diff options
Diffstat (limited to 'posix/regcomp.c')
| -rw-r--r-- | posix/regcomp.c | 272 |
1 files changed, 77 insertions, 195 deletions
diff --git a/posix/regcomp.c b/posix/regcomp.c index fde262b83c..353689e186 100644 --- a/posix/regcomp.c +++ b/posix/regcomp.c @@ -325,10 +325,8 @@ re_set_fastmap (char *fastmap, int icase, int ch) Compile fastmap for the initial_state INIT_STATE. */ static void -re_compile_fastmap_iter (bufp, init_state, fastmap) - regex_t *bufp; - const re_dfastate_t *init_state; - char *fastmap; +re_compile_fastmap_iter (regex_t *bufp, const re_dfastate_t *init_state, + char *fastmap) { re_dfa_t *dfa = (re_dfa_t *) bufp->buffer; int node_cnt; @@ -735,11 +733,8 @@ libc_freeres_fn (free_mem) SYNTAX indicate regular expression's syntax. */ static reg_errcode_t -re_compile_internal (preg, pattern, length, syntax) - regex_t *preg; - const char * pattern; - size_t length; - reg_syntax_t syntax; +re_compile_internal (regex_t *preg, const char * pattern, size_t length, + reg_syntax_t syntax) { reg_errcode_t err = REG_NOERROR; re_dfa_t *dfa; @@ -837,9 +832,7 @@ re_compile_internal (preg, pattern, length, syntax) as the initial length of some arrays. */ static reg_errcode_t -init_dfa (dfa, pat_len) - re_dfa_t *dfa; - size_t pat_len; +init_dfa (re_dfa_t *dfa, size_t pat_len) { unsigned int table_size; #ifndef _LIBC @@ -936,8 +929,7 @@ init_dfa (dfa, pat_len) character used by some operators like "\<", "\>", etc. */ static void -init_word_char (dfa) - re_dfa_t *dfa; +init_word_char (re_dfa_t *dfa) { int i, j, ch; dfa->word_ops_used = 1; @@ -950,8 +942,7 @@ init_word_char (dfa) /* Free the work area which are only used while compiling. */ static void -free_workarea_compile (preg) - regex_t *preg; +free_workarea_compile (regex_t *preg) { re_dfa_t *dfa = (re_dfa_t *) preg->buffer; bin_tree_storage_t *storage, *next; @@ -970,8 +961,7 @@ free_workarea_compile (preg) /* Create initial states for all contexts. */ static reg_errcode_t -create_initial_state (dfa) - re_dfa_t *dfa; +create_initial_state (re_dfa_t *dfa) { int first, i; reg_errcode_t err; @@ -1053,8 +1043,7 @@ create_initial_state (dfa) DFA nodes where needed. */ static void -optimize_utf8 (dfa) - re_dfa_t *dfa; +optimize_utf8 (re_dfa_t *dfa) { int node, i, mb_chars = 0, has_period = 0; @@ -1122,8 +1111,7 @@ optimize_utf8 (dfa) "eclosure", and "inveclosure". */ static reg_errcode_t -analyze (preg) - regex_t *preg; +analyze (regex_t *preg) { re_dfa_t *dfa = (re_dfa_t *) preg->buffer; reg_errcode_t ret; @@ -1186,10 +1174,8 @@ analyze (preg) implement parse tree visits. Instead, we use parent pointers and some hairy code in these two functions. */ static reg_errcode_t -postorder (root, fn, extra) - bin_tree_t *root; - reg_errcode_t (fn (void *, bin_tree_t *)); - void *extra; +postorder (bin_tree_t *root, reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra) { bin_tree_t *node, *prev; @@ -1220,10 +1206,8 @@ postorder (root, fn, extra) } static reg_errcode_t -preorder (root, fn, extra) - bin_tree_t *root; - reg_errcode_t (fn (void *, bin_tree_t *)); - void *extra; +preorder (bin_tree_t *root, reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra) { bin_tree_t *node; @@ -1255,9 +1239,7 @@ preorder (root, fn, extra) re_search_internal to map the inner one's opr.idx to this one's. Adjust backreferences as well. Requires a preorder visit. */ static reg_errcode_t -optimize_subexps (extra, node) - void *extra; - bin_tree_t *node; +optimize_subexps (void *extra, bin_tree_t *node) { re_dfa_t *dfa = (re_dfa_t *) extra; @@ -1288,9 +1270,7 @@ optimize_subexps (extra, node) /* Lowering pass: Turn each SUBEXP node into the appropriate concatenation of OP_OPEN_SUBEXP, the body of the SUBEXP (if any) and OP_CLOSE_SUBEXP. */ static reg_errcode_t -lower_subexps (extra, node) - void *extra; - bin_tree_t *node; +lower_subexps (void *extra, bin_tree_t *node) { regex_t *preg = (regex_t *) extra; reg_errcode_t err = REG_NOERROR; @@ -1312,10 +1292,7 @@ lower_subexps (extra, node) } static bin_tree_t * -lower_subexp (err, preg, node) - reg_errcode_t *err; - regex_t *preg; - bin_tree_t *node; +lower_subexp (reg_errcode_t *err, regex_t *preg, bin_tree_t *node) { re_dfa_t *dfa = (re_dfa_t *) preg->buffer; bin_tree_t *body = node->left; @@ -1352,9 +1329,7 @@ lower_subexp (err, preg, node) /* Pass 1 in building the NFA: compute FIRST and create unlinked automaton nodes. Requires a postorder visit. */ static reg_errcode_t -calc_first (extra, node) - void *extra; - bin_tree_t *node; +calc_first (void *extra, bin_tree_t *node) { re_dfa_t *dfa = (re_dfa_t *) extra; if (node->token.type == CONCAT) @@ -1374,9 +1349,7 @@ calc_first (extra, node) /* Pass 2: compute NEXT on the tree. Preorder visit. */ static reg_errcode_t -calc_next (extra, node) - void *extra; - bin_tree_t *node; +calc_next (void *extra, bin_tree_t *node) { switch (node->token.type) { @@ -1399,9 +1372,7 @@ calc_next (extra, node) /* Pass 3: link all DFA nodes to their NEXT node (any order will do). */ static reg_errcode_t -link_nfa_nodes (extra, node) - void *extra; - bin_tree_t *node; +link_nfa_nodes (void *extra, bin_tree_t *node) { re_dfa_t *dfa = (re_dfa_t *) extra; int idx = node->node_idx; @@ -1461,11 +1432,8 @@ link_nfa_nodes (extra, node) to their own constraint. */ static reg_errcode_t -duplicate_node_closure (dfa, top_org_node, top_clone_node, root_node, - init_constraint) - re_dfa_t *dfa; - int top_org_node, top_clone_node, root_node; - unsigned int init_constraint; +duplicate_node_closure (re_dfa_t *dfa, int top_org_node, int top_clone_node, + int root_node, unsigned int init_constraint) { int org_node, clone_node, ret; unsigned int constraint = init_constraint; @@ -1575,10 +1543,8 @@ duplicate_node_closure (dfa, top_org_node, top_clone_node, root_node, satisfies the constraint CONSTRAINT. */ static int -search_duplicated_node (dfa, org_node, constraint) - const re_dfa_t *dfa; - int org_node; - unsigned int constraint; +search_duplicated_node (const re_dfa_t *dfa, int org_node, + unsigned int constraint) { int idx; for (idx = dfa->nodes_len - 1; dfa->nodes[idx].duplicated && idx > 0; --idx) @@ -1595,10 +1561,7 @@ search_duplicated_node (dfa, org_node, constraint) available. */ static int -duplicate_node (dfa, org_idx, constraint) - re_dfa_t *dfa; - int org_idx; - unsigned int constraint; +duplicate_node (re_dfa_t *dfa, int org_idx, unsigned int constraint) { int dup_idx = re_dfa_add_node (dfa, dfa->nodes[org_idx]); if (BE (dup_idx != -1, 1)) @@ -1615,8 +1578,7 @@ duplicate_node (dfa, org_idx, constraint) } static reg_errcode_t -calc_inveclosure (dfa) - re_dfa_t *dfa; +calc_inveclosure (re_dfa_t *dfa) { int src, idx, ret; for (idx = 0; idx < dfa->nodes_len; ++idx) @@ -1639,8 +1601,7 @@ calc_inveclosure (dfa) /* Calculate "eclosure" for all the node in DFA. */ static reg_errcode_t -calc_eclosure (dfa) - re_dfa_t *dfa; +calc_eclosure (re_dfa_t *dfa) { int node_idx, incomplete; #ifdef DEBUG @@ -1684,10 +1645,7 @@ calc_eclosure (dfa) /* Calculate epsilon closure of NODE. */ static reg_errcode_t -calc_eclosure_iter (new_set, dfa, node, root) - re_node_set *new_set; - re_dfa_t *dfa; - int node, root; +calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa, int node, int root) { reg_errcode_t err; unsigned int constraint; @@ -1767,10 +1725,7 @@ calc_eclosure_iter (new_set, dfa, node, root) We must not use this function inside bracket expressions. */ static void -fetch_token (result, input, syntax) - re_token_t *result; - re_string_t *input; - reg_syntax_t syntax; +fetch_token (re_token_t *result, re_string_t *input, reg_syntax_t syntax) { re_string_skip_bytes (input, peek_token (result, input, syntax)); } @@ -1779,10 +1734,7 @@ fetch_token (result, input, syntax) We must not use this function inside bracket expressions. */ static int -peek_token (token, input, syntax) - re_token_t *token; - re_string_t *input; - reg_syntax_t syntax; +peek_token (re_token_t *token, re_string_t *input, reg_syntax_t syntax) { unsigned char c; @@ -2020,10 +1972,7 @@ peek_token (token, input, syntax) We must not use this function out of bracket expressions. */ static int -peek_token_bracket (token, input, syntax) - re_token_t *token; - re_string_t *input; - reg_syntax_t syntax; +peek_token_bracket (re_token_t *token, re_string_t *input, reg_syntax_t syntax) { unsigned char c; if (re_string_eoi (input)) @@ -2119,11 +2068,8 @@ peek_token_bracket (token, input, syntax) EOR means end of regular expression. */ static bin_tree_t * -parse (regexp, preg, syntax, err) - re_string_t *regexp; - regex_t *preg; - reg_syntax_t syntax; - reg_errcode_t *err; +parse (re_string_t *regexp, regex_t *preg, reg_syntax_t syntax, + reg_errcode_t *err) { re_dfa_t *dfa = (re_dfa_t *) preg->buffer; bin_tree_t *tree, *eor, *root; @@ -2156,13 +2102,8 @@ parse (regexp, preg, syntax, err) ALT means alternative, which represents the operator `|'. */ static bin_tree_t * -parse_reg_exp (regexp, preg, token, syntax, nest, err) - re_string_t *regexp; - regex_t *preg; - re_token_t *token; - reg_syntax_t syntax; - int nest; - reg_errcode_t *err; +parse_reg_exp (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, int nest, reg_errcode_t *err) { re_dfa_t *dfa = (re_dfa_t *) preg->buffer; bin_tree_t *tree, *branch = NULL; @@ -2202,13 +2143,8 @@ parse_reg_exp (regexp, preg, token, syntax, nest, err) CAT means concatenation. */ static bin_tree_t * -parse_branch (regexp, preg, token, syntax, nest, err) - re_string_t *regexp; - regex_t *preg; - re_token_t *token; - reg_syntax_t syntax; - int nest; - reg_errcode_t *err; +parse_branch (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, int nest, reg_errcode_t *err) { bin_tree_t *tree, *exp; re_dfa_t *dfa = (re_dfa_t *) preg->buffer; @@ -2247,13 +2183,8 @@ parse_branch (regexp, preg, token, syntax, nest, err) */ static bin_tree_t * -parse_expression (regexp, preg, token, syntax, nest, err) - re_string_t *regexp; - regex_t *preg; - re_token_t *token; - reg_syntax_t syntax; - int nest; - reg_errcode_t *err; +parse_expression (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, int nest, reg_errcode_t *err) { re_dfa_t *dfa = (re_dfa_t *) preg->buffer; bin_tree_t *tree; @@ -2468,13 +2399,8 @@ parse_expression (regexp, preg, token, syntax, nest, err) */ static bin_tree_t * -parse_sub_exp (regexp, preg, token, syntax, nest, err) - re_string_t *regexp; - regex_t *preg; - re_token_t *token; - reg_syntax_t syntax; - int nest; - reg_errcode_t *err; +parse_sub_exp (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, int nest, reg_errcode_t *err) { re_dfa_t *dfa = (re_dfa_t *) preg->buffer; bin_tree_t *tree; @@ -2511,13 +2437,8 @@ parse_sub_exp (regexp, preg, token, syntax, nest, err) /* This function parse repetition operators like "*", "+", "{1,3}" etc. */ static bin_tree_t * -parse_dup_op (elem, regexp, dfa, token, syntax, err) - bin_tree_t *elem; - re_string_t *regexp; - re_dfa_t *dfa; - re_token_t *token; - reg_syntax_t syntax; - reg_errcode_t *err; +parse_dup_op (bin_tree_t *elem, re_string_t *regexp, re_dfa_t *dfa, + re_token_t *token, reg_syntax_t syntax, reg_errcode_t *err) { bin_tree_t *tree = NULL, *old_tree = NULL; int i, start, end, start_idx = re_string_cur_idx (regexp); @@ -2657,14 +2578,12 @@ parse_dup_op (elem, regexp, dfa, token, syntax, err) static reg_errcode_t # ifdef RE_ENABLE_I18N -build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem) - re_charset_t *mbcset; - int *range_alloc; +build_range_exp (bitset_t sbcset, re_charset_t *mbcset, int *range_alloc, + bracket_elem_t *start_elem, bracket_elem_t *end_elem) # else /* not RE_ENABLE_I18N */ -build_range_exp (sbcset, start_elem, end_elem) +build_range_exp (bitset_t sbcset, bracket_elem_t *start_elem, + bracket_elem_t *end_elem) # endif /* not RE_ENABLE_I18N */ - bitset_t sbcset; - bracket_elem_t *start_elem, *end_elem; { unsigned int start_ch, end_ch; /* Equivalence Classes and Character Classes can't be a range start/end. */ @@ -2779,14 +2698,11 @@ build_range_exp (sbcset, start_elem, end_elem) static reg_errcode_t # ifdef RE_ENABLE_I18N -build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name) - re_charset_t *mbcset; - int *coll_sym_alloc; +build_collating_symbol (bitset_t sbcset, re_charset_t *mbcset, + int *coll_sym_alloc, const unsigned char *name) # else /* not RE_ENABLE_I18N */ -build_collating_symbol (sbcset, name) +build_collating_symbol (bitset_t sbcset, const unsigned char *name) # endif /* not RE_ENABLE_I18N */ - bitset_t sbcset; - const unsigned char *name; { size_t name_len = strlen ((const char *) name); if (BE (name_len != 1, 0)) @@ -2803,12 +2719,8 @@ build_collating_symbol (sbcset, name) "[[.a-a.]]" etc. */ static bin_tree_t * -parse_bracket_exp (regexp, dfa, token, syntax, err) - re_string_t *regexp; - re_dfa_t *dfa; - re_token_t *token; - reg_syntax_t syntax; - reg_errcode_t *err; +parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa, re_token_t *token, + reg_syntax_t syntax, reg_errcode_t *err) { #ifdef _LIBC const unsigned char *collseqmb; @@ -3359,15 +3271,9 @@ parse_bracket_exp (regexp, dfa, token, syntax, err) /* Parse an element in the bracket expression. */ static reg_errcode_t -parse_bracket_element (elem, regexp, token, token_len, dfa, syntax, - accept_hyphen) - bracket_elem_t *elem; - re_string_t *regexp; - re_token_t *token; - int token_len; - re_dfa_t *dfa; - reg_syntax_t syntax; - int accept_hyphen; +parse_bracket_element (bracket_elem_t *elem, re_string_t *regexp, + re_token_t *token, int token_len, re_dfa_t *dfa, + reg_syntax_t syntax, int accept_hyphen) { #ifdef RE_ENABLE_I18N int cur_char_size; @@ -3405,10 +3311,8 @@ parse_bracket_element (elem, regexp, token, token_len, dfa, syntax, [=<equivalent_class>=]. */ static reg_errcode_t -parse_bracket_symbol (elem, regexp, token) - bracket_elem_t *elem; - re_string_t *regexp; - re_token_t *token; +parse_bracket_symbol (bracket_elem_t *elem, re_string_t *regexp, + re_token_t *token) { unsigned char ch, delim = token->opr.c; int i = 0; @@ -3455,16 +3359,13 @@ parse_bracket_symbol (elem, regexp, token) static reg_errcode_t #ifdef RE_ENABLE_I18N -build_equiv_class (sbcset, mbcset, equiv_class_alloc, name) - re_charset_t *mbcset; - int *equiv_class_alloc; +build_equiv_class (bitset_t sbcset, re_charset_t *mbcset, + int *equiv_class_alloc, const unsigned char *name) #else /* not RE_ENABLE_I18N */ -build_equiv_class (sbcset, name) +build_equiv_class (bitset_t sbcset, const unsigned char *name) #endif /* not RE_ENABLE_I18N */ - bitset_t sbcset; - const unsigned char *name; { -#if defined _LIBC +#ifdef _LIBC uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); if (nrules != 0) { @@ -3550,16 +3451,13 @@ build_equiv_class (sbcset, name) static reg_errcode_t #ifdef RE_ENABLE_I18N -build_charclass (trans, sbcset, mbcset, char_class_alloc, class_name, syntax) - re_charset_t *mbcset; - int *char_class_alloc; +build_charclass (RE_TRANSLATE_TYPE trans, bitset_t sbcset, + re_charset_t *mbcset, int *char_class_alloc, + const unsigned char *class_name, reg_syntax_t syntax) #else /* not RE_ENABLE_I18N */ -build_charclass (trans, sbcset, class_name, syntax) +build_charclass (RE_TRANSLATE_TYPE trans, bitset_t sbcset, + const unsigned char *class_name, reg_syntax_t syntax) #endif /* not RE_ENABLE_I18N */ - RE_TRANSLATE_TYPE trans; - bitset_t sbcset; - const unsigned char *class_name; - reg_syntax_t syntax; { int i; const char *name = (const char *) class_name; @@ -3629,13 +3527,10 @@ build_charclass (trans, sbcset, class_name, syntax) } static bin_tree_t * -build_charclass_op (dfa, trans, class_name, extra, non_match, err) - re_dfa_t *dfa; - RE_TRANSLATE_TYPE trans; - const unsigned char *class_name; - const unsigned char *extra; - int non_match; - reg_errcode_t *err; +build_charclass_op (re_dfa_t *dfa, RE_TRANSLATE_TYPE trans, + const unsigned char *class_name, + const unsigned char *extra, int non_match, + reg_errcode_t *err) { re_bitset_ptr_t sbcset; #ifdef RE_ENABLE_I18N @@ -3749,10 +3644,7 @@ build_charclass_op (dfa, trans, class_name, extra, non_match, err) Return -2, If an error is occured. */ static int -fetch_number (input, token, syntax) - re_string_t *input; - re_token_t *token; - reg_syntax_t syntax; +fetch_number (re_string_t *input, re_token_t *token, reg_syntax_t syntax) { int num = -1; unsigned char c; @@ -3792,11 +3684,8 @@ free_charset (re_charset_t *cset) /* Create a tree node. */ static bin_tree_t * -create_tree (dfa, left, right, type) - re_dfa_t *dfa; - bin_tree_t *left; - bin_tree_t *right; - re_token_type_t type; +create_tree (re_dfa_t *dfa, bin_tree_t *left, bin_tree_t *right, + re_token_type_t type) { re_token_t t; t.type = type; @@ -3804,11 +3693,8 @@ create_tree (dfa, left, right, type) } static bin_tree_t * -create_token_tree (dfa, left, right, token) - re_dfa_t *dfa; - bin_tree_t *left; - bin_tree_t *right; - const re_token_t *token; +create_token_tree (re_dfa_t *dfa, bin_tree_t *left, bin_tree_t *right, + const re_token_t *token) { bin_tree_t *tree; if (BE (dfa->str_tree_storage_idx == BIN_TREE_STORAGE_SIZE, 0)) @@ -3844,9 +3730,7 @@ create_token_tree (dfa, left, right, token) To be called from preorder or postorder. */ static reg_errcode_t -mark_opt_subexp (extra, node) - void *extra; - bin_tree_t *node; +mark_opt_subexp (void *extra, bin_tree_t *node) { int idx = (int) (long) extra; if (node->token.type == SUBEXP && node->token.opr.idx == idx) @@ -3886,9 +3770,7 @@ free_tree (void *extra, bin_tree_t *node) it's easier to duplicate. */ static bin_tree_t * -duplicate_tree (root, dfa) - const bin_tree_t *root; - re_dfa_t *dfa; +duplicate_tree (const bin_tree_t *root, re_dfa_t *dfa) { const bin_tree_t *node; bin_tree_t *dup_root; |