Context Navigation

string_res.cfa@ 804bf677

Visit:

ADT ast-experimental enum pthread-emulation qualifiedEnum

Last change on this file since 804bf677 was 804bf677, checked in by Michael Brooks <mlbrooks@…>, 4 years ago
String hybrid: Basic cases of solo alloc now working
Property mode set to `100644`
File size: 35.0 KB

Line
1	//
2	// Cforall Version 1.0.0 Copyright (C) 2016 University of Waterloo
3	//
4	// The contents of this file are covered under the licence agreement in the
5	// file "LICENCE" distributed with Cforall.
6	//
7	// string_res -- variable-length, mutable run of text, with resource semantics
8	//
9	// Author : Michael L. Brooks
10	// Created On : Fri Sep 03 11:00:00 2021
11	// Last Modified By : Michael L. Brooks
12	// Last Modified On : Fri Sep 03 11:00:00 2021
13	// Update Count : 1
14	//
15
16	#include "string_res.hfa"
17	#include "string_sharectx.hfa"
18
19	#include <stdlib.hfa> // e.g. malloc
20	#include <string.h> // e.g. strlen
21	#include <assert.h>
22
23	//######################### VbyteHeap "header" #########################
24
25	#ifdef VbyteDebug
26	HandleNode *HeaderPtr;
27	#endif // VbyteDebug
28
29	struct VbyteHeap {
30
31	int NoOfCompactions; // number of compactions of the byte area
32	int NoOfExtensions; // number of extensions in the size of the byte area
33	int NoOfReductions; // number of reductions in the size of the byte area
34
35	int InitSize; // initial number of bytes in the byte-string area
36	int CurrSize; // current number of bytes in the byte-string area
37	char *StartVbyte; // pointer to the `st byte of the start of the byte-string area
38	char *EndVbyte; // pointer to the next byte after the end of the currently used portion of byte-string area
39	void *ExtVbyte; // pointer to the next byte after the end of the byte-string area
40
41	HandleNode Header; // header node for handle list
42	}; // VbyteHeap
43
44
45	static inline void compaction( VbyteHeap & ); // compaction of the byte area
46	static inline void garbage( VbyteHeap & ); // garbage collect the byte area
47	static inline void extend( VbyteHeap &, int ); // extend the size of the byte area
48	static inline void reduce( VbyteHeap &, int ); // reduce the size of the byte area
49
50	static inline void ?{}( VbyteHeap &, size_t = 1000 );
51	static inline void ^?{}( VbyteHeap & );
52	static inline void ByteCopy( char , int, int, char , int, int ); // copy a block of bytes from one location in the heap to another
53	static inline int ByteCmp( char , int, int, char , int, int ); // compare 2 blocks of bytes
54	static inline char *VbyteAlloc( VbyteHeap &, int ); // allocate a block bytes in the heap
55
56
57	static inline void AddThisAfter( HandleNode &, HandleNode & );
58	static inline void DeleteNode( HandleNode & );
59	static inline void MoveThisAfter( HandleNode &, const HandleNode & ); // move current handle after parameter handle
60
61
62	// Allocate the storage for the variable sized area and intialize the heap variables.
63
64	static inline void ?{}( VbyteHeap & this, size_t Size ) with(this) {
65	#ifdef VbyteDebug
66	serr \| "enter:VbyteHeap::VbyteHeap, this:" \| &this \| " Size:" \| Size;
67	#endif // VbyteDebug
68	NoOfCompactions = NoOfExtensions = NoOfReductions = 0;
69	InitSize = CurrSize = Size;
70	StartVbyte = EndVbyte = alloc(CurrSize);
71	ExtVbyte = (void *)( StartVbyte + CurrSize );
72	Header.flink = Header.blink = &Header;
73	Header.ulink = & this;
74	#ifdef VbyteDebug
75	HeaderPtr = &Header;
76	serr \| "exit:VbyteHeap::VbyteHeap, this:" \| &this;
77	#endif // VbyteDebug
78	} // VbyteHeap
79
80
81	// Release the dynamically allocated storage for the byte area.
82
83	static inline void ^?{}( VbyteHeap & this ) with(this) {
84	free( StartVbyte );
85	} // ~VbyteHeap
86
87
88	//######################### HandleNode #########################
89
90
91	// Create a handle node. The handle is not linked into the handle list. This is the responsibilitiy of the handle
92	// creator.
93
94	static inline void ?{}( HandleNode & this ) with(this) {
95	#ifdef VbyteDebug
96	serr \| "enter:HandleNode::HandleNode, this:" \| &this;
97	#endif // VbyteDebug
98	s = 0;
99	lnth = 0;
100	#ifdef VbyteDebug
101	serr \| "exit:HandleNode::HandleNode, this:" \| &this;
102	#endif // VbyteDebug
103	} // HandleNode
104
105	// Create a handle node. The handle is linked into the handle list at the end. This means that this handle will NOT be
106	// in order by string address, but this is not a problem because a string with length zero does nothing during garbage
107	// collection.
108
109	static inline void ?{}( HandleNode & this, VbyteHeap & vh ) with(this) {
110	#ifdef VbyteDebug
111	serr \| "enter:HandleNode::HandleNode, this:" \| &this;
112	#endif // VbyteDebug
113	s = 0;
114	lnth = 0;
115	ulink = &vh;
116	AddThisAfter( this, *vh.Header.blink );
117	#ifdef VbyteDebug
118	serr \| "exit:HandleNode::HandleNode, this:" \| &this;
119	#endif // VbyteDebug
120	} // HandleNode
121
122
123	// Delete a node from the handle list by unchaining it from the list. If the handle node was allocated dynamically, it
124	// is the responsibility of the creator to destroy it.
