diff options
Diffstat (limited to 'src/syntax.c')
| -rw-r--r-- | src/syntax.c | 6790 |
1 files changed, 6790 insertions, 0 deletions
diff --git a/src/syntax.c b/src/syntax.c new file mode 100644 index 0000000..4e35956 --- /dev/null +++ b/src/syntax.c @@ -0,0 +1,6790 @@ +/* vi:set ts=8 sts=4 sw=4 noet: + * + * VIM - Vi IMproved by Bram Moolenaar + * + * Do ":help uganda" in Vim to read copying and usage conditions. + * Do ":help credits" in Vim to see a list of people who contributed. + * See README.txt for an overview of the Vim source code. + */ + +/* + * syntax.c: code for syntax highlighting + */ + +#include "vim.h" + +#if defined(FEAT_SYN_HL) || defined(PROTO) + +#define SYN_NAMELEN 50 // maximum length of a syntax name + +// different types of offsets that are possible +#define SPO_MS_OFF 0 // match start offset +#define SPO_ME_OFF 1 // match end offset +#define SPO_HS_OFF 2 // highl. start offset +#define SPO_HE_OFF 3 // highl. end offset +#define SPO_RS_OFF 4 // region start offset +#define SPO_RE_OFF 5 // region end offset +#define SPO_LC_OFF 6 // leading context offset +#define SPO_COUNT 7 + +static char *(spo_name_tab[SPO_COUNT]) = + {"ms=", "me=", "hs=", "he=", "rs=", "re=", "lc="}; + +/* + * The patterns that are being searched for are stored in a syn_pattern. + * A match item consists of one pattern. + * A start/end item consists of n start patterns and m end patterns. + * A start/skip/end item consists of n start patterns, one skip pattern and m + * end patterns. + * For the latter two, the patterns are always consecutive: start-skip-end. + * + * A character offset can be given for the matched text (_m_start and _m_end) + * and for the actually highlighted text (_h_start and _h_end). + * + * Note that ordering of members is optimized to reduce padding. + */ +typedef struct syn_pattern +{ + char sp_type; // see SPTYPE_ defines below + char sp_syncing; // this item used for syncing + short sp_syn_match_id; // highlight group ID of pattern + short sp_off_flags; // see below + int sp_offsets[SPO_COUNT]; // offsets + int sp_flags; // see HL_ defines below +#ifdef FEAT_CONCEAL + int sp_cchar; // conceal substitute character +#endif + int sp_ic; // ignore-case flag for sp_prog + int sp_sync_idx; // sync item index (syncing only) + int sp_line_id; // ID of last line where tried + int sp_startcol; // next match in sp_line_id line + short *sp_cont_list; // cont. group IDs, if non-zero + short *sp_next_list; // next group IDs, if non-zero + struct sp_syn sp_syn; // struct passed to in_id_list() + char_u *sp_pattern; // regexp to match, pattern + regprog_T *sp_prog; // regexp to match, program +#ifdef FEAT_PROFILE + syn_time_T sp_time; +#endif +} synpat_T; + +// The sp_off_flags are computed like this: +// offset from the start of the matched text: (1 << SPO_XX_OFF) +// offset from the end of the matched text: (1 << (SPO_XX_OFF + SPO_COUNT)) +// When both are present, only one is used. + +#define SPTYPE_MATCH 1 // match keyword with this group ID +#define SPTYPE_START 2 // match a regexp, start of item +#define SPTYPE_END 3 // match a regexp, end of item +#define SPTYPE_SKIP 4 // match a regexp, skip within item + + +#define SYN_ITEMS(buf) ((synpat_T *)((buf)->b_syn_patterns.ga_data)) + +#define NONE_IDX (-2) // value of sp_sync_idx for "NONE" + +/* + * Flags for b_syn_sync_flags: + */ +#define SF_CCOMMENT 0x01 // sync on a C-style comment +#define SF_MATCH 0x02 // sync by matching a pattern + +#define SYN_STATE_P(ssp) ((bufstate_T *)((ssp)->ga_data)) + +#define MAXKEYWLEN 80 // maximum length of a keyword + +/* + * The attributes of the syntax item that has been recognized. + */ +static int current_attr = 0; // attr of current syntax word +#ifdef FEAT_EVAL +static int current_id = 0; // ID of current char for syn_get_id() +static int current_trans_id = 0; // idem, transparency removed +#endif +#ifdef FEAT_CONCEAL +static int current_flags = 0; +static int current_seqnr = 0; +static int current_sub_char = 0; +#endif + +typedef struct syn_cluster_S +{ + char_u *scl_name; // syntax cluster name + char_u *scl_name_u; // uppercase of scl_name + short *scl_list; // IDs in this syntax cluster +} syn_cluster_T; + +/* + * Methods of combining two clusters + */ +#define CLUSTER_REPLACE 1 // replace first list with second +#define CLUSTER_ADD 2 // add second list to first +#define CLUSTER_SUBTRACT 3 // subtract second list from first + +#define SYN_CLSTR(buf) ((syn_cluster_T *)((buf)->b_syn_clusters.ga_data)) + +/* + * Syntax group IDs have different types: + * 0 - 19999 normal syntax groups + * 20000 - 20999 ALLBUT indicator (current_syn_inc_tag added) + * 21000 - 21999 TOP indicator (current_syn_inc_tag added) + * 22000 - 22999 CONTAINED indicator (current_syn_inc_tag added) + * 23000 - 32767 cluster IDs (subtract SYNID_CLUSTER for the cluster ID) + */ +#define SYNID_ALLBUT MAX_HL_ID // syntax group ID for contains=ALLBUT +#define SYNID_TOP 21000 // syntax group ID for contains=TOP +#define SYNID_CONTAINED 22000 // syntax group ID for contains=CONTAINED +#define SYNID_CLUSTER 23000 // first syntax group ID for clusters + +#define MAX_SYN_INC_TAG 999 // maximum before the above overflow +#define MAX_CLUSTER_ID (32767 - SYNID_CLUSTER) + +/* + * Annoying Hack(TM): ":syn include" needs this pointer to pass to + * expand_filename(). Most of the other syntax commands don't need it, so + * instead of passing it to them, we stow it here. + */ +static char_u **syn_cmdlinep; + +/* + * Another Annoying Hack(TM): To prevent rules from other ":syn include"'d + * files from leaking into ALLBUT lists, we assign a unique ID to the + * rules in each ":syn include"'d file. + */ +static int current_syn_inc_tag = 0; +static int running_syn_inc_tag = 0; + +/* + * In a hashtable item "hi_key" points to "keyword" in a keyentry. + * This avoids adding a pointer to the hashtable item. + * KE2HIKEY() converts a var pointer to a hashitem key pointer. + * HIKEY2KE() converts a hashitem key pointer to a var pointer. + * HI2KE() converts a hashitem pointer to a var pointer. + */ +static keyentry_T dumkey; +#define KE2HIKEY(kp) ((kp)->keyword) +#define HIKEY2KE(p) ((keyentry_T *)((p) - (dumkey.keyword - (char_u *)&dumkey))) +#define HI2KE(hi) HIKEY2KE((hi)->hi_key) + +/* + * To reduce the time spent in keepend(), remember at which level in the state + * stack the first item with "keepend" is present. When "-1", there is no + * "keepend" on the stack. + */ +static int keepend_level = -1; + +static char msg_no_items[] = N_("No Syntax items defined for this buffer"); + +/* + * For the current state we need to remember more than just the idx. + * When si_m_endpos.lnum is 0, the items other than si_idx are unknown. + * (The end positions have the column number of the next char) + */ +typedef struct state_item +{ + int si_idx; // index of syntax pattern or + // KEYWORD_IDX + int si_id; // highlight group ID for keywords + int si_trans_id; // idem, transparency removed + int si_m_lnum; // lnum of the match + int si_m_startcol; // starting column of the match + lpos_T si_m_endpos; // just after end posn of the match + lpos_T si_h_startpos; // start position of the highlighting + lpos_T si_h_endpos; // end position of the highlighting + lpos_T si_eoe_pos; // end position of end pattern + int si_end_idx; // group ID for end pattern or zero + int si_ends; // if match ends before si_m_endpos + int si_attr; // attributes in this state + long si_flags; // HL_HAS_EOL flag in this state, and + // HL_SKIP* for si_next_list +#ifdef FEAT_CONCEAL + int si_seqnr; // sequence number + int si_cchar; // substitution character for conceal +#endif + short *si_cont_list; // list of contained groups + short *si_next_list; // nextgroup IDs after this item ends + reg_extmatch_T *si_extmatch; // \z(...\) matches from start + // pattern +} stateitem_T; + +#define KEYWORD_IDX (-1) // value of si_idx for keywords +#define ID_LIST_ALL ((short *)-1) // valid of si_cont_list for containing all + // but contained groups + +#ifdef FEAT_CONCEAL +static int next_seqnr = 1; // value to use for si_seqnr +#endif + +/* + * Struct to reduce the number of arguments to get_syn_options(), it's used + * very often. + */ +typedef struct +{ + int flags; // flags for contained and transparent + int keyword; // TRUE for ":syn keyword" + int *sync_idx; // syntax item for "grouphere" argument, NULL + // if not allowed + char has_cont_list; // TRUE if "cont_list" can be used + short *cont_list; // group IDs for "contains" argument + short *cont_in_list; // group IDs for "containedin" argument + short *next_list; // group IDs for "nextgroup" argument +} syn_opt_arg_T; + +/* + * The next possible match in the current line for any pattern is remembered, + * to avoid having to try for a match in each column. + * If next_match_idx == -1, not tried (in this line) yet. + * If next_match_col == MAXCOL, no match found in this line. + * (All end positions have the column of the char after the end) + */ +static int next_match_col; // column for start of next match +static lpos_T next_match_m_endpos; // position for end of next match +static lpos_T next_match_h_startpos; // pos. for highl. start of next match +static lpos_T next_match_h_endpos; // pos. for highl. end of next match +static int next_match_idx; // index of matched item +static long next_match_flags; // flags for next match +static lpos_T next_match_eos_pos; // end of start pattn (start region) +static lpos_T next_match_eoe_pos; // pos. for end of end pattern +static int next_match_end_idx; // ID of group for end pattn or zero +static reg_extmatch_T *next_match_extmatch = NULL; + +/* + * A state stack is an array of integers or stateitem_T, stored in a + * garray_T. A state stack is invalid if its itemsize entry is zero. + */ +#define INVALID_STATE(ssp) ((ssp)->ga_itemsize == 0) +#define VALID_STATE(ssp) ((ssp)->ga_itemsize != 0) + +#define FOR_ALL_SYNSTATES(sb, sst) \ + for ((sst) = (sb)->b_sst_first; (sst) != NULL; (sst) = (sst)->sst_next) + +/* + * The current state (within the line) of the recognition engine. + * When current_state.ga_itemsize is 0 the current state is invalid. + */ +static win_T *syn_win; // current window for highlighting +static buf_T *syn_buf; // current buffer for highlighting +static synblock_T *syn_block; // current buffer for highlighting +static linenr_T current_lnum = 0; // lnum of current state +static colnr_T current_col = 0; // column of current state +static int current_state_stored = 0; // TRUE if stored current state + // after setting current_finished +static int current_finished = 0; // current line has been finished +static garray_T current_state // current stack of state_items + = {0, 0, 0, 0, NULL}; +static short *current_next_list = NULL; // when non-zero, nextgroup list +static int current_next_flags = 0; // flags for current_next_list +static int current_line_id = 0; // unique number for current line + +#define CUR_STATE(idx) ((stateitem_T *)(current_state.ga_data))[idx] + +static void syn_sync(win_T *wp, linenr_T lnum, synstate_T *last_valid); +static int syn_match_linecont(linenr_T lnum); +static void syn_start_line(void); +static void syn_update_ends(int startofline); +static void syn_stack_alloc(void); +static int syn_stack_cleanup(void); +static void syn_stack_free_entry(synblock_T *block, synstate_T *p); +static synstate_T *syn_stack_find_entry(linenr_T lnum); +static synstate_T *store_current_state(void); +static void load_current_state(synstate_T *from); +static void invalidate_current_state(void); +static int syn_stack_equal(synstate_T *sp); +static void validate_current_state(void); +static int syn_finish_line(int syncing); +static int syn_current_attr(int syncing, int displaying, int *can_spell, int keep_state); +static int did_match_already(int idx, garray_T *gap); +static stateitem_T *push_next_match(stateitem_T *cur_si); +static void check_state_ends(void); +static void update_si_attr(int idx); +static void check_keepend(void); +static void update_si_end(stateitem_T *sip, int startcol, int force); +static short *copy_id_list(short *list); +static int in_id_list(stateitem_T *item, short *cont_list, struct sp_syn *ssp, int contained); +static int push_current_state(int idx); +static void pop_current_state(void); +#ifdef FEAT_PROFILE +static void syn_clear_time(syn_time_T *tt); +static void syntime_clear(void); +static void syntime_report(void); +static int syn_time_on = FALSE; +# define IF_SYN_TIME(p) (p) +#else +# define IF_SYN_TIME(p) NULL +typedef int syn_time_T; +#endif + +static void syn_stack_apply_changes_block(synblock_T *block, buf_T *buf); +static void find_endpos(int idx, lpos_T *startpos, lpos_T *m_endpos, lpos_T *hl_endpos, long *flagsp, lpos_T *end_endpos, int *end_idx, reg_extmatch_T *start_ext); + +static void limit_pos(lpos_T *pos, lpos_T *limit); +static void limit_pos_zero(lpos_T *pos, lpos_T *limit); +static void syn_add_end_off(lpos_T *result, regmmatch_T *regmatch, synpat_T *spp, int idx, int extra); +static void syn_add_start_off(lpos_T *result, regmmatch_T *regmatch, synpat_T *spp, int idx, int extra); +static char_u *syn_getcurline(void); +static int syn_regexec(regmmatch_T *rmp, linenr_T lnum, colnr_T col, syn_time_T *st); +static int check_keyword_id(char_u *line, int startcol, int *endcol, long *flags, short **next_list, stateitem_T *cur_si, int *ccharp); +static void syn_remove_pattern(synblock_T *block, int idx); +static void syn_clear_pattern(synblock_T *block, int i); +static void syn_clear_cluster(synblock_T *block, int i); +static void syn_clear_one(int id, int syncing); +static void syn_cmd_onoff(exarg_T *eap, char *name); +static void syn_lines_msg(void); +static void syn_match_msg(void); +static void syn_list_one(int id, int syncing, int link_only); +static void syn_list_cluster(int id); +static void put_id_list(char_u *name, short *list, int attr); +static void put_pattern(char *s, int c, synpat_T *spp, int attr); +static int syn_list_keywords(int id, hashtab_T *ht, int did_header, int attr); +static void syn_clear_keyword(int id, hashtab_T *ht); +static void clear_keywtab(hashtab_T *ht); +static int syn_scl_namen2id(char_u *linep, int len); +static int syn_check_cluster(char_u *pp, int len); +static int syn_add_cluster(char_u *name); +static void init_syn_patterns(void); +static char_u *get_syn_pattern(char_u *arg, synpat_T *ci); +static int get_id_list(char_u **arg, int keylen, short **list, int skip); +static void syn_combine_list(short **clstr1, short **clstr2, int list_op); + +/* + * Start the syntax recognition for a line. This function is normally called + * from the screen updating, once for each displayed line. + * The buffer is remembered in syn_buf, because get_syntax_attr() doesn't get + * it. Careful: curbuf and curwin are likely to point to another buffer and + * window. + */ + void +syntax_start(win_T *wp, linenr_T lnum) +{ + synstate_T *p; + synstate_T *last_valid = NULL; + synstate_T *last_min_valid = NULL; + synstate_T *sp, *prev = NULL; + linenr_T parsed_lnum; + linenr_T first_stored; + int dist; + static varnumber_T changedtick = 0; // remember the last change ID + +#ifdef FEAT_CONCEAL + current_sub_char = NUL; +#endif + + /* + * After switching buffers, invalidate current_state. + * Also do this when a change was made, the current state may be invalid + * then. + */ + if (syn_block != wp->w_s + || syn_buf != wp->w_buffer + || changedtick != CHANGEDTICK(syn_buf)) + { + invalidate_current_state(); + syn_buf = wp->w_buffer; + syn_block = wp->w_s; + } + changedtick = CHANGEDTICK(syn_buf); + syn_win = wp; + + /* + * Allocate syntax stack when needed. + */ + syn_stack_alloc(); + if (syn_block->b_sst_array == NULL) + return; // out of memory + syn_block->b_sst_lasttick = display_tick; + + /* + * If the state of the end of the previous line is useful, store it. + */ + if (VALID_STATE(¤t_state) + && current_lnum < lnum + && current_lnum < syn_buf->b_ml.ml_line_count) + { + (void)syn_finish_line(FALSE); + if (!current_state_stored) + { + ++current_lnum; + (void)store_current_state(); + } + + /* + * If the current_lnum is now the same as "lnum", keep the current + * state (this happens very often!). Otherwise invalidate + * current_state and figure it out below. + */ + if (current_lnum != lnum) + invalidate_current_state(); + } + else + invalidate_current_state(); + + /* + * Try to synchronize from a saved state in b_sst_array[]. + * Only do this if lnum is not before and not to far beyond a saved state. + */ + if (INVALID_STATE(¤t_state) && syn_block->b_sst_array != NULL) + { + // Find last valid saved state before start_lnum. + FOR_ALL_SYNSTATES(syn_block, p) + { + if (p->sst_lnum > lnum) + break; + if (p->sst_lnum <= lnum && p->sst_change_lnum == 0) + { + last_valid = p; + if (p->sst_lnum >= lnum - syn_block->b_syn_sync_minlines) + last_min_valid = p; + } + } + if (last_min_valid != NULL) + load_current_state(last_min_valid); + } + + /* + * If "lnum" is before or far beyond a line with a saved state, need to + * re-synchronize. + */ + if (INVALID_STATE(¤t_state)) + { + syn_sync(wp, lnum, last_valid); + if (current_lnum == 1) + // First line is always valid, no matter "minlines". + first_stored = 1; + else + // Need to parse "minlines" lines before state can be considered + // valid to store. + first_stored = current_lnum + syn_block->b_syn_sync_minlines; + } + else + first_stored = current_lnum; + + /* + * Advance from the sync point or saved state until the current line. + * Save some entries for syncing with later on. + */ + if (syn_block->b_sst_len <= Rows) + dist = 999999; + else + dist = syn_buf->b_ml.ml_line_count / (syn_block->b_sst_len - Rows) + 1; + while (current_lnum < lnum) + { + syn_start_line(); + (void)syn_finish_line(FALSE); + ++current_lnum; + + // If we parsed at least "minlines" lines or started at a valid + // state, the current state is considered valid. + if (current_lnum >= first_stored) + { + // Check if the saved state entry is for the current line and is + // equal to the current state. If so, then validate all saved + // states that depended on a change before the parsed line. + if (prev == NULL) + prev = syn_stack_find_entry(current_lnum - 1); + if (prev == NULL) + sp = syn_block->b_sst_first; + else + sp = prev; + while (sp != NULL && sp->sst_lnum < current_lnum) + sp = sp->sst_next; + if (sp != NULL + && sp->sst_lnum == current_lnum + && syn_stack_equal(sp)) + { + parsed_lnum = current_lnum; + prev = sp; + while (sp != NULL && sp->sst_change_lnum <= parsed_lnum) + { + if (sp->sst_lnum <= lnum) + // valid state before desired line, use this one + prev = sp; + else if (sp->sst_change_lnum == 0) + // past saved states depending on change, break here. + break; + sp->sst_change_lnum = 0; + sp = sp->sst_next; + } + load_current_state(prev); + } + // Store the state at this line when it's the first one, the line + // where we start parsing, or some distance from the previously + // saved state. But only when parsed at least 'minlines'. + else if (prev == NULL + || current_lnum == lnum + || current_lnum >= prev->sst_lnum + dist) + prev = store_current_state(); + } + + // This can take a long time: break when CTRL-C pressed. The current + // state will be wrong then. + line_breakcheck(); + if (got_int) + { + current_lnum = lnum; + break; + } + } + + syn_start_line(); +} + +/* + * We cannot simply discard growarrays full of state_items or buf_states; we + * have to manually release their extmatch pointers first. + */ + static void +clear_syn_state(synstate_T *p) +{ + int i; + garray_T *gap; + + if (p->sst_stacksize > SST_FIX_STATES) + { + gap = &(p->sst_union.sst_ga); + for (i = 0; i < gap->ga_len; i++) + unref_extmatch(SYN_STATE_P(gap)[i].bs_extmatch); + ga_clear(gap); + } + else + { + for (i = 0; i < p->sst_stacksize; i++) + unref_extmatch(p->sst_union.sst_stack[i].bs_extmatch); + } +} + +/* + * Cleanup the current_state stack. + */ + static void +clear_current_state(void) +{ + int i; + stateitem_T *sip; + + sip = (stateitem_T *)(current_state.ga_data); + for (i = 0; i < current_state.ga_len; i++) + unref_extmatch(sip[i].si_extmatch); + ga_clear(¤t_state); +} + +/* + * Try to find a synchronisation point for line "lnum". + * + * This sets current_lnum and the current state. One of three methods is + * used: + * 1. Search backwards for the end of a C-comment. + * 2. Search backwards for given sync patterns. + * 3. Simply start on a given number of lines above "lnum". + */ + static void +syn_sync( + win_T *wp, + linenr_T start_lnum, + synstate_T *last_valid) +{ + buf_T *curbuf_save; + win_T *curwin_save; + pos_T cursor_save; + int idx; + linenr_T lnum; + linenr_T end_lnum; + linenr_T break_lnum; + int had_sync_point; + stateitem_T *cur_si; + synpat_T *spp; + char_u *line; + int found_flags = 0; + int found_match_idx = 0; + linenr_T found_current_lnum = 0; + int found_current_col= 0; + lpos_T found_m_endpos; + colnr_T prev_current_col; + + /* + * Clear any current state that might be hanging around. + */ + invalidate_current_state(); + + /* + * Start at least "minlines" back. Default starting point for parsing is + * there. + * Start further back, to avoid that scrolling backwards will result in + * resyncing for every