i) ... } -
50: * examine each character in a cord. CORD_pos_fetch(i) is the char.
51: * CORD_fetch(int i) - Retrieve i'th character (slowly).
52: * CORD_cmp(cord1, cord2) - compare two cords.
53: * CORD_from_file(FILE * f) - turn a read-only file into a cord.
54: * CORD_to_char_star(cord) - convert to C string.
55: * (Non-NULL C constant strings are cords.)
56: * CORD_printf (etc.) - cord version of printf. Use %r for cords.
57: */
58: # ifndef CORD_H
59:
60: # define CORD_H
61: # include <stddef.h>
62: # include <stdio.h>
63: /* Cords have type const char *. This is cheating quite a bit, and not */
64: /* 100% portable. But it means that nonempty character string */
65: /* constants may be used as cords directly, provided the string is */
66: /* never modified in place. The empty cord is represented by, and */
67: /* can be written as, 0. */
68:
69: typedef const char * CORD;
70:
71: /* An empty cord is always represented as nil */
72: # define CORD_EMPTY 0
73:
74: /* Is a nonempty cord represented as a C string? */
75: #define CORD_IS_STRING(s) (*(s) != '\0')
76:
77: /* Concatenate two cords. If the arguments are C strings, they may */
78: /* not be subsequently altered. */
79: CORD CORD_cat(CORD x, CORD y);
80:
81: /* Concatenate a cord and a C string with known length. Except for the */
82: /* empty string case, this is a special case of CORD_cat. Since the */
83: /* length is known, it can be faster. */
84: /* The string y is shared with the resulting CORD. Hence it should */
85: /* not be altered by the caller. */
86: CORD CORD_cat_char_star(CORD x, const char * y, size_t leny);
87:
88: /* Compute the length of a cord */
89: size_t CORD_len(CORD x);
90:
91: /* Cords may be represented by functions defining the ith character */
92: typedef char (* CORD_fn)(size_t i, void * client_data);
93:
94: /* Turn a functional description into a cord. */
95: CORD CORD_from_fn(CORD_fn fn, void * client_data, size_t len);
96:
97: /* Return the substring (subcord really) of x with length at most n, */
98: /* starting at position i. (The initial character has position 0.) */
99: CORD CORD_substr(CORD x, size_t i, size_t n);
100:
101: /* Return the argument, but rebalanced to allow more efficient */
102: /* character retrieval, substring operations, and comparisons. */
103: /* This is useful only for cords that were built using repeated */
104: /* concatenation. Guarantees log time access to the result, unless */
105: /* x was obtained through a large number of repeated substring ops */
106: /* or the embedded functional descriptions take longer to evaluate. */
107: /* May reallocate significant parts of the cord. The argument is not */
108: /* modified; only the result is balanced. */
109: CORD CORD_balance(CORD x);
110:
111: /* The following traverse a cord by applying a function to each */
112: /* character. This is occasionally appropriate, especially where */
113: /* speed is crucial. But, since C doesn't have nested functions, */
114: /* clients of this sort of traversal are clumsy to write. Consider */
115: /* the functions that operate on cord positions instead. */
116:
117: /* Function to iteratively apply to individual characters in cord. */
118: typedef int (* CORD_iter_fn)(char c, void * client_data);
119:
120: /* Function to apply to substrings of a cord. Each substring is a */
121: /* a C character string, not a general cord. */
122: typedef int (* CORD_batched_iter_fn)(const char * s, void * client_data);
123: # define CORD_NO_FN ((CORD_batched_iter_fn)0)
124:
125: /* Apply f1 to each character in the cord, in ascending order, */
126: /* starting at position i. If */
127: /* f2 is not CORD_NO_FN, then multiple calls to f1 may be replaced by */
128: /* a single call to f2. The parameter f2 is provided only to allow */
129: /* some optimization by the client. This terminates when the right */
130: /* end of this string is reached, or when f1 or f2 return != 0. In the */
131: /* latter case CORD_iter returns != 0. Otherwise it returns 0. */
132: /* The specified value of i must be < CORD_len(x). */
133: int CORD_iter5(CORD x, size_t i, CORD_iter_fn f1,
134: CORD_batched_iter_fn f2, void * client_data);
135:
136: /* A simpler version that starts at 0, and without f2: */
137: int CORD_iter(CORD x, CORD_iter_fn f1, void * client_data);
138: # define CORD_iter(x, f1, cd) CORD_iter5(x, 0, f1, CORD_NO_FN, cd)