125
126	static inline void ^?{}( HandleNode & this ) with(this) {
127	#ifdef VbyteDebug
128	serr \| "enter:HandleNode::~HandleNode, this:" \| & this;
129	{
130	serr \| nlOff;
131	serr \| " lnth:" \| lnth \| " s:" \| (void *)s \| ",\"";
132	for ( int i = 0; i < lnth; i += 1 ) {
133	serr \| s[i];
134	} // for
135	serr \| "\" flink:" \| flink \| " blink:" \| blink \| nl;
136	serr \| nlOn;
137	}
138	#endif // VbyteDebug
139	DeleteNode( this );
140	} // ~HandleNode
141
142
143	//######################### String Sharing Context #########################
144
145	static string_sharectx * ambient_string_sharectx; // fickle top of stack
146	static string_sharectx default_string_sharectx = {NEW_SHARING}; // stable bottom of stack
147
148	void ?{}( string_sharectx & this, StringSharectx_Mode mode ) with( this ) {
149	(older){ ambient_string_sharectx };
150	if ( mode == NEW_SHARING ) {
151	(activeHeap){ new( (size_t) 1000 ) };
152	} else {
153	verify( mode == NO_SHARING );
154	(activeHeap){ 0p };
155	}
156	ambient_string_sharectx = & this;
157	}
158
159	void ^?{}( string_sharectx & this ) with( this ) {
160	if ( activeHeap ) delete( activeHeap );
161
162	// unlink this from older-list starting from ambient_string_sharectx
163	// usually, this==ambient_string_sharectx and the loop runs zero times
164	string_sharectx *& c = ambient_string_sharectx;
165	while ( c != &this ) &c = &c->older; // find this
166	c = this.older; // unlink
167	}
168
169	//######################### String Resource #########################
170
171
172	VbyteHeap * DEBUG_string_heap() {
173	assert( ambient_string_sharectx->activeHeap && "No sharing context is active" );
174	return ambient_string_sharectx->activeHeap;
175	}
176
177	size_t DEBUG_string_bytes_avail_until_gc( VbyteHeap * heap ) {
178	return ((char*)heap->ExtVbyte) - heap->EndVbyte;
179	}
180
181	const char * DEBUG_string_heap_start( VbyteHeap * heap ) {
182	return heap->StartVbyte;
183	}
184
185
186	// Returns the size of the string in bytes
187	size_t size(const string_res &s) with(s) {
188	return Handle.lnth;
189	}
190
191	// Output operator
192	ofstream & ?\|?(ofstream &out, const string_res &s) {
193	// Store auto-newline state so it can be restored
194	bool anl = getANL$(out);
195	nlOff(out);
196	for (size_t i = 0; i < s.Handle.lnth; i++) {
197	out \| s[i];
198	}
199	out \| sep;
200	// Re-apply newlines after done, for chaining version
201	if (anl) nlOn(out);
202	return out;
203	}
204
205	void ?\|?(ofstream &out, const string_res &s) {
206	// Store auto-newline state so it can be restored
207	bool anl = getANL$(out);
208	nlOff(out);
209	for (size_t i = 0; i < s.Handle.lnth; i++) {
210	// Need to re-apply on the last output operator, for whole-statement version
211	if (anl && i == s.Handle.lnth-1) nlOn(out);
212	out \| s[i];
213	}
214	return out;
215	}
216
217	// Empty constructor
218	void ?{}(string_res &s) with(s) {
219	if( ambient_string_sharectx->activeHeap ) {
220	(Handle){ * ambient_string_sharectx->activeHeap };
221	(shareEditSet_owns_ulink){ false };
222	} else {
223	(Handle){ * new( (size_t) 10 ) }; // TODO: can I lazily avoid allocating for empty string
224	(shareEditSet_owns_ulink){ true };
225	}
226	s.shareEditSet_prev = &s;
227	s.shareEditSet_next = &s;
228	}
229
230	static inline void eagerCopyCtorHelper(string_res &s, const char* rhs, size_t rhslnth) with(s) {
231	if( ambient_string_sharectx->activeHeap ) {
232	(Handle){ * ambient_string_sharectx->activeHeap };
233	(shareEditSet_owns_ulink){ false };
234	} else {
235	(Handle){ * new( rhslnth ) };
236	(shareEditSet_owns_ulink){ true };
237	}
238	Handle.s = VbyteAlloc(*Handle.ulink, rhslnth);
239	Handle.lnth = rhslnth;
240	for ( int i = 0; i < rhslnth; i += 1 ) { // copy characters
241	Handle.s[i] = rhs[i];
242	} // for
243	s.shareEditSet_prev = &s;
244	s.shareEditSet_next = &s;
245	}
246
247	// Constructor from a raw buffer and size
248	void ?{}(string_res &s, const char* rhs, size_t rhslnth) with(s) {
249	eagerCopyCtorHelper(s, rhs, rhslnth);
250	}
251
252	// String literal constructor
253	void ?{}(string_res &s, const char* rhs) {
254	(s){ rhs, strlen(rhs) };
255	}
256
257	// General copy constructor
258	void ?{}(string_res &s, const string_res & s2, StrResInitMode mode, size_t start, size_t end ) {
259
260	verify( start <= end && end <= s2.Handle.lnth );
261
262	if (s2.Handle.ulink == ambient_string_sharectx->activeHeap) {
263	// same heap: allow overlap
264	(s.Handle){};
