Annotation of parser3/src/main/pa_common.C, revision 1.80
1.15 paf 1: /** @file
1.16 paf 2: Parser: commonly functions.
3:
1.8 paf 4: Copyright(c) 2001 ArtLebedev Group(http://www.artlebedev.com)
1.68 parser 5: Author: Alexander Petrosyan <paf@design.ru>(http://design.ru/paf)
1.16 paf 6:
1.80 ! paf 7: $Id: pa_common.C,v 1.79 2001/10/30 16:30:08 paf Exp $
1.1 paf 8: */
9:
10: #include "pa_common.h"
1.2 paf 11: #include "pa_types.h"
1.4 paf 12: #include "pa_exception.h"
1.14 paf 13: #include "pa_pool.h"
14: #include "pa_globals.h"
15: #include "pa_value.h"
16: #include "pa_hash.h"
17: #include "pa_string.h"
1.1 paf 18:
1.80 ! paf 19: #ifdef WIN32
1.47 paf 20: # ifndef _O_TEXT
21: # define _O_TEXT 0
22: # endif
23: # ifndef _O_BINARY
24: # define _O_BINARY 0
25: # endif
26: #endif
27:
1.80 ! paf 28: #ifndef O_TRUNC
! 29: # define O_TRUNC 0
! 30: #endif
! 31:
1.22 paf 32: #if _MSC_VER
1.1 paf 33:
34: int __vsnprintf(char *b, size_t s, const char *f, va_list l) {
35: int r=_vsnprintf(b, --s, f, l);
36: b[s]=0;
37: return r;
38: }
39: int __snprintf(char *b, size_t s, const char *f, ...) {
40: va_list l;
41: va_start(l, f);
42: int r=__vsnprintf(b, s, f, l);
43: va_end(l);
44: return r;
45: }
46:
47: #endif
1.2 paf 48:
1.79 paf 49: void fix_line_breaks(char *cstr, size_t& size) {
50: char *dest=cstr;
1.72 parser 51: // fix DOS: \r\n -> \n
52: // fix Macintosh: \r -> \n
1.79 paf 53: char *bol=cstr;
1.72 parser 54: while(char *eol=strchr(bol, '\r')) {
55: size_t len=eol-bol;
56: if(dest!=bol)
57: memcpy(dest, bol, len);
58: dest+=len;
59: *dest++='\n';
60:
61: if(eol[1]=='\n') { // \r,\n = DOS
62: bol=eol+2;
63: size--;
64: } else // \r,not \n = Macintosh
65: bol=eol+1;
66: }
67: // last piece without \r, including terminating 0
68: if(dest!=bol)
69: strcpy(dest, bol);
70: }
1.18 paf 71:
1.28 paf 72: char *file_read_text(Pool& pool, const String& file_spec, bool fail_on_read_problem) {
1.72 parser 73: void *result; size_t size;
74: return file_read(pool, file_spec, result, size, true, fail_on_read_problem)?(char *)result:0;
1.34 paf 75: }
76: bool file_read(Pool& pool, const String& file_spec,
1.44 paf 77: void*& data, size_t& read_size, bool as_text,
78: bool fail_on_read_problem,
79: size_t offset, size_t limit) {
1.64 parser 80: const char *fname=file_spec.cstr(String::UL_FILE_SPEC);
1.62 parser 81: //printf("file_read(%s)\n", fname);
1.33 paf 82: int f;
1.2 paf 83: struct stat finfo;
1.33 paf 84:
85: // first open, next stat:
1.45 paf 86: // directory update of NTFS hard links performed on open.
1.33 paf 87: // ex:
88: // a.html:^test[] and b.html hardlink to a.html
89: // user inserts ! before ^test in a.html
90: // directory entry of b.html in NTFS not updated at once,
1.35 paf 91: // they delay update till open, so we would receive "!^test[" string
92: // if would do stat, next open.