line. Now it resyncs only one out of N lines, + * where N is minlines * 1.5, or minlines * 2 if minlines is small. + * Watch out for overflow when minlines is MAXLNUM. + */ + if (syn_block->b_syn_sync_minlines > start_lnum) + start_lnum = 1; + else + { + if (syn_block->b_syn_sync_minlines == 1) + lnum = 1; + else if (syn_block->b_syn_sync_minlines < 10) + lnum = syn_block->b_syn_sync_minlines * 2; + else + lnum = syn_block->b_syn_sync_minlines * 3 / 2; + if (syn_block->b_syn_sync_maxlines != 0 + && lnum > syn_block->b_syn_sync_maxlines) + lnum = syn_block->b_syn_sync_maxlines; + if (lnum >= start_lnum) + start_lnum = 1; + else + start_lnum -= lnum; + } + current_lnum = start_lnum; + + /* + * 1. Search backwards for the end of a C-style comment. + */ + if (syn_block->b_syn_sync_flags & SF_CCOMMENT) + { + // Need to make syn_buf the current buffer for a moment, to be able to + // use find_start_comment(). + curwin_save = curwin; + curwin = wp; + curbuf_save = curbuf; + curbuf = syn_buf; + + /* + * Skip lines that end in a backslash. + */ + for ( ; start_lnum > 1; --start_lnum) + { + line = ml_get(start_lnum - 1); + if (*line == NUL || *(line + STRLEN(line) - 1) != '\\') + break; + } + current_lnum = start_lnum; + + // set cursor to start of search + cursor_save = wp->w_cursor; + wp->w_cursor.lnum = start_lnum; + wp->w_cursor.col = 0; + + /* + * If the line is inside a comment, need to find the syntax item that + * defines the comment. + * Restrict the search for the end of a comment to b_syn_sync_maxlines. + */ + if (find_start_comment((int)syn_block->b_syn_sync_maxlines) != NULL) + { + for (idx = syn_block->b_syn_patterns.ga_len; --idx >= 0; ) + if (SYN_ITEMS(syn_block)[idx].sp_syn.id + == syn_block->b_syn_sync_id + && SYN_ITEMS(syn_block)[idx].sp_type == SPTYPE_START) + { + validate_current_state(); + if (push_current_state(idx) == OK) + update_si_attr(current_state.ga_len - 1); + break; + } + } + + // restore cursor and buffer + wp->w_cursor = cursor_save; + curwin = curwin_save; + curbuf = curbuf_save; + } + + /* + * 2. Search backwards for given sync patterns. + */ + else if (syn_block->b_syn_sync_flags & SF_MATCH) + { + if (syn_block->b_syn_sync_maxlines != 0 + && start_lnum > syn_block->b_syn_sync_maxlines) + break_lnum = start_lnum - syn_block->b_syn_sync_maxlines; + else + break_lnum = 0; + + found_m_endpos.lnum = 0; + found_m_endpos.col = 0; + end_lnum = start_lnum; + lnum = start_lnum; + while (--lnum > break_lnum) + { + // This can take a long time: break when CTRL-C pressed. + line_breakcheck(); + if (got_int) + { + invalidate_current_state(); + current_lnum = start_lnum; + break; + } + + // Check if we have run into a valid saved state stack now. + if (last_valid != NULL && lnum == last_valid->sst_lnum) + { + load_current_state(last_valid); + break; + } + + /* + * Check if the previous line has the line-continuation pattern. + */ + if (lnum > 1 && syn_match_linecont(lnum - 1)) + continue; + + /* + * Start with nothing on the state stack + */ + validate_current_state(); + + for (current_lnum = lnum; current_lnum < end_lnum; ++current_lnum) + { + syn_start_line(); + for (;;) + { + had_sync_point = syn_finish_line(TRUE); + /* + * When a sync point has been found, remember where, and + * continue to look for another one, further on in the line. + */ + if (had_sync_point && current_state.ga_len) + { + cur_si = &CUR_STATE(current_state.ga_len - 1); + if (cur_si->si_m_endpos.lnum > start_lnum) + { + // ignore match that goes to after where started + current_lnum = end_lnum; + break; + } + if (cur_si->si_idx < 0) + { + // Cannot happen? + found_flags = 0; + found_match_idx = KEYWORD_IDX; + } + else + { + spp = &(SYN_ITEMS(syn_block)[cur_si->si_idx]); + found_flags = spp->sp_flags; + found_match_idx = spp->sp_sync_idx; + } + found_current_lnum = current_lnum; + found_current_col = current_col; + found_m_endpos = cur_si->si_m_endpos; + /* + * Continue after the match (be aware of a zero-length + * match). + */ + if (found_m_endpos.lnum > current_lnum) + { + current_lnum = found_m_endpos.lnum; + current_col = found_m_endpos.col; + if (current_lnum >= end_lnum) + break; + } + else if (found_m_endpos.col > current_col) + current_col = found_m_endpos.col; + else + ++current_col; + + // syn_current_attr() will have skipped the check for + // an item that ends here, need to do that now. Be + // careful not to go past the NUL. + prev_current_col = current_col; + if (syn_getcurline()[current_col] != NUL) + ++current_col; + check_state_ends(); + current_col = prev_current_col; + } + else + break; + } + } + + /* + * If a sync point was encountered, break here. + */ + if (found_flags) + { + /* + * Put the item that was specified by the sync point on the + * state stack. If there was no item specified, make the + * state stack empty. + */ + clear_current_state(); + if (found_match_idx >= 0 + && push_current_state(found_match_idx) == OK) + update_si_attr(current_state.ga_len - 1); + + /* + * When using "grouphere", continue from the sync point + * match, until the end of the line. Parsing starts at + * the next line. + * For "groupthere" the parsing starts at start_lnum. + */ + if (found_flags & HL_SYNC_HERE) + { + if (current_state.ga_len) + { + cur_si = &CUR_STATE(current_state.ga_len - 1); + cur_si->si_h_startpos.lnum = found_current_lnum; + cur_si->si_h_startpos.col = found_current_col; + update_si_end(cur_si, (int)current_col, TRUE); + check_keepend(); + } + current_col = found_m_endpos.col; + current_lnum = found_m_endpos.lnum; + (void)syn_finish_line(FALSE); + ++current_lnum; + } + else + current_lnum = start_lnum; + + break; + } + + end_lnum = lnum; + invalidate_current_state(); + } + + // Ran into start of the file or exceeded maximum number of lines + if (lnum <= break_lnum) + { + invalidate_current_state(); + current_lnum = break_lnum + 1; + } + } + + validate_current_state(); +} + + static void +save_chartab(char_u *chartab) +{ + if (syn_block->b_syn_isk == empty_option) + return; + + mch_memmove(chartab, syn_buf->b_chartab, (size_t)32); + mch_memmove(syn_buf->b_chartab, syn_win->w_s->b_syn_chartab, (size_t)32); +} + + static void +restore_chartab(char_u *chartab) +{ + if (syn_win->w_s->b_syn_isk != empty_option) + mch_memmove(syn_buf->b_chartab, chartab, (size_t)32); +} + +/* + * Return TRUE if the line-continuation pattern matches in line "lnum". + */ + static int +syn_match_linecont(linenr_T lnum) +{ + regmmatch_T regmatch; + int r; + char_u buf_chartab[32]; // chartab array for syn iskyeyword + + if (syn_block->b_syn_linecont_prog == NULL) + return FALSE; + + // use syntax iskeyword option + save_chartab(buf_chartab); + regmatch.rmm_ic = syn_block->b_syn_linecont_ic; + regmatch.regprog = syn_block->b_syn_linecont_prog; + r = syn_regexec(®match, lnum, (colnr_T)0, + IF_SYN_TIME(&syn_block->b_syn_linecont_time)); + syn_block->b_syn_linecont_prog = regmatch.regprog; + restore_chartab(buf_chartab); + return r; +} + +/* + * Prepare the current state for the start of a line. + */ + static void +syn_start_line(void) +{ + current_finished = FALSE; + current_col = 0; + + /* + * Need to update the end of a start/skip/end that continues from the + * previous line and regions that have "keepend". + */ + if (current_state.ga_len > 0) + { + syn_update_ends(TRUE); + check_state_ends(); + } + + next_match_idx = -1; + ++current_line_id; +#ifdef FEAT_CONCEAL + next_seqnr = 1; +#endif +} + +/* + * Check for items in the stack that need their end updated. + * When "startofline" is TRUE the last item is always updated. + * When "startofline" is FALSE the item with "keepend" is forcefully updated. + */ + static void +syn_update_ends(int startofline) +{ + stateitem_T *cur_si; + int i; + int seen_keepend; + + if (startofline) + { + // Check for a match carried over from a previous line with a + // contained region. The match ends as soon as the region ends. + for (i = 0; i < current_state.ga_len; ++i) + { + cur_si = &CUR_STATE(i); + if (cur_si->si_idx >= 0 + && (SYN_ITEMS(syn_block)[cur_si->si_idx]).sp_type + == SPTYPE_MATCH + && cur_si->si_m_endpos.lnum < current_lnum) + { + cur_si->si_flags |= HL_MATCHCONT; + cur_si->si_m_endpos.lnum = 0; + cur_si->si_m_endpos.col = 0; + cur_si->si_h_endpos = cur_si->si_m_endpos; + cur_si->si_ends = TRUE; + } + } + } + + /* + * Need to update the end of a start/skip/end that continues from the + * previous line. And regions that have "keepend", because they may + * influence contained items. If we've just removed "extend" + * (startofline == 0) then we should update ends of normal regions + * contained inside "keepend" because "extend" could have extended + * these "keepend" regions as well as contained normal regions. + * Then check for items ending in column 0. + */ + i = current_state.ga_len - 1; + if (keepend_level >= 0) + for ( ; i > keepend_level; --i) + if (CUR_STATE(i).si_flags & HL_EXTEND) + break; + + seen_keepend = FALSE; + for ( ; i < current_state.ga_len; ++i) + { + cur_si = &CUR_STATE(i); + if ((cur_si->si_flags & HL_KEEPEND) + || (seen_keepend && !startofline) + || (i == current_state.ga_len - 1 && startofline)) + { + cur_si->si_h_startpos.col = 0; // start highl. in col 0 + cur_si->si_h_startpos.lnum = current_lnum; + + if (!(cur_si->si_flags & HL_MATCHCONT)) + update_si_end(cur_si, (int)current_col, !startofline); + + if (!startofline && (cur_si->si_flags & HL_KEEPEND)) + seen_keepend = TRUE; + } + } + check_keepend(); +} + +///////////////////////////////////////// +// Handling of the state stack cache. + +/* + * EXPLANATION OF THE SYNTAX STATE STACK CACHE + * + * To speed up syntax highlighting, the state stack for the start of some + * lines is cached. These entries can be used to start parsing at that point. + * + * The stack is kept in b_sst_array[] for each buffer. There is a list of + * valid entries. b_sst_first points to the first one, then follow sst_next. + * The entries are sorted on line number. The first entry is often for line 2 + * (line 1 always starts with an empty stack). + * There is also a list for free entries. This construction is used to avoid + * having to allocate and free memory blocks too often. + * + * When making changes to the buffer, this is logged in b_mod_*. When calling + * update_screen() to update the display, it will call + * syn_stack_apply_changes() for each displayed buffer to adjust the cached + * entries. The entries which are inside the changed area are removed, + * because they must be recomputed. Entries below the changed have their line + * number adjusted for deleted/inserted lines, and have their sst_change_lnum + * set to indicate that a check must be made if the changed lines would change + * the cached entry. + * + * When later displaying lines, an entry is stored for each line. Displayed + * lines are likely to be displayed again, in which case the state at the + * start of the line is needed. + * For not displayed lines, an entry is stored for every so many lines. These + * entries will be used e.g., when scrolling backwards. The distance between + * entries depends on the number of lines in the buffer. For small buffers + * the distance is fixed at SST_DIST, for large buffers there is a fixed + * number of entries SST_MAX_ENTRIES, and the distance is computed. + */ + + static void +syn_stack_free_block(synblock_T *block) +{ + synstate_T *p; + + if (block->b_sst_array == NULL) + return; + + FOR_ALL_SYNSTATES(block, p) + clear_syn_state(p); + VIM_CLEAR(block->b_sst_array); + block->b_sst_first = NULL; + block->b_sst_len = 0; +} +/* + * Free b_sst_array[] for buffer "buf". + * Used when syntax items changed to force resyncing everywhere. + */ + void +syn_stack_free_all(synblock_T *block) +{ +#ifdef FEAT_FOLDING + win_T *wp; +#endif + + syn_stack_free_block(block); + +#ifdef FEAT_FOLDING + // When using "syntax" fold method, must update all folds. + FOR_ALL_WINDOWS(wp) + { + if (wp->w_s == block && foldmethodIsSyntax(wp)) + foldUpdateAll(wp); + } +#endif +} + +/* + * Allocate the syntax state stack for syn_buf when needed. + * If the number of entries in b_sst_array[] is much too big or a bit too + * small, reallocate it. + * Also used to allocate b_sst_array[] for the first time. + */ + static void +syn_stack_alloc(void) +{ + long len; + synstate_T *to, *from; + synstate_T *sstp; + + len = syn_buf->b_ml.ml_line_count / SST_DIST + Rows * 2; + if (len < SST_MIN_ENTRIES) + len = SST_MIN_ENTRIES; + else if (len > SST_MAX_ENTRIES) + len = SST_MAX_ENTRIES; + if (syn_block->b_sst_len > len * 2 || syn_block->b_sst_len < len) + { + // Allocate 50% too much, to avoid reallocating too often. + len = syn_buf->b_ml.ml_line_count; + len = (len + len / 2) / SST_DIST + Rows * 2; + if (len < SST_MIN_ENTRIES) + len = SST_MIN_ENTRIES; + else if (len > SST_MAX_ENTRIES) + len = SST_MAX_ENTRIES; + + if (syn_block->b_sst_array != NULL) + { + // When shrinking the array, cleanup the existing stack. + // Make sure that all valid entries fit in the new array. + while (syn_block->b_sst_len - syn_block->b_sst_freecount + 2 > len + && syn_stack_cleanup()) + ; + if (len < syn_block->b_sst_len - syn_block->b_sst_freecount + 2) + len = syn_block->b_sst_len - syn_block->b_sst_freecount + 2; + } + + sstp = ALLOC_CLEAR_MULT(synstate_T, len); + if (sstp == NULL) // out of memory! + return; + + to = sstp - 1; + if (syn_block->b_sst_array != NULL) + { + // Move the states from the old array to the new one. + for (from = syn_block->b_sst_first; from != NULL; + from = from->sst_next) + { + ++to; + *to = *from; + to->sst_next = to + 1; + } + } + if (to != sstp - 1) + { + to->sst_next = NULL; + syn_block->b_sst_first = sstp; + syn_block->b_sst_freecount = len - (int)(to - sstp) - 1; + } + else + { + syn_block->b_sst_first = NULL; + syn_block->b_sst_freecount = len; + } + + // Create the list of free entries. + syn_block->b_sst_firstfree = to + 1; + while (++to < sstp + len) + to->sst_next = to + 1; + (sstp + len - 1)->sst_next = NULL; + + vim_free(syn_block->b_sst_array); + syn_block->b_sst_array = sstp; + syn_block->b_sst_len = len; + } +} + +/* + * Check for changes in a buffer to affect stored syntax states. Uses the + * b_mod_* fields. + * Called from update_screen(), before screen is being updated, once for each + * displayed buffer. + */ + void +syn_stack_apply_changes(buf_T *buf) +{ + win_T *wp; + + syn_stack_apply_changes_block(&buf->b_s, buf); + + FOR_ALL_WINDOWS(wp) + { + if ((wp->w_buffer == buf) && (wp->w_s != &buf->b_s)) + syn_stack_apply_changes_block(wp->w_s, buf); + } +} + + static void +syn_stack_apply_changes_block(synblock_T *block, buf_T *buf) +{ + synstate_T *p, *prev, *np; + linenr_T n; + + prev = NULL; + for (p = block->b_sst_first; p != NULL; ) + { + if (p->sst_lnum + block->b_syn_sync_linebreaks > buf->b_mod_top) + { + n = p->sst_lnum + buf->b_mod_xlines; + if (n <= buf->b_mod_bot) + { + // this state is inside the changed area, remove it + np = p->sst_next; + if (prev == NULL) + block->b_sst_first = np; + else + prev->sst_next = np; + syn_stack_free_entry(block, p); + p = np; + continue; + } + // This state is below the changed area. Remember the line + // that needs to be parsed before this entry can be made valid + // again. + if (p->sst_change_lnum != 0 && p->sst_change_lnum > buf->b_mod_top) + { + if (p->sst_change_lnum + buf->b_mod_xlines > buf->b_mod_top) + p->sst_change_lnum += buf->b_mod_xlines; + else + p->sst_change_lnum = buf->b_mod_top; + } + if (p->sst_change_lnum == 0 + || p->sst_change_lnum < buf->b_mod_bot) + p->sst_change_lnum = buf->b_mod_bot; + + p->sst_lnum = n; + } + prev = p; + p = p->sst_next; + } +} + +/* + * Reduce the number of entries in the state stack for syn_buf. + * Returns TRUE if at least one entry was freed. + */ + static int +syn_stack_cleanup(void) +{ + synstate_T *p, *prev; + disptick_T tick; + int above; + int dist; + int retval = FALSE; + + if (syn_block->b_sst_first == NULL) + return retval; + + // Compute normal distance between non-displayed entries. + if (syn_block->b_sst_len <= Rows) + dist = 999999; + else + dist = syn_buf->b_ml.ml_line_count / (syn_block->b_sst_len - Rows) + 1; + + /* + * Go through the list to find the "tick" for the oldest entry that can + * be removed. Set "above" when the "tick" for the oldest entry is above + * "b_sst_lasttick" (the display tick wraps around). + */ + tick = syn_block->b_sst_lasttick; + above = FALSE; + prev = syn_block->b_sst_first; + for (p = prev->sst_next; p != NULL; prev = p, p = p->sst_next) + { + if (prev->sst_lnum + dist > p->sst_lnum) + { + if (p->sst_tick > syn_block->b_sst_lasttick) + { + if (!above || p->sst_tick < tick) + tick = p->sst_tick; + above = TRUE; + } + else if (!above && p->sst_tick < tick) + tick = p->sst_tick; + } + } + + /* + * Go through the list to make the entries for the oldest tick at an + * interval of several lines. + */ + prev = syn_block->b_sst_first; + for (p = prev->sst_next; p != NULL; prev = p, p = p->sst_next) + { + if (p->sst_tick == tick && prev->sst_lnum + dist > p->sst_lnum) + { + // Move this entry from used list to free list + prev->sst_next = p->sst_next; + syn_stack_free_entry(syn_block, p); + p = prev; + retval = TRUE; + } + } + return retval; +} + +/* + * Free the allocated memory for a syn_state item. + * Move the entry into the free list. + */ + static void +syn_stack_free_entry(synblock_T *block, synstate_T *p) +{ + clear_syn_state(p); + p->sst_next = block->b_sst_firstfree; + block->b_sst_firstfree = p; + ++block->b_sst_freecount; +} + +/* + * Find an entry in the list of state stacks at or before "lnum". + * Returns NULL when there is no entry or the first entry is after "lnum". + */ + static synstate_T * +syn_stack_find_entry(linenr_T lnum) +{ + synstate_T *p, *prev; + + prev = NULL; + for (p = syn_block->b_sst_first; p != NULL; prev = p, p = p->sst_next) + { + if (p->sst_lnum == lnum) + return p; + if (p->sst_lnum > lnum) + break; + } + return prev; +} + +/* + * Try saving the current state in b_sst_array[]. + * The current state must be valid for the start of the current_lnum line! + */ + static synstate_T * +store_current_state(void) +{ + int i; + synstate_T *p; + bufstate_T *bp; + stateitem_T *cur_si; + synstate_T *sp = syn_stack_find_entry(current_lnum); + + /* + * If the current state contains a start or end pattern that continues + * from the previous line, we can't use it. Don't store it then. + */ + for (i = current_state.ga_len - 1; i >= 0; --i) + { + cur_si = &CUR_STATE(i); + if (cur_si->si_h_startpos.lnum >= current_lnum + || cur_si->si_m_endpos.lnum >= current_lnum + || cur_si->si_h_endpos.lnum >= current_lnum + || (cur_si->si_end_idx + && cur_si->si_eoe_pos.lnum >= current_lnum)) + break; + } + if (i >= 0) + { + if (sp != NULL) + { + // find "sp" in the list and remove it + if (syn_block->b_sst_first == sp) + // it's the first entry + syn_block->b_sst_first = sp->sst_next; + else + { + // find the entry just before this one to adjust sst_next + FOR_ALL_SYNSTATES(syn_block, p) + if (p->sst_next == sp) + break; + if (p != NULL) // just in case + p->sst_next = sp->sst_next; + } + syn_stack_free_entry(syn_block, sp); + sp = NULL; + } + } + else if (sp == NULL || sp->sst_lnum != current_lnum) + { + /* + * Add a new entry + */ + // If no free items, cleanup the array first. + if (syn_block->b_sst_freecount == 0) + { + (void)syn_stack_cleanup(); + // "sp" may have been moved to the freelist now + sp = syn_stack_find_entry(current_lnum); + } + // Still no free items? Must be a strange problem... + if (syn_block->b_sst_freecount == 0) + sp = NULL; + else + { + // Take the first item from the free list and put it in the used + // list, after *sp + p = syn_block->b_sst_firstfree; + syn_block->b_sst_firstfree = p->sst_next; + --syn_block->b_sst_freecount; + if (sp == NULL) + { + // Insert in front of the list + p->sst_next = syn_block->b_sst_first; + syn_block->b_sst_first = p; + } + else + { + // insert in list after *sp + p->sst_next = sp->sst_next; + sp->sst_next = p; + } + sp = p; + sp->sst_stacksize = 0; + sp->sst_lnum = current_lnum; + } + } + if (sp != NULL) + { + // When overwriting an existing state stack, clear it first + clear_syn_state(sp); + sp->sst_stacksize = current_state.ga_len; + if (current_state.ga_len > SST_FIX_STATES) + { + // Need to clear it, might be something remaining from when the + // length was less than SST_FIX_STATES. + ga_init2(&sp->sst_union.sst_ga, sizeof(bufstate_T), 1); + if (ga_grow(&sp->sst_union.sst_ga, current_state.ga_len) == FAIL) + sp->sst_stacksize = 0; + else + sp->sst_union.sst_ga.ga_len = current_state.ga_len; + bp = SYN_STATE_P(&(sp->sst_union.sst_ga)); + } + else + bp = sp->sst_union.sst_stack; + for (i = 0; i < sp->sst_stacksize; ++i) + { + bp[i].bs_idx = CUR_STATE(i).si_idx; + bp[i].bs_flags = CUR_STATE(i).si_flags; +#ifdef FEAT_CONCEAL + bp[i].bs_seqnr = CUR_STATE(i).si_seqnr; + bp[i].bs_cchar = CUR_STATE(i).si_cchar; +#endif + bp[i].bs_extmatch = ref_extmatch(CUR_STATE(i).si_extmatch); + } + sp->sst_next_flags = current_next_flags; + sp->sst_next_list = current_next_list; + sp->sst_tick = display_tick; + sp->sst_change_lnum = 0; + } + current_state_stored = TRUE; + return sp; +} + +/* + * Copy a state stack from "from" in b_sst_array[] to