265	s.Handle.s = s2.Handle.s + start;
266	s.Handle.lnth = end - start;
267	s.Handle.ulink = ambient_string_sharectx->activeHeap;
268	(s.shareEditSet_owns_ulink){ false };
269	AddThisAfter(s.Handle, s2.Handle ); // insert this handle after rhs handle
270	// ^ bug? skip others at early point in string
271
272	if (mode == COPY_VALUE) {
273	// make s alone in its shareEditSet
274	s.shareEditSet_prev = &s;
275	s.shareEditSet_next = &s;
276	} else {
277	verify( mode == SHARE_EDITS );
278
279	// s2 is logically const but not implementation const
280	string_res & s2mod = (string_res &) s2;
281
282	// insert s after s2 on shareEditSet
283	s.shareEditSet_next = s2mod.shareEditSet_next;
284	s.shareEditSet_prev = &s2mod;
285	s.shareEditSet_next->shareEditSet_prev = &s;
286	s.shareEditSet_prev->shareEditSet_next = &s;
287	}
288	} else {
289	// crossing heaps: eager copy
290	assert( mode == COPY_VALUE && "need to solidify context-crossing rules for requesting shared edits");
291	eagerCopyCtorHelper(s, s2.Handle.s + start, end - start);
292	verify(s.shareEditSet_prev == &s);
293	verify(s.shareEditSet_next == &s);
294	}
295	}
296
297	static void assignEditSet(string_res & this, string_res * shareEditSetStartPeer, string_res * shareEditSetEndPeer,
298	char * resultSesStart,
299	size_t resultSesLnth,
300	HandleNode * resultPadPosition, size_t bsize ) {
301
302	char * beforeBegin = shareEditSetStartPeer->Handle.s;
303	size_t beforeLen = this.Handle.s - beforeBegin;
304
305	char * afterBegin = this.Handle.s + this.Handle.lnth;
306	size_t afterLen = shareEditSetEndPeer->Handle.s + shareEditSetEndPeer->Handle.lnth - afterBegin;
307
308	size_t oldLnth = this.Handle.lnth;
309
310	this.Handle.s = resultSesStart + beforeLen;
311	this.Handle.lnth = bsize;
312	MoveThisAfter( this.Handle, *resultPadPosition );
313
314	// adjust all substring string and handle locations, and check if any substring strings are outside the new base string
315	char *limit = resultSesStart + resultSesLnth;
316	for (string_res * p = this.shareEditSet_next; p != &this; p = p->shareEditSet_next) {
317	verify (p->Handle.s >= beforeBegin);
318	if ( p->Handle.s >= afterBegin ) {
319	verify ( p->Handle.s <= afterBegin + afterLen );
320	verify ( p->Handle.s + p->Handle.lnth <= afterBegin + afterLen );
321	// p starts after the edit
322	// take start and end as end-anchored
323	size_t startOffsetFromEnd = afterBegin + afterLen - p->Handle.s;
324	p->Handle.s = limit - startOffsetFromEnd;
325	// p->Handle.lnth unaffected
326	} else if ( p->Handle.s <= beforeBegin + beforeLen ) {
327	// p starts before, or at the start of, the edit
328	if ( p->Handle.s + p->Handle.lnth <= beforeBegin + beforeLen ) {
329	// p ends before the edit
330	// take end as start-anchored too
331	// p->Handle.lnth unaffected
332	} else if ( p->Handle.s + p->Handle.lnth < afterBegin ) {
333	// p ends during the edit; p does not include the last character replaced
334	// clip end of p to end at start of edit
335	p->Handle.lnth = beforeLen - ( p->Handle.s - beforeBegin );
336	} else {
337	// p ends after the edit
338	verify ( p->Handle.s + p->Handle.lnth <= afterBegin + afterLen );
339	// take end as end-anchored
340	// stretch-shrink p according to the edit
341	p->Handle.lnth += this.Handle.lnth;
342	p->Handle.lnth -= oldLnth;
343	}
344	// take start as start-anchored
345	size_t startOffsetFromStart = p->Handle.s - beforeBegin;
346	p->Handle.s = resultSesStart + startOffsetFromStart;
347	} else {
348	verify ( p->Handle.s < afterBegin );
349	// p starts during the edit
350	verify( p->Handle.s + p->Handle.lnth >= beforeBegin + beforeLen );
351	if ( p->Handle.s + p->Handle.lnth < afterBegin ) {
352	// p ends during the edit; p does not include the last character replaced
353	// set p to empty string at start of edit
354	p->Handle.s = this.Handle.s;
355	p->Handle.lnth = 0;
356	} else {
357	// p includes the end of the edit
358	// clip start of p to start at end of edit
359	int charsToClip = afterBegin - p->Handle.s;
360	p->Handle.s = this.Handle.s + this.Handle.lnth;
361	p->Handle.lnth -= charsToClip;
362	}
363	}
364	MoveThisAfter( p->Handle, *resultPadPosition ); // move substring handle to maintain sorted order by string position
365	}
366	}
367
368	static void assign_(string_res &this, const char* buffer, size_t bsize, const string_res & valSrc) {
369
370	// traverse the incumbent share-edit set (SES) to recover the range of a base string to which `this` belongs
