17 ** language. The code for the "sqlite3" command-line shell is also in a
21 ** 2019.09.02-Complete codec logic for encryption and decryption.
39 ** ("International Components for Unicode", an open-source library
43 **   * An implementation of the SQL regexp() function (and hence REGEXP
52 **     provide case-independent matching.
56 #include <string.h>
87 # define SQLITE_EXTENSION_INIT1     /*no-op*/
89 # define SQLITE_EXTENSION_INIT3     /*no-op*/
131   sqlite3_result_error(pCtx, zBuf, -1);  in icuFunctionError()
153 ** a multi-byte UTF8 character. It is copied here from SQLite source
170     c = icuUtf8Trans1[c-0xc0];                             \
184 ** Compare two UTF-8 strings for equality where the first string is
190   const uint8_t *zString,    /* The UTF-8 string to compare against */  in icuLikeCompare()
191   const UChar32 uEsc         /* The escape character */  in icuLikeCompare()
207     **     1. uPattern is an unescaped match-all character "%",  in icuLikeCompare()
208     **     2. uPattern is an unescaped match-one character "_",  in icuLikeCompare()
209     **     3. uPattern is an unescaped escape character, or  in icuLikeCompare()
218       ** test string.  in icuLikeCompare()
265 ** the build-in LIKE operator.  The first argument to the function is the
266 ** pattern and the second argument is the string.  So, the SQL statements:
270 ** is implemented as like(B, A). If there is an escape character E,
272 **       A LIKE B ESCAPE E
289     sqlite3_result_error(context, "LIKE or GLOB pattern too complex", -1);  in icuLikeFunc()
295     /* The escape character string must consist of a single UTF-8 character.  in icuLikeFunc()
305           "ESCAPE expression must be a single character", -1);  in icuLikeFunc()
316 ** Function to delete compiled regexp objects. Registered as
325 ** Implementation of SQLite REGEXP operator. This scalar function takes
327 ** the second is a string to match against that pattern. If either
329 ** is 1 if the string matches the pattern, or 0 otherwise.
331 ** SQLite maps the regexp() function to the regexp() operator such
334 **     zString REGEXP zPattern
335 **     regexp(zPattern, zString)
337 ** Uses the following ICU regexp APIs:
351   /* If the left hand side of the regexp operator is NULL,  in icuRegexpFunc()
364     pExpr = uregex_open(zPattern, -1, 0, 0, &status);  in icuRegexpFunc()
377   uregex_setText(pExpr, zString, -1, &status);  in icuRegexpFunc()
402 ** Implementations of scalar functions for case mapping - upper() and
403 ** lower(). Function upper() converts its input to upper-case (ABC).
404 ** Function lower() converts to lower-case (abc).
413 **     upper('ABC') -> 'abc'
414 **     lower('abc') -> 'ABC'
418 ** of the locale to use. Passing an empty string ("") or SQL NULL value
422 **     lower('I', 'en_us') -> 'i'
423 **     lower('I', 'tr_tr') -> '\u131' (small dotless i)
425 ** http://www.icu-project.org/userguide/posix.html#case_mappings
428   const UChar *zInput;            /* Pointer to input string */  in icuCaseFunc16()
430   int nInput;                     /* Size of utf-16 input string in bytes */  in icuCaseFunc16()
507     case UCOL_LESS:    return -1;  in icuCollationColl()
522 **     SELECT icu_load_collation(<locale>, <collation-name>);
524 ** Where <locale> is a string containing an ICU locale identifier (i.e.
525 ** "en_AU", "tr_TR" etc.) and <collation-name> is the name of the
535   const char *zLocale;      /* Locale identifier - (eg. "jp_JP") */  in icuLoadCollation()
561     sqlite3_result_error(p, "Error registering collation function", -1);  in icuLoadCollation()
579     {"regexp", 2, SQLITE_ANY|SQLITEICU_EXTRAFLAGS,         0, icuRegexpFunc},  in sqlite3IcuInit()
602         db, p->zName, p->nArg, p->enc,  in sqlite3IcuInit()
603         p->iContext ? (void*)db : (void*)0,  in sqlite3IcuInit()
604         p->xFunc, 0, 0  in sqlite3IcuInit()
647 /* #include <string.h> */
666   UBreakIterator *pIter;      /* ICU break-iterator object */
668   UChar *aChar;               /* Copy of input using utf-16 encoding */
669   int *aOffset;               /* Offsets of each character in utf-8 input */
698     p->zLocale = (char *)&p[1];  in icuCreate()
699     memcpy(p->zLocale, argv[0], n);  in icuCreate()
717 ** Prepare to begin tokenizing a particular string.  The input
718 ** string to be tokenized is pInput[0..nBytes-1].  A cursor
719 ** used to incrementally tokenize this string is returned in
724   const char *zInput,                    /* Input string */  in icuOpen()
757   pCsr->aChar = (UChar *)&pCsr[1];  in icuOpen()
758   pCsr->aOffset = (int *)&pCsr->aChar[(nChar+3)&~3];  in icuOpen()
760   pCsr->aOffset[iOut] = iInput;  in icuOpen()
765     U16_APPEND(pCsr->aChar, iOut, nChar, c, isError);  in icuOpen()
770     pCsr->aOffset[iOut] = iInput;  in icuOpen()
779   pCsr->pIter = ubrk_open(UBRK_WORD, p->zLocale, pCsr->aChar, iOut, &status);  in icuOpen()
784   pCsr->nChar = iOut;  in icuOpen()
786   ubrk_first(pCsr->pIter);  in icuOpen()
796   ubrk_close(pCsr->pIter);  in icuClose()
797   sqlite3_free(pCsr->zBuffer);  in icuClose()
822     iStart = ubrk_current(pCsr->pIter);  in icuNext()
823     iEnd = ubrk_next(pCsr->pIter);  in icuNext()
830       U16_NEXT(pCsr->aChar, iWhite, pCsr->nChar, c);  in icuNext()
843       char *zNew = sqlite3_realloc(pCsr->zBuffer, nByte);  in icuNext()
847       pCsr->zBuffer = zNew;  in icuNext()
848       pCsr->nBuffer = nByte;  in icuNext()
852         pCsr->zBuffer, pCsr->nBuffer, &nByte,    /* Output vars */  in icuNext()
853         &pCsr->aChar[iStart], iEnd-iStart,       /* Input vars */  in icuNext()
856   } while( nByte>pCsr->nBuffer );  in icuNext()
858   *ppToken = pCsr->zBuffer;  in icuNext()
860   *piStartOffset = pCsr->aOffset[iStart];  in icuNext()
861   *piEndOffset = pCsr->aOffset[iEnd];  in icuNext()
862   *piPosition = pCsr->iToken++;  in icuNext()