current_state; + */ + static void +load_current_state(synstate_T *from) +{ + int i; + bufstate_T *bp; + + clear_current_state(); + validate_current_state(); + keepend_level = -1; + if (from->sst_stacksize + && ga_grow(¤t_state, from->sst_stacksize) == OK) + { + if (from->sst_stacksize > SST_FIX_STATES) + bp = SYN_STATE_P(&(from->sst_union.sst_ga)); + else + bp = from->sst_union.sst_stack; + for (i = 0; i < from->sst_stacksize; ++i) + { + CUR_STATE(i).si_idx = bp[i].bs_idx; + CUR_STATE(i).si_flags = bp[i].bs_flags; +#ifdef FEAT_CONCEAL + CUR_STATE(i).si_seqnr = bp[i].bs_seqnr; + CUR_STATE(i).si_cchar = bp[i].bs_cchar; +#endif + CUR_STATE(i).si_extmatch = ref_extmatch(bp[i].bs_extmatch); + if (keepend_level < 0 && (CUR_STATE(i).si_flags & HL_KEEPEND)) + keepend_level = i; + CUR_STATE(i).si_ends = FALSE; + CUR_STATE(i).si_m_lnum = 0; + if (CUR_STATE(i).si_idx >= 0) + CUR_STATE(i).si_next_list = + (SYN_ITEMS(syn_block)[CUR_STATE(i).si_idx]).sp_next_list; + else + CUR_STATE(i).si_next_list = NULL; + update_si_attr(i); + } + current_state.ga_len = from->sst_stacksize; + } + current_next_list = from->sst_next_list; + current_next_flags = from->sst_next_flags; + current_lnum = from->sst_lnum; +} + +/* + * Compare saved state stack "*sp" with the current state. + * Return TRUE when they are equal. + */ + static int +syn_stack_equal(synstate_T *sp) +{ + int i, j; + bufstate_T *bp; + reg_extmatch_T *six, *bsx; + + // First a quick check if the stacks have the same size end nextlist. + if (sp->sst_stacksize != current_state.ga_len + || sp->sst_next_list != current_next_list) + return FALSE; + + // Need to compare all states on both stacks. + if (sp->sst_stacksize > SST_FIX_STATES) + bp = SYN_STATE_P(&(sp->sst_union.sst_ga)); + else + bp = sp->sst_union.sst_stack; + + for (i = current_state.ga_len; --i >= 0; ) + { + // If the item has another index the state is different. + if (bp[i].bs_idx != CUR_STATE(i).si_idx) + break; + if (bp[i].bs_extmatch == CUR_STATE(i).si_extmatch) + continue; + // When the extmatch pointers are different, the strings in them can + // still be the same. Check if the extmatch references are equal. + bsx = bp[i].bs_extmatch; + six = CUR_STATE(i).si_extmatch; + // If one of the extmatch pointers is NULL the states are different. + if (bsx == NULL || six == NULL) + break; + for (j = 0; j < NSUBEXP; ++j) + { + // Check each referenced match string. They must all be equal. + if (bsx->matches[j] != six->matches[j]) + { + // If the pointer is different it can still be the same text. + // Compare the strings, ignore case when the start item has the + // sp_ic flag set. + if (bsx->matches[j] == NULL || six->matches[j] == NULL) + break; + if ((SYN_ITEMS(syn_block)[CUR_STATE(i).si_idx]).sp_ic + ? MB_STRICMP(bsx->matches[j], six->matches[j]) != 0 + : STRCMP(bsx->matches[j], six->matches[j]) != 0) + break; + } + } + if (j != NSUBEXP) + break; + } + return i < 0 ? TRUE : FALSE; +} + +/* + * We stop parsing syntax above line "lnum". If the stored state at or below + * this line depended on a change before it, it now depends on the line below + * the last parsed line. + * The window looks like this: + * line which changed + * displayed line + * displayed line + * lnum -> line below window + */ + void +syntax_end_parsing(win_T *wp, linenr_T lnum) +{ + synstate_T *sp; + + if (syn_block != wp->w_s) + return; // not the right window + sp = syn_stack_find_entry(lnum); + if (sp != NULL && sp->sst_lnum < lnum) + sp = sp->sst_next; + + if (sp != NULL && sp->sst_change_lnum != 0) + sp->sst_change_lnum = lnum; +} + +/* + * End of handling of the state stack. + ****************************************/ + + static void +invalidate_current_state(void) +{ + clear_current_state(); + current_state.ga_itemsize = 0; // mark current_state invalid + current_next_list = NULL; + keepend_level = -1; +} + + static void +validate_current_state(void) +{ + current_state.ga_itemsize = sizeof(stateitem_T); + current_state.ga_growsize = 3; +} + +/* + * Return TRUE if the syntax at start of lnum changed since last time. + * This will only be called just after get_syntax_attr() for the previous + * line, to check if the next line needs to be redrawn too. + */ + int +syntax_check_changed(linenr_T lnum) +{ + int retval = TRUE; + synstate_T *sp; + + /* + * Check the state stack when: + * - lnum is just below the previously syntaxed line. + * - lnum is not before the lines with saved states. + * - lnum is not past the lines with saved states. + * - lnum is at or before the last changed line. + */ + if (VALID_STATE(¤t_state) && lnum == current_lnum + 1) + { + sp = syn_stack_find_entry(lnum); + if (sp != NULL && sp->sst_lnum == lnum) + { + /* + * finish the previous line (needed when not all of the line was + * drawn) + */ + (void)syn_finish_line(FALSE); + + /* + * Compare the current state with the previously saved state of + * the line. + */ + if (syn_stack_equal(sp)) + retval = FALSE; + + /* + * Store the current state in b_sst_array[] for later use. + */ + ++current_lnum; + (void)store_current_state(); + } + } + + return retval; +} + +/* + * Finish the current line. + * This doesn't return any attributes, it only gets the state at the end of + * the line. It can start anywhere in the line, as long as the current state + * is valid. + */ + static int +syn_finish_line( + int syncing) // called for syncing +{ + stateitem_T *cur_si; + colnr_T prev_current_col; + + while (!current_finished) + { + (void)syn_current_attr(syncing, FALSE, NULL, FALSE); + /* + * When syncing, and found some item, need to check the item. + */ + if (syncing && current_state.ga_len) + { + /* + * Check for match with sync item. + */ + cur_si = &CUR_STATE(current_state.ga_len - 1); + if (cur_si->si_idx >= 0 + && (SYN_ITEMS(syn_block)[cur_si->si_idx].sp_flags + & (HL_SYNC_HERE|HL_SYNC_THERE))) + return TRUE; + + // syn_current_attr() will have skipped the check for an item + // that ends here, need to do that now. Be careful not to go + // past the NUL. + prev_current_col = current_col; + if (syn_getcurline()[current_col] != NUL) + ++current_col; + check_state_ends(); + current_col = prev_current_col; + } + ++current_col; + } + return FALSE; +} + +/* + * Return highlight attributes for next character. + * Must first call syntax_start() once for the line. + * "col" is normally 0 for the first use in a line, and increments by one each + * time. It's allowed to skip characters and to stop before the end of the + * line. But only a "col" after a previously used column is allowed. + * When "can_spell" is not NULL set it to TRUE when spell-checking should be + * done. + */ + int +get_syntax_attr( + colnr_T col, + int *can_spell, + int keep_state) // keep state of char at "col" +{ + int attr = 0; + + if (can_spell != NULL) + // Default: Only do spelling when there is no @Spell cluster or when + // ":syn spell toplevel" was used. + *can_spell = syn_block->b_syn_spell == SYNSPL_DEFAULT + ? (syn_block->b_spell_cluster_id == 0) + : (syn_block->b_syn_spell == SYNSPL_TOP); + + // check for out of memory situation + if (syn_block->b_sst_array == NULL) + return 0; + + // After 'synmaxcol' the attribute is always zero. + if (syn_buf->b_p_smc > 0 && col >= (colnr_T)syn_buf->b_p_smc) + { + clear_current_state(); +#ifdef FEAT_EVAL + current_id = 0; + current_trans_id = 0; +#endif +#ifdef FEAT_CONCEAL + current_flags = 0; + current_seqnr = 0; +#endif + return 0; + } + + // Make sure current_state is valid + if (INVALID_STATE(¤t_state)) + validate_current_state(); + + /* + * Skip from the current column to "col", get the attributes for "col". + */ + while (current_col <= col) + { + attr = syn_current_attr(FALSE, TRUE, can_spell, + current_col == col ? keep_state : FALSE); + ++current_col; + } + + return attr; +} + +/* + * Get syntax attributes for current_lnum, current_col. + */ + static int +syn_current_attr( + int syncing, // When 1: called for syncing + int displaying, // result will be displayed + int *can_spell, // return: do spell checking + int keep_state) // keep syntax stack afterwards +{ + int syn_id; + lpos_T endpos; // was: char_u *endp; + lpos_T hl_startpos; // was: int hl_startcol; + lpos_T hl_endpos; + lpos_T eos_pos; // end-of-start match (start region) + lpos_T eoe_pos; // end-of-end pattern + int end_idx; // group ID for end pattern + int idx; + synpat_T *spp; + stateitem_T *cur_si, *sip = NULL; + int startcol; + int endcol; + long flags; + int cchar; + short *next_list; + int found_match; // found usable match + static int try_next_column = FALSE; // must try in next col + int do_keywords; + regmmatch_T regmatch; + lpos_T pos; + int lc_col; + reg_extmatch_T *cur_extmatch = NULL; + char_u buf_chartab[32]; // chartab array for syn iskyeyword + char_u *line; // current line. NOTE: becomes invalid after + // looking for a pattern match! + + // variables for zero-width matches that have a "nextgroup" argument + int keep_next_list; + int zero_width_next_list = FALSE; + garray_T zero_width_next_ga; + + /* + * No character, no attributes! Past end of line? + * Do try matching with an empty line (could be the start of a region). + */ + line = syn_getcurline(); + if (line[current_col] == NUL && current_col != 0) + { + /* + * If we found a match after the last column, use it. + */ + if (next_match_idx >= 0 && next_match_col >= (int)current_col + && next_match_col != MAXCOL) + (void)push_next_match(NULL); + + current_finished = TRUE; + current_state_stored = FALSE; + return 0; + } + + // if the current or next character is NUL, we will finish the line now + if (line[current_col] == NUL || line[current_col + 1] == NUL) + { + current_finished = TRUE; + current_state_stored = FALSE; + } + + /* + * When in the previous column there was a match but it could not be used + * (empty match or already matched in this column) need to try again in + * the next column. + */ + if (try_next_column) + { + next_match_idx = -1; + try_next_column = FALSE; + } + + // Only check for keywords when not syncing and there are some. + do_keywords = !syncing + && (syn_block->b_keywtab.ht_used > 0 + || syn_block->b_keywtab_ic.ht_used > 0); + + // Init the list of zero-width matches with a nextlist. This is used to + // avoid matching the same item in the same position twice. + ga_init2(&zero_width_next_ga, sizeof(int), 10); + + // use syntax iskeyword option + save_chartab(buf_chartab); + + /* + * Repeat matching keywords and patterns, to find contained items at the + * same column. This stops when there are no extra matches at the current + * column. + */ + do + { + found_match = FALSE; + keep_next_list = FALSE; + syn_id = 0; + + + /* + * 1. Check for a current state. + * Only when there is no current state, or if the current state may + * contain other things, we need to check for keywords and patterns. + * Always need to check for contained items if some item has the + * "containedin" argument (takes extra time!). + */ + if (current_state.ga_len) + cur_si = &CUR_STATE(current_state.ga_len - 1); + else + cur_si = NULL; + + if (syn_block->b_syn_containedin || cur_si == NULL + || cur_si->si_cont_list != NULL) + { + /* + * 2. Check for keywords, if on a keyword char after a non-keyword + * char. Don't do this when syncing. + */ + if (do_keywords) + { + line = syn_getcurline(); + if (vim_iswordp_buf(line + current_col, syn_buf) + && (current_col == 0 + || !vim_iswordp_buf(line + current_col - 1 + - (has_mbyte + ? (*mb_head_off)(line, line + current_col - 1) + : 0) , syn_buf))) + { + syn_id = check_keyword_id(line, (int)current_col, + &endcol, &flags, &next_list, cur_si, + &cchar); + if (syn_id != 0) + { + if (push_current_state(KEYWORD_IDX) == OK) + { + cur_si = &CUR_STATE(current_state.ga_len - 1); + cur_si->si_m_startcol = current_col; + cur_si->si_h_startpos.lnum = current_lnum; + cur_si->si_h_startpos.col = 0; // starts right away + cur_si->si_m_endpos.lnum = current_lnum; + cur_si->si_m_endpos.col = endcol; + cur_si->si_h_endpos.lnum = current_lnum; + cur_si->si_h_endpos.col = endcol; + cur_si->si_ends = TRUE; + cur_si->si_end_idx = 0; + cur_si->si_flags = flags; +#ifdef FEAT_CONCEAL + cur_si->si_seqnr = next_seqnr++; + cur_si->si_cchar = cchar; + if (current_state.ga_len > 1) + cur_si->si_flags |= + CUR_STATE(current_state.ga_len - 2).si_flags + & HL_CONCEAL; +#endif + cur_si->si_id = syn_id; + cur_si->si_trans_id = syn_id; + if (flags & HL_TRANSP) + { + if (current_state.ga_len < 2) + { + cur_si->si_attr = 0; + cur_si->si_trans_id = 0; + } + else + { + cur_si->si_attr = CUR_STATE( + current_state.ga_len - 2).si_attr; + cur_si->si_trans_id = CUR_STATE( + current_state.ga_len - 2).si_trans_id; + } + } + else + cur_si->si_attr = syn_id2attr(syn_id); + cur_si->si_cont_list = NULL; + cur_si->si_next_list = next_list; + check_keepend(); + } + else + vim_free(next_list); + } + } + } + + /* + * 3. Check for patterns (only if no keyword found). + */ + if (syn_id == 0 && syn_block->b_syn_patterns.ga_len) + { + /* + * If we didn't check for a match yet, or we are past it, check + * for any match with a pattern. + */ + if (next_match_idx < 0 || next_match_col < (int)current_col) + { + /* + * Check all relevant patterns for a match at this + * position. This is complicated, because matching with a + * pattern takes quite a bit of time, thus we want to + * avoid doing it when it's not needed. + */ + next_match_idx = 0; // no match in this line yet + next_match_col = MAXCOL; + for (idx = syn_block->b_syn_patterns.ga_len; --idx >= 0; ) + { + spp = &(SYN_ITEMS(syn_block)[idx]); + if ( spp->sp_syncing == syncing + && (displaying || !(spp->sp_flags & HL_DISPLAY)) + && (spp->sp_type == SPTYPE_MATCH + || spp->sp_type == SPTYPE_START) + && (current_next_list != NULL + ? in_id_list(NULL, current_next_list, + &spp->sp_syn, 0) + : (cur_si == NULL + ? !(spp->sp_flags & HL_CONTAINED) + : in_id_list(cur_si, + cur_si->si_cont_list, &spp->sp_syn, + spp->sp_flags & HL_CONTAINED)))) + { + int r; + + // If we already tried matching in this line, and + // there isn't a match before next_match_col, skip + // this item. + if (spp->sp_line_id == current_line_id + && spp->sp_startcol >= next_match_col) + continue; + spp->sp_line_id = current_line_id; + + lc_col = current_col - spp->sp_offsets[SPO_LC_OFF]; + if (lc_col < 0) + lc_col = 0; + + regmatch.rmm_ic = spp->sp_ic; + regmatch.regprog = spp->sp_prog; + r = syn_regexec(®match, + current_lnum, + (colnr_T)lc_col, + IF_SYN_TIME(&spp->sp_time)); + spp->sp_prog = regmatch.regprog; + if (!r) + { + // no match in this line, try another one + spp->sp_startcol = MAXCOL; + continue; + } + + /* + * Compute the first column of the match. + */ + syn_add_start_off(&pos, ®match, + spp, SPO_MS_OFF, -1); + if (pos.lnum > current_lnum) + { + // must have used end of match in a next line, + // we can't handle that + spp->sp_startcol = MAXCOL; + continue; + } + startcol = pos.col; + + // remember the next column where this pattern + // matches in the current line + spp->sp_startcol = startcol; + + /* + * If a previously found match starts at a lower + * column number, don't use this one. + */ + if (startcol >= next_match_col) + continue; + + /* + * If we matched this pattern at this position + * before, skip it. Must retry in the next + * column, because it may match from there. + */ + if (did_match_already(idx, &zero_width_next_ga)) + { + try_next_column = TRUE; + continue; + } + + endpos.lnum = regmatch.endpos[0].lnum; + endpos.col = regmatch.endpos[0].col; + + // Compute the highlight start. + syn_add_start_off(&hl_startpos, ®match, + spp, SPO_HS_OFF, -1); + + // Compute the region start. + // Default is to use the end of the match. + syn_add_end_off(&eos_pos, ®match, + spp, SPO_RS_OFF, 0); + + /* + * Grab the external submatches before they get + * overwritten. Reference count doesn't change. + */ + unref_extmatch(cur_extmatch); + cur_extmatch = re_extmatch_out; + re_extmatch_out = NULL; + + flags = 0; + eoe_pos.lnum = 0; // avoid warning + eoe_pos.col = 0; + end_idx = 0; + hl_endpos.lnum = 0; + + /* + * For a "oneline" the end must be found in the + * same line too. Search for it after the end of + * the match with the start pattern. Set the + * resulting end positions at the same time. + */ + if (spp->sp_type == SPTYPE_START + && (spp->sp_flags & HL_ONELINE)) + { + lpos_T startpos; + + startpos = endpos; + find_endpos(idx, &startpos, &endpos, &hl_endpos, + &flags, &eoe_pos, &end_idx, cur_extmatch); + if (endpos.lnum == 0) + continue; // not found + } + + /* + * For a "match" the size must be > 0 after the + * end offset needs has been added. Except when + * syncing. + */ + else if (spp->sp_type == SPTYPE_MATCH) + { + syn_add_end_off(&hl_endpos, ®match, spp, + SPO_HE_OFF, 0); + syn_add_end_off(&endpos, ®match, spp, + SPO_ME_OFF, 0); + if (endpos.lnum == current_lnum + && (int)endpos.col + syncing < startcol) + { + /* + * If an empty string is matched, may need + * to try matching again at next column. + */ + if (regmatch.startpos[0].col + == regmatch.endpos[0].col) + try_next_column = TRUE; + continue; + } + } + + /* + * keep the best match so far in next_match_* + */ + // Highlighting must start after startpos and end + // before endpos. + if (hl_startpos.lnum == current_lnum + && (int)hl_startpos.col < startcol) + hl_startpos.col = startcol; + limit_pos_zero(&hl_endpos, &endpos); + + next_match_idx = idx; + next_match_col = startcol; + next_match_m_endpos = endpos; + next_match_h_endpos = hl_endpos; + next_match_h_startpos = hl_startpos; + next_match_flags = flags; + next_match_eos_pos = eos_pos; + next_match_eoe_pos = eoe_pos; + next_match_end_idx = end_idx; + unref_extmatch(next_match_extmatch); + next_match_extmatch = cur_extmatch; + cur_extmatch = NULL; + } + } + } + + /* + * If we found a match at the current column, use it. + */ + if (next_match_idx >= 0 && next_match_col == (int)current_col) + { + synpat_T *lspp; + + // When a zero-width item matched which has a nextgroup, + // don't push the item but set nextgroup. + lspp = &(SYN_ITEMS(syn_block)[next_match_idx]); + if (next_match_m_endpos.lnum == current_lnum + && next_match_m_endpos.col == current_col + && lspp->sp_next_list != NULL) + { + current_next_list = lspp->sp_next_list; + current_next_flags = lspp->sp_flags; + keep_next_list = TRUE; + zero_width_next_list = TRUE; + + // Add the index to a list, so that we can check + // later that we don't match it again (and cause an + // endless loop). + if (ga_grow(&zero_width_next_ga, 1) == OK) + { + ((int *)(zero_width_next_ga.ga_data)) + [zero_width_next_ga.ga_len++] = next_match_idx; + } + next_match_idx = -1; + } + else + cur_si = push_next_match(cur_si); + found_match = TRUE; + } + } + } + + /* + * Handle searching for nextgroup match. + */ + if (current_next_list != NULL && !keep_next_list) + { + /* + * If a nextgroup was not found, continue looking for one if: + * - this is an empty line and the "skipempty" option was given + * - we are on white space and the "skipwhite" option was given + */ + if (!found_match) + { + line = syn_getcurline(); + if (((current_next_flags & HL_SKIPWHITE) + && VIM_ISWHITE(line[current_col])) + || ((current_next_flags & HL_SKIPEMPTY) + && *line == NUL)) + break; + } + + /* + * If a nextgroup was found: Use it, and continue looking for + * contained matches. + * If a nextgroup was not found: Continue looking for a normal + * match. + * When did set current_next_list for a zero-width item and no + * match was found don't loop (would get stuck). + */ + current_next_list = NULL; + next_match_idx = -1; + if (!zero_width_next_list) + found_match = TRUE; + } + + } while (found_match); + + restore_chartab(buf_chartab); + + /* + * Use attributes from the current state, if within its highlighting. + * If not, use attributes from the current-but-one state, etc. + */ + current_attr = 0; +#ifdef FEAT_EVAL + current_id = 0; + current_trans_id = 0; +#endif +#ifdef FEAT_CONCEAL + current_flags = 0; + current_seqnr = 0; +#endif + if (cur_si != NULL) + { +#ifndef FEAT_EVAL + int current_trans_id = 0; +#endif + for (idx = current_state.ga_len - 1; idx >= 0; --idx) + { + sip = &CUR_STATE(idx); + if ((current_lnum > sip->si_h_startpos.lnum + || (current_lnum == sip->si_h_startpos.lnum + && current_col >= sip->si_h_startpos.col)) + && (sip->si_h_endpos.lnum == 0 + || current_lnum < sip->si_h_endpos.lnum + || (current_lnum == sip->si_h_endpos.lnum + && current_col < sip->si_h_endpos.col))) + { + current_attr = sip->si_attr; +#ifdef FEAT_EVAL + current_id = sip->si_id; +#endif + current_trans_id = sip->si_trans_id; +#ifdef FEAT_CONCEAL + current_flags = sip->si_flags; + current_seqnr = sip->si_seqnr; + current_sub_char = sip->si_cchar; +#endif + break; + } + } + + if (can_spell != NULL) + { + struct sp_syn sps; + + /* + * set "can_spell" to TRUE if spell checking is supposed to be + * done in the current item. + */ + if (syn_block->b_spell_cluster_id == 0) + { + // There is no @Spell cluster: Do spelling for items without + // @NoSpell cluster. + if (syn_block->b_nospell_cluster_id == 0 + || current_trans_id == 0) + *can_spell = (syn_block->b_syn_spell != SYNSPL_NOTOP); + else + { + sps.inc_tag = 0; + sps.id = syn_block->b_nospell_cluster_id; + sps.cont_in_list = NULL; + *can_spell = !in_id_list(sip, sip->si_cont_list, &sps, 0); + } + } + else + { + // The @Spell cluster is defined: Do spelling in items with + // the @Spell cluster. But not when @NoSpell is also there. + // At the toplevel only spell check when ":syn spell toplevel" + // was used. + if (current_trans_id == 0) + *can_spell = (syn_block->b_syn_spell == SYNSPL_TOP); + else + { + sps.inc_tag = 0; + sps.id = syn_block->b_spell_cluster_id; + sps.cont_in_list = NULL; + *can_spell = in_id_list(sip, sip->si_cont_list, &sps, 0); + + if (syn_block->b_nospell_cluster_id != 0) + { + sps.id = syn_block->b_nospell_cluster_id; + if (in_id_list(sip, sip->si_cont_list, &sps, 0)) + *can_spell = FALSE; + } + } + } + } + + + /* + * Check for end of current state (and the states before it) at the + * next column. Don't do this for syncing, because we would miss a + * single character match. + * First check if the current state ends at the current column. It + * may be for an empty match and a containing item might end in the + * current column. + */ + if (!syncing && !keep_state) + { + check_state_ends(); + if (current_state.ga_len > 0 + && syn_getcurline()[current_col] != NUL) + { + ++current_col; + check_state_ends(); + --current_col; + } + } + } + else if (can_spell != NULL) + // Default: Only do spelling when there is no @Spell cluster or when + // ":syn spell toplevel" was used. + *can_spell = syn_block->b_syn_spell == SYNSPL_DEFAULT + ? (syn_block->b_spell_cluster_id == 0) + : (syn_block->b_syn_spell == SYNSPL_TOP); + + // nextgroup ends at end of line, unless "skipnl" or "skipempty" present + if (current_next_list != NULL + && (line = syn_getcurline())[current_col] != NUL + && line[current_col + 1] == NUL + && !