371	string_res * shareEditSetStartPeer = & this;
372	string_res * shareEditSetEndPeer = & this;
373	for (string_res * editPeer = this.shareEditSet_next; editPeer != &this; editPeer = editPeer->shareEditSet_next) {
374	if ( editPeer->Handle.s < shareEditSetStartPeer->Handle.s ) {
375	shareEditSetStartPeer = editPeer;
376	}
377	if ( shareEditSetEndPeer->Handle.s + shareEditSetEndPeer->Handle.lnth < editPeer->Handle.s + editPeer->Handle.lnth) {
378	shareEditSetEndPeer = editPeer;
379	}
380	}
381
382	verify( shareEditSetEndPeer->Handle.s >= shareEditSetStartPeer->Handle.s );
383	size_t editSetLength = shareEditSetEndPeer->Handle.s + shareEditSetEndPeer->Handle.lnth - shareEditSetStartPeer->Handle.s;
384	verify( editSetLength >= this.Handle.lnth );
385
386
387	if ( this.Handle.lnth == editSetLength // the entire run of the share-edit set is being overwritten: SES's result will only use characters from the source string
388	&& & valSrc // sourcing from a managed string
389	&& valSrc.Handle.ulink == this.Handle.ulink ) { // sourcing from same heap
390
391	assignEditSet(this, shareEditSetStartPeer, shareEditSetEndPeer,
392	valSrc.Handle.s,
393	valSrc.Handle.lnth,
394	&((string_res&)valSrc).Handle, bsize);
395
396	} else {
397
398	// full string is from start of shareEditSetStartPeer thru end of shareEditSetEndPeer
399	// `this` occurs in the middle of it, to be replaced
400	// build up the new text in `pasting`
401
402	// super private string ctor: ignore ambient context
403	void ?{}( string_res &s, VbyteHeap & heap, const char* rhs, size_t rhslnth ) with(s) {
404	(Handle){ heap };
405	Handle.s = VbyteAlloc(*Handle.ulink, rhslnth);
406	Handle.lnth = rhslnth;
407	(s.shareEditSet_owns_ulink){ false };
408	for ( int i = 0; i < rhslnth; i += 1 ) { // copy characters
409	Handle.s[i] = rhs[i];
410	} // for
411	s.shareEditSet_prev = &s;
412	s.shareEditSet_next = &s;
413	}
414
415	// we are only overwriting a proper substring of some string: need to mash characters from old and new together
416	// OR we are importing characters: need to copy eagerly
417	string_res pasting = {
418	* this.Handle.ulink, // maintain same heap, regardless of context
419	shareEditSetStartPeer->Handle.s, // start of SES
420	this.Handle.s - shareEditSetStartPeer->Handle.s }; // length of SES, before this
421	append( pasting,
422	buffer, // start of replacement for this
423	bsize ); // length of replacement for this
424	append( pasting,
425	this.Handle.s + this.Handle.lnth, // start of SES after this
426	shareEditSetEndPeer->Handle.s + shareEditSetEndPeer->Handle.lnth -
427	(this.Handle.s + this.Handle.lnth) ); // length of SES, after this
428
429	// The above string building can trigger compaction.
430	// The reference points (that are arguments of the string building) may move during that building.
431	// From this point on, they are stable.
432
433	assignEditSet(this, shareEditSetStartPeer, shareEditSetEndPeer,
434	pasting.Handle.s,
435	pasting.Handle.lnth,
436	&pasting.Handle, bsize);
437	}
438
439	// So now, capture their values for use in the overlap cases, below.
440	// Do not factor these definitions with the arguments used in string building above.
441
442	}
443
444	void assign(string_res &this, const char* buffer, size_t bsize) {
445	assign_(this, buffer, bsize, *0p);
446	}
447
448	void ?=?(string_res &s, const char* other) {
449	assign(s, other, strlen(other));
450	}
451
452	void ?=?(string_res &s, char other) {
453	assign(s, &other, 1);
454	}
455
456	// Copy assignment operator
457	void ?=?(string_res & this, const string_res & rhs) with( this ) {
458	assign_(this, rhs.Handle.s, rhs.Handle.lnth, rhs);
459	}
460
461	void ?=?(string_res & this, string_res & rhs) with( this ) {
462	const string_res & rhs2 = rhs;
463	this = rhs2;
464	}
465
466
467	// Destructor
468	void ^?{}(string_res &s) with(s) {
469	// much delegated to implied ^VbyteSM
470
471	// sever s from its share-edit peers, if any (four no-ops when already solo)
472	s.shareEditSet_prev->shareEditSet_next = s.shareEditSet_next;
473	s.shareEditSet_next->shareEditSet_prev = s.shareEditSet_prev;
474	// s.shareEditSet_next = &s;
475	// s.shareEditSet_prev = &s;
476
477	if (shareEditSet_owns_ulink && s.shareEditSet_next == &s) { // last one out
478	delete( s.Handle.ulink );
479	}
480	}
481
482
483	// Returns the character at the given index
484	// With unicode support, this may be different from just the byte at the given
485	// offset from the start of the string.
