Commit 8994c38288b51d44611f81182a06ea59ed36a67c
2019-04-05T08:19:01
md_is_code_span: Follow new spec rules for space trimming.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
diff --git a/md4c/md4c.c b/md4c/md4c.c
index 2edc9ff..e4ffd7b 100644
--- a/md4c/md4c.c
+++ b/md4c/md4c.c
@@ -2760,21 +2760,27 @@ md_build_mark_char_map(MD_CTX* ctx)
#define CODESPAN_MARK_MAXLEN 32
static int
-md_is_code_span(MD_CTX* ctx, OFF beg, OFF max_end,
+md_is_code_span(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg,
OFF* p_opener_beg, OFF* p_opener_end,
OFF* p_closer_beg, OFF* p_closer_end,
OFF last_potential_closers[CODESPAN_MARK_MAXLEN],
- int* p_reached_max_end)
+ int* p_reached_paragraph_end)
{
OFF opener_beg = beg;
OFF opener_end;
OFF closer_beg;
OFF closer_end;
SZ mark_len;
+ OFF line_end;
+ int has_space_after_opener;
+ int has_space_before_closer;
+ int has_only_space = TRUE;
+ line_end = lines[0].end;
opener_end = opener_beg;
- while(opener_end < max_end && CH(opener_end) == _T('`'))
+ while(opener_end < line_end && CH(opener_end) == _T('`'))
opener_end++;
+ has_space_after_opener = (opener_end >= line_end || CH(opener_end) == _T(' '));
/* The caller needs to know end of the opening mark even if we fail. */
*p_opener_end = opener_end;
@@ -2785,53 +2791,72 @@ md_is_code_span(MD_CTX* ctx, OFF beg, OFF max_end,
/* Check whether we already know there is no closer of this length.
* If so, re-scan does no sense. This fixes issue #59. */
- if(last_potential_closers[mark_len-1] >= max_end ||
- (*p_reached_max_end && last_potential_closers[mark_len-1] < opener_end))
+ if(last_potential_closers[mark_len-1] >= lines[n_lines-1].end ||
+ (*p_reached_paragraph_end && last_potential_closers[mark_len-1] < opener_end))
return FALSE;
closer_beg = opener_end;
closer_end = opener_end;
- /* Find closer mark. Note we rely on the fact that '`' cannot be in the
- * "gaps" between lines here so we scan as in a simple string. */
+ /* Find closer mark. */
while(TRUE) {
- while(closer_beg < max_end && CH(closer_beg) != _T('`'))
+ while(closer_beg < line_end && CH(closer_beg) != _T('`')) {
+ if(CH(closer_beg) != _T(' '))
+ has_only_space = FALSE;
closer_beg++;
+ }
closer_end = closer_beg;
- while(closer_end < max_end && CH(closer_end) == _T('`'))
+ while(closer_end < line_end && CH(closer_end) == _T('`'))
closer_end++;
- if(closer_end - closer_beg == mark_len)
+ if(closer_end - closer_beg == mark_len) {
+ /* Success. */
+ has_space_before_closer = (closer_beg == lines[0].beg || CH(closer_beg-1) == _T(' '));
break;
-
- if(closer_end - closer_beg > 0 && closer_end - closer_beg < CODESPAN_MARK_MAXLEN) {
- if(closer_beg > last_potential_closers[closer_end - closer_beg - 1])
- last_potential_closers[closer_end - closer_beg - 1] = closer_beg;
}
- if(closer_end >= max_end) {
- *p_reached_max_end = TRUE;
- return FALSE;
+ if(closer_end - closer_beg > 0) {
+ /* We have found a back-tick which is not part of the closer. */
+ has_only_space = FALSE;
+
+ /* But if we eventually fail, remember it as a potential closer
+ * of its own length for future attempts. This mitigates needs for
+ * rescans. */
+ if(closer_end - closer_beg < CODESPAN_MARK_MAXLEN) {
+ if(closer_beg > last_potential_closers[closer_end - closer_beg - 1])
+ last_potential_closers[closer_end - closer_beg - 1] = closer_beg;
+ }
}
- closer_beg = closer_end;
+ if(closer_end >= line_end) {
+ lines++;
+ n_lines--;
+ if(n_lines == 0) {
+ /* Reached end of the paragraph and still nothing. */
+ *p_reached_paragraph_end = TRUE;
+ return FALSE;
+ }
+ /* Try on the next line. */
+ line_end = lines[0].end;
+ closer_beg = lines[0].beg;
+ } else {
+ closer_beg = closer_end;
+ }
}
- /* If there is a space ore new line both after and before the opener,
- * consume it. It may be tricky as the new line may be digraph "\r\n". */
- if((CH(opener_end) == _T(' ') || ISNEWLINE(opener_end)) &&
- (CH(closer_beg-1) == _T(' ') || ISNEWLINE(closer_beg-1))) {
+ /* If there is a space or a new line both after and before the opener
+ * (and if the code span is not made of spaces only), consume one initial
+ * and one trailing space as part of the marks. */
+ if(has_space_after_opener && has_space_before_closer && !has_only_space) {
if(CH(opener_end) == _T('\r') && CH(opener_end+1) == _T('\n'))
opener_end += 2;
else
opener_end += 1;
- if(closer_beg > opener_end) {
- if(CH(closer_beg-2) == _T('\r') && CH(closer_beg-1) == _T('\n'))
- closer_beg -= 2;
- else
- closer_beg -= 1;
- }
+ if(CH(closer_beg-2) == _T('\r') && CH(closer_beg-1) == _T('\n'))
+ closer_beg -= 2;
+ else
+ closer_beg -= 1;
}
*p_opener_beg = opener_beg;
@@ -2953,7 +2978,7 @@ md_collect_marks(MD_CTX* ctx, const MD_LINE* lines, int n_lines, int table_mode)
int ret = 0;
MD_MARK* mark;
OFF codespan_last_potential_closers[CODESPAN_MARK_MAXLEN] = { 0 };
- int codespan_scanned_till_end = FALSE;
+ int codespan_scanned_till_paragraph_end = FALSE;
for(i = 0; i < n_lines; i++) {
const MD_LINE* line = &lines[i];
@@ -3068,10 +3093,10 @@ md_collect_marks(MD_CTX* ctx, const MD_LINE* lines, int n_lines, int table_mode)
OFF closer_beg, closer_end;
int is_code_span;
- is_code_span = md_is_code_span(ctx, off, lines[n_lines-1].end,
+ is_code_span = md_is_code_span(ctx, lines + i, n_lines - i, off,
&opener_beg, &opener_end, &closer_beg, &closer_end,
codespan_last_potential_closers,
- &codespan_scanned_till_end);
+ &codespan_scanned_till_paragraph_end);
if(is_code_span) {
PUSH_MARK(_T('`'), opener_beg, opener_end, MD_MARK_OPENER | MD_MARK_RESOLVED);
PUSH_MARK(_T('`'), closer_beg, closer_end, MD_MARK_CLOSER | MD_MARK_RESOLVED);