(current_next_flags & (HL_SKIPNL | HL_SKIPEMPTY))) + current_next_list = NULL; + + if (zero_width_next_ga.ga_len > 0) + ga_clear(&zero_width_next_ga); + + // No longer need external matches. But keep next_match_extmatch. + unref_extmatch(re_extmatch_out); + re_extmatch_out = NULL; + unref_extmatch(cur_extmatch); + + return current_attr; +} + + +/* + * Check if we already matched pattern "idx" at the current column. + */ + static int +did_match_already(int idx, garray_T *gap) +{ + int i; + + for (i = current_state.ga_len; --i >= 0; ) + if (CUR_STATE(i).si_m_startcol == (int)current_col + && CUR_STATE(i).si_m_lnum == (int)current_lnum + && CUR_STATE(i).si_idx == idx) + return TRUE; + + // Zero-width matches with a nextgroup argument are not put on the syntax + // stack, and can only be matched once anyway. + for (i = gap->ga_len; --i >= 0; ) + if (((int *)(gap->ga_data))[i] == idx) + return TRUE; + + return FALSE; +} + +/* + * Push the next match onto the stack. + */ + static stateitem_T * +push_next_match(stateitem_T *cur_si) +{ + synpat_T *spp; +#ifdef FEAT_CONCEAL + int save_flags; +#endif + + spp = &(SYN_ITEMS(syn_block)[next_match_idx]); + + /* + * Push the item in current_state stack; + */ + if (push_current_state(next_match_idx) == OK) + { + /* + * If it's a start-skip-end type that crosses lines, figure out how + * much it continues in this line. Otherwise just fill in the length. + */ + cur_si = &CUR_STATE(current_state.ga_len - 1); + cur_si->si_h_startpos = next_match_h_startpos; + cur_si->si_m_startcol = current_col; + cur_si->si_m_lnum = current_lnum; + cur_si->si_flags = spp->sp_flags; +#ifdef FEAT_CONCEAL + cur_si->si_seqnr = next_seqnr++; + cur_si->si_cchar = spp->sp_cchar; + if (current_state.ga_len > 1) + cur_si->si_flags |= + CUR_STATE(current_state.ga_len - 2).si_flags & HL_CONCEAL; +#endif + cur_si->si_next_list = spp->sp_next_list; + cur_si->si_extmatch = ref_extmatch(next_match_extmatch); + if (spp->sp_type == SPTYPE_START && !(spp->sp_flags & HL_ONELINE)) + { + // Try to find the end pattern in the current line + update_si_end(cur_si, (int)(next_match_m_endpos.col), TRUE); + check_keepend(); + } + else + { + cur_si->si_m_endpos = next_match_m_endpos; + cur_si->si_h_endpos = next_match_h_endpos; + cur_si->si_ends = TRUE; + cur_si->si_flags |= next_match_flags; + cur_si->si_eoe_pos = next_match_eoe_pos; + cur_si->si_end_idx = next_match_end_idx; + } + if (keepend_level < 0 && (cur_si->si_flags & HL_KEEPEND)) + keepend_level = current_state.ga_len - 1; + check_keepend(); + update_si_attr(current_state.ga_len - 1); + +#ifdef FEAT_CONCEAL + save_flags = cur_si->si_flags & (HL_CONCEAL | HL_CONCEALENDS); +#endif + /* + * If the start pattern has another highlight group, push another item + * on the stack for the start pattern. + */ + if ( spp->sp_type == SPTYPE_START + && spp->sp_syn_match_id != 0 + && push_current_state(next_match_idx) == OK) + { + cur_si = &CUR_STATE(current_state.ga_len - 1); + cur_si->si_h_startpos = next_match_h_startpos; + cur_si->si_m_startcol = current_col; + cur_si->si_m_lnum = current_lnum; + cur_si->si_m_endpos = next_match_eos_pos; + cur_si->si_h_endpos = next_match_eos_pos; + cur_si->si_ends = TRUE; + cur_si->si_end_idx = 0; + cur_si->si_flags = HL_MATCH; +#ifdef FEAT_CONCEAL + cur_si->si_seqnr = next_seqnr++; + cur_si->si_flags |= save_flags; + if (cur_si->si_flags & HL_CONCEALENDS) + cur_si->si_flags |= HL_CONCEAL; +#endif + cur_si->si_next_list = NULL; + check_keepend(); + update_si_attr(current_state.ga_len - 1); + } + } + + next_match_idx = -1; // try other match next time + + return cur_si; +} + +/* + * Check for end of current state (and the states before it). + */ + static void +check_state_ends(void) +{ + stateitem_T *cur_si; + int had_extend; + + cur_si = &CUR_STATE(current_state.ga_len - 1); + for (;;) + { + if (cur_si->si_ends + && (cur_si->si_m_endpos.lnum < current_lnum + || (cur_si->si_m_endpos.lnum == current_lnum + && cur_si->si_m_endpos.col <= current_col))) + { + /* + * If there is an end pattern group ID, highlight the end pattern + * now. No need to pop the current item from the stack. + * Only do this if the end pattern continues beyond the current + * position. + */ + if (cur_si->si_end_idx + && (cur_si->si_eoe_pos.lnum > current_lnum + || (cur_si->si_eoe_pos.lnum == current_lnum + && cur_si->si_eoe_pos.col > current_col))) + { + cur_si->si_idx = cur_si->si_end_idx; + cur_si->si_end_idx = 0; + cur_si->si_m_endpos = cur_si->si_eoe_pos; + cur_si->si_h_endpos = cur_si->si_eoe_pos; + cur_si->si_flags |= HL_MATCH; +#ifdef FEAT_CONCEAL + cur_si->si_seqnr = next_seqnr++; + if (cur_si->si_flags & HL_CONCEALENDS) + cur_si->si_flags |= HL_CONCEAL; +#endif + update_si_attr(current_state.ga_len - 1); + + // nextgroup= should not match in the end pattern + current_next_list = NULL; + + // what matches next may be different now, clear it + next_match_idx = 0; + next_match_col = MAXCOL; + break; + } + + // handle next_list, unless at end of line and no "skipnl" or + // "skipempty" + current_next_list = cur_si->si_next_list; + current_next_flags = cur_si->si_flags; + if (!(current_next_flags & (HL_SKIPNL | HL_SKIPEMPTY)) + && syn_getcurline()[current_col] == NUL) + current_next_list = NULL; + + // When the ended item has "extend", another item with + // "keepend" now needs to check for its end. + had_extend = (cur_si->si_flags & HL_EXTEND); + + pop_current_state(); + + if (current_state.ga_len == 0) + break; + + if (had_extend && keepend_level >= 0) + { + syn_update_ends(FALSE); + if (current_state.ga_len == 0) + break; + } + + cur_si = &CUR_STATE(current_state.ga_len - 1); + + /* + * Only for a region the search for the end continues after + * the end of the contained item. If the contained match + * included the end-of-line, break here, the region continues. + * Don't do this when: + * - "keepend" is used for the contained item + * - not at the end of the line (could be end="x$"me=e-1). + * - "excludenl" is used (HL_HAS_EOL won't be set) + */ + if (cur_si->si_idx >= 0 + && SYN_ITEMS(syn_block)[cur_si->si_idx].sp_type + == SPTYPE_START + && !(cur_si->si_flags & (HL_MATCH | HL_KEEPEND))) + { + update_si_end(cur_si, (int)current_col, TRUE); + check_keepend(); + if ((current_next_flags & HL_HAS_EOL) + && keepend_level < 0 + && syn_getcurline()[current_col] == NUL) + break; + } + } + else + break; + } +} + +/* + * Update an entry in the current_state stack for a match or region. This + * fills in si_attr, si_next_list and si_cont_list. + */ + static void +update_si_attr(int idx) +{ + stateitem_T *sip = &CUR_STATE(idx); + synpat_T *spp; + + // This should not happen... + if (sip->si_idx < 0) + return; + + spp = &(SYN_ITEMS(syn_block)[sip->si_idx]); + if (sip->si_flags & HL_MATCH) + sip->si_id = spp->sp_syn_match_id; + else + sip->si_id = spp->sp_syn.id; + sip->si_attr = syn_id2attr(sip->si_id); + sip->si_trans_id = sip->si_id; + if (sip->si_flags & HL_MATCH) + sip->si_cont_list = NULL; + else + sip->si_cont_list = spp->sp_cont_list; + + /* + * For transparent items, take attr from outer item. + * Also take cont_list, if there is none. + * Don't do this for the matchgroup of a start or end pattern. + */ + if ((spp->sp_flags & HL_TRANSP) && !(sip->si_flags & HL_MATCH)) + { + if (idx == 0) + { + sip->si_attr = 0; + sip->si_trans_id = 0; + if (sip->si_cont_list == NULL) + sip->si_cont_list = ID_LIST_ALL; + } + else + { + sip->si_attr = CUR_STATE(idx - 1).si_attr; + sip->si_trans_id = CUR_STATE(idx - 1).si_trans_id; + if (sip->si_cont_list == NULL) + { + sip->si_flags |= HL_TRANS_CONT; + sip->si_cont_list = CUR_STATE(idx - 1).si_cont_list; + } + } + } +} + +/* + * Check the current stack for patterns with "keepend" flag. + * Propagate the match-end to contained items, until a "skipend" item is found. + */ + static void +check_keepend(void) +{ + int i; + lpos_T maxpos; + lpos_T maxpos_h; + stateitem_T *sip; + + /* + * This check can consume a lot of time; only do it from the level where + * there really is a keepend. + */ + if (keepend_level < 0) + return; + + /* + * Find the last index of an "extend" item. "keepend" items before that + * won't do anything. If there is no "extend" item "i" will be + * "keepend_level" and all "keepend" items will work normally. + */ + for (i = current_state.ga_len - 1; i > keepend_level; --i) + if (CUR_STATE(i).si_flags & HL_EXTEND) + break; + + maxpos.lnum = 0; + maxpos.col = 0; + maxpos_h.lnum = 0; + maxpos_h.col = 0; + for ( ; i < current_state.ga_len; ++i) + { + sip = &CUR_STATE(i); + if (maxpos.lnum != 0) + { + limit_pos_zero(&sip->si_m_endpos, &maxpos); + limit_pos_zero(&sip->si_h_endpos, &maxpos_h); + limit_pos_zero(&sip->si_eoe_pos, &maxpos); + sip->si_ends = TRUE; + } + if (sip->si_ends && (sip->si_flags & HL_KEEPEND)) + { + if (maxpos.lnum == 0 + || maxpos.lnum > sip->si_m_endpos.lnum + || (maxpos.lnum == sip->si_m_endpos.lnum + && maxpos.col > sip->si_m_endpos.col)) + maxpos = sip->si_m_endpos; + if (maxpos_h.lnum == 0 + || maxpos_h.lnum > sip->si_h_endpos.lnum + || (maxpos_h.lnum == sip->si_h_endpos.lnum + && maxpos_h.col > sip->si_h_endpos.col)) + maxpos_h = sip->si_h_endpos; + } + } +} + +/* + * Update an entry in the current_state stack for a start-skip-end pattern. + * This finds the end of the current item, if it's in the current line. + * + * Return the flags for the matched END. + */ + static void +update_si_end( + stateitem_T *sip, + int startcol, // where to start searching for the end + int force) // when TRUE overrule a previous end +{ + lpos_T startpos; + lpos_T endpos; + lpos_T hl_endpos; + lpos_T end_endpos; + int end_idx; + + // return quickly for a keyword + if (sip->si_idx < 0) + return; + + // Don't update when it's already done. Can be a match of an end pattern + // that started in a previous line. Watch out: can also be a "keepend" + // from a containing item. + if (!force && sip->si_m_endpos.lnum >= current_lnum) + return; + + /* + * We need to find the end of the region. It may continue in the next + * line. + */ + end_idx = 0; + startpos.lnum = current_lnum; + startpos.col = startcol; + find_endpos(sip->si_idx, &startpos, &endpos, &hl_endpos, + &(sip->si_flags), &end_endpos, &end_idx, sip->si_extmatch); + + if (endpos.lnum == 0) + { + // No end pattern matched. + if (SYN_ITEMS(syn_block)[sip->si_idx].sp_flags & HL_ONELINE) + { + // a "oneline" never continues in the next line + sip->si_ends = TRUE; + sip->si_m_endpos.lnum = current_lnum; + sip->si_m_endpos.col = (colnr_T)STRLEN(syn_getcurline()); + } + else + { + // continues in the next line + sip->si_ends = FALSE; + sip->si_m_endpos.lnum = 0; + } + sip->si_h_endpos = sip->si_m_endpos; + } + else + { + // match within this line + sip->si_m_endpos = endpos; + sip->si_h_endpos = hl_endpos; + sip->si_eoe_pos = end_endpos; + sip->si_ends = TRUE; + sip->si_end_idx = end_idx; + } +} + +/* + * Add a new state to the current state stack. + * It is cleared and the index set to "idx". + * Return FAIL if it's not possible (out of memory). + */ + static int +push_current_state(int idx) +{ + if (ga_grow(¤t_state, 1) == FAIL) + return FAIL; + CLEAR_POINTER(&CUR_STATE(current_state.ga_len)); + CUR_STATE(current_state.ga_len).si_idx = idx; + ++current_state.ga_len; + return OK; +} + +/* + * Remove a state from the current_state stack. + */ + static void +pop_current_state(void) +{ + if (current_state.ga_len) + { + unref_extmatch(CUR_STATE(current_state.ga_len - 1).si_extmatch); + --current_state.ga_len; + } + // after the end of a pattern, try matching a keyword or pattern + next_match_idx = -1; + + // if first state with "keepend" is popped, reset keepend_level + if (keepend_level >= current_state.ga_len) + keepend_level = -1; +} + +/* + * Find the end of a start/skip/end syntax region after "startpos". + * Only checks one line. + * Also handles a match item that continued from a previous line. + * If not found, the syntax item continues in the next line. m_endpos->lnum + * will be 0. + * If found, the end of the region and the end of the highlighting is + * computed. + */ + static void +find_endpos( + int idx, // index of the pattern + lpos_T *startpos, // where to start looking for an END match + lpos_T *m_endpos, // return: end of match + lpos_T *hl_endpos, // return: end of highlighting + long *flagsp, // return: flags of matching END + lpos_T *end_endpos, // return: end of end pattern match + int *end_idx, // return: group ID for end pat. match, or 0 + reg_extmatch_T *start_ext) // submatches from the start pattern +{ + colnr_T matchcol; + synpat_T *spp, *spp_skip; + int start_idx; + int best_idx; + regmmatch_T regmatch; + regmmatch_T best_regmatch; // startpos/endpos of best match + lpos_T pos; + char_u *line; + int had_match = FALSE; + char_u buf_chartab[32]; // chartab array for syn option iskyeyword + + // just in case we are invoked for a keyword + if (idx < 0) + return; + + /* + * Check for being called with a START pattern. + * Can happen with a match that continues to the next line, because it + * contained a region. + */ + spp = &(SYN_ITEMS(syn_block)[idx]); + if (spp->sp_type != SPTYPE_START) + { + *hl_endpos = *startpos; + return; + } + + /* + * Find the SKIP or first END pattern after the last START pattern. + */ + for (;;) + { + spp = &(SYN_ITEMS(syn_block)[idx]); + if (spp->sp_type != SPTYPE_START) + break; + ++idx; + } + + /* + * Lookup the SKIP pattern (if present) + */ + if (spp->sp_type == SPTYPE_SKIP) + { + spp_skip = spp; + ++idx; + } + else + spp_skip = NULL; + + // Setup external matches for syn_regexec(). + unref_extmatch(re_extmatch_in); + re_extmatch_in = ref_extmatch(start_ext); + + matchcol = startpos->col; // start looking for a match at sstart + start_idx = idx; // remember the first END pattern. + best_regmatch.startpos[0].col = 0; // avoid compiler warning + + // use syntax iskeyword option + save_chartab(buf_chartab); + + for (;;) + { + /* + * Find end pattern that matches first after "matchcol". + */ + best_idx = -1; + for (idx = start_idx; idx < syn_block->b_syn_patterns.ga_len; ++idx) + { + int lc_col = matchcol; + int r; + + spp = &(SYN_ITEMS(syn_block)[idx]); + if (spp->sp_type != SPTYPE_END) // past last END pattern + break; + lc_col -= spp->sp_offsets[SPO_LC_OFF]; + if (lc_col < 0) + lc_col = 0; + + regmatch.rmm_ic = spp->sp_ic; + regmatch.regprog = spp->sp_prog; + r = syn_regexec(®match, startpos->lnum, lc_col, + IF_SYN_TIME(&spp->sp_time)); + spp->sp_prog = regmatch.regprog; + if (r) + { + if (best_idx == -1 || regmatch.startpos[0].col + < best_regmatch.startpos[0].col) + { + best_idx = idx; + best_regmatch.startpos[0] = regmatch.startpos[0]; + best_regmatch.endpos[0] = regmatch.endpos[0]; + } + } + } + + /* + * If all end patterns have been tried, and there is no match, the + * item continues until end-of-line. + */ + if (best_idx == -1) + break; + + /* + * If the skip pattern matches before the end pattern, + * continue searching after the skip pattern. + */ + if (spp_skip != NULL) + { + int lc_col = matchcol - spp_skip->sp_offsets[SPO_LC_OFF]; + int r; + + if (lc_col < 0) + lc_col = 0; + regmatch.rmm_ic = spp_skip->sp_ic; + regmatch.regprog = spp_skip->sp_prog; + r = syn_regexec(®match, startpos->lnum, lc_col, + IF_SYN_TIME(&spp_skip->sp_time)); + spp_skip->sp_prog = regmatch.regprog; + if (r && regmatch.startpos[0].col + <= best_regmatch.startpos[0].col) + { + int line_len; + + // Add offset to skip pattern match + syn_add_end_off(&pos, ®match, spp_skip, SPO_ME_OFF, 1); + + // If the skip pattern goes on to the next line, there is no + // match with an end pattern in this line. + if (pos.lnum > startpos->lnum) + break; + + line = ml_get_buf(syn_buf, startpos->lnum, FALSE); + line_len = (int)STRLEN(line); + + // take care of an empty match or negative offset + if (pos.col <= matchcol) + ++matchcol; + else if (pos.col <= regmatch.endpos[0].col) + matchcol = pos.col; + else + // Be careful not to jump over the NUL at the end-of-line + for (matchcol = regmatch.endpos[0].col; + matchcol < line_len && matchcol < pos.col; + ++matchcol) + ; + + // if the skip pattern includes end-of-line, break here + if (matchcol >= line_len) + break; + + continue; // start with first end pattern again + } + } + + /* + * Match from start pattern to end pattern. + * Correct for match and highlight offset of end pattern. + */ + spp = &(SYN_ITEMS(syn_block)[best_idx]); + syn_add_end_off(m_endpos, &best_regmatch, spp, SPO_ME_OFF, 1); + // can't end before the start + if (m_endpos->lnum == startpos->lnum && m_endpos->col < startpos->col) + m_endpos->col = startpos->col; + + syn_add_end_off(end_endpos, &best_regmatch, spp, SPO_HE_OFF, 1); + // can't end before the start + if (end_endpos->lnum == startpos->lnum + && end_endpos->col < startpos->col) + end_endpos->col = startpos->col; + // can't end after the match + limit_pos(end_endpos, m_endpos); + + /* + * If the end group is highlighted differently, adjust the pointers. + */ + if (spp->sp_syn_match_id != spp->sp_syn.id && spp->sp_syn_match_id != 0) + { + *end_idx = best_idx; + if (spp->sp_off_flags & (1 << (SPO_RE_OFF + SPO_COUNT))) + { + hl_endpos->lnum = best_regmatch.endpos[0].lnum; + hl_endpos->col = best_regmatch.endpos[0].col; + } + else + { + hl_endpos->lnum = best_regmatch.startpos[0].lnum; + hl_endpos->col = best_regmatch.startpos[0].col; + } + hl_endpos->col += spp->sp_offsets[SPO_RE_OFF]; + + // can't end before the start + if (hl_endpos->lnum == startpos->lnum + && hl_endpos->col < startpos->col) + hl_endpos->col = startpos->col; + limit_pos(hl_endpos, m_endpos); + + // now the match ends where the highlighting ends, it is turned + // into the matchgroup for the end + *m_endpos = *hl_endpos; + } + else + { + *end_idx = 0; + *hl_endpos = *end_endpos; + } + + *flagsp = spp->sp_flags; + + had_match = TRUE; + break; + } + + // no match for an END pattern in this line + if (!had_match) + m_endpos->lnum = 0; + + restore_chartab(buf_chartab); + + // Remove external matches. + unref_extmatch(re_extmatch_in); + re_extmatch_in = NULL; +} + +/* + * Limit "pos" not to be after "limit". + */ + static void +limit_pos(lpos_T *pos, lpos_T *limit) +{ + if (pos->lnum > limit->lnum) + *pos = *limit; + else if (pos->lnum == limit->lnum && pos->col > limit->col) + pos->col = limit->col; +} + +/* + * Limit "pos" not to be after "limit", unless pos->lnum is zero. + */ + static void +limit_pos_zero( + lpos_T *pos, + lpos_T *limit) +{ + if (pos->lnum == 0) + *pos = *limit; + else + limit_pos(pos, limit); +} + +/* + * Add offset to matched text for end of match or highlight. + */ + static void +syn_add_end_off( + lpos_T *result, // returned position + regmmatch_T *regmatch, // start/end of match + synpat_T *spp, // matched pattern + int idx, // index of offset + int extra) // extra chars for offset to start +{ + int col; + int off; + char_u *base; + char_u *p; + + if (spp->sp_off_flags & (1 << idx)) + { + result->lnum = regmatch->startpos[0].lnum; + col = regmatch->startpos[0].col; + off = spp->sp_offsets[idx] + extra; + } + else + { + result->lnum = regmatch->endpos[0].lnum; + col = regmatch->endpos[0].col; + off = spp->sp_offsets[idx]; + } + // Don't go past the end of the line. Matters for "rs=e+2" when there + // is a matchgroup. Watch out for match with last NL in the buffer. + if (result->lnum > syn_buf->b_ml.ml_line_count) + col = 0; + else if (off != 0) + { + base = ml_get_buf(syn_buf, result->lnum, FALSE); + p = base + col; + if (off > 0) + { + while (off-- > 0 && *p != NUL) + MB_PTR_ADV(p); + } + else if (off < 0) + { + while (off++ < 0 && base < p) + MB_PTR_BACK(base, p); + } + col = (int)(p - base); + } + result->col = col; +} + +/* + * Add offset to matched text for start of match or highlight. + * Avoid resulting column to become negative. + */ + static void +syn_add_start_off( + lpos_T *result, // returned position + regmmatch_T *regmatch, // start/end of match + synpat_T *spp, + int idx, + int extra) // extra chars for offset to end +{ + int col; + int off; + char_u *base; + char_u *p; + + if (spp->sp_off_flags & (1 << (idx + SPO_COUNT))) + { + result->lnum = regmatch->endpos[0].lnum; + col = regmatch->endpos[0].col; + off = spp->sp_offsets[idx] + extra; + } + else + { + result->lnum = regmatch->startpos[0].lnum; + col = regmatch->startpos[0].col; + off = spp->sp_offsets[idx]; + } + if (result->lnum > syn_buf->b_ml.ml_line_count) + { + // a "\n" at the end of the pattern may take us below the last line + result->lnum = syn_buf->b_ml.ml_line_count; + col = (int)STRLEN(ml_get_buf(syn_buf, result->lnum, FALSE)); + } + if (off != 0) + { + base = ml_get_buf(syn_buf, result->lnum, FALSE); + p = base + col; + if (off > 0) + { + while (off-- && *p != NUL) + MB_PTR_ADV(p); + } + else if (off < 0) + { + while (off++ && base < p) + MB_PTR_BACK(base, p); + } + col = (int)(p - base); + } + result->col = col; +} + +/* + * Get current line in syntax buffer. + */ + static char_u * +syn_getcurline(void) +{ + return ml_get_buf(syn_buf, current_lnum, FALSE); +} + +/* + * Call vim_regexec() to find a match with "rmp" in "syn_buf". + * Returns TRUE when there is a match. + */ + static int +syn_regexec( + regmmatch_T *rmp, + linenr_T lnum, + colnr_T col, + syn_time_T *st UNUSED) +{ + int r; + int timed_out = FALSE; +#ifdef FEAT_PROFILE + proftime_T pt; + + if (syn_time_on) + profile_start(&pt); +#endif + + if (rmp->regprog == NULL) + // This can happen if a previous call to vim_regexec_multi() tried to + // use the NFA engine, which resulted in NFA_TOO_EXPENSIVE, and + // compiling the pattern with the other engine fails. + return FALSE; + + rmp->rmm_maxcol = syn_buf->b_p_smc; + r = vim_regexec_multi(rmp, syn_win, syn_buf, lnum, col, &timed_out); + +#ifdef FEAT_PROFILE + if (syn_time_on) + { + profile_end(&pt); + profile_add(&st->total, &pt); + if (profile_cmp(&pt, &st->slowest) < 0) + st->slowest = pt; + ++st->count; + if (r > 0) + ++st->match; + } +#endif +#ifdef FEAT_RELTIME + if (timed_out && redrawtime_limit_set && !syn_win->w_s->b_syn_slow) + { + syn_win->w_s->b_syn_slow = TRUE; + msg(_("'redrawtime' exceeded, syntax highlighting disabled")); + } +#endif + + if (r > 0) + { + rmp->startpos[0].lnum += lnum; + rmp->endpos[0].lnum += lnum; + return TRUE; + } + return FALSE; +} + +/* + * Check one position in a line for a matching keyword. + * The caller must check if a keyword can start at startcol. + * Return its ID if found, 0 otherwise. + */ + static int +check_keyword_id( + char_u *line, + int startcol, // position in line to check for keyword + int *endcolp, // return: character after found keyword + long *flagsp, // return: flags of matching keyword + short **next_listp, // return: next_list of matching keyword + stateitem_T *cur_si, // item at the top of the stack + int *ccharp UNUSED) // conceal substitution char +{ + keyentry_T *kp; + char_u *kwp; + int round; + int kwlen; + char_u keyword[MAXKEYWLEN + 1]; // assume max. keyword len is 80 + hashtab_T *ht; + hashitem_T *hi; + + // Find first character after the keyword. First character was already + // checked. + kwp = line + startcol; + kwlen = 0; + do + { + if (has_mbyte) + kwlen += (*mb_ptr2len)(kwp + kwlen); + else + ++kwlen; + } + while (vim_iswordp_buf(kwp + kwlen, syn_buf)); + + if (kwlen > MAXKEYWLEN) + return 0; + + /* + * Must make a copy of the keyword, so we can add a NUL and make it + * lowercase. + */ + vim_strncpy(keyword, kwp, kwlen); + + /* + * Try twice: + * 1. matching case + * 2. ignoring case + */ + for (round = 1; round <= 2; ++round) + { + ht = round == 1 ? &syn_block->b_keywtab : &syn_block->b_keywtab_ic; + if (ht->ht_used == 0) + continue; + if (round == 2) // ignore case + (void)str_foldcase(kwp, kwlen, keyword, MAXKEYWLEN + 1); + + /* + * Find keywords that match. There can be several with different + * attributes. + * When current_next_list is non-zero accept only that group, otherwise: + * Accept a not-contained keyword at toplevel. + * Accept a keyword at other levels only if it is in the contains list. + */ + hi = hash_find(ht, keyword); + if (!HASHITEM_EMPTY(hi)) + for (kp = HI2KE(hi); kp != NULL; kp = kp->ke_next) + { + if (current_next_list != 0 + ? in_id_list(NULL, current_next_list, &kp->k_syn, 0) + : (cur_si == NULL + ? !(kp->flags & HL_CONTAINED) + : in_id_list(cur_si, cur_si->si_cont_list, + &kp->k_syn, kp->flags & HL_CONTAINED))) + { + *endcolp = startcol + kwlen; + *flagsp = kp->flags; + *next_listp = kp->next_list; +#ifdef FEAT_CONCEAL + *ccharp = kp->k_char; +#endif + return kp->k_syn.id; + } + } + } + return 0; +} + +/* + * Handle ":syntax conceal" command. + */ + static void +syn_cmd_conceal(exarg_T *eap UNUSED, int syncing UNUSED) +{ +#ifdef FEAT_CONCEAL + char_u *arg = eap->arg; + char_u *next; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + next = skiptowhite(arg); + if (*arg == NUL) + { + if (curwin->w_s->b_syn_conceal) + msg("syntax conceal on"); + else + msg("syntax conceal off"); + } + else if (STRNICMP(arg, "on", 2) == 0 && next - arg == 2) + curwin->w_s->b_syn_conceal = TRUE; + else if (STRNICMP(arg, "off", 3) == 0 && next - arg == 3) + curwin->w_s->b_syn_conceal = FALSE; + else + semsg(_(e_illegal_argument_str_2), arg); +#endif +} + +/* + * Handle ":syntax case" command. + */ + static void +syn_cmd_case(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg = eap->arg; + char_u *next; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + next = skiptowhite(arg); + if (*arg == NUL) + { + if (curwin->w_s->b_syn_ic) + msg("syntax case ignore"); + else + msg("syntax case match"); + } + else if (STRNICMP(arg, "match", 5) == 0 && next - arg == 5) + curwin->w_s->b_syn_ic = FALSE; + else if (STRNICMP(arg, "ignore", 6) == 0 && next - arg == 6) + curwin->w_s->b_syn_ic = TRUE; + else + semsg(_(e_illegal_argument_str_2), arg); +} + +/* + * Handle ":syntax foldlevel" command. + */ + static void +syn_cmd_foldlevel(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg = eap->arg; + char_u *arg_end; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + if (*arg == NUL) + { + switch (curwin->w_s->b_syn_foldlevel) + { + case SYNFLD_START: msg("syntax foldlevel start"); break; + case SYNFLD_MINIMUM: msg("syntax foldlevel minimum"); break; + default: break; + } + return; + } + + arg_end = skiptowhite(arg); + if (STRNICMP(arg, "start", 5) == 0 && arg_end - arg == 5) + curwin->w_s->b_syn_foldlevel = SYNFLD_START; + else if (STRNICMP(arg, "minimum", 7) == 0 && arg_end - arg == 7) + curwin->w_s->b_syn_foldlevel = SYNFLD_MINIMUM; + else + { + semsg(_(e_illegal_argument_str_2), arg); + return; + } + + arg = skipwhite(arg_end); + if (*arg != NUL) + { + semsg(_(e_illegal_argument_str_2), arg); + } +} + +/* + * Handle ":syntax spell" command. + */ + static void +syn_cmd_spell(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg = eap->arg; + char_u *next; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + next = skiptowhite(arg); + if (*arg == NUL) + { + if (curwin->w_s->b_syn_spell == SYNSPL_TOP) + msg("syntax spell toplevel"); + else if (curwin->w_s->b_syn_spell == SYNSPL_NOTOP) + msg("syntax spell notoplevel"); + else + msg("syntax spell default"); + } + else if (STRNICMP(arg, "toplevel", 8) == 0 && next - arg == 8) + curwin->w_s->b_syn_spell = SYNSPL_TOP; + else if (STRNICMP(arg, "notoplevel", 10) == 0 && next - arg == 10) + curwin->w_s->b_syn_spell = SYNSPL_NOTOP; + else if (STRNICMP(arg, "default", 7) == 0 && next - arg == 7) + curwin->w_s->b_syn_spell = SYNSPL_DEFAULT; + else + { + semsg(_(e_illegal_argument_str_2), arg); + return; + } + + // assume spell checking changed, force a redraw + redraw_win_later(curwin, UPD_NOT_VALID); +} + +/* + * Handle ":syntax iskeyword" command. + */ + static void +syn_cmd_iskeyword(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg = eap->arg; + char_u save_chartab[32]; + char_u *save_isk; + + if (eap->skip) + return; + + arg = skipwhite(arg); + if (*arg == NUL) + { + msg_puts("\n"); + if (curwin->w_s->b_syn_isk != empty_option) + { + msg_puts("syntax iskeyword "); + msg_outtrans(curwin->w_s->b_syn_isk); + } + else + msg_outtrans((char_u *)_("syntax iskeyword not set")); + } + else + { + if (STRNICMP(arg, "clear", 5) == 0) + { + mch_memmove(curwin->w_s->b_syn_chartab, curbuf->b_chartab, + (size_t)32); + clear_string_option(&curwin->w_s->b_syn_isk); + } + else + { + mch_memmove(save_chartab, curbuf->b_chartab, (size_t)32); + save_isk = curbuf->b_p_isk; + curbuf->b_p_isk = vim_strsave(arg); + + buf_init_chartab(curbuf, FALSE); + mch_memmove(curwin->w_s->b_syn_chartab, curbuf->b_chartab, + (size_t)32); + mch_memmove(curbuf->b_chartab, save_chartab, (size_t)32); + clear_string_option(&curwin->w_s->b_syn_isk); + curwin->w_s->b_syn_isk = curbuf->b_p_isk; + curbuf->b_p_isk = save_isk; + } + } + redraw_win_later(curwin, UPD_NOT_VALID); +} + +/* + * Clear all syntax info for one buffer. + */ + void +syntax_clear(synblock_T *block) +{ + int i; + + block->b_syn_error = FALSE; // clear previous error +#ifdef FEAT_RELTIME + block->b_syn_slow = FALSE; // clear previous timeout +#endif + block->b_syn_ic = FALSE; // Use case, by default + block->b_syn_foldlevel = SYNFLD_START; + block->b_syn_spell = SYNSPL_DEFAULT; // default spell checking + block->b_syn_containedin = FALSE; +#ifdef FEAT_CONCEAL + block->b_syn_conceal = FALSE; +#endif + + // free the keywords + clear_keywtab(&block->b_keywtab); + clear_keywtab(&block->b_keywtab_ic); + + // free the syntax patterns + for (i = block->b_syn_patterns.ga_len; --i >= 0; ) + syn_clear_pattern(block, i); + ga_clear(&block->b_syn_patterns); + + // free the syntax clusters + for (i = block->b_syn_clusters.ga_len; --i >= 0; ) + syn_clear_cluster(block, i); + ga_clear(&block->b_syn_clusters); + block->b_spell_cluster_id = 0; + block->b_nospell_cluster_id = 0; + + block->b_syn_sync_flags = 0; + block->b_syn_sync_minlines = 0; + block->b_syn_sync_maxlines = 0; + block->b_syn_sync_linebreaks = 0; + + vim_regfree(block->b_syn_linecont_prog); + block->b_syn_linecont_prog = NULL; + VIM_CLEAR(block->b_syn_linecont_pat); +#ifdef FEAT_FOLDING + block->b_syn_folditems = 0; +#endif + clear_string_option(&block->b_syn_isk); + + // free the stored states + syn_stack_free_all(block); + invalidate_current_state(); + + // Reset the counter for ":syn include" + running_syn_inc_tag = 0; +} + +/* + * Get rid of ownsyntax for window "wp". + */ + void +reset_synblock(win_T *wp) +{ + if (wp->w_s != &wp->w_buffer->b_s) + { + syntax_clear(wp->w_s); + vim_free(wp->w_s); + wp->w_s = &wp->w_buffer->b_s; + } +} + +/* + * Clear syncing info for one buffer. + */ + static void +syntax_sync_clear(void) +{ + int i; + + // free the syntax patterns + for (i = curwin->w_s->b_syn_patterns.ga_len; --i >= 0; ) + if (SYN_ITEMS(curwin->w_s)[i].sp_syncing) + syn_remove_pattern(curwin->w_s, i); + + curwin->w_s->b_syn_sync_flags = 0; + curwin->w_s->b_syn_sync_minlines = 0; + curwin->w_s->b_syn_sync_maxlines = 0; + curwin->w_s->b_syn_sync_linebreaks = 0; + + vim_regfree(curwin->w_s->b_syn_linecont_prog); + curwin->w_s->b_syn_linecont_prog = NULL; + VIM_CLEAR(curwin->w_s->b_syn_linecont_pat); + clear_string_option(&curwin->w_s->b_syn_isk); + + syn_stack_free_all(curwin->w_s); // Need to recompute all syntax. +} + +/* + * Remove one pattern from the buffer's pattern list. + */ + static void +syn_remove_pattern( + synblock_T *block, + int idx) +{ + synpat_T *spp; + + spp = &(SYN_ITEMS(block)[idx]); +#ifdef FEAT_FOLDING + if (spp->sp_flags & HL_FOLD) + --block->b_syn_folditems; +#endif + syn_clear_pattern(block, idx); + mch_memmove(spp, spp + 1, + sizeof(synpat_T) * (block->b_syn_patterns.ga_len - idx - 1)); + --block->b_syn_patterns.ga_len; +} + +/* + * Clear and free one syntax pattern. When clearing all, must be called from + * last to first! + */ + static void +syn_clear_pattern(synblock_T *block, int i) +{ + vim_free(SYN_ITEMS(block)[i].sp_pattern); + vim_regfree(SYN_ITEMS(block)[i].sp_prog); + // Only free sp_cont_list and sp_next_list of first start pattern + if (i == 0 || SYN_ITEMS(block)[i - 1].sp_type != SPTYPE_START) + { + vim_free(SYN_ITEMS(block)[i].sp_cont_list); + vim_free(SYN_ITEMS(block)[i].sp_next_list); + vim_free(SYN_ITEMS(block)[i].sp_syn.cont_in_list); + } +} + +/* + * Clear and free one syntax cluster. + */ + static void +syn_clear_cluster(synblock_T *block, int i) +{ + vim_free(SYN_CLSTR(block)[i].scl_name); + vim_free(SYN_CLSTR(block)[i].scl_name_u); + vim_free(SYN_CLSTR(block)[i].scl_list); +} + +/* + * Handle ":syntax clear" command. + */ + static void +syn_cmd_clear(exarg_T *eap, int syncing) +{ + char_u *arg = eap->arg; + char_u *arg_end; + int id; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + /* + * We have to disable this within ":syn include @group filename", + * because otherwise @group would get deleted. + * Only required for Vim 5.x syntax files, 6.0 ones don't contain ":syn + * clear". + */ + if (curwin->w_s->b_syn_topgrp != 0) + return; + + if (ends_excmd2(eap->cmd, arg)) + { + /* + * No argument: Clear all syntax items. + */ + if (syncing) + syntax_sync_clear(); + else + { + syntax_clear(curwin->w_s); + if (curwin->w_s == &curwin->w_buffer->b_s) + do_unlet((char_u *)"b:current_syntax", TRUE); + do_unlet((char_u *)"w:current_syntax", TRUE); + } + } + else + { + /* + * Clear the group IDs that are in the argument. + */ + while (!ends_excmd2(eap->cmd, arg)) + { + arg_end = skiptowhite(arg); + if (*arg == '@') + { + id = syn_scl_namen2id(arg + 1, (int)(arg_end - arg - 1)); + if (id == 0) + { + semsg(_(e_no_such_syntax_cluster_str_1), arg); + break; + } + else + { + /* + * We can't physically delete a cluster without changing + * the IDs of other clusters, so we do the next best thing + * and make it empty. + */ + short scl_id = id - SYNID_CLUSTER; + + VIM_CLEAR(SYN_CLSTR(curwin->w_s)[scl_id].scl_list); + } + } + else + { + id = syn_namen2id(arg, (int)(arg_end - arg)); + if (id == 0) + { + semsg(_(e_no_such_highlight_group_name_str), arg); + break; + } + else + syn_clear_one(id, syncing); + } + arg = skipwhite(arg_end); + } + } + redraw_curbuf_later(UPD_SOME_VALID); + syn_stack_free_all(curwin->w_s); // Need to recompute all syntax. +} + +/* + * Clear one syntax group for the current buffer. + */ + static void +syn_clear_one(int id, int syncing) +{ + synpat_T *spp; + int idx; + + // Clear keywords only when not ":syn sync clear group-name" + if (!syncing) + { + (void)syn_clear_keyword(id, &curwin->w_s->b_keywtab); + (void)syn_clear_keyword(id, &curwin->w_s->b_keywtab_ic); + } + + // clear the patterns for "id" + for (idx = curwin->w_s->b_syn_patterns.ga_len; --idx >= 0; ) + { + spp = &(SYN_ITEMS(curwin->w_s)[idx]); + if (spp->sp_syn.id != id || spp->sp_syncing != syncing) + continue; + syn_remove_pattern(curwin->w_s, idx); + } +} + +/* + * Handle ":syntax on" command. + */ + static void +syn_cmd_on(exarg_T *eap, int syncing UNUSED) +{ + syn_cmd_onoff(eap, "syntax"); +} + +/* + * Handle ":syntax enable" command. + */ + static void +syn_cmd_enable(exarg_T *eap, int syncing UNUSED) +{ + set_internal_string_var((char_u *)"g:syntax_cmd", (char_u *)"enable"); + syn_cmd_onoff(eap, "syntax"); + do_unlet((char_u *)"g:syntax_cmd", TRUE); +} + +/* + * Handle ":syntax reset" command. + * It actually resets highlighting, not syntax. + */ + static void +syn_cmd_reset(exarg_T *eap, int syncing UNUSED) +{ + set_nextcmd(eap, eap->arg); + if (!eap->skip) + { + set_internal_string_var((char_u *)"g:syntax_cmd", (char_u *)"reset"); + do_cmdline_cmd((char_u *)"runtime! syntax/syncolor.vim"); + do_unlet((char_u *)"g:syntax_cmd", TRUE); + } +} + +/* + * Handle ":syntax manual" command. + */ + static void +syn_cmd_manual(exarg_T *eap, int syncing UNUSED) +{ + syn_cmd_onoff(eap, "manual"); +} + +/* + * Handle ":syntax off" command. + */ + static void +syn_cmd_off(exarg_T *eap, int syncing UNUSED) +{ + syn_cmd_onoff(eap, "nosyntax"); +} + + static void +syn_cmd_onoff(exarg_T *eap, char *name) +{ + char_u buf[100]; + + set_nextcmd(eap, eap->arg); + if (!eap->skip) + { + STRCPY(buf, "so "); + vim_snprintf((char *)buf + 3, sizeof(buf) - 3, SYNTAX_FNAME, name); + do_cmdline_cmd(buf); + } +} + +/* + * Handle ":syntax [list]" command: list current syntax words. + */ + static void +syn_cmd_list( + exarg_T *eap, + int syncing) // when TRUE: list syncing items +{ + char_u *arg = eap->arg; + int id; + char_u *arg_end; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + if (!syntax_present(curwin)) + { + msg(_(msg_no_items)); + return; + } + + if (syncing) + { + if (curwin->w_s->b_syn_sync_flags & SF_CCOMMENT) + { + msg_puts(_("syncing on C-style comments")); + syn_lines_msg(); + syn_match_msg(); + return; + } + else if (!(curwin->w_s->b_syn_sync_flags & SF_MATCH)) + { + if (curwin->w_s->b_syn_sync_minlines == 0) + msg_puts(_("no syncing")); + else + { + if (curwin->w_s->b_syn_sync_minlines == MAXLNUM) + msg_puts(_("syncing starts at the first line")); + else + { + msg_puts(_("syncing starts ")); + msg_outnum(curwin->w_s->b_syn_sync_minlines); + msg_puts(_(" lines before top line")); + } + syn_match_msg(); + } + return; + } + msg_puts_title(_("\n--- Syntax sync items ---")); + if (curwin->w_s->b_syn_sync_minlines > 0 + || curwin->w_s->b_syn_sync_maxlines > 0 + || curwin->w_s->b_syn_sync_linebreaks > 0) + { + msg_puts(_("\nsyncing on items")); + syn_lines_msg(); + syn_match_msg(); + } + } + else + msg_puts_title(_("\n--- Syntax items ---")); + if (ends_excmd2(eap->cmd, arg)) + { + /* + * No argument: List all group IDs and all syntax clusters. + */ + for (id = 1; id <= highlight_num_groups() && !got_int; ++id) + syn_list_one(id, syncing, FALSE); + for (id = 0; id < curwin->w_s->b_syn_clusters.ga_len && !got_int; ++id) + syn_list_cluster(id); + } + else + { + /* + * List the group IDs and syntax clusters that are in the argument. + */ + while (!ends_excmd2(eap->cmd, arg) && !got_int) + { + arg_end = skiptowhite(arg); + if (*arg == '@') + { + id = syn_scl_namen2id(arg + 1, (int)(arg_end - arg - 1)); + if (id == 0) + semsg(_(e_no_such_syntax_cluster_str_2), arg); + else + syn_list_cluster(id - SYNID_CLUSTER); + } + else + { + id = syn_namen2id(arg, (int)(arg_end - arg)); + if (id == 0) + semsg(_(e_no_such_highlight_group_name_str), arg); + else + syn_list_one(id, syncing, TRUE); + } + arg = skipwhite(arg_end); + } + } + set_nextcmd(eap, arg); +} + + static void +syn_lines_msg(void) +{ + if (curwin->w_s->b_syn_sync_maxlines > 0 + || curwin->w_s->b_syn_sync_minlines > 0) + { + msg_puts("; "); + if (curwin->w_s->b_syn_sync_minlines == MAXLNUM) + msg_puts(_("from the first line")); + else + { + if (curwin->w_s->b_syn_sync_minlines > 0) + { + msg_puts(_("minimal ")); + msg_outnum(curwin->w_s->b_syn_sync_minlines); + if (curwin->w_s->b_syn_sync_maxlines) + msg_puts(", "); + } + if (curwin->w_s->b_syn_sync_maxlines > 0) + { + msg_puts(_("maximal ")); + msg_outnum(curwin->w_s->b_syn_sync_maxlines); + } + msg_puts(_(" lines before top line")); + } + } +} + + static void +syn_match_msg(void) +{ + if (curwin->w_s->b_syn_sync_linebreaks > 0) + { + msg_puts(_("; match ")); + msg_outnum(curwin->w_s->b_syn_sync_linebreaks); + msg_puts(_(" line breaks")); + } +} + +static int last_matchgroup; + +struct name_list +{ + int flag; + char *name; +}; + +static void syn_list_flags(struct name_list *nl, int flags, int attr); + +/* + * List one syntax item, for ":syntax" or "syntax list syntax_name". + */ + static void +syn_list_one( + int id, + int syncing, // when TRUE: list syncing items + int link_only) // when TRUE; list link-only too +{ + int attr; + int idx; + int did_header = FALSE; + synpat_T *spp; + static struct name_list namelist1[] = + { + {HL_DISPLAY, "display"}, + {HL_CONTAINED, "contained"}, + {HL_ONELINE, "oneline"}, + {HL_KEEPEND, "keepend"}, + {HL_EXTEND, "extend"}, + {HL_EXCLUDENL, "excludenl"}, + {HL_TRANSP, "transparent"}, + {HL_FOLD, "fold"}, +#ifdef FEAT_CONCEAL + {HL_CONCEAL, "conceal"}, + {HL_CONCEALENDS, "concealends"}, +#endif + {0, NULL} + }; + static struct name_list namelist2[] = + { + {HL_SKIPWHITE, "skipwhite"}, + {HL_SKIPNL, "skipnl"}, + {HL_SKIPEMPTY, "skipempty"}, + {0, NULL} + }; + + attr = HL_ATTR(HLF_D); // highlight like directories + + // list the keywords for "id" + if (!syncing) + { + did_header = syn_list_keywords(id, &curwin->w_s->b_keywtab, FALSE, attr); + did_header = syn_list_keywords(id, &curwin->w_s->b_keywtab_ic, + did_header, attr); + } + + // list the patterns for "id" + for (idx = 0; idx < curwin->w_s->b_syn_patterns.ga_len && !got_int; ++idx) + { + spp = &(SYN_ITEMS(curwin->w_s)[idx]); + if (spp->sp_syn.id != id || spp->sp_syncing != syncing) + continue; + + (void)syn_list_header(did_header, 999, id); + did_header = TRUE; + last_matchgroup = 0; + if (spp->sp_type == SPTYPE_MATCH) + { + put_pattern("match", ' ', spp, attr); + msg_putchar(' '); + } + else if (spp->sp_type == SPTYPE_START) + { + while (SYN_ITEMS(curwin->w_s)[idx].sp_type == SPTYPE_START) + put_pattern("start", '=', &SYN_ITEMS(curwin->w_s)[idx++], attr); + if (SYN_ITEMS(curwin->w_s)[idx].sp_type == SPTYPE_SKIP) + put_pattern("skip", '=', &SYN_ITEMS(curwin->w_s)[idx++], attr); + while (idx < curwin->w_s->b_syn_patterns.ga_len + && SYN_ITEMS(curwin->w_s)[idx].sp_type == SPTYPE_END) + put_pattern("end", '=', &SYN_ITEMS(curwin->w_s)[idx++], attr); + --idx; + msg_putchar(' '); + } + syn_list_flags(namelist1, spp->sp_flags, attr); + + if (spp->sp_cont_list != NULL) + put_id_list((char_u *)"contains", spp->sp_cont_list, attr); + + if (spp->sp_syn.cont_in_list != NULL) + put_id_list((char_u *)"containedin", + spp->sp_syn.cont_in_list, attr); + + if (spp->sp_next_list != NULL) + { + put_id_list((char_u *)"nextgroup", spp->sp_next_list, attr); + syn_list_flags(namelist2, spp->sp_flags, attr); + } + if (spp->sp_flags & (HL_SYNC_HERE|HL_SYNC_THERE)) + { + if (spp->sp_flags & HL_SYNC_HERE) + msg_puts_attr("grouphere", attr); + else + msg_puts_attr("groupthere", attr); + msg_putchar(' '); + if (spp->sp_sync_idx >= 0) + msg_outtrans(highlight_group_name(SYN_ITEMS(curwin->w_s) + [spp->sp_sync_idx].sp_syn.id - 1)); + else + msg_puts("NONE"); + msg_putchar(' '); + } + } + + // list the link, if there is one + if (highlight_link_id(id - 1) && (did_header || link_only) && !got_int) + { + (void)syn_list_header(did_header, 999, id); + msg_puts_attr("links to", attr); + msg_putchar(' '); + msg_outtrans(highlight_group_name(highlight_link_id(id - 1) - 1)); + } +} + + static void +syn_list_flags(struct name_list *nlist, int flags, int attr) +{ + int i; + + for (i = 0; nlist[i].flag != 0; ++i) + if (flags & nlist[i].flag) + { + msg_puts_attr(nlist[i].name, attr); + msg_putchar(' '); + } +} + +/* + * List one syntax cluster, for ":syntax" or "syntax list syntax_name". + */ + static void +syn_list_cluster(int id) +{ + int endcol = 15; + + // slight hack: roughly duplicate the guts of syn_list_header() + msg_putchar('\n'); + msg_outtrans(SYN_CLSTR(curwin->w_s)[id].scl_name); + + if (msg_col >= endcol) // output at least one space + endcol = msg_col + 1; + if (Columns <= endcol) // avoid hang for tiny window + endcol = Columns - 1; + + msg_advance(endcol); + if (SYN_CLSTR(curwin->w_s)[id].scl_list != NULL) + { + put_id_list((char_u *)"cluster", SYN_CLSTR(curwin->w_s)[id].scl_list, + HL_ATTR(HLF_D)); + } + else + { + msg_puts_attr("cluster", HL_ATTR(HLF_D)); + msg_puts("=NONE"); + } +} + + static void +put_id_list(char_u *name, short *list, int attr) +{ + short *p; + + msg_puts_attr((char *)name, attr); + msg_putchar('='); + for (p = list; *p; ++p) + { + if (*p >= SYNID_ALLBUT && *p < SYNID_TOP) + { + if (p[1]) + msg_puts("ALLBUT"); + else + msg_puts("ALL"); + } + else if (*p >= SYNID_TOP && *p < SYNID_CONTAINED) + { + msg_puts("TOP"); + } + else if (*p >= SYNID_CONTAINED && *p < SYNID_CLUSTER) + { + msg_puts("CONTAINED"); + } + else if (*p >= SYNID_CLUSTER) + { + short scl_id = *p - SYNID_CLUSTER; + + msg_putchar('@'); + msg_outtrans(SYN_CLSTR(curwin->w_s)[scl_id].scl_name); + } + else + msg_outtrans(highlight_group_name(*p - 1)); + if (p[1]) + msg_putchar(','); + } + msg_putchar(' '); +} + + static void +put_pattern( + char *s, + int c, + synpat_T *spp, + int attr) +{ + long n; + int mask; + int first; + static char *sepchars = "/+=-#@\"|'^&"; + int i; + + // May have to write "matchgroup=group" + if (last_matchgroup != spp->sp_syn_match_id) + { + last_matchgroup = spp->sp_syn_match_id; + msg_puts_attr("matchgroup", attr); + msg_putchar('='); + if (last_matchgroup == 0) + msg_outtrans((char_u *)"NONE"); + else + msg_outtrans(highlight_group_name(last_matchgroup - 1)); + msg_putchar(' '); + } + + // output the name of the pattern and an '=' or ' ' + msg_puts_attr(s, attr); + msg_putchar(c); + + // output the pattern, in between a char that is not in the pattern + for (i = 0; vim_strchr(spp->sp_pattern, sepchars[i]) != NULL; ) + if (sepchars[++i] == NUL) + { + i = 0; // no good char found, just use the first one + break; + } + msg_putchar(sepchars[i]); + msg_outtrans(spp->sp_pattern); + msg_putchar(sepchars[i]); + + // output any pattern options + first = TRUE; + for (i = 0; i < SPO_COUNT; ++i) + { + mask = (1 << i); + if (spp->sp_off_flags & (mask + (mask << SPO_COUNT))) + { + if (!first) + msg_putchar(','); // separate with commas + msg_puts(spo_name_tab[i]); + n = spp->sp_offsets[i]; + if (i != SPO_LC_OFF) + { + if (spp->sp_off_flags & mask) + msg_putchar('s'); + else + msg_putchar('e'); + if (n > 0) + msg_putchar('+'); + } + if (n || i == SPO_LC_OFF) + msg_outnum(n); + first = FALSE; + } + } + msg_putchar(' '); +} + +/* + * List or clear the keywords for one syntax group. + * Return TRUE if the header has been printed. + */ + static int +syn_list_keywords( + int id, + hashtab_T *ht, + int did_header, // header has already been printed + int attr) +{ + int outlen; + hashitem_T *hi; + keyentry_T *kp; + int todo; + int prev_contained = 0; + short *prev_next_list = NULL; + short *prev_cont_in_list = NULL; + int prev_skipnl = 0; + int prev_skipwhite = 0; + int prev_skipempty = 0; + + /* + * Unfortunately, this list of keywords is not sorted on alphabet but on + * hash value... + */ + todo = (int)ht->ht_used; + for (hi = ht->ht_array; todo > 0 && !got_int; ++hi) + { + if (!HASHITEM_EMPTY(hi)) + { + --todo; + for (kp = HI2KE(hi); kp != NULL && !got_int; kp = kp->ke_next) + { + if (kp->k_syn.id == id) + { + if (prev_contained != (kp->flags & HL_CONTAINED) + || prev_skipnl != (kp->flags & HL_SKIPNL) + || prev_skipwhite != (kp->flags & HL_SKIPWHITE) + || prev_skipempty != (kp->flags & HL_SKIPEMPTY) + || prev_cont_in_list != kp->k_syn.cont_in_list + || prev_next_list != kp->next_list) + outlen = 9999; + else + outlen = (int)STRLEN(kp->keyword); + // output "contained" and "nextgroup" on each line + if (syn_list_header(did_header, outlen, id)) + { + prev_contained = 0; + prev_next_list = NULL; + prev_cont_in_list = NULL; + prev_skipnl = 0; + prev_skipwhite = 0; + prev_skipempty = 0; + } + did_header = TRUE; + if (prev_contained != (kp->flags & HL_CONTAINED)) + { + msg_puts_attr("contained", attr); + msg_putchar(' '); + prev_contained = (kp->flags & HL_CONTAINED); + } + if (kp->k_syn.cont_in_list != prev_cont_in_list) + { + put_id_list((char_u *)"containedin", + kp->k_syn.cont_in_list, attr); + msg_putchar(' '); + prev_cont_in_list = kp->k_syn.cont_in_list; + } + if (kp->next_list != prev_next_list) + { + put_id_list((char_u *)"nextgroup", kp->next_list, attr); + msg_putchar(' '); + prev_next_list = kp->next_list; + if (kp->flags & HL_SKIPNL) + { + msg_puts_attr("skipnl", attr); + msg_putchar(' '); + prev_skipnl = (kp->flags & HL_SKIPNL); + } + if (kp->flags & HL_SKIPWHITE) + { + msg_puts_attr("skipwhite", attr); + msg_putchar(' '); + prev_skipwhite = (kp->flags & HL_SKIPWHITE); + } + if (kp->flags & HL_SKIPEMPTY) + { + msg_puts_attr("skipempty", attr); + msg_putchar(' '); + prev_skipempty = (kp->flags & HL_SKIPEMPTY); + } + } + msg_outtrans(kp->keyword); + } + } + } + } + + return did_header; +} + + static void +syn_clear_keyword(int id, hashtab_T *ht) +{ + hashitem_T *hi; + keyentry_T *kp; + keyentry_T *kp_prev; + keyentry_T *kp_next; + int todo; + + hash_lock(ht); + todo = (int)ht->ht_used; + FOR_ALL_HASHTAB_ITEMS(ht, hi, todo) + { + if (!HASHITEM_EMPTY(hi)) + { + --todo; + kp_prev = NULL; + for (kp = HI2KE(hi); kp != NULL; ) + { + if (kp->k_syn.id == id) + { + kp_next = kp->ke_next; + if (kp_prev == NULL) + { + if (kp_next == NULL) + hash_remove(ht, hi, "syntax clear keyword"); + else + hi->hi_key = KE2HIKEY(kp_next); + } + else + kp_prev->ke_next = kp_next; + vim_free(kp->next_list); + vim_free(kp->k_syn.cont_in_list); + vim_free(kp); + kp = kp_next; + } + else + { + kp_prev = kp; + kp = kp->ke_next; + } + } + } + } + hash_unlock(ht); +} + +/* + * Clear a whole keyword table. + */ + static void +clear_keywtab(hashtab_T *ht) +{ + hashitem_T *hi; + int todo; + keyentry_T *kp; + keyentry_T *kp_next; + + todo = (int)ht->ht_used; + FOR_ALL_HASHTAB_ITEMS(ht, hi, todo) + { + if (!HASHITEM_EMPTY(hi)) + { + --todo; + for (kp = HI2KE(hi); kp != NULL; kp = kp_next) + { + kp_next = kp->ke_next; + vim_free(kp->next_list); + vim_free(kp->k_syn.cont_in_list); + vim_free(kp); + } + } + } + hash_clear(ht); + hash_init(ht); +} + +/* + * Add a keyword to the list of keywords. + */ + static void +add_keyword( + char_u *name, // name of keyword + int id, // group ID for this keyword + int flags, // flags for this keyword + short *cont_in_list, // containedin for this keyword + short *next_list, // nextgroup for this keyword + int conceal_char) +{ + keyentry_T *kp; + hashtab_T *ht; + hashitem_T *hi; + char_u *name_ic; + long_u hash; + char_u name_folded[MAXKEYWLEN + 1]; + + if (curwin->w_s->b_syn_ic) + name_ic = str_foldcase(name, (int)STRLEN(name), + name_folded, MAXKEYWLEN + 1); + else + name_ic = name; + kp = alloc(offsetof(keyentry_T, keyword) + STRLEN(name_ic) + 1); + if (kp == NULL) + return; + STRCPY(kp->keyword, name_ic); + kp->k_syn.id = id; + kp->k_syn.inc_tag = current_syn_inc_tag; + kp->flags = flags; + kp->k_char = conceal_char; + kp->k_syn.cont_in_list = copy_id_list(cont_in_list); + if (cont_in_list != NULL) + curwin->w_s->b_syn_containedin = TRUE; + kp->next_list = copy_id_list(next_list); + + if (curwin->w_s->b_syn_ic) + ht = &curwin->w_s->b_keywtab_ic; + else + ht = &curwin->w_s->b_keywtab; + + hash = hash_hash(kp->keyword); + hi = hash_lookup(ht, kp->keyword, hash); + if (HASHITEM_EMPTY(hi)) + { + // new keyword, add to hashtable + kp->ke_next = NULL; + hash_add_item(ht, hi, kp->keyword, hash); + } + else + { + // keyword already exists, prepend to list + kp->ke_next = HI2KE(hi); + hi->hi_key = KE2HIKEY(kp); + } +} + +/* + * Get the start and end of the group name argument. + * Return a pointer to the first argument. + * Return NULL if the end of the command was found instead of further args. + */ + static char_u * +get_group_name( + char_u *arg, // start of the argument + char_u **name_end) // pointer to end of the name +{ + char_u *rest; + + *name_end = skiptowhite(arg); + rest = skipwhite(*name_end); + + /* + * Check if there are enough arguments. The first argument may be a + * pattern, where '|' is allowed, so only check for NUL. + */ + if (ends_excmd(*arg) || *rest == NUL) + return NULL; + return rest; +} + +/* + * Check for syntax command option arguments. + * This can be called at any place in the list of arguments, and just picks + * out the arguments that are known. Can be called several times in a row to + * collect all options in between other arguments. + * Return a pointer to the next argument (which isn't an option). + * Return NULL for any error; + */ + static char_u * +get_syn_options( + char_u *start, // next argument to be checked + syn_opt_arg_T *opt, // various things + int *conceal_char UNUSED, + int skip) // TRUE if skipping over command +{ + char_u *arg = start; + char_u *gname_start, *gname; + int syn_id; + int len; + char *p; + int i; + int fidx; + static struct flag + { + char *name; + int argtype; + int flags; + } flagtab[] = { {"cCoOnNtTaAiInNeEdD", 0, HL_CONTAINED}, + {"oOnNeElLiInNeE", 0, HL_ONELINE}, + {"kKeEeEpPeEnNdD", 0, HL_KEEPEND}, + {"eExXtTeEnNdD", 0, HL_EXTEND}, + {"eExXcClLuUdDeEnNlL", 0, HL_EXCLUDENL}, + {"tTrRaAnNsSpPaArReEnNtT", 0, HL_TRANSP}, + {"sSkKiIpPnNlL", 0, HL_SKIPNL}, + {"sSkKiIpPwWhHiItTeE", 0, HL_SKIPWHITE}, + {"sSkKiIpPeEmMpPtTyY", 0, HL_SKIPEMPTY}, + {"gGrRoOuUpPhHeErReE", 0, HL_SYNC_HERE}, + {"gGrRoOuUpPtThHeErReE", 0, HL_SYNC_THERE}, + {"dDiIsSpPlLaAyY", 0, HL_DISPLAY}, + {"fFoOlLdD", 0, HL_FOLD}, + {"cCoOnNcCeEaAlL", 0, HL_CONCEAL}, + {"cCoOnNcCeEaAlLeEnNdDsS", 0, HL_CONCEALENDS}, + {"cCcChHaArR", 11, 0}, + {"cCoOnNtTaAiInNsS", 1, 0}, + {"cCoOnNtTaAiInNeEdDiInN", 2, 0}, + {"nNeExXtTgGrRoOuUpP", 3, 0}, + }; + static char *first_letters = "cCoOkKeEtTsSgGdDfFnN"; + + if (arg == NULL) // already detected error + return NULL; + +#ifdef FEAT_CONCEAL + if (curwin->w_s->b_syn_conceal) + opt->flags |= HL_CONCEAL; +#endif + + for (;;) + { + /* + * This is used very often when a large number of keywords is defined. + * Need to skip quickly when no option name is found. + * Also avoid tolower(), it's slow. + */ + if (strchr(first_letters, *arg) == NULL) + break; + + for (fidx = ARRAY_LENGTH(flagtab); --fidx >= 0; ) + { + p = flagtab[fidx].name; + for (i = 0, len = 0; p[i] != NUL; i += 2, ++len) + if (arg[len] != p[i] && arg[len] != p[i + 1]) + break; + if (p[i] == NUL && (VIM_ISWHITE(arg[len]) + || (flagtab[fidx].argtype > 0 + ? arg[len] == '=' + : ends_excmd2(start, arg + len)))) + { + if (opt->keyword + && (flagtab[fidx].flags == HL_DISPLAY + || flagtab[fidx].flags == HL_FOLD + || flagtab[fidx].flags == HL_EXTEND)) + // treat "display", "fold" and "extend" as a keyword + fidx = -1; + break; + } + } + if (fidx < 0) // no match found + break; + + if (flagtab[fidx].argtype == 1) + { + if (!opt->has_cont_list) + { + emsg(_(e_contains_argument_not_accepted_here)); + return NULL; + } + if (get_id_list(&arg, 8, &opt->cont_list, skip) == FAIL) + return NULL; + } + else if (flagtab[fidx].argtype == 2) + { + if (get_id_list(&arg, 11, &opt->cont_in_list, skip) == FAIL) + return NULL; + } + else if (flagtab[fidx].argtype == 3) + { + if (get_id_list(&arg, 9, &opt->next_list, skip) == FAIL) + return NULL; + } + else if (flagtab[fidx].argtype == 11 && arg[5] == '=') + { + // cchar=? + if (has_mbyte) + { +#ifdef FEAT_CONCEAL + *conceal_char = mb_ptr2char(arg + 6); +#endif + arg += mb_ptr2len(arg + 6) - 1; + } + else + { +#ifdef FEAT_CONCEAL + *conceal_char = arg[6]; +#else + ; +#endif + } +#ifdef FEAT_CONCEAL + if (!vim_isprintc_strict(*conceal_char)) + { + emsg(_(e_invalid_cchar_value)); + return NULL; + } +#endif + arg = skipwhite(arg + 7); + } + else + { + opt->flags |= flagtab[fidx].flags; + arg = skipwhite(arg + len); + + if (flagtab[fidx].flags == HL_SYNC_HERE + || flagtab[fidx].flags == HL_SYNC_THERE) + { + if (opt->sync_idx == NULL) + { + emsg(_(e_groupthere_not_accepted_here)); + return NULL; + } + gname_start = arg; + arg = skiptowhite(arg); + if (gname_start == arg) + return NULL; + gname = vim_strnsave(gname_start, arg - gname_start); + if (gname == NULL) + return NULL; + if (STRCMP(gname, "NONE") == 0) + *opt->sync_idx = NONE_IDX; + else + { + syn_id = syn_name2id(gname); + for (i = curwin->w_s->b_syn_patterns.ga_len; --i >= 0; ) + if (SYN_ITEMS(curwin->w_s)[i].sp_syn.id == syn_id + && SYN_ITEMS(curwin->w_s)[i].sp_type + == SPTYPE_START) + { + *opt->sync_idx = i; + break; + } + if (i < 0) + { + semsg(_(e_didnt_find_region_item_for_str), gname); + vim_free(gname); + return NULL; + } + } + + vim_free(gname); + arg = skipwhite(arg); + } +#ifdef FEAT_FOLDING + else if (flagtab[fidx].flags == HL_FOLD + && foldmethodIsSyntax(curwin)) + // Need to update folds later. + foldUpdateAll(curwin); +#endif + } + } + + return arg; +} + +/* + * Adjustments to syntax item when declared in a ":syn include"'d file. + * Set the contained flag, and if the item is not already contained, add it + * to the specified top-level group, if any. + */ + static void +syn_incl_toplevel(int id, int *flagsp) +{ + if ((*flagsp & HL_CONTAINED) || curwin->w_s->b_syn_topgrp == 0) + return; + *flagsp |= HL_CONTAINED; + if (curwin->w_s->b_syn_topgrp >= SYNID_CLUSTER) + { + // We have to alloc this, because syn_combine_list() will free it. + short *grp_list = ALLOC_MULT(short, 2); + int tlg_id = curwin->w_s->b_syn_topgrp - SYNID_CLUSTER; + + if (grp_list != NULL) + { + grp_list[0] = id; + grp_list[1] = 0; + syn_combine_list(&SYN_CLSTR(curwin->w_s)[tlg_id].scl_list, + &grp_list, CLUSTER_ADD); + } + } +} + +/* + * Handle ":syntax include [@{group-name}] filename" command. + */ + static void +syn_cmd_include(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg = eap->arg; + int sgl_id = 1; + char_u *group_name_end; + char_u *rest; + char *errormsg = NULL; + int prev_toplvl_grp; + int prev_syn_inc_tag; + int source = FALSE; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + if (arg[0] == '@') + { + ++arg; + rest = get_group_name(arg, &group_name_end); + if (rest == NULL) + { + emsg(_(e_filename_required)); + return; + } + sgl_id = syn_check_cluster(arg, (int)(group_name_end - arg)); + if (sgl_id == 0) + return; + // separate_nextcmd() and expand_filename() depend on this + eap->arg = rest; + } + + /* + * Everything that's left, up to the next command, should be the + * filename to include. + */ + eap->argt |= (EX_XFILE | EX_NOSPC); + separate_nextcmd(eap, FALSE); + if (*eap->arg == '<' || *eap->arg == '$' || mch_isFullName(eap->arg)) + { + // For an absolute path, "$VIM/..." or "<sfile>.." we ":source" the + // file. Need to expand the file name first. In other cases + // ":runtime!" is used. + source = TRUE; + if (expand_filename(eap, syn_cmdlinep, &errormsg) == FAIL) + { + if (errormsg != NULL) + emsg(errormsg); + return; + } + } + + /* + * Save and restore the existing top-level grouplist id and ":syn + * include" tag around the actual inclusion. + */ + if (running_syn_inc_tag >= MAX_SYN_INC_TAG) + { + emsg(_(e_too_many_syntax_includes)); + return; + } + prev_syn_inc_tag = current_syn_inc_tag; + current_syn_inc_tag = ++running_syn_inc_tag; + prev_toplvl_grp = curwin->w_s->b_syn_topgrp; + curwin->w_s->b_syn_topgrp = sgl_id; + if (source ? do_source(eap->arg, FALSE, DOSO_NONE, NULL) == FAIL + : source_runtime(eap->arg, DIP_ALL) == FAIL) + semsg(_(e_cant_open_file_str), eap->arg); + curwin->w_s->b_syn_topgrp = prev_toplvl_grp; + current_syn_inc_tag = prev_syn_inc_tag; +} + +/* + * Handle ":syntax keyword {group-name} [{option}] keyword .." command. + */ + static void +syn_cmd_keyword(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg = eap->arg; + char_u *group_name_end; + int syn_id; + char_u *rest; + char_u *keyword_copy = NULL; + char_u *p; + char_u *kw; + syn_opt_arg_T syn_opt_arg; + int cnt; + int conceal_char = NUL; + + rest = get_group_name(arg, &group_name_end); + + if (rest != NULL) + { + if (eap->skip) + syn_id = -1; + else + syn_id = syn_check_group(arg, (int)(group_name_end - arg)); + if (syn_id != 0) + // allocate a buffer, for removing backslashes in the keyword + keyword_copy = alloc(STRLEN(rest) + 1); + if (keyword_copy != NULL) + { + syn_opt_arg.flags = 0; + syn_opt_arg.keyword = TRUE; + syn_opt_arg.sync_idx = NULL; + syn_opt_arg.has_cont_list = FALSE; + syn_opt_arg.cont_in_list = NULL; + syn_opt_arg.next_list = NULL; + + /* + * The options given apply to ALL keywords, so all options must be + * found before keywords can be created. + * 1: collect the options and copy the keywords to keyword_copy. + */ + cnt = 0; + p = keyword_copy; + for ( ; rest != NULL && !ends_excmd2(eap->arg, rest); + rest = skipwhite(rest)) + { + rest = get_syn_options(rest, &syn_opt_arg, &conceal_char, + eap->skip); + if (rest == NULL || ends_excmd2(eap->arg, rest)) + break; + // Copy the keyword, removing backslashes, and add a NUL. + while (*rest != NUL && !VIM_ISWHITE(*rest)) + { + if (*rest == '\\' && rest[1] != NUL) + ++rest; + *p++ = *rest++; + } + *p++ = NUL; + ++cnt; + } + + if (!eap->skip) + { + // Adjust flags for use of ":syn include". + syn_incl_toplevel(syn_id, &syn_opt_arg.flags); + + /* + * 2: Add an entry for each keyword. + */ + for (kw = keyword_copy; --cnt >= 0; kw += STRLEN(kw) + 1) + { + for (p = vim_strchr(kw, '['); ; ) + { + if (p != NULL) + *p = NUL; + add_keyword(kw, syn_id, syn_opt_arg.flags, + syn_opt_arg.cont_in_list, + syn_opt_arg.next_list, conceal_char); + if (p == NULL) + break; + if (p[1] == NUL) + { + semsg(_(e_error_missing_rsb_str), kw); + goto error; + } + if (p[1] == ']') + { + if (p[2] != NUL) + { + semsg(_(e_trailing_char_after_rsb_str_str), + kw, &p[2]); + goto error; + } + kw = p + 1; // skip over the "]" + break; + } + if (has_mbyte) + { + int l = (*mb_ptr2len)(p + 1); + + mch_memmove(p, p + 1, l); + p += l; + } + else + { + p[0] = p[1]; + ++p; + } + } + } + } +error: + vim_free(keyword_copy); + vim_free(syn_opt_arg.cont_in_list); + vim_free(syn_opt_arg.next_list); + } + } + + if (rest != NULL) + set_nextcmd(eap, rest); + else + semsg(_(e_invalid_argument_str), arg); + + redraw_curbuf_later(UPD_SOME_VALID); + syn_stack_free_all(curwin->w_s); // Need to recompute all syntax. +} + +/* + * Handle ":syntax match {name} [{options}] {pattern} [{options}]". + * + * Also ":syntax sync match {name} [[grouphere | groupthere] {group-name}] .." + */ + static void +syn_cmd_match( + exarg_T *eap, + int syncing) // TRUE for ":syntax sync match .. " +{ + char_u *arg = eap->arg; + char_u *group_name_end; + char_u *rest; + synpat_T item; // the item found in the line + int syn_id; + int idx; + syn_opt_arg_T syn_opt_arg; + int sync_idx = 0; + int conceal_char = NUL; + int orig_called_emsg = called_emsg; + + // Isolate the group name, check for validity + rest = get_group_name(arg, &group_name_end); + + // Get options before the pattern + syn_opt_arg.flags = 0; + syn_opt_arg.keyword = FALSE; + syn_opt_arg.sync_idx = syncing ? &sync_idx : NULL; + syn_opt_arg.has_cont_list = TRUE; + syn_opt_arg.cont_list = NULL; + syn_opt_arg.cont_in_list = NULL; + syn_opt_arg.next_list = NULL; + rest = get_syn_options(rest, &syn_opt_arg, &conceal_char, eap->skip); + + // get the pattern. + init_syn_patterns(); + CLEAR_FIELD(item); + rest = get_syn_pattern(rest, &item); + if (vim_regcomp_had_eol() && !