486	char ?[?](const string_res &s, size_t index) with(s) {
487	//TODO: Check if index is valid (no exceptions yet)
488	return Handle.s[index];
489	}
490
491	void assignAt(const string_res &s, size_t index, char val) {
492	string_res editZone = { s, SHARE_EDITS, index, index+1 };
493	assign(editZone, &val, 1);
494	}
495
496
497	///////////////////////////////////////////////////////////////////
498	// Concatenation
499
500	void append(string_res &str1, const char * buffer, size_t bsize) {
501	size_t clnth = size(str1) + bsize;
502	if ( str1.Handle.s + size(str1) == buffer ) { // already juxtapose ?
503	// no-op
504	} else { // must copy some text
505	if ( str1.Handle.s + size(str1) == VbyteAlloc(*str1.Handle.ulink, 0) ) { // str1 at end of string area ?
506	VbyteAlloc( *str1.Handle.ulink, bsize ); // create room for 2nd part at the end of string area
507	} else { // copy the two parts
508	char * str1oldBuf = str1.Handle.s;
509	str1.Handle.s = VbyteAlloc( *str1.Handle.ulink, clnth );
510	ByteCopy( str1.Handle.s, 0, str1.Handle.lnth, str1oldBuf, 0, str1.Handle.lnth);
511	} // if
512	ByteCopy( str1.Handle.s, str1.Handle.lnth, bsize, (char*)buffer, 0, (int)bsize);
513	// VbyteHeap & this, char Dst, int DstStart, int DstLnth, char Src, int SrcStart, int SrcLnth
514	} // if
515	str1.Handle.lnth = clnth;
516	}
517
518	void ?+=?(string_res &str1, const string_res &str2) {
519	append( str1, str2.Handle.s, str2.Handle.lnth );
520	}
521
522	void ?+=?(string_res &s, char other) {
523	append( s, &other, 1 );
524	}
525
526	void ?+=?(string_res &s, const char* other) {
527	append( s, other, strlen(other) );
528	}
529
530
531
532
533	//////////////////////////////////////////////////////////
534	// Comparisons
535
536
537	bool ?==?(const string_res &s1, const string_res &s2) {
538	return ByteCmp( s1.Handle.s, 0, s1.Handle.lnth, s2.Handle.s, 0, s2.Handle.lnth) == 0;
539	}
540
541	bool ?!=?(const string_res &s1, const string_res &s2) {
542	return !(s1 == s2);
543	}
544	bool ?==?(const string_res &s, const char* other) {
545	string_res sother = other;
546	return s == sother;
547	}
548	bool ?!=?(const string_res &s, const char* other) {
549	return !(s == other);
550	}
551
552
553	//////////////////////////////////////////////////////////
554	// Search
555
556	bool contains(const string_res &s, char ch) {
557	for (i; size(s)) {
558	if (s[i] == ch) return true;
559	}
560	return false;
561	}
562
563	int find(const string_res &s, char search) {
564	for (i; size(s)) {
565	if (s[i] == search) return i;
566	}
567	return size(s);
568	}
569
570	/* Remaining implementations essentially ported from Sunjay's work */
571
572	int find(const string_res &s, const string_res &search) {
573	return find(s, search.Handle.s, search.Handle.lnth);
574	}
575
576	int find(const string_res &s, const char* search) {
577	return find(s, search, strlen(search));
578	}
579
580	int find(const string_res &s, const char* search, size_t searchsize) {
581	// FIXME: This is a naive algorithm. We probably want to switch to someting
582	// like Boyer-Moore in the future.
583	// https://en.wikipedia.org/wiki/String_searching_algorithm
584
585	// Always find the empty string
586	if (searchsize == 0) {
587	return 0;
588	}
589
590	for (size_t i = 0; i < s.Handle.lnth; i++) {
591	size_t remaining = s.Handle.lnth - i;
592	// Never going to find the search string if the remaining string is
593	// smaller than search
594	if (remaining < searchsize) {
595	break;
596	}
597
598	bool matched = true;
599	for (size_t j = 0; j < searchsize; j++) {
600	if (search[j] != s.Handle.s[i + j]) {
601	matched = false;
602	break;
603	}
604	}
605	if (matched) {
606	return i;
607	}
608	}
609
610	return s.Handle.lnth;
611	}
612
613	bool includes(const string_res &s, const string_res &search) {
614	return includes(s, search.Handle.s, search.Handle.lnth);
615	}
616
617	bool includes(const string_res &s, const char* search) {
618	return includes(s, search, strlen(search));
619	}
620
621	bool includes(const string_res &s, const char* search, size_t searchsize) {
622	return find(s, search, searchsize) < s.Handle.lnth;
623	}
624
625	bool startsWith(const string_res &s, const string_res &prefix) {
626	return startsWith(s, prefix.Handle.s, prefix.Handle.lnth);
627	}
628
629	bool startsWith(const string_res &s, const char* prefix) {
630	return startsWith(s, prefix, strlen(prefix));
631	}
632
633	bool startsWith(const string_res &s, const char* prefix, size_t prefixsize) {
634	if (s.Handle.lnth < prefixsize) {
635	return false;
636	}
637	return memcmp(s.Handle.s, prefix, prefixsize) == 0;
638	}
639
640	bool endsWith(const string_res &s, const string_res &suffix) {
641	return endsWith(s, suffix.Handle.s, suffix.Handle.lnth);
642	}
643
644	bool endsWith(const string_res &s, const char* suffix) {
645	return endsWith(s, suffix, strlen(suffix));
646	}
647
648	bool endsWith(const string_res &s, const char* suffix, size_t suffixsize) {
649	if (s.Handle.lnth < suffixsize) {
650	return false;
651	}
652	// Amount to offset the bytes pointer so that we are comparing the end of s
653	// to suffix. s.bytes + offset should be the first byte to compare against suffix
654	size_t offset = s.Handle.lnth - suffixsize;
655	return memcmp(s.Handle.s + offset, suffix, suffixsize) == 0;
656	}
657
658	/* Back to Mike's work */
659
660
661	///////////////////////////////////////////////////////////////////////////
662	// charclass, include, exclude
663
664	void ?{}( charclass_res & this, const string_res & chars) {
665	(this){ chars.Handle.s, chars.Handle.lnth };
666	}
667
668	void ?{}( charclass_res & this, const char * chars ) {
669	(this){ chars, strlen(chars) };
670	}
671
672	void ?{}( charclass_res & this, const char * chars, size_t charssize ) {
673	(this.chars){ chars, charssize };
674	// now sort it ?