1.34 paf 93: if(
1.72 parser 94: (f=open(fname, O_RDONLY|(as_text?_O_BINARY/*_O_TEXT*/:_O_BINARY)))>=0 &&
1.34 paf 95: stat(fname, &finfo)==0) {
1.2 paf 96: /*if(exclusive)
97: flock(f, LOCK_EX);*/
1.44 paf 98: size_t max_size=limit?min(offset+limit, finfo.st_size)-offset:finfo.st_size;
1.47 paf 99: if(!max_size) { // eof
100: data=0;
101: read_size=0;
1.59 parser 102: } else {
1.78 paf 103: data=pool.malloc(max_size+(as_text?1:0), 3);
1.59 parser 104: if(offset)
105: lseek(f, offset, SEEK_SET);
106: read_size=read(f, data, max_size);
1.47 paf 107: }
1.2 paf 108: /*if(exclusive)
109: flock(f, LOCK_UN);*/
110: close(f);
1.59 parser 111: if(!max_size) // eof
112: return true;
1.32 paf 113:
1.44 paf 114: if(read_size>=0 && read_size<=max_size) {
1.34 paf 115: if(as_text)
1.46 paf 116: ((char*&)data)[read_size]=0;
1.34 paf 117: } else
1.73 parser 118: throw Exception(0, 0,
1.33 paf 119: &file_spec,
1.44 paf 120: "read failed: actually read %d bytes count not in [0..%lu] valid range",
121: read_size, (unsigned long)max_size); //never
1.32 paf 122:
1.72 parser 123: if(as_text)
124: fix_line_breaks((char *)data, read_size);
125: return true;
1.2 paf 126: }
1.4 paf 127: if(fail_on_read_problem)
1.73 parser 128: throw Exception(0, 0,
1.33 paf 129: &file_spec,
1.54 parser 130: "read failed: %s (%d), actual filename '%s'",
131: strerror(errno), errno, fname);
1.34 paf 132: return false;
1.8 paf 133: }
134:
1.63 parser 135: static void create_dir_for_file(const String& file_spec) {
136: size_t pos_after=1;
137: int pos_before;
138: while((pos_before=file_spec.pos("/", 1, pos_after))>=0) {
1.64 parser 139: mkdir(file_spec.mid(0, pos_before).cstr(String::UL_FILE_SPEC), 0775);
1.63 parser 140: pos_after=pos_before+1;
141: }
142: }
143:
1.18 paf 144: void file_write(Pool& pool,
1.28 paf 145: const String& file_spec,
1.34 paf 146: const void *data, size_t size,
1.80 ! paf 147: bool as_text,
! 148: bool do_append/*,
1.20 paf 149: bool exclusive*/) {
1.64 parser 150: const char *fname=file_spec.cstr(String::UL_FILE_SPEC);
1.28 paf 151: int f;
1.80 ! paf 152: if(access(fname, W_OK)!=0) // no
1.63 parser 153: create_dir_for_file(file_spec);
1.50 paf 154:
1.80 ! paf 155: if((f=open(fname,
! 156: O_CREAT|O_RDWR
! 157: |(as_text?_O_TEXT:_O_BINARY)
! 158: |(do_append?O_APPEND:O_TRUNC), 0666))>=0) {
! 159: /*if(exclusive)
! 160: flock(f, LOCK_EX);*/
! 161:
! 162: if(size) write(f, data, size);
! 163: #if O_TRUNC==0
! 164: ftruncate(f, size);
1.18 paf 165: #endif
1.80 ! paf 166: /*if(exclusive)
! 167: flock(f, LOCK_UN);*/
! 168: close(f);
! 169: } else
! 170: throw Exception(0, 0,
! 171: &file_spec,
! 172: "write failed: %s (%d), actual filename '%s'",
! 173: strerror(errno), errno, fname);
1.30 paf 174: }
175:
1.63 parser 176: // throws nothing! [this is required in file_move & file_delete]
1.50 paf 177: static void rmdir(const String& file_spec, size_t pos_after) {
178: int pos_before;
179: if((pos_before=file_spec.pos("/", 1, pos_after))>=0)
180: rmdir(file_spec, pos_before+1);
181:
1.64 parser 182: rmdir(file_spec.mid(0, pos_after-1/* / */).cstr(String::UL_FILE_SPEC));
1.50 paf 183: }
1.30 paf 184: void file_delete(Pool& pool, const String& file_spec) {
1.64 parser 185: const char *fname=file_spec.cstr(String::UL_FILE_SPEC);
1.54 parser 186: if(unlink(fname)!=0)
1.73 parser 187: throw Exception(0, 0,
1.33 paf 188: &file_spec,
1.54 parser 189: "unlink failed: %s (%d), actual filename '%s'",
190: strerror(errno), errno, fname);
1.50 paf 191:
192: rmdir(file_spec, 1);
1.60 parser 193: }
194: void file_move(Pool& pool, const String& old_spec, const String& new_spec) {
1.64 parser 195: const char *old_spec_cstr=old_spec.cstr(String::UL_FILE_SPEC);
196: const char *new_spec_cstr=new_spec.cstr(String::UL_FILE_SPEC);
1.63 parser 197:
198: create_dir_for_file(new_spec);
199:
1.60 parser 200: if(rename(old_spec_cstr, new_spec_cstr)!=0)
1.73 parser 201: throw Exception(0, 0,
1.60 parser 202: &old_spec,