(syn_opt_arg.flags & HL_EXCLUDENL)) + syn_opt_arg.flags |= HL_HAS_EOL; + + // Get options after the pattern + rest = get_syn_options(rest, &syn_opt_arg, &conceal_char, eap->skip); + + if (rest != NULL) // all arguments are valid + { + /* + * Check for trailing command and illegal trailing arguments. + */ + set_nextcmd(eap, rest); + if (!ends_excmd2(eap->cmd, rest) || eap->skip) + rest = NULL; + else if (ga_grow(&curwin->w_s->b_syn_patterns, 1) == OK + && (syn_id = syn_check_group(arg, + (int)(group_name_end - arg))) != 0) + { + syn_incl_toplevel(syn_id, &syn_opt_arg.flags); + /* + * Store the pattern in the syn_items list + */ + idx = curwin->w_s->b_syn_patterns.ga_len; + SYN_ITEMS(curwin->w_s)[idx] = item; + SYN_ITEMS(curwin->w_s)[idx].sp_syncing = syncing; + SYN_ITEMS(curwin->w_s)[idx].sp_type = SPTYPE_MATCH; + SYN_ITEMS(curwin->w_s)[idx].sp_syn.id = syn_id; + SYN_ITEMS(curwin->w_s)[idx].sp_syn.inc_tag = current_syn_inc_tag; + SYN_ITEMS(curwin->w_s)[idx].sp_flags = syn_opt_arg.flags; + SYN_ITEMS(curwin->w_s)[idx].sp_sync_idx = sync_idx; + SYN_ITEMS(curwin->w_s)[idx].sp_cont_list = syn_opt_arg.cont_list; + SYN_ITEMS(curwin->w_s)[idx].sp_syn.cont_in_list = + syn_opt_arg.cont_in_list; +#ifdef FEAT_CONCEAL + SYN_ITEMS(curwin->w_s)[idx].sp_cchar = conceal_char; +#endif + if (syn_opt_arg.cont_in_list != NULL) + curwin->w_s->b_syn_containedin = TRUE; + SYN_ITEMS(curwin->w_s)[idx].sp_next_list = syn_opt_arg.next_list; + ++curwin->w_s->b_syn_patterns.ga_len; + + // remember that we found a match for syncing on + if (syn_opt_arg.flags & (HL_SYNC_HERE|HL_SYNC_THERE)) + curwin->w_s->b_syn_sync_flags |= SF_MATCH; +#ifdef FEAT_FOLDING + if (syn_opt_arg.flags & HL_FOLD) + ++curwin->w_s->b_syn_folditems; +#endif + + redraw_curbuf_later(UPD_SOME_VALID); + syn_stack_free_all(curwin->w_s); // Need to recompute all syntax. + return; // don't free the progs and patterns now + } + } + + /* + * Something failed, free the allocated memory. + */ + vim_regfree(item.sp_prog); + vim_free(item.sp_pattern); + vim_free(syn_opt_arg.cont_list); + vim_free(syn_opt_arg.cont_in_list); + vim_free(syn_opt_arg.next_list); + + if (rest == NULL && called_emsg == orig_called_emsg) + semsg(_(e_invalid_argument_str), arg); +} + +/* + * Handle ":syntax region {group-name} [matchgroup={group-name}] + * start {start} .. [skip {skip}] end {end} .. [{options}]". + */ + static void +syn_cmd_region( + exarg_T *eap, + int syncing) // TRUE for ":syntax sync region .." +{ + char_u *arg = eap->arg; + char_u *group_name_end; + char_u *rest; // next arg, NULL on error + char_u *key_end; + char_u *key = NULL; + char_u *p; + int item; +#define ITEM_START 0 +#define ITEM_SKIP 1 +#define ITEM_END 2 +#define ITEM_MATCHGROUP 3 + struct pat_ptr + { + synpat_T *pp_synp; // pointer to syn_pattern + int pp_matchgroup_id; // matchgroup ID + struct pat_ptr *pp_next; // pointer to next pat_ptr + } *(pat_ptrs[3]); + // patterns found in the line + struct pat_ptr *ppp; + struct pat_ptr *ppp_next; + int pat_count = 0; // nr of syn_patterns found + int syn_id; + int matchgroup_id = 0; + int not_enough = FALSE; // not enough arguments + int illegal = FALSE; // illegal arguments + int success = FALSE; + int idx; + syn_opt_arg_T syn_opt_arg; + int conceal_char = NUL; + + // Isolate the group name, check for validity + rest = get_group_name(arg, &group_name_end); + + pat_ptrs[0] = NULL; + pat_ptrs[1] = NULL; + pat_ptrs[2] = NULL; + + init_syn_patterns(); + + syn_opt_arg.flags = 0; + syn_opt_arg.keyword = FALSE; + syn_opt_arg.sync_idx = NULL; + syn_opt_arg.has_cont_list = TRUE; + syn_opt_arg.cont_list = NULL; + syn_opt_arg.cont_in_list = NULL; + syn_opt_arg.next_list = NULL; + + /* + * get the options, patterns and matchgroup. + */ + while (rest != NULL && !ends_excmd2(eap->cmd, rest)) + { + // Check for option arguments + rest = get_syn_options(rest, &syn_opt_arg, &conceal_char, eap->skip); + if (rest == NULL || ends_excmd2(eap->cmd, rest)) + break; + + // must be a pattern or matchgroup then + key_end = rest; + while (*key_end && !VIM_ISWHITE(*key_end) && *key_end != '=') + ++key_end; + vim_free(key); + key = vim_strnsave_up(rest, key_end - rest); + if (key == NULL) // out of memory + { + rest = NULL; + break; + } + if (STRCMP(key, "MATCHGROUP") == 0) + item = ITEM_MATCHGROUP; + else if (STRCMP(key, "START") == 0) + item = ITEM_START; + else if (STRCMP(key, "END") == 0) + item = ITEM_END; + else if (STRCMP(key, "SKIP") == 0) + { + if (pat_ptrs[ITEM_SKIP] != NULL) // one skip pattern allowed + { + illegal = TRUE; + break; + } + item = ITEM_SKIP; + } + else + break; + rest = skipwhite(key_end); + if (*rest != '=') + { + rest = NULL; + semsg(_(e_missing_equal_str), arg); + break; + } + rest = skipwhite(rest + 1); + if (*rest == NUL) + { + not_enough = TRUE; + break; + } + + if (item == ITEM_MATCHGROUP) + { + p = skiptowhite(rest); + if ((p - rest == 4 && STRNCMP(rest, "NONE", 4) == 0) || eap->skip) + matchgroup_id = 0; + else + { + matchgroup_id = syn_check_group(rest, (int)(p - rest)); + if (matchgroup_id == 0) + { + illegal = TRUE; + break; + } + } + rest = skipwhite(p); + } + else + { + /* + * Allocate room for a syn_pattern, and link it in the list of + * syn_patterns for this item, at the start (because the list is + * used from end to start). + */ + ppp = ALLOC_ONE(struct pat_ptr); + if (ppp == NULL) + { + rest = NULL; + break; + } + ppp->pp_next = pat_ptrs[item]; + pat_ptrs[item] = ppp; + ppp->pp_synp = ALLOC_CLEAR_ONE(synpat_T); + if (ppp->pp_synp == NULL) + { + rest = NULL; + break; + } + + /* + * Get the syntax pattern and the following offset(s). + */ + // Enable the appropriate \z specials. + if (item == ITEM_START) + reg_do_extmatch = REX_SET; + else if (item == ITEM_SKIP || item == ITEM_END) + reg_do_extmatch = REX_USE; + rest = get_syn_pattern(rest, ppp->pp_synp); + reg_do_extmatch = 0; + if (item == ITEM_END && vim_regcomp_had_eol() + && !(syn_opt_arg.flags & HL_EXCLUDENL)) + ppp->pp_synp->sp_flags |= HL_HAS_EOL; + ppp->pp_matchgroup_id = matchgroup_id; + ++pat_count; + } + } + vim_free(key); + if (illegal || not_enough) + rest = NULL; + + /* + * Must have a "start" and "end" pattern. + */ + if (rest != NULL && (pat_ptrs[ITEM_START] == NULL || + pat_ptrs[ITEM_END] == NULL)) + { + not_enough = TRUE; + rest = NULL; + } + + if (rest != NULL) + { + /* + * Check for trailing garbage or command. + * If OK, add the item. + */ + set_nextcmd(eap, rest); + if (!ends_excmd(*rest) || eap->skip) + rest = NULL; + else if (ga_grow(&(curwin->w_s->b_syn_patterns), pat_count) == OK + && (syn_id = syn_check_group(arg, + (int)(group_name_end - arg))) != 0) + { + syn_incl_toplevel(syn_id, &syn_opt_arg.flags); + /* + * Store the start/skip/end in the syn_items list + */ + idx = curwin->w_s->b_syn_patterns.ga_len; + for (item = ITEM_START; item <= ITEM_END; ++item) + { + for (ppp = pat_ptrs[item]; ppp != NULL; ppp = ppp->pp_next) + { + SYN_ITEMS(curwin->w_s)[idx] = *(ppp->pp_synp); + SYN_ITEMS(curwin->w_s)[idx].sp_syncing = syncing; + SYN_ITEMS(curwin->w_s)[idx].sp_type = + (item == ITEM_START) ? SPTYPE_START : + (item == ITEM_SKIP) ? SPTYPE_SKIP : SPTYPE_END; + SYN_ITEMS(curwin->w_s)[idx].sp_flags |= syn_opt_arg.flags; + SYN_ITEMS(curwin->w_s)[idx].sp_syn.id = syn_id; + SYN_ITEMS(curwin->w_s)[idx].sp_syn.inc_tag = + current_syn_inc_tag; + SYN_ITEMS(curwin->w_s)[idx].sp_syn_match_id = + ppp->pp_matchgroup_id; +#ifdef FEAT_CONCEAL + SYN_ITEMS(curwin->w_s)[idx].sp_cchar = conceal_char; +#endif + if (item == ITEM_START) + { + SYN_ITEMS(curwin->w_s)[idx].sp_cont_list = + syn_opt_arg.cont_list; + SYN_ITEMS(curwin->w_s)[idx].sp_syn.cont_in_list = + syn_opt_arg.cont_in_list; + if (syn_opt_arg.cont_in_list != NULL) + curwin->w_s->b_syn_containedin = TRUE; + SYN_ITEMS(curwin->w_s)[idx].sp_next_list = + syn_opt_arg.next_list; + } + ++curwin->w_s->b_syn_patterns.ga_len; + ++idx; +#ifdef FEAT_FOLDING + if (syn_opt_arg.flags & HL_FOLD) + ++curwin->w_s->b_syn_folditems; +#endif + } + } + + redraw_curbuf_later(UPD_SOME_VALID); + syn_stack_free_all(curwin->w_s); // Need to recompute all syntax. + success = TRUE; // don't free the progs and patterns now + } + } + + /* + * Free the allocated memory. + */ + for (item = ITEM_START; item <= ITEM_END; ++item) + for (ppp = pat_ptrs[item]; ppp != NULL; ppp = ppp_next) + { + if (!success && ppp->pp_synp != NULL) + { + vim_regfree(ppp->pp_synp->sp_prog); + vim_free(ppp->pp_synp->sp_pattern); + } + vim_free(ppp->pp_synp); + ppp_next = ppp->pp_next; + vim_free(ppp); + } + + if (!success) + { + vim_free(syn_opt_arg.cont_list); + vim_free(syn_opt_arg.cont_in_list); + vim_free(syn_opt_arg.next_list); + if (not_enough) + semsg(_(e_not_enough_arguments_syntax_region_str), arg); + else if (illegal || rest == NULL) + semsg(_(e_invalid_argument_str), arg); + } +} + +/* + * A simple syntax group ID comparison function suitable for use in qsort() + */ + static int +syn_compare_stub(const void *v1, const void *v2) +{ + const short *s1 = v1; + const short *s2 = v2; + + return (*s1 > *s2 ? 1 : *s1 < *s2 ? -1 : 0); +} + +/* + * Combines lists of syntax clusters. + * *clstr1 and *clstr2 must both be allocated memory; they will be consumed. + */ + static void +syn_combine_list(short **clstr1, short **clstr2, int list_op) +{ + int count1 = 0; + int count2 = 0; + short *g1; + short *g2; + short *clstr = NULL; + int count; + int round; + + /* + * Handle degenerate cases. + */ + if (*clstr2 == NULL) + return; + if (*clstr1 == NULL || list_op == CLUSTER_REPLACE) + { + if (list_op == CLUSTER_REPLACE) + vim_free(*clstr1); + if (list_op == CLUSTER_REPLACE || list_op == CLUSTER_ADD) + *clstr1 = *clstr2; + else + vim_free(*clstr2); + return; + } + + for (g1 = *clstr1; *g1; g1++) + ++count1; + for (g2 = *clstr2; *g2; g2++) + ++count2; + + /* + * For speed purposes, sort both lists. + */ + qsort(*clstr1, (size_t)count1, sizeof(short), syn_compare_stub); + qsort(*clstr2, (size_t)count2, sizeof(short), syn_compare_stub); + + /* + * We proceed in two passes; in round 1, we count the elements to place + * in the new list, and in round 2, we allocate and populate the new + * list. For speed, we use a mergesort-like method, adding the smaller + * of the current elements in each list to the new list. + */ + for (round = 1; round <= 2; round++) + { + g1 = *clstr1; + g2 = *clstr2; + count = 0; + + /* + * First, loop through the lists until one of them is empty. + */ + while (*g1 && *g2) + { + /* + * We always want to add from the first list. + */ + if (*g1 < *g2) + { + if (round == 2) + clstr[count] = *g1; + count++; + g1++; + continue; + } + /* + * We only want to add from the second list if we're adding the + * lists. + */ + if (list_op == CLUSTER_ADD) + { + if (round == 2) + clstr[count] = *g2; + count++; + } + if (*g1 == *g2) + g1++; + g2++; + } + + /* + * Now add the leftovers from whichever list didn't get finished + * first. As before, we only want to add from the second list if + * we're adding the lists. + */ + for (; *g1; g1++, count++) + if (round == 2) + clstr[count] = *g1; + if (list_op == CLUSTER_ADD) + for (; *g2; g2++, count++) + if (round == 2) + clstr[count] = *g2; + + if (round == 1) + { + /* + * If the group ended up empty, we don't need to allocate any + * space for it. + */ + if (count == 0) + { + clstr = NULL; + break; + } + clstr = ALLOC_MULT(short, count + 1); + if (clstr == NULL) + break; + clstr[count] = 0; + } + } + + /* + * Finally, put the new list in place. + */ + vim_free(*clstr1); + vim_free(*clstr2); + *clstr1 = clstr; +} + +/* + * Lookup a syntax cluster name and return its ID. + * If it is not found, 0 is returned. + */ + static int +syn_scl_name2id(char_u *name) +{ + int i; + char_u *name_u; + + // Avoid using stricmp() too much, it's slow on some systems + name_u = vim_strsave_up(name); + if (name_u == NULL) + return 0; + for (i = curwin->w_s->b_syn_clusters.ga_len; --i >= 0; ) + if (SYN_CLSTR(curwin->w_s)[i].scl_name_u != NULL + && STRCMP(name_u, SYN_CLSTR(curwin->w_s)[i].scl_name_u) == 0) + break; + vim_free(name_u); + return (i < 0 ? 0 : i + SYNID_CLUSTER); +} + +/* + * Like syn_scl_name2id(), but take a pointer + length argument. + */ + static int +syn_scl_namen2id(char_u *linep, int len) +{ + char_u *name; + int id = 0; + + name = vim_strnsave(linep, len); + if (name == NULL) + return 0; + + id = syn_scl_name2id(name); + vim_free(name); + return id; +} + +/* + * Find syntax cluster name in the table and return its ID. + * The argument is a pointer to the name and the length of the name. + * If it doesn't exist yet, a new entry is created. + * Return 0 for failure. + */ + static int +syn_check_cluster(char_u *pp, int len) +{ + int id; + char_u *name; + + name = vim_strnsave(pp, len); + if (name == NULL) + return 0; + + id = syn_scl_name2id(name); + if (id == 0) // doesn't exist yet + id = syn_add_cluster(name); + else + vim_free(name); + return id; +} + +/* + * Add new syntax cluster and return its ID. + * "name" must be an allocated string, it will be consumed. + * Return 0 for failure. + */ + static int +syn_add_cluster(char_u *name) +{ + int len; + + /* + * First call for this growarray: init growing array. + */ + if (curwin->w_s->b_syn_clusters.ga_data == NULL) + { + curwin->w_s->b_syn_clusters.ga_itemsize = sizeof(syn_cluster_T); + curwin->w_s->b_syn_clusters.ga_growsize = 10; + } + + len = curwin->w_s->b_syn_clusters.ga_len; + if (len >= MAX_CLUSTER_ID) + { + emsg(_(e_too_many_syntax_clusters)); + vim_free(name); + return 0; + } + + /* + * Make room for at least one other cluster entry. + */ + if (ga_grow(&curwin->w_s->b_syn_clusters, 1) == FAIL) + { + vim_free(name); + return 0; + } + + CLEAR_POINTER(&(SYN_CLSTR(curwin->w_s)[len])); + SYN_CLSTR(curwin->w_s)[len].scl_name = name; + SYN_CLSTR(curwin->w_s)[len].scl_name_u = vim_strsave_up(name); + SYN_CLSTR(curwin->w_s)[len].scl_list = NULL; + ++curwin->w_s->b_syn_clusters.ga_len; + + if (STRICMP(name, "Spell") == 0) + curwin->w_s->b_spell_cluster_id = len + SYNID_CLUSTER; + if (STRICMP(name, "NoSpell") == 0) + curwin->w_s->b_nospell_cluster_id = len + SYNID_CLUSTER; + + return len + SYNID_CLUSTER; +} + +/* + * Handle ":syntax cluster {cluster-name} [contains={groupname},..] + * [add={groupname},..] [remove={groupname},..]". + */ + static void +syn_cmd_cluster(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg = eap->arg; + char_u *group_name_end; + char_u *rest; + int scl_id; + short *clstr_list; + int got_clstr = FALSE; + int opt_len; + int list_op; + + eap->nextcmd = find_nextcmd(arg); + if (eap->skip) + return; + + rest = get_group_name(arg, &group_name_end); + + if (rest != NULL) + { + scl_id = syn_check_cluster(arg, (int)(group_name_end - arg)); + if (scl_id == 0) + return; + scl_id -= SYNID_CLUSTER; + + for (;;) + { + if (STRNICMP(rest, "add", 3) == 0 + && (VIM_ISWHITE(rest[3]) || rest[3] == '=')) + { + opt_len = 3; + list_op = CLUSTER_ADD; + } + else if (STRNICMP(rest, "remove", 6) == 0 + && (VIM_ISWHITE(rest[6]) || rest[6] == '=')) + { + opt_len = 6; + list_op = CLUSTER_SUBTRACT; + } + else if (STRNICMP(rest, "contains", 8) == 0 + && (VIM_ISWHITE(rest[8]) || rest[8] == '=')) + { + opt_len = 8; + list_op = CLUSTER_REPLACE; + } + else + break; + + clstr_list = NULL; + if (get_id_list(&rest, opt_len, &clstr_list, eap->skip) == FAIL) + { + semsg(_(e_invalid_argument_str), rest); + break; + } + if (scl_id >= 0) + syn_combine_list(&SYN_CLSTR(curwin->w_s)[scl_id].scl_list, + &clstr_list, list_op); + else + vim_free(clstr_list); + got_clstr = TRUE; + } + + if (got_clstr) + { + redraw_curbuf_later(UPD_SOME_VALID); + syn_stack_free_all(curwin->w_s); // Need to recompute all. + } + } + + if (!got_clstr) + emsg(_(e_no_cluster_specified)); + if (rest == NULL || !ends_excmd2(eap->cmd, rest)) + semsg(_(e_invalid_argument_str), arg); +} + +/* + * On first call for current buffer: Init growing array. + */ + static void +init_syn_patterns(void) +{ + curwin->w_s->b_syn_patterns.ga_itemsize = sizeof(synpat_T); + curwin->w_s->b_syn_patterns.ga_growsize = 10; +} + +/* + * Get one pattern for a ":syntax match" or ":syntax region" command. + * Stores the pattern and program in a synpat_T. + * Returns a pointer to the next argument, or NULL in case of an error. + */ + static char_u * +get_syn_pattern(char_u *arg, synpat_T *ci) +{ + char_u *end; + int *p; + int idx; + char_u *cpo_save; + + // need at least three chars + if (arg == NULL || arg[0] == NUL || arg[1] == NUL || arg[2] == NUL) + return NULL; + + end = skip_regexp(arg + 1, *arg, TRUE); + if (*end != *arg) // end delimiter not found + { + semsg(_(e_pattern_delimiter_not_found_str), arg); + return NULL; + } + // store the pattern and compiled regexp program + if ((ci->sp_pattern = vim_strnsave(arg + 1, end - arg - 1)) == NULL) + return NULL; + + // Make 'cpoptions' empty, to avoid the 'l' flag + cpo_save = p_cpo; + p_cpo = empty_option; + ci->sp_prog = vim_regcomp(ci->sp_pattern, RE_MAGIC); + p_cpo = cpo_save; + + if (ci->sp_prog == NULL) + return NULL; + ci->sp_ic = curwin->w_s->b_syn_ic; +#ifdef FEAT_PROFILE + syn_clear_time(&ci->sp_time); +#endif + + /* + * Check for a match, highlight or region offset. + */ + ++end; + do + { + for (idx = SPO_COUNT; --idx >= 0; ) + if (STRNCMP(end, spo_name_tab[idx], 3) == 0) + break; + if (idx >= 0) + { + p = &(ci->sp_offsets[idx]); + if (idx != SPO_LC_OFF) + switch (end[3]) + { + case 's': break; + case 'b': break; + case 'e': idx += SPO_COUNT; break; + default: idx = -1; break; + } + if (idx >= 0) + { + ci->sp_off_flags |= (1 << idx); + if (idx == SPO_LC_OFF) // lc=99 + { + end += 3; + *p = getdigits(&end); + + // "lc=" offset automatically sets "ms=" offset + if (!(ci->sp_off_flags & (1 << SPO_MS_OFF))) + { + ci->sp_off_flags |= (1 << SPO_MS_OFF); + ci->sp_offsets[SPO_MS_OFF] = *p; + } + } + else // yy=x+99 + { + end += 4; + if (*end == '+') + { + ++end; + *p = getdigits(&end); // positive offset + } + else if (*end == '-') + { + ++end; + *p = -getdigits(&end); // negative offset + } + } + if (*end != ',') + break; + ++end; + } + } + } while (idx >= 0); + + if (!ends_excmd2(arg, end) && !VIM_ISWHITE(*end)) + { + semsg(_(e_garbage_after_pattern_str), arg); + return NULL; + } + return skipwhite(end); +} + +/* + * Handle ":syntax sync .." command. + */ + static void +syn_cmd_sync(exarg_T *eap, int syncing UNUSED) +{ + char_u *arg_start = eap->arg; + char_u *arg_end; + char_u *key = NULL; + char_u *next_arg; + int illegal = FALSE; + int finished = FALSE; + long n; + char_u *cpo_save; + + if (ends_excmd2(eap->cmd, arg_start)) + { + syn_cmd_list(eap, TRUE); + return; + } + + while (!ends_excmd2(eap->cmd, arg_start)) + { + arg_end = skiptowhite(arg_start); + next_arg = skipwhite(arg_end); + vim_free(key); + key = vim_strnsave_up(arg_start, arg_end - arg_start); + if (key == NULL) + break; + if (STRCMP(key, "CCOMMENT") == 0) + { + if (!eap->skip) + curwin->w_s->b_syn_sync_flags |= SF_CCOMMENT; + if (!ends_excmd2(eap->cmd, next_arg)) + { + arg_end = skiptowhite(next_arg); + if (!eap->skip) + curwin->w_s->b_syn_sync_id = syn_check_group(next_arg, + (int)(arg_end - next_arg)); + next_arg = skipwhite(arg_end); + } + else if (!eap->skip) + curwin->w_s->b_syn_sync_id = syn_name2id((char_u *)"Comment"); + } + else if ( STRNCMP(key, "LINES", 5) == 0 + || STRNCMP(key, "MINLINES", 8) == 0 + || STRNCMP(key, "MAXLINES", 8) == 0 + || STRNCMP(key, "LINEBREAKS", 10) == 0) + { + if (key[4] == 'S') + arg_end = key + 6; + else if (key[0] == 'L') + arg_end = key + 11; + else + arg_end = key + 9; + if (arg_end[-1] != '=' || !VIM_ISDIGIT(*arg_end)) + { + illegal = TRUE; + break; + } + n = getdigits(&arg_end); + if (!eap->skip) + { + if (key[4] == 'B') + curwin->w_s->b_syn_sync_linebreaks = n; + else if (key[1] == 'A') + curwin->w_s->b_syn_sync_maxlines = n; + else + curwin->w_s->b_syn_sync_minlines = n; + } + } + else if (STRCMP(key, "FROMSTART") == 0) + { + if (!eap->skip) + { + curwin->w_s->b_syn_sync_minlines = MAXLNUM; + curwin->w_s->b_syn_sync_maxlines = 0; + } + } + else if (STRCMP(key, "LINECONT") == 0) + { + if (*next_arg == NUL) // missing pattern + { + illegal = TRUE; + break; + } + if (curwin->w_s->b_syn_linecont_pat != NULL) + { + emsg(_(e_syntax_sync_line_continuations_pattern_specified_twice)); + finished = TRUE; + break; + } + arg_end = skip_regexp(next_arg + 1, *next_arg, TRUE); + if (*arg_end != *next_arg) // end delimiter not found + { + illegal = TRUE; + break; + } + + if (!eap->skip) + { + // store the pattern and compiled regexp program + if ((curwin->w_s->b_syn_linecont_pat = + vim_strnsave(next_arg + 1, + arg_end - next_arg - 1)) == NULL) + { + finished = TRUE; + break; + } + curwin->w_s->b_syn_linecont_ic = curwin->w_s->b_syn_ic; + + // Make 'cpoptions' empty, to avoid the 'l' flag + cpo_save = p_cpo; + p_cpo = empty_option; + curwin->w_s->b_syn_linecont_prog = + vim_regcomp(curwin->w_s->b_syn_linecont_pat, RE_MAGIC); + p_cpo = cpo_save; +#ifdef FEAT_PROFILE + syn_clear_time(&curwin->w_s->b_syn_linecont_time); +#endif + + if (curwin->w_s->b_syn_linecont_prog == NULL) + { + VIM_CLEAR(curwin->w_s->b_syn_linecont_pat); + finished = TRUE; + break; + } + } + next_arg = skipwhite(arg_end + 1); + } + else + { + eap->arg = next_arg; + if (STRCMP(key, "MATCH") == 0) + syn_cmd_match(eap, TRUE); + else if (STRCMP(key, "REGION") == 0) + syn_cmd_region(eap, TRUE); + else if (STRCMP(key, "CLEAR") == 0) + syn_cmd_clear(eap, TRUE); + else + illegal = TRUE; + finished = TRUE; + break; + } + arg_start = next_arg; + } + vim_free(key); + if (illegal) + semsg(_(e_illegal_arguments_str), arg_start); + else if (!finished) + { + set_nextcmd(eap, arg_start); + redraw_curbuf_later(UPD_SOME_VALID); + syn_stack_free_all(curwin->w_s); // Need to recompute all syntax. + } +} + +/* + * Convert a line of highlight group names into a list of group ID numbers. + * "arg" should point to the "contains" or "nextgroup" keyword. + * "arg" is advanced to after the last group name. + * Careful: the argument is modified (NULs added). + * returns FAIL for some error, OK for success. + */ + static int +get_id_list( + char_u **arg, + int keylen, // length of keyword + short **list, // where to store the resulting list, if not + // NULL, the list is silently skipped! + int skip) +{ + char_u *p = NULL; + char_u *end; + int round; + int count; + int total_count = 0; + short *retval = NULL; + char_u *name; + regmatch_T regmatch; + int id; + int i; + int failed = FALSE; + + /* + * We parse the list twice: + * round == 1: count the number of items, allocate the array. + * round == 2: fill the array with the items. + * In round 1 new groups may be added, causing the number of items to + * grow when a regexp is used. In that case round 1 is done once again. + */ + for (round = 1; round <= 2; ++round) + { + /* + * skip "contains" + */ + p = skipwhite(*arg + keylen); + if (*p != '=') + { + semsg(_(e_missing_equal_sign_str), *arg); + break; + } + p = skipwhite(p + 1); + if (ends_excmd2(*arg, p)) + { + semsg(_(e_empty_argument_str), *arg); + break; + } + + /* + * parse the arguments after "contains" + */ + count = 0; + while (!ends_excmd2(*arg, p)) + { + for (end = p; *end && !VIM_ISWHITE(*end) && *end != ','; ++end) + ; + name = alloc(end - p + 3); // leave room for "^$" + if (name == NULL) + { + failed = TRUE; + break; + } + vim_strncpy(name + 1, p, end - p); + if ( STRCMP(name + 1, "ALLBUT") == 0 + || STRCMP(name + 1, "ALL") == 0 + || STRCMP(name + 1, "TOP") == 0 + || STRCMP(name + 1, "CONTAINED") == 0) + { + if (TOUPPER_ASC(**arg) != 'C') + { + semsg(_(e_str_not_allowed_here), name + 1); + failed = TRUE; + vim_free(name); + break; + } + if (count != 0) + { + semsg(_(e_str_must_be_first_in_contains_list), name + 1); + failed = TRUE; + vim_free(name); + break; + } + if (name[1] == 'A') + id = SYNID_ALLBUT + current_syn_inc_tag; + else if (name[1] == 'T') + { + if (curwin->w_s->b_syn_topgrp >= SYNID_CLUSTER) + id = curwin->w_s->b_syn_topgrp; + else + id = SYNID_TOP + current_syn_inc_tag; + } + else + id = SYNID_CONTAINED + current_syn_inc_tag; + + } + else if (name[1] == '@') + { + if (skip) + id = -1; + else + id = syn_check_cluster(name + 2, (int)(end - p - 1)); + } + else + { + /* + * Handle full group name. + */ + if (vim_strpbrk(name + 1, (char_u *)"\\.