675	}
676
677	void ^?{}( charclass_res & this ) {
678	^(this.chars){};
679	}
680
681	static bool test( const charclass_res & mask, char c ) {
682	// instead, use sorted char list?
683	return contains( mask.chars, c );
684	}
685
686	int exclude(const string_res &s, const charclass_res &mask) {
687	for (int i = 0; i < size(s); i++) {
688	if ( test(mask, s[i]) ) return i;
689	}
690	return size(s);
691	}
692
693	int include(const string_res &s, const charclass_res &mask) {
694	for (int i = 0; i < size(s); i++) {
695	if ( ! test(mask, s[i]) ) return i;
696	}
697	return size(s);
698	}
699
700	//######################### VbyteHeap "implementation" #########################
701
702
703	// Add a new HandleNode node n after the current HandleNode node.
704
705	static inline void AddThisAfter( HandleNode & this, HandleNode & n ) with(this) {
706	#ifdef VbyteDebug
707	serr \| "enter:AddThisAfter, this:" \| &this \| " n:" \| &n;
708	#endif // VbyteDebug
709	verify( n.ulink != 0p );
710	verify( this.ulink == n.ulink );
711	flink = n.flink;
712	blink = &n;
713	n.flink->blink = &this;
714	n.flink = &this;
715	#ifdef VbyteDebug
716	{
717	serr \| "HandleList:";
718	serr \| nlOff;
719	for ( HandleNode *ni = HeaderPtr->flink; ni != HeaderPtr; ni = ni->flink ) {
720	serr \| "\tnode:" \| ni \| " lnth:" \| ni->lnth \| " s:" \| (void *)ni->s \| ",\"";
721	for ( int i = 0; i < ni->lnth; i += 1 ) {
722	serr \| ni->s[i];
723	} // for
724	serr \| "\" flink:" \| ni->flink \| " blink:" \| ni->blink \| nl;
725	} // for
726	serr \| nlOn;
727	}
728	serr \| "exit:AddThisAfter";
729	#endif // VbyteDebug
730	} // AddThisAfter
731
732
733	// Delete the current HandleNode node.
734
735	static inline void DeleteNode( HandleNode & this ) with(this) {
736	#ifdef VbyteDebug
737	serr \| "enter:DeleteNode, this:" \| &this;
738	#endif // VbyteDebug
739	flink->blink = blink;
740	blink->flink = flink;
741	#ifdef VbyteDebug
742	serr \| "exit:DeleteNode";
743	#endif // VbyteDebug
744	} // DeleteNode
745
746
747
748	// Allocates specified storage for a string from byte-string area. If not enough space remains to perform the
749	// allocation, the garbage collection routine is called and a second attempt is made to allocate the space. If the
750	// second attempt fails, a further attempt is made to create a new, larger byte-string area.
751
752	static inline char * VbyteAlloc( VbyteHeap & this, int size ) with(this) {
753	#ifdef VbyteDebug
754	serr \| "enter:VbyteAlloc, size:" \| size;
755	#endif // VbyteDebug
756	uintptr_t NoBytes;
757	char *r;
758
759	NoBytes = ( uintptr_t )EndVbyte + size;
760	if ( NoBytes > ( uintptr_t )ExtVbyte ) { // enough room for new byte-string ?
761	garbage( this ); // firer up the garbage collector
762	NoBytes = ( uintptr_t )EndVbyte + size; // try again
763	if ( NoBytes > ( uintptr_t )ExtVbyte ) { // enough room for new byte-string ?
764	// unimplemented feature - assert, not verify
765	assert( 0 && "need to implement actual growth" );
766	// extend( size ); // extend the byte-string area
767	} // if
768	} // if
769	r = EndVbyte;
770	EndVbyte += size;
771	#ifdef VbyteDebug
772	serr \| "exit:VbyteAlloc, r:" \| (void )r \| " EndVbyte:" \| (void )EndVbyte \| " ExtVbyte:" \| ExtVbyte;
773	#endif // VbyteDebug
774	return r;
775	} // VbyteAlloc
776
777
778	// Move an existing HandleNode node h somewhere after the current HandleNode node so that it is in ascending order by
779	// the address in the byte string area.