139:
140: /* Similar to CORD_iter5, but end-to-beginning. No provisions for */
141: /* CORD_batched_iter_fn. */
142: int CORD_riter4(CORD x, size_t i, CORD_iter_fn f1, void * client_data);
143:
144: /* A simpler version that starts at the end: */
145: int CORD_riter(CORD x, CORD_iter_fn f1, void * client_data);
146:
147: /* Functions that operate on cord positions. The easy way to traverse */
148: /* cords. A cord position is logically a pair consisting of a cord */
149: /* and an index into that cord. But it is much faster to retrieve a */
150: /* charcter based on a position than on an index. Unfortunately, */
151: /* positions are big (order of a few 100 bytes), so allocate them with */
152: /* caution. */
153: /* Things in cord_pos.h should be treated as opaque, except as */
154: /* described below. Also note that */
155: /* CORD_pos_fetch, CORD_next and CORD_prev have both macro and function */
156: /* definitions. The former may evaluate their argument more than once. */
157: # include "private/cord_pos.h"
158:
159: /*
160: Visible definitions from above:
161:
162: typedef <OPAQUE but fairly big> CORD_pos[1];
163:
164: * Extract the cord from a position:
165: CORD CORD_pos_to_cord(CORD_pos p);
166:
167: * Extract the current index from a position:
168: size_t CORD_pos_to_index(CORD_pos p);
169:
170: * Fetch the character located at the given position:
171: char CORD_pos_fetch(CORD_pos p);
172:
173: * Initialize the position to refer to the given cord and index.
174: * Note that this is the most expensive function on positions:
175: void CORD_set_pos(CORD_pos p, CORD x, size_t i);
176:
177: * Advance the position to the next character.
178: * P must be initialized and valid.
179: * Invalidates p if past end:
180: void CORD_next(CORD_pos p);
181:
182: * Move the position to the preceding character.
183: * P must be initialized and valid.
184: * Invalidates p if past beginning:
185: void CORD_prev(CORD_pos p);
186:
187: * Is the position valid, i.e. inside the cord?
188: int CORD_pos_valid(CORD_pos p);
189: */
190: # define CORD_FOR(pos, cord) \
191: for (CORD_set_pos(pos, cord, 0); CORD_pos_valid(pos); CORD_next(pos))
192:
193:
194: /* An out of memory handler to call. May be supplied by client. */
195: /* Must not return. */
196: extern void (* CORD_oom_fn)(void);
197:
198: /* Dump the representation of x to stdout in an implementation defined */
199: /* manner. Intended for debugging only. */
200: void CORD_dump(CORD x);
201:
202: /* The following could easily be implemented by the client. They are */
203: /* provided in cordxtra.c for convenience. */
204:
205: /* Concatenate a character to the end of a cord. */
206: CORD CORD_cat_char(CORD x, char c);
207:
208: /* Concatenate n cords. */
209: CORD CORD_catn(int n, /* CORD */ ...);
210:
211: /* Return the character in CORD_substr(x, i, 1) */
212: char CORD_fetch(CORD x, size_t i);
213:
214: /* Return < 0, 0, or > 0, depending on whether x < y, x = y, x > y */
215: int CORD_cmp(CORD x, CORD y);
216:
217: /* A generalization that takes both starting positions for the */
218: /* comparison, and a limit on the number of characters to be compared. */
219: int CORD_ncmp(CORD x, size_t x_start, CORD y, size_t y_start, size_t len);
220:
221: /* Find the first occurrence of s in x at position start or later. */
222: /* Return the position of the first character of s in x, or */
223: /* CORD_NOT_FOUND if there is none. */
224: size_t CORD_str(CORD x, size_t start, CORD s);
225:
226: /* Return a cord consisting of i copies of (possibly NUL) c. Dangerous */
227: /* in conjunction with CORD_to_char_star. */
228: /* The resulting representation takes constant space, independent of i. */
229: CORD CORD_chars(char c, size_t i);
230: # define CORD_nul(i) CORD_chars('\0', (i))
231:
232: /* Turn a file into cord. The file must be seekable. Its contents */
233: /* must remain constant. The file may be accessed as an immediate */
234: /* result of this call and/or as a result of subsequent accesses to */
235: /* the cord. Short files are likely to be immediately read, but */
236: /* long files are likely to be read on demand, possibly relying on */
237: /* stdio for buffering. */
238: /* We must have exclusive access to the descriptor f, i.e. we may */
239: /* read it at any time, and expect the file pointer to be */
240: /* where we left it. Normally this should be invoked as */