203: "rename failed: %s (%d), actual filename '%s' to '%s'",
204: strerror(errno), errno, old_spec_cstr, new_spec_cstr);
1.63 parser 205:
206: rmdir(old_spec, 1);
1.31 paf 207: }
208:
1.51 paf 209:
210: static bool entry_readable(const String& file_spec, bool need_dir) {
1.64 parser 211: const char *fname=file_spec.cstr(String::UL_FILE_SPEC);
1.51 paf 212: struct stat finfo;
213: if(access(fname, R_OK)==0 && stat(fname, &finfo)==0) {
1.73 parser 214: bool is_dir=finfo.st_mode&S_IFDIR != 0;
1.51 paf 215: return is_dir==need_dir;
216: }
217: return false;
218: }
1.31 paf 219: bool file_readable(const String& file_spec) {
1.51 paf 220: return entry_readable(file_spec, false);
221: }
222: bool dir_readable(const String& file_spec) {
223: return entry_readable(file_spec, true);
1.65 parser 224: }
225: String *file_readable(const String& path, const String& name) {
226: String *result=new(path.pool()) String(path);
227: *result << "/";
228: *result << name;
229: return file_readable(*result)?result:0;
1.43 paf 230: }
231: bool file_executable(const String& file_spec) {
1.64 parser 232: return access(file_spec.cstr(String::UL_FILE_SPEC), X_OK)==0;
1.44 paf 233: }
234:
1.64 parser 235: bool file_stat(const String& file_spec,
1.58 parser 236: size_t& rsize,
237: time_t& ratime,
238: time_t& rmtime,
1.64 parser 239: time_t& rctime,
240: bool fail_on_read_problem) {
1.44 paf 241: Pool& pool=file_spec.pool();
1.64 parser 242: const char *fname=file_spec.cstr(String::UL_FILE_SPEC);
1.44 paf 243: struct stat finfo;
244: if(stat(fname, &finfo)!=0)
1.64 parser 245: if(fail_on_read_problem)
1.73 parser 246: throw Exception(0, 0,
1.67 parser 247: &file_spec,
248: "getting file size failed: %s (%d), real filename '%s'",
249: strerror(errno), errno, fname);
1.64 parser 250: else
251: return false;
1.58 parser 252: rsize=finfo.st_size;
253: ratime=finfo.st_atime;
254: rmtime=finfo.st_mtime;
255: rctime=finfo.st_ctime;
1.64 parser 256: return true;
1.18 paf 257: }
258:
1.8 paf 259: char *getrow(char **row_ref, char delim) {
260: char *result=*row_ref;
261: if(result) {
262: *row_ref=strchr(result, delim);
263: if(*row_ref)
264: *((*row_ref)++)=0;
265: else if(!*result)
266: return 0;
267: }
268: return result;
269: }
270:
1.23 paf 271: char *lsplit(char *string, char delim) {
272: if(string) {
273: char *v=strchr(string, delim);
1.8 paf 274: if(v) {
275: *v=0;
276: return v+1;
277: }
278: }
279: return 0;
280: }
281:
282: char *lsplit(char **string_ref, char delim) {
283: char *result=*string_ref;
284: char *next=lsplit(*string_ref, delim);
285: *string_ref=next;
286: return result;
1.9 paf 287: }
288:
289: char *rsplit(char *string, char delim) {
1.18 paf 290: if(string) {
1.9 paf 291: char *v=strrchr(string, delim);
1.18 paf 292: if(v) {
1.9 paf 293: *v=0;
294: return v+1;
295: }
296: }
297: return NULL;
1.10 paf 298: }
299:
1.37 paf 300: /// @todo less stupid type detection
1.10 paf 301: char *format(Pool& pool, double value, char *fmt) {
1.78 paf 302: char *result=(char *)pool.malloc(MAX_NUMBER, 4);
1.10 paf 303: if(fmt)
304: if(strpbrk(fmt, "diouxX"))
305: if(strpbrk(fmt, "ouxX"))
1.33 paf 306: snprintf(result, MAX_NUMBER, fmt, (uint)value );
1.10 paf 307: else
1.33 paf 308: snprintf(result, MAX_NUMBER, fmt, (int)value );
1.10 paf 309: else
310: snprintf(result, MAX_NUMBER, fmt, value);
311: else
1.33 paf 312: snprintf(result, MAX_NUMBER, "%d", (int)value);
1.10 paf 313:
314: return result;
1.12 paf 315: }
316:
1.36 paf 317: size_t stdout_write(const void *buf, size_t size) {
1.12 paf 318: #ifdef WIN32
319: do{
320: int chunk_written=fwrite(buf, 1, min(8*0x400, size), stdout);
321: if(chunk_written<=0)
322: break;
323: size-=chunk_written;
1.36 paf 324: buf=((const char*)buf)+chunk_written;
1.12 paf 325: } while(size>0);
326:
327: return size;
328: #else
1.13 paf 329: return fwrite(buf, 1, size, stdout);
1.12 paf 330: #endif
1.2 paf 331: }
1.14 paf 332:
1.77 parser 333: char *unescape_chars(Pool& pool, const char *cp, int len) {
1.78 paf 334: char *s=(char *)pool.malloc(len + 1, 5);
1.14 paf 335: enum EscapeState {
1.33 paf 336: EscapeRest,
337: EscapeFirst,
1.14 paf 338: EscapeSecond
339: } escapeState=EscapeRest;
340: int escapedValue=0;
341: int srcPos=0;
342: int dstPos=0;
343: while(srcPos < len) {
344: int ch=cp[srcPos];
345: switch(escapeState) {
346: case EscapeRest:
347: if(ch=='%') {
348: escapeState=EscapeFirst;
349: } else if(ch=='+') {
350: s[dstPos++]=' ';
351: } else {
352: s[dstPos++]=ch;
353: }
354: break;
355: case EscapeFirst:
356: escapedValue=hex_value[ch] << 4;
357: escapeState=EscapeSecond;
358: break;
359: case EscapeSecond:
360: escapedValue +=hex_value[ch];
361: s[dstPos++]=escapedValue;
362: escapeState=EscapeRest;
363: break;
364: }
365: srcPos++;
366: }
367: s[dstPos]=0;
368: return s;
369: }
370:
1.52 paf 371: /// used by attributed_meaning_to_string / append_attribute_subattribute
1.40 paf 372: struct Attributed_meaning_info {
1.52 paf 373: String *header; // header line being constructed
374: String::Untaint_lang lang; // language in which to append to that line
1.40 paf 375: };
1.17 paf 376: static void append_attribute_subattribute(const Hash::Key& akey, Hash::Val *avalue,
1.14 paf 377: void *info) {
378: if(akey==VALUE_NAME)
379: return;
380:
1.40 paf 381: Attributed_meaning_info& ami=*static_cast<Attributed_meaning_info *>(info);
382:
1.14 paf 383: // ...; charset=windows1251
1.48 paf 384: *ami.header << "; ";
1.70 parser 385: ami.header->append(akey, ami.lang);
1.48 paf 386: *ami.header << "=";
1.70 parser 387: ami.header->append(static_cast<Value *>(avalue)->as_string(), ami.lang);
1.14 paf 388: }
1.49 paf 389: const String& attributed_meaning_to_string(Value& meaning,
390: String::Untaint_lang lang) {
1.20 paf 391: String &result=*new(meaning.pool()) String(meaning.pool());
1.75 parser 392: if(Hash *hash=meaning.get_hash(0)) {
1.14 paf 393: // $value(value) $subattribute(subattribute value)
394: if(Value *value=static_cast<Value *>(hash->get(*value_name)))
1.40 paf 395: result.append(value->as_string(), lang, true);
1.14 paf 396:
1.40 paf 397: Attributed_meaning_info attributed_meaning_info={
1.52 paf 398: &result,
399: lang
1.40 paf 400: };
401: hash->for_each(append_attribute_subattribute, &attributed_meaning_info);
1.14 paf 402: } else // result value
1.40 paf 403: result.append(meaning.as_string(), lang, true);
1.14 paf 404:
405: return result;
1.24 paf 406: }
407:
408: #ifdef WIN32
409: void back_slashes_to_slashes(char *s) {
410: if(s)
411: for(; *s; s++)
412: if(*s=='\\')
413: *s='/';
414: }
1.42 paf 415: /*
416: void slashes_to_back_slashes(char *s) {
417: if(s)
418: for(; *s; s++)
419: if(*s=='/')
420: *s='\\';
421: }
422: */
1.24 paf 423: #endif
1.41 paf 424:
425: bool StrEqNc(const char *s1, const char *s2, bool strict) {
426: while(true) {
427: if(!(*s1)) {
428: if(!(*s2))
429: return true;
430: else
431: return !strict;
432: } else if(!(*s2))
433: return !strict;
434: if(isalpha(*s1)) {
435: if(tolower(*s1) !=tolower(*s2))
436: return false;
437: } else if((*s1) !=(*s2))
438: return false;
439: s1++;
440: s2++;
441: }
1.57 parser 442: }
443:
444: static int isLeap(int year) {
445: return !(
446: (year % 4) || ((year % 400) && !(year % 100))
447: );
448: }
449:
450: int getMonthDays(int year, int month) {
451: int monthDays[]={
452: 31,
453: isLeap(year) ? 29 : 28,
454: 31,
455: 30,
456: 31,
457: 30,
458: 31,
459: 31,
460: 30,
461: 31,
462: 30,
463: 31
464: };
465: return monthDays[month];
1.41 paf 466: }
1.69 parser 467:
468: void remove_crlf(char *start, char *end) {
469: for(char *p=start; p<end; p++)
470: switch(*p) {
471: case '\n': *p='|'; break;
472: case '\r': *p=' '; break;
473: }
1.74 parser 474: }
E-mail:
![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to pa_common.C CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [parser3project] / parser3 / src / main