780
781	static inline void MoveThisAfter( HandleNode & this, const HandleNode & h ) with(this) {
782	#ifdef VbyteDebug
783	serr \| "enter:MoveThisAfter, this:" \| & this \| " h:" \| & h;
784	#endif // VbyteDebug
785	verify( h.ulink != 0p );
786	verify( this.ulink == h.ulink );
787	if ( s < h.s ) { // check argument values
788	// serr \| "VbyteSM: Error - Cannot move byte string starting at:" \| s \| " after byte string starting at:"
789	// \| ( h->s ) \| " and keep handles in ascending order";
790	// exit(-1 );
791	verify( 0 && "VbyteSM: Error - Cannot move byte strings as requested and keep handles in ascending order");
792	} // if
793
794	HandleNode *i;
795	for ( i = h.flink; i->s != 0 && s > ( i->s ); i = i->flink ); // find the position for this node after h
796	if ( & this != i->blink ) {
797	DeleteNode( this );
798	AddThisAfter( this, *i->blink );
799	} // if
800	#ifdef VbyteDebug
801	{
802	serr \| "HandleList:";
803	serr \| nlOff;
804	for ( HandleNode *n = HeaderPtr->flink; n != HeaderPtr; n = n->flink ) {
805	serr \| "\tnode:" \| n \| " lnth:" \| n->lnth \| " s:" \| (void *)n->s \| ",\"";
806	for ( int i = 0; i < n->lnth; i += 1 ) {
807	serr \| n->s[i];
808	} // for
809	serr \| "\" flink:" \| n->flink \| " blink:" \| n->blink \| nl;
810	} // for
811	serr \| nlOn;
812	}
813	serr \| "exit:MoveThisAfter";
814	#endif // VbyteDebug
815	} // MoveThisAfter
816
817
818
819
820
821	//######################### VbyteHeap #########################
822
823	// Move characters from one location in the byte-string area to another. The routine handles the following situations:
824	//
825	// if the \|Src\| > \|Dst\| => truncate
826	// if the \|Dst\| > \|Src\| => pad Dst with blanks
827
828	void ByteCopy( char Dst, int DstStart, int DstLnth, char Src, int SrcStart, int SrcLnth ) {
829	for ( int i = 0; i < DstLnth; i += 1 ) {
830	if ( i == SrcLnth ) { // \|Dst\| > \|Src\|
831	for ( ; i < DstLnth; i += 1 ) { // pad Dst with blanks
832	Dst[DstStart + i] = ' ';
833	} // for
834	break;
835	} // exit
836	Dst[DstStart + i] = Src[SrcStart + i];
837	} // for
838	} // ByteCopy
839
840	// Compare two byte strings in the byte-string area. The routine returns the following values:
841	//
842	// 1 => Src1-byte-string > Src2-byte-string
843	// 0 => Src1-byte-string = Src2-byte-string
844	// -1 => Src1-byte-string < Src2-byte-string
845
846	int ByteCmp( char Src1, int Src1Start, int Src1Lnth, char Src2, int Src2Start, int Src2Lnth ) {
847	#ifdef VbyteDebug
848	serr \| "enter:ByteCmp, Src1Start:" \| Src1Start \| " Src1Lnth:" \| Src1Lnth \| " Src2Start:" \| Src2Start \| " Src2Lnth:" \| Src2Lnth;
849	#endif // VbyteDebug
850	int cmp;
851
852	CharZip: for ( int i = 0; ; i += 1 ) {
853	if ( i == Src2Lnth - 1 ) {
854	for ( ; ; i += 1 ) {
855	if ( i == Src1Lnth - 1 ) {
856	cmp = 0;
857	break CharZip;
858	} // exit
859	if ( Src1[Src1Start + i] != ' ') {
860	// SUSPECTED BUG: this could be be why Peter got the bug report about == " " (why is this case here at all?)
861	cmp = 1;
862	break CharZip;
863	} // exit
864	} // for
865	} // exit
866	if ( i == Src1Lnth - 1 ) {
867	for ( ; ; i += 1 ) {
868	if ( i == Src2Lnth - 1 ) {
869	cmp = 0;
870	break CharZip;
871	} // exit
872	if ( Src2[Src2Start + i] != ' ') {
873	cmp = -1;
874	break CharZip;
875	} // exit
876	} // for
877	} // exit
878	if ( Src2[Src2Start + i] != Src1[Src1Start+ i]) {
879	cmp = Src1[Src1Start + i] > Src2[Src2Start + i] ? 1 : -1;
880	break CharZip;
881	} // exit
882	} // for
883	#ifdef VbyteDebug
884	serr \| "exit:ByteCmp, cmp:" \| cmp;
885	#endif // VbyteDebug
886	return cmp;
887	} // ByteCmp
888
889
890	// The compaction moves all of the byte strings currently in use to the beginning of the byte-string area and modifies
891	// the handles to reflect the new positions of the byte strings. Compaction assumes that the handle list is in ascending
892	// order by pointers into the byte-string area. The strings associated with substrings do not have to be moved because
893	// the containing string has been moved. Hence, they only require that their string pointers be adjusted.
894
895	void compaction(VbyteHeap & this) with(this) {
896	HandleNode *h;
897	char obase, nbase, *limit;
898
899	NoOfCompactions += 1;
900	EndVbyte = StartVbyte;
901	h = Header.flink; // ignore header node
902	for (;;) {
903	ByteCopy( EndVbyte, 0, h->lnth, h->s, 0, h->lnth );
904	obase = h->s;
905	h->s = EndVbyte;
906	nbase = h->s;
907	EndVbyte += h->lnth;
908	limit = obase + h->lnth;
909	h = h->flink;
910
911	// check if any substrings are allocated within a string
912
913	for (;;) {
914	if ( h == &Header ) break; // end of header list ?
