X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=dict%2Flookgrep.c;h=733be365a9226ac0ea19da8cccced267b0b68cbf;hb=edf09fc5529eae3e8214a432058b4c07b2b8d2f9;hp=a3cc6b77b706b7bec54deb06a2ff5f66fbd2d7bf;hpb=b7f00ecbe71200814196edcb38c378dcd4acaacc;p=idzebra-moved-to-github.git diff --git a/dict/lookgrep.c b/dict/lookgrep.c index a3cc6b7..733be36 100644 --- a/dict/lookgrep.c +++ b/dict/lookgrep.c @@ -1,10 +1,30 @@ /* - * Copyright (C) 1994, Index Data I/S + * Copyright (C) 1994-1996, Index Data I/S * All rights reserved. * Sebastian Hammer, Adam Dickmeiss * * $Log: lookgrep.c,v $ - * Revision 1.11 1995-12-06 14:43:02 adam + * Revision 1.17 1996-06-04 10:20:06 adam + * Added support for character mapping. + * + * Revision 1.16 1996/05/24 14:46:04 adam + * Added dict_grep_cmap function to define user-mapping in grep lookups. + * + * Revision 1.15 1996/03/20 09:35:18 adam + * Function dict_lookup_grep got extra parameter, init_pos, which marks + * from which position in pattern approximate pattern matching should occur. + * + * Revision 1.14 1996/02/02 13:43:51 adam + * The public functions simply use char instead of Dict_char to represent + * search strings. Dict_char is used internally only. + * + * Revision 1.13 1996/01/08 09:09:30 adam + * Function dfa_parse got 'const' string argument. + * + * Revision 1.12 1995/12/11 09:04:48 adam + * Bug fix: the lookup/scan/lookgrep didn't handle empty dictionary. + * + * Revision 1.11 1995/12/06 14:43:02 adam * New function: dict_delete. * * Revision 1.10 1995/11/16 17:00:44 adam @@ -222,7 +242,8 @@ static void or (MatchContext *mc, MatchWord *Rdst, } static INLINE int move (MatchContext *mc, MatchWord *Rj1, MatchWord *Rj, - Dict_char ch, struct DFA *dfa, MatchWord *Rtmp) + Dict_char ch, struct DFA *dfa, MatchWord *Rtmp, + int range) { int d; MatchWord *Rtmp_2 = Rtmp + mc->n; @@ -251,9 +272,9 @@ static INLINE int move (MatchContext *mc, MatchWord *Rj1, MatchWord *Rj, static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, MatchWord *Rj, int pos, void *client, - int (*userfunc)(Dict_char *, const char *, void *), + int (*userfunc)(char *, const char *, void *), Dict_char *prefix, struct DFA *dfa, - int *max_pos) + int *max_pos, int init_pos) { int lo, hi, d; void *p; @@ -281,6 +302,7 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, MatchWord *Rj0 = Rj + j *mc->fact; MatchWord *Rj1 = Rj + (j+1)*mc->fact; MatchWord *Rj_tmp = Rj + (j+2)*mc->fact; + int range; memcpy (&ch, info+j*sizeof(Dict_char), sizeof(Dict_char)); prefix[pos+j] = ch; @@ -289,20 +311,24 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, if (ch == DICT_EOS) { if (was_match) - if ((*userfunc)(prefix, info+(j+1)*sizeof(Dict_char), - client)) + if ((*userfunc)((char*) prefix, + info+(j+1)*sizeof(Dict_char), client)) return 1; break; } - move (mc, Rj1, Rj0, ch, dfa, Rj_tmp); + if (pos+j >= init_pos) + range = mc->range; + else + range = 0; + move (mc, Rj1, Rj0, ch, dfa, Rj_tmp, range); for (d = mc->n; --d >= 0; ) - if (Rj1[mc->range*mc->n + d]) + if (Rj1[range*mc->n + d]) break; if (d < 0) break; was_match = 0; for (d = mc->n; --d >= 0; ) - if (Rj1[mc->range*mc->n + d] & mc->match_mask[d]) + if (Rj1[range*mc->n + d] & mc->match_mask[d]) { was_match = 1; break; @@ -314,6 +340,7 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, MatchWord *Rj1 = Rj+ mc->fact; MatchWord *Rj_tmp = Rj+2*mc->fact; Dict_char ch; + int range; /* Dict_ptr subptr */ /* Dict_char sub char */ @@ -325,9 +352,13 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, if (pos > *max_pos) *max_pos = pos; - move (mc, Rj1, Rj, ch, dfa, Rj_tmp); + if (pos >= init_pos) + range = mc->range; + else + range = 0; + move (mc, Rj1, Rj, ch, dfa, Rj_tmp, range); for (d = mc->n; --d >= 0; ) - if (Rj1[mc->range*mc->n + d]) + if (Rj1[range*mc->n + d]) break; if (d >= 0) { @@ -335,10 +366,11 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, if (info[sizeof(Dict_ptr)+sizeof(Dict_char)]) { for (d = mc->n; --d >= 0; ) - if (Rj1[mc->range*mc->n + d] & mc->match_mask[d]) + if (Rj1[range*mc->n + d] & mc->match_mask[d]) { prefix[pos+1] = DICT_EOS; - if ((*userfunc)(prefix, info+sizeof(Dict_ptr)+ + if ((*userfunc)((char*) prefix, + info+sizeof(Dict_ptr)+ sizeof(Dict_char), client)) return 1; break; @@ -348,7 +380,8 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, if (subptr) { if (dict_grep (dict, subptr, mc, Rj1, pos+1, - client, userfunc, prefix, dfa, max_pos)) + client, userfunc, prefix, dfa, max_pos, + init_pos)) return 1; dict_bf_readp (dict->dbf, ptr, &p); indxp = (short*) ((char*) p+DICT_pagesize(dict) @@ -361,19 +394,22 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, return 0; } -int dict_lookup_grep (Dict dict, Dict_char *pattern, int range, void *client, - int *max_pos, - int (*userfunc)(Dict_char *name, const char *info, +int dict_lookup_grep (Dict dict, const char *pattern, int range, void *client, + int *max_pos, int init_pos, + int (*userfunc)(char *name, const char *info, void *client)) { MatchWord *Rj; Dict_char prefix[MAX_LENGTH+1]; - char *this_pattern = pattern; + const char *this_pattern = pattern; MatchContext *mc; struct DFA *dfa = dfa_init(); int i, d; logf (LOG_DEBUG, "dict_lookup_grep '%s' range=%d", pattern, range); + + dfa_set_cmap (dfa, dict->grep_cmap); + i = dfa_parse (dfa, &this_pattern); if (i || *this_pattern) { @@ -403,11 +439,19 @@ int dict_lookup_grep (Dict dict, Dict_char *pattern, int range, void *client, } } *max_pos = 0; - i = dict_grep (dict, 1, mc, Rj, 0, client, userfunc, prefix, dfa, - max_pos); + if (dict->head.last > 1) + i = dict_grep (dict, 1, mc, Rj, 0, client, userfunc, prefix, + dfa, max_pos, init_pos); + else + i = 0; logf (LOG_DEBUG, "max_pos = %d", *max_pos); dfa_delete (&dfa); xfree (Rj); rm_MatchContext (&mc); return i; } + +void dict_grep_cmap (Dict dict, char **(*cmap)(const char **from, int len)) +{ + dict->grep_cmap = cmap; +}