241: /* CORD_from_file(fopen(...)) */
242: /* CORD_from_file arranges to close the file descriptor when it is no */
243: /* longer needed (e.g. when the result becomes inaccessible). */
244: /* The file f must be such that ftell reflects the actual character */
245: /* position in the file, i.e. the number of characters that can be */
246: /* or were read with fread. On UNIX systems this is always true. On */
247: /* MS Windows systems, f must be opened in binary mode. */
248: CORD CORD_from_file(FILE * f);
249:
250: /* Equivalent to the above, except that the entire file will be read */
251: /* and the file pointer will be closed immediately. */
252: /* The binary mode restriction from above does not apply. */
253: CORD CORD_from_file_eager(FILE * f);
254:
255: /* Equivalent to the above, except that the file will be read on demand.*/
256: /* The binary mode restriction applies. */
257: CORD CORD_from_file_lazy(FILE * f);
258:
259: /* Turn a cord into a C string. The result shares no structure with */
260: /* x, and is thus modifiable. */
261: char * CORD_to_char_star(CORD x);
262:
263: /* Turn a C string into a CORD. The C string is copied, and so may */
264: /* subsequently be modified. */
265: CORD CORD_from_char_star(const char *s);
266:
267: /* Identical to the above, but the result may share structure with */
268: /* the argument and is thus not modifiable. */
269: const char * CORD_to_const_char_star(CORD x);
270:
271: /* Write a cord to a file, starting at the current position. No */
272: /* trailing NULs are newlines are added. */
273: /* Returns EOF if a write error occurs, 1 otherwise. */
274: int CORD_put(CORD x, FILE * f);
275:
276: /* "Not found" result for the following two functions. */
277: # define CORD_NOT_FOUND ((size_t)(-1))
278:
279: /* A vague analog of strchr. Returns the position (an integer, not */
280: /* a pointer) of the first occurrence of (char) c inside x at position */
281: /* i or later. The value i must be < CORD_len(x). */
282: size_t CORD_chr(CORD x, size_t i, int c);
283:
284: /* A vague analog of strrchr. Returns index of the last occurrence */
285: /* of (char) c inside x at position i or earlier. The value i */
286: /* must be < CORD_len(x). */
287: size_t CORD_rchr(CORD x, size_t i, int c);
288:
289:
290: /* The following are also not primitive, but are implemented in */
291: /* cordprnt.c. They provide functionality similar to the ANSI C */
292: /* functions with corresponding names, but with the following */
293: /* additions and changes: */
294: /* 1. A %r conversion specification specifies a CORD argument. Field */
295: /* width, precision, etc. have the same semantics as for %s. */
296: /* (Note that %c,%C, and %S were already taken.) */
297: /* 2. The format string is represented as a CORD. */
298: /* 3. CORD_sprintf and CORD_vsprintf assign the result through the 1st */ /* argument. Unlike their ANSI C versions, there is no need to guess */
299: /* the correct buffer size. */
300: /* 4. Most of the conversions are implement through the native */
301: /* vsprintf. Hence they are usually no faster, and */
302: /* idiosyncracies of the native printf are preserved. However, */
303: /* CORD arguments to CORD_sprintf and CORD_vsprintf are NOT copied; */
304: /* the result shares the original structure. This may make them */
305: /* very efficient in some unusual applications. */
306: /* The format string is copied. */
307: /* All functions return the number of characters generated or -1 on */
308: /* error. This complies with the ANSI standard, but is inconsistent */
309: /* with some older implementations of sprintf. */
310:
311: /* The implementation of these is probably less portable than the rest */
312: /* of this package. */
313:
314: #ifndef CORD_NO_IO
315:
316: #include <stdarg.h>
317:
318: int CORD_sprintf(CORD * out, CORD format, ...);
319: int CORD_vsprintf(CORD * out, CORD format, va_list args);
320: int CORD_fprintf(FILE * f, CORD format, ...);
321: int CORD_vfprintf(FILE * f, CORD format, va_list args);
322: int CORD_printf(CORD format, ...);
323: int CORD_vprintf(CORD format, va_list args);
324:
325: #endif /* CORD_NO_IO */
326:
327: # endif /* CORD_H */
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to cord.h CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [local] / OpenXM_contrib / gc / include