915	if ( h->s >= limit ) break; // outside of current string ?
916	h->s = nbase + (( uintptr_t )h->s - ( uintptr_t )obase );
917	h = h->flink;
918	} // for
919	if ( h == &Header ) break; // end of header list ?
920	} // for
921	} // compaction
922
923
924	// Garbage determines the amount of free space left in the heap and then reduces, leave the same, or extends the size of
925	// the heap. The heap is then compacted in the existing heap or into the newly allocated heap.
926
927	void garbage(VbyteHeap & this ) with(this) {
928	#ifdef VbyteDebug
929	serr \| "enter:garbage";
930	{
931	serr \| "HandleList:";
932	for ( HandleNode *n = Header.flink; n != &Header; n = n->flink ) {
933	serr \| nlOff;
934	serr \| "\tnode:" \| n \| " lnth:" \| n->lnth \| " s:" \| (void *)n->s \| ",\"";
935	for ( int i = 0; i < n->lnth; i += 1 ) {
936	serr \| n->s[i];
937	} // for
938	serr \| nlOn;
939	serr \| "\" flink:" \| n->flink \| " blink:" \| n->blink;
940	} // for
941	}
942	#endif // VbyteDebug
943	int AmountUsed, AmountFree;
944
945	AmountUsed = 0;
946	for ( HandleNode *i = Header.flink; i != &Header; i = i->flink ) { // calculate amount of byte area used
947	AmountUsed += i->lnth;
948	} // for
949	AmountFree = ( uintptr_t )ExtVbyte - ( uintptr_t )StartVbyte - AmountUsed;
950
951	if ( AmountFree < ( int )( CurrSize * 0.1 )) { // free space less than 10% ?
952
953	// unimplemented feature - assert, not verify
954	assert( 0 && "need to implement actual growth" );
955	// extend( CurrSize ); // extend the heap
956
957	// Peter says, "This needs work before it should be used."
958	// } else if ( AmountFree > CurrSize / 2 ) { // free space greater than 3 times the initial allocation ?
959	// reduce(( AmountFree / CurrSize - 3 ) * CurrSize ); // reduce the memory
960
961	} // if
962	compaction(this); // compact the byte area, in the same or new heap area
963	#ifdef VbyteDebug
964	{
965	serr \| "HandleList:";
966	for ( HandleNode *n = Header.flink; n != &Header; n = n->flink ) {
967	serr \| nlOff;
968	serr \| "\tnode:" \| n \| " lnth:" \| n->lnth \| " s:" \| (void *)n->s \| ",\"";
969	for ( int i = 0; i < n->lnth; i += 1 ) {
970	serr \| n->s[i];
971	} // for
972	serr \| nlOn;
973	serr \| "\" flink:" \| n->flink \| " blink:" \| n->blink;
974	} // for
975	}
976	serr \| "exit:garbage";
977	#endif // VbyteDebug
978	} // garbage
979
980	#undef VbyteDebug
981
982	//WIP
983	#if 0
984
985
986	// Extend the size of the byte-string area by creating a new area and copying the old area into it. The old byte-string
987	// area is deleted.
988
989	void VbyteHeap::extend( int size ) {
990	#ifdef VbyteDebug
991	serr \| "enter:extend, size:" \| size;
992	#endif // VbyteDebug
993	char *OldStartVbyte;
994
995	NoOfExtensions += 1;
996	OldStartVbyte = StartVbyte; // save previous byte area
997
998	CurrSize += size > InitSize ? size : InitSize; // minimum extension, initial size
999	StartVbyte = EndVbyte = new char[CurrSize];
1000	ExtVbyte = (void *)( StartVbyte + CurrSize );
1001	compaction(); // copy from old heap to new & adjust pointers to new heap
1002	delete OldStartVbyte; // release old heap
1003	#ifdef VbyteDebug
1004	serr \| "exit:extend, CurrSize:" \| CurrSize;
1005	#endif // VbyteDebug
1006	} // extend
1007
1008
1009	// Extend the size of the byte-string area by creating a new area and copying the old area into it. The old byte-string
1010	// area is deleted.
1011
1012	void VbyteHeap::reduce( int size ) {
1013	#ifdef VbyteDebug
1014	serr \| "enter:reduce, size:" \| size;
1015	#endif // VbyteDebug
1016	char *OldStartVbyte;
1017
1018	NoOfReductions += 1;
1019	OldStartVbyte = StartVbyte; // save previous byte area
1020
1021	CurrSize -= size;
1022	StartVbyte = EndVbyte = new char[CurrSize];
1023	ExtVbyte = (void *)( StartVbyte + CurrSize );
1024	compaction(); // copy from old heap to new & adjust pointers to new heap
1025	delete OldStartVbyte; // release old heap
1026	#ifdef VbyteDebug
1027	serr \| "exit:reduce, CurrSize:" \| CurrSize;
1028	#endif // VbyteDebug
1029	} // reduce
1030
1031
1032	#endif

Note: See TracBrowser for help on using the repository browser.

Download in other formats:

Original Format