*^$~[") == NULL) + id = syn_check_group(name + 1, (int)(end - p)); + else + { + /* + * Handle match of regexp with group names. + */ + *name = '^'; + STRCAT(name, "$"); + regmatch.regprog = vim_regcomp(name, RE_MAGIC); + if (regmatch.regprog == NULL) + { + failed = TRUE; + vim_free(name); + break; + } + + regmatch.rm_ic = TRUE; + id = 0; + for (i = highlight_num_groups(); --i >= 0; ) + { + if (vim_regexec(®match, highlight_group_name(i), + (colnr_T)0)) + { + if (round == 2) + { + // Got more items than expected; can happen + // when adding items that match: + // "contains=a.*b,axb". + // Go back to first round + if (count >= total_count) + { + vim_free(retval); + round = 1; + } + else + retval[count] = i + 1; + } + ++count; + id = -1; // remember that we found one + } + } + vim_regfree(regmatch.regprog); + } + } + vim_free(name); + if (id == 0) + { + semsg(_(e_unknown_group_name_str), p); + failed = TRUE; + break; + } + if (id > 0) + { + if (round == 2) + { + // Got more items than expected, go back to first round + if (count >= total_count) + { + vim_free(retval); + round = 1; + } + else + retval[count] = id; + } + ++count; + } + p = skipwhite(end); + if (*p != ',') + break; + p = skipwhite(p + 1); // skip comma in between arguments + } + if (failed) + break; + if (round == 1) + { + retval = ALLOC_MULT(short, count + 1); + if (retval == NULL) + break; + retval[count] = 0; // zero means end of the list + total_count = count; + } + } + + *arg = p; + if (failed || retval == NULL) + { + vim_free(retval); + return FAIL; + } + + if (*list == NULL) + *list = retval; + else + vim_free(retval); // list already found, don't overwrite it + + return OK; +} + +/* + * Make a copy of an ID list. + */ + static short * +copy_id_list(short *list) +{ + int len; + int count; + short *retval; + + if (list == NULL) + return NULL; + + for (count = 0; list[count]; ++count) + ; + len = (count + 1) * sizeof(short); + retval = alloc(len); + if (retval != NULL) + mch_memmove(retval, list, (size_t)len); + + return retval; +} + +/* + * Check if syntax group "ssp" is in the ID list "list" of "cur_si". + * "cur_si" can be NULL if not checking the "containedin" list. + * Used to check if a syntax item is in the "contains" or "nextgroup" list of + * the current item. + * This function is called very often, keep it fast!! + */ + static int +in_id_list( + stateitem_T *cur_si, // current item or NULL + short *list, // id list + struct sp_syn *ssp, // group id and ":syn include" tag of group + int contained) // group id is contained +{ + int retval; + short *scl_list; + short item; + short id = ssp->id; + static int depth = 0; + int r; + + // If ssp has a "containedin" list and "cur_si" is in it, return TRUE. + if (cur_si != NULL && ssp->cont_in_list != NULL + && !(cur_si->si_flags & HL_MATCH)) + { + // Ignore transparent items without a contains argument. Double check + // that we don't go back past the first one. + while ((cur_si->si_flags & HL_TRANS_CONT) + && cur_si > (stateitem_T *)(current_state.ga_data)) + --cur_si; + // cur_si->si_idx is -1 for keywords, these never contain anything. + if (cur_si->si_idx >= 0 && in_id_list(NULL, ssp->cont_in_list, + &(SYN_ITEMS(syn_block)[cur_si->si_idx].sp_syn), + SYN_ITEMS(syn_block)[cur_si->si_idx].sp_flags & HL_CONTAINED)) + return TRUE; + } + + if (list == NULL) + return FALSE; + + /* + * If list is ID_LIST_ALL, we are in a transparent item that isn't + * inside anything. Only allow not-contained groups. + */ + if (list == ID_LIST_ALL) + return !contained; + + /* + * If the first item is "ALLBUT", return TRUE if "id" is NOT in the + * contains list. We also require that "id" is at the same ":syn include" + * level as the list. + */ + item = *list; + if (item >= SYNID_ALLBUT && item < SYNID_CLUSTER) + { + if (item < SYNID_TOP) + { + // ALL or ALLBUT: accept all groups in the same file + if (item - SYNID_ALLBUT != ssp->inc_tag) + return FALSE; + } + else if (item < SYNID_CONTAINED) + { + // TOP: accept all not-contained groups in the same file + if (item - SYNID_TOP != ssp->inc_tag || contained) + return FALSE; + } + else + { + // CONTAINED: accept all contained groups in the same file + if (item - SYNID_CONTAINED != ssp->inc_tag || !contained) + return FALSE; + } + item = *++list; + retval = FALSE; + } + else + retval = TRUE; + + /* + * Return "retval" if id is in the contains list. + */ + while (item != 0) + { + if (item == id) + return retval; + if (item >= SYNID_CLUSTER) + { + scl_list = SYN_CLSTR(syn_block)[item - SYNID_CLUSTER].scl_list; + // restrict recursiveness to 30 to avoid an endless loop for a + // cluster that includes itself (indirectly) + if (scl_list != NULL && depth < 30) + { + ++depth; + r = in_id_list(NULL, scl_list, ssp, contained); + --depth; + if (r) + return retval; + } + } + item = *++list; + } + return !retval; +} + +struct subcommand +{ + char *name; // subcommand name + void (*func)(exarg_T *, int); // function to call +}; + +static struct subcommand subcommands[] = +{ + {"case", syn_cmd_case}, + {"clear", syn_cmd_clear}, + {"cluster", syn_cmd_cluster}, + {"conceal", syn_cmd_conceal}, + {"enable", syn_cmd_enable}, + {"foldlevel", syn_cmd_foldlevel}, + {"include", syn_cmd_include}, + {"iskeyword", syn_cmd_iskeyword}, + {"keyword", syn_cmd_keyword}, + {"list", syn_cmd_list}, + {"manual", syn_cmd_manual}, + {"match", syn_cmd_match}, + {"on", syn_cmd_on}, + {"off", syn_cmd_off}, + {"region", syn_cmd_region}, + {"reset", syn_cmd_reset}, + {"spell", syn_cmd_spell}, + {"sync", syn_cmd_sync}, + {"", syn_cmd_list}, + {NULL, NULL} +}; + +/* + * ":syntax". + * This searches the subcommands[] table for the subcommand name, and calls a + * syntax_subcommand() function to do the rest. + */ + void +ex_syntax(exarg_T *eap) +{ + char_u *arg = eap->arg; + char_u *subcmd_end; + char_u *subcmd_name; + int i; + + syn_cmdlinep = eap->cmdlinep; + + // isolate subcommand name + for (subcmd_end = arg; ASCII_ISALPHA(*subcmd_end); ++subcmd_end) + ; + subcmd_name = vim_strnsave(arg, subcmd_end - arg); + if (subcmd_name == NULL) + return; + + if (eap->skip) // skip error messages for all subcommands + ++emsg_skip; + for (i = 0; ; ++i) + { + if (subcommands[i].name == NULL) + { + semsg(_(e_invalid_syntax_subcommand_str), subcmd_name); + break; + } + if (STRCMP(subcmd_name, (char_u *)subcommands[i].name) == 0) + { + eap->arg = skipwhite(subcmd_end); + (subcommands[i].func)(eap, FALSE); + break; + } + } + vim_free(subcmd_name); + if (eap->skip) + --emsg_skip; +} + + void +ex_ownsyntax(exarg_T *eap) +{ + char_u *old_value; + char_u *new_value; + + if (curwin->w_s == &curwin->w_buffer->b_s) + { + curwin->w_s = ALLOC_ONE(synblock_T); + CLEAR_POINTER(curwin->w_s); + hash_init(&curwin->w_s->b_keywtab); + hash_init(&curwin->w_s->b_keywtab_ic); +#ifdef FEAT_SPELL + // TODO: keep the spell checking as it was. + curwin->w_p_spell = FALSE; // No spell checking + // make sure option values are "empty_option" instead of NULL + clear_string_option(&curwin->w_s->b_p_spc); + clear_string_option(&curwin->w_s->b_p_spf); + clear_string_option(&curwin->w_s->b_p_spl); + clear_string_option(&curwin->w_s->b_p_spo); +#endif + clear_string_option(&curwin->w_s->b_syn_isk); + } + + // save value of b:current_syntax + old_value = get_var_value((char_u *)"b:current_syntax"); + if (old_value != NULL) + old_value = vim_strsave(old_value); + + // Apply the "syntax" autocommand event, this finds and loads the syntax + // file. + apply_autocmds(EVENT_SYNTAX, eap->arg, curbuf->b_fname, TRUE, curbuf); + + // move value of b:current_syntax to w:current_syntax + new_value = get_var_value((char_u *)"b:current_syntax"); + if (new_value != NULL) + set_internal_string_var((char_u *)"w:current_syntax", new_value); + + // restore value of b:current_syntax + if (old_value == NULL) + do_unlet((char_u *)"b:current_syntax", TRUE); + else + { + set_internal_string_var((char_u *)"b:current_syntax", old_value); + vim_free(old_value); + } +} + + int +syntax_present(win_T *win) +{ + return (win->w_s->b_syn_patterns.ga_len != 0 + || win->w_s->b_syn_clusters.ga_len != 0 + || win->w_s->b_keywtab.ht_used > 0 + || win->w_s->b_keywtab_ic.ht_used > 0); +} + + +static enum +{ + EXP_SUBCMD, // expand ":syn" sub-commands + EXP_CASE, // expand ":syn case" arguments + EXP_SPELL, // expand ":syn spell" arguments + EXP_SYNC, // expand ":syn sync" arguments + EXP_CLUSTER // expand ":syn list @cluster" arguments +} expand_what; + +/* + * Reset include_link, include_default, include_none to 0. + * Called when we are done expanding. + */ + void +reset_expand_highlight(void) +{ + include_link = include_default = include_none = 0; +} + +/* + * Handle command line completion for :match and :echohl command: Add "None" + * as highlight group. + */ + void +set_context_in_echohl_cmd(expand_T *xp, char_u *arg) +{ + xp->xp_context = EXPAND_HIGHLIGHT; + xp->xp_pattern = arg; + include_none = 1; +} + +/* + * Handle command line completion for :syntax command. + */ + void +set_context_in_syntax_cmd(expand_T *xp, char_u *arg) +{ + char_u *p; + + // Default: expand subcommands + xp->xp_context = EXPAND_SYNTAX; + expand_what = EXP_SUBCMD; + xp->xp_pattern = arg; + include_link = 0; + include_default = 0; + + if (*arg == NUL) + return; + + // (part of) subcommand already typed + p = skiptowhite(arg); + if (*p == NUL) + return; + + // past first word + xp->xp_pattern = skipwhite(p); + if (*skiptowhite(xp->xp_pattern) != NUL) + xp->xp_context = EXPAND_NOTHING; + else if (STRNICMP(arg, "case", p - arg) == 0) + expand_what = EXP_CASE; + else if (STRNICMP(arg, "spell", p - arg) == 0) + expand_what = EXP_SPELL; + else if (STRNICMP(arg, "sync", p - arg) == 0) + expand_what = EXP_SYNC; + else if (STRNICMP(arg, "list", p - arg) == 0) + { + p = skipwhite(p); + if (*p == '@') + expand_what = EXP_CLUSTER; + else + xp->xp_context = EXPAND_HIGHLIGHT; + } + else if (STRNICMP(arg, "keyword", p - arg) == 0 + || STRNICMP(arg, "region", p - arg) == 0 + || STRNICMP(arg, "match", p - arg) == 0) + xp->xp_context = EXPAND_HIGHLIGHT; + else + xp->xp_context = EXPAND_NOTHING; +} + +/* + * Function given to ExpandGeneric() to obtain the list syntax names for + * expansion. + */ + char_u * +get_syntax_name(expand_T *xp, int idx) +{ + switch (expand_what) + { + case EXP_SUBCMD: + return (char_u *)subcommands[idx].name; + case EXP_CASE: + { + static char *case_args[] = {"match", "ignore", NULL}; + return (char_u *)case_args[idx]; + } + case EXP_SPELL: + { + static char *spell_args[] = + {"toplevel", "notoplevel", "default", NULL}; + return (char_u *)spell_args[idx]; + } + case EXP_SYNC: + { + static char *sync_args[] = + {"ccomment", "clear", "fromstart", + "linebreaks=", "linecont", "lines=", "match", + "maxlines=", "minlines=", "region", NULL}; + return (char_u *)sync_args[idx]; + } + case EXP_CLUSTER: + { + if (idx < curwin->w_s->b_syn_clusters.ga_len) + { + vim_snprintf((char *)xp->xp_buf, EXPAND_BUF_LEN, "@%s", + SYN_CLSTR(curwin->w_s)[idx].scl_name); + return xp->xp_buf; + } + else + return NULL; + } + } + return NULL; +} + + +/* + * Function called for expression evaluation: get syntax ID at file position. + */ + int +syn_get_id( + win_T *wp, + long lnum, + colnr_T col, + int trans, // remove transparency + int *spellp, // return: can do spell checking + int keep_state) // keep state of char at "col" +{ + // When the position is not after the current position and in the same + // line of the same window with the same buffer, need to restart parsing. + if (wp != syn_win + || wp->w_buffer != syn_buf + || lnum != current_lnum + || col < current_col) + syntax_start(wp, lnum); + else if (wp->w_buffer == syn_buf + && lnum == current_lnum + && col > current_col) + // next_match may not be correct when moving around, e.g. with the + // "skip" expression in searchpair() + next_match_idx = -1; + + (void)get_syntax_attr(col, spellp, keep_state); + + return (trans ? current_trans_id : current_id); +} + +#if defined(FEAT_CONCEAL) || defined(PROTO) +/* + * Get extra information about the syntax item. Must be called right after + * get_syntax_attr(). + * Stores the current item sequence nr in "*seqnrp". + * Returns the current flags. + */ + int +get_syntax_info(int *seqnrp) +{ + *seqnrp = current_seqnr; + return current_flags; +} + +/* + * Return conceal substitution character + */ + int +syn_get_sub_char(void) +{ + return current_sub_char; +} +#endif + +#if defined(FEAT_EVAL) || defined(PROTO) +/* + * Return the syntax ID at position "i" in the current stack. + * The caller must have called syn_get_id() before to fill the stack. + * Returns -1 when "i" is out of range. + */ + int +syn_get_stack_item(int i) +{ + if (i >= current_state.ga_len) + { + // Need to invalidate the state, because we didn't properly finish it + // for the last character, "keep_state" was TRUE. + invalidate_current_state(); + current_col = MAXCOL; + return -1; + } + return CUR_STATE(i).si_id; +} +#endif + +#if defined(FEAT_FOLDING) || defined(PROTO) + static int +syn_cur_foldlevel(void) +{ + int level = 0; + int i; + + for (i = 0; i < current_state.ga_len; ++i) + if (CUR_STATE(i).si_flags & HL_FOLD) + ++level; + return level; +} + +/* + * Function called to get folding level for line "lnum" in window "wp". + */ + int +syn_get_foldlevel(win_T *wp, long lnum) +{ + int level = 0; + int low_level; + int cur_level; + + // Return quickly when there are no fold items at all. + if (wp->w_s->b_syn_folditems != 0 + && !wp->w_s->b_syn_error +# ifdef SYN_TIME_LIMIT + && !wp->w_s->b_syn_slow +# endif + ) + { + syntax_start(wp, lnum); + + // Start with the fold level at the start of the line. + level = syn_cur_foldlevel(); + + if (wp->w_s->b_syn_foldlevel == SYNFLD_MINIMUM) + { + // Find the lowest fold level that is followed by a higher one. + cur_level = level; + low_level = cur_level; + while (!current_finished) + { + (void)syn_current_attr(FALSE, FALSE, NULL, FALSE); + cur_level = syn_cur_foldlevel(); + if (cur_level < low_level) + low_level = cur_level; + else if (cur_level > low_level) + level = low_level; + ++current_col; + } + } + } + if (level > wp->w_p_fdn) + { + level = wp->w_p_fdn; + if (level < 0) + level = 0; + } + return level; +} +#endif + +#if defined(FEAT_PROFILE) || defined(PROTO) +/* + * ":syntime". + */ + void +ex_syntime(exarg_T *eap) +{ + if (STRCMP(eap->arg, "on") == 0) + syn_time_on = TRUE; + else if (STRCMP(eap->arg, "off") == 0) + syn_time_on = FALSE; + else if (STRCMP(eap->arg, "clear") == 0) + syntime_clear(); + else if (STRCMP(eap->arg, "report") == 0) + syntime_report(); + else + semsg(_(e_invalid_argument_str), eap->arg); +} + + static void +syn_clear_time(syn_time_T *st) +{ + profile_zero(&st->total); + profile_zero(&st->slowest); + st->count = 0; + st->match = 0; +} + +/* + * Clear the syntax timing for the current buffer. + */ + static void +syntime_clear(void) +{ + int idx; + synpat_T *spp; + + if (!syntax_present(curwin)) + { + msg(_(msg_no_items)); + return; + } + for (idx = 0; idx < curwin->w_s->b_syn_patterns.ga_len; ++idx) + { + spp = &(SYN_ITEMS(curwin->w_s)[idx]); + syn_clear_time(&spp->sp_time); + } +} + +/* + * Function given to ExpandGeneric() to obtain the possible arguments of the + * ":syntime {on,off,clear,report}" command. + */ + char_u * +get_syntime_arg(expand_T *xp UNUSED, int idx) +{ + switch (idx) + { + case 0: return (char_u *)"on"; + case 1: return (char_u *)"off"; + case 2: return (char_u *)"clear"; + case 3: return (char_u *)"report"; + } + return NULL; +} + +typedef struct +{ + proftime_T total; + int count; + int match; + proftime_T slowest; + proftime_T average; + int id; + char_u *pattern; +} time_entry_T; + + static int +syn_compare_syntime(const void *v1, const void *v2) +{ + const time_entry_T *s1 = v1; + const time_entry_T *s2 = v2; + + return profile_cmp(&s1->total, &s2->total); +} + +/* + * Clear the syntax timing for the current buffer. + */ + static void +syntime_report(void) +{ + int idx; + synpat_T *spp; +# if defined(FEAT_RELTIME) + proftime_T tm; +# endif + int len; + proftime_T total_total; + int total_count = 0; + garray_T ga; + time_entry_T *p; + + if (!syntax_present(curwin)) + { + msg(_(msg_no_items)); + return; + } + + ga_init2(&ga, sizeof(time_entry_T), 50); + profile_zero(&total_total); + for (idx = 0; idx < curwin->w_s->b_syn_patterns.ga_len; ++idx) + { + spp = &(SYN_ITEMS(curwin->w_s)[idx]); + if (spp->sp_time.count > 0) + { + (void)ga_grow(&ga, 1); + p = ((time_entry_T *)ga.ga_data) + ga.ga_len; + p->total = spp->sp_time.total; + profile_add(&total_total, &spp->sp_time.total); + p->count = spp->sp_time.count; + p->match = spp->sp_time.match; + total_count += spp->sp_time.count; + p->slowest = spp->sp_time.slowest; +# if defined(FEAT_RELTIME) + profile_divide(&spp->sp_time.total, spp->sp_time.count, &tm); + p->average = tm; +# endif + p->id = spp->sp_syn.id; + p->pattern = spp->sp_pattern; + ++ga.ga_len; + } + } + + // Sort on total time. Skip if there are no items to avoid passing NULL + // pointer to qsort(). + if (ga.ga_len > 1) + qsort(ga.ga_data, (size_t)ga.ga_len, sizeof(time_entry_T), + syn_compare_syntime); + + msg_puts_title(_(" TOTAL COUNT MATCH SLOWEST AVERAGE NAME PATTERN")); + msg_puts("\n"); + for (idx = 0; idx < ga.ga_len && !got_int; ++idx) + { + p = ((time_entry_T *)ga.ga_data) + idx; + + msg_puts(profile_msg(&p->total)); + msg_puts(" "); // make sure there is always a separating space + msg_advance(13); + msg_outnum(p->count); + msg_puts(" "); + msg_advance(20); + msg_outnum(p->match); + msg_puts(" "); + msg_advance(26); + msg_puts(profile_msg(&p->slowest)); + msg_puts(" "); + msg_advance(38); + msg_puts(profile_msg(&p->average)); + msg_puts(" "); + msg_advance(50); + msg_outtrans(highlight_group_name(p->id - 1)); + msg_puts(" "); + + msg_advance(69); + if (Columns < 80) + len = 20; // will wrap anyway + else + len = Columns - 70; + if (len > (int)STRLEN(p->pattern)) + len = (int)STRLEN(p->pattern); + msg_outtrans_len(p->pattern, len); + msg_puts("\n"); + } + ga_clear(&ga); + if (!got_int) + { + msg_puts("\n"); + msg_puts(profile_msg(&total_total)); + msg_advance(13); + msg_outnum(total_count); + msg_puts("\n"); + } +} +#endif + +#endif // FEAT_SYN_HL |