aboutsummaryrefslogtreecommitdiffstats
path: root/src
diff options
context:
space:
mode:
Diffstat (limited to 'src')
-rw-r--r--src/template-parser.c99
-rw-r--r--src/template-parser.h5
-rw-r--r--src/utils/strings.c45
-rw-r--r--src/utils/utils.h2
4 files changed, 129 insertions, 22 deletions
diff --git a/src/template-parser.c b/src/template-parser.c
index 0783254..1d9046e 100644
--- a/src/template-parser.c
+++ b/src/template-parser.c
@@ -22,6 +22,7 @@ typedef enum {
TEMPLATE_START = 1,
TEMPLATE_OPEN_BRACKET,
TEMPLATE_BLOCK_START,
+ TEMPLATE_BLOCK_START_WHITESPACE_CLEANER,
TEMPLATE_BLOCK_TYPE,
TEMPLATE_BLOCK_BLOCK_TYPE_START,
TEMPLATE_BLOCK_BLOCK_TYPE,
@@ -34,6 +35,7 @@ typedef enum {
TEMPLATE_BLOCK_IF_VARIABLE_OPERAND,
TEMPLATE_BLOCK_FOREACH_START,
TEMPLATE_BLOCK_FOREACH_VARIABLE,
+ TEMPLATE_BLOCK_END_WHITESPACE_CLEANER,
TEMPLATE_BLOCK_END,
TEMPLATE_VARIABLE_START,
TEMPLATE_VARIABLE,
@@ -72,6 +74,20 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
b_slist_t *stmts = NULL;
blogc_template_stmt_t *stmt = NULL;
+ /*
+ * this is a reference to the content of previous node in the singly-linked
+ * list. The "correct" solution here would be implement a doubly-linked
+ * list, but here are a few reasons to avoid it:
+ *
+ * - i'm too tired to implement it :P
+ * - template parser never walk backwards, then the list itself does not
+ * need to know its previous node.
+ */
+ blogc_template_stmt_t *previous = NULL;
+
+ bool lstrip_next = false;
+ char *tmp = NULL;
+
blogc_template_parser_state_t state = TEMPLATE_START;
blogc_template_parser_block_state_t block_state = BLOCK_CLOSED;
blogc_template_stmt_type_t type = BLOGC_TEMPLATE_CONTENT_STMT;
@@ -86,10 +102,20 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
if (last) {
stmt = b_malloc(sizeof(blogc_template_stmt_t));
stmt->type = type;
- stmt->value = b_strndup(src + start, src_len - start);
+ if (lstrip_next) {
+ tmp = b_strndup(src + start, src_len - start);
+ stmt->value = b_strdup(b_str_lstrip(tmp));
+ free(tmp);
+ tmp = NULL;
+ lstrip_next = false;
+ }
+ else {
+ stmt->value = b_strndup(src + start, src_len - start);
+ }
stmt->op = 0;
stmt->value2 = NULL;
stmts = b_slist_append(stmts, stmt);
+ previous = stmt;
stmt = NULL;
}
if (c == '{') {
@@ -101,16 +127,26 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
case TEMPLATE_OPEN_BRACKET:
if (c == '%' || c == '{') {
if (c == '%')
- state = TEMPLATE_BLOCK_START;
+ state = TEMPLATE_BLOCK_START_WHITESPACE_CLEANER;
else
state = TEMPLATE_VARIABLE_START;
if (end > start) {
stmt = b_malloc(sizeof(blogc_template_stmt_t));
stmt->type = type;
- stmt->value = b_strndup(src + start, end - start);
+ if (lstrip_next) {
+ tmp = b_strndup(src + start, end - start);
+ stmt->value = b_strdup(b_str_lstrip(tmp));
+ free(tmp);
+ tmp = NULL;
+ lstrip_next = false;
+ }
+ else {
+ stmt->value = b_strndup(src + start, end - start);
+ }
stmt->op = 0;
stmt->value2 = NULL;
stmts = b_slist_append(stmts, stmt);
+ previous = stmt;
stmt = NULL;
}
break;
@@ -118,6 +154,18 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
state = TEMPLATE_START;
break;
+ case TEMPLATE_BLOCK_START_WHITESPACE_CLEANER:
+ if (c == '-') {
+ if ((previous != NULL) &&
+ (previous->type == BLOGC_TEMPLATE_CONTENT_STMT))
+ {
+ previous->value = b_str_rstrip(previous->value); // does not need copy
+ }
+ state = TEMPLATE_BLOCK_START;
+ break;
+ }
+ state = TEMPLATE_BLOCK_START;
+
case TEMPLATE_BLOCK_START:
if (c == ' ')
break;
@@ -126,6 +174,13 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
start = current;
break;
}
+ if (c == '-') {
+ *err = blogc_error_parser(BLOGC_ERROR_TEMPLATE_PARSER, src,
+ src_len, current,
+ "Invalid statement syntax. Duplicated whitespace "
+ "cleaner before statement.");
+ break;
+ }
*err = blogc_error_parser(BLOGC_ERROR_TEMPLATE_PARSER, src,
src_len, current,
"Invalid statement syntax. Must begin with lowercase letter.");
@@ -152,7 +207,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
(0 == strncmp("endblock", src + start, 8)))
{
if (block_state != BLOCK_CLOSED) {
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
type = BLOGC_TEMPLATE_ENDBLOCK_STMT;
block_state = BLOCK_CLOSED;
break;
@@ -193,7 +248,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
(0 == strncmp("endif", src + start, 5)))
{
if (if_count > 0) {
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
type = BLOGC_TEMPLATE_ENDIF_STMT;
if_count--;
break;
@@ -223,7 +278,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
(0 == strncmp("endforeach", src + start, 10)))
{
if (foreach_open) {
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
type = BLOGC_TEMPLATE_ENDFOREACH_STMT;
foreach_open = false;
break;
@@ -264,7 +319,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
{
block_state = BLOCK_ENTRY;
end = current;
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
break;
}
else if ((current - start == 7) &&
@@ -272,7 +327,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
{
block_state = BLOCK_LISTING;
end = current;
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
break;
}
else if ((current - start == 12) &&
@@ -280,7 +335,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
{
block_state = BLOCK_LISTING_ONCE;
end = current;
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
break;
}
}
@@ -311,7 +366,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
if (type == BLOGC_TEMPLATE_IF_STMT)
state = TEMPLATE_BLOCK_IF_OPERATOR_START;
else
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
break;
}
*err = blogc_error_parser(BLOGC_ERROR_TEMPLATE_PARSER, src,
@@ -361,14 +416,14 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
break;
if (c == '"' && src[current - 1] == '\\')
break;
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
end2 = current + 1;
break;
case TEMPLATE_BLOCK_IF_VARIABLE_OPERAND:
if ((c >= 'A' && c <= 'Z') || (c >= '0' && c <= '9') || c == '_')
break;
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
end2 = current;
break;
@@ -391,7 +446,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
break;
if (c == ' ') {
end = current;
- state = TEMPLATE_BLOCK_END;
+ state = TEMPLATE_BLOCK_END_WHITESPACE_CLEANER;
break;
}
*err = blogc_error_parser(BLOGC_ERROR_TEMPLATE_PARSER, src,
@@ -400,13 +455,28 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
"number or '_'.");
break;
- case TEMPLATE_BLOCK_END:
+ case TEMPLATE_BLOCK_END_WHITESPACE_CLEANER:
if (c == ' ')
break;
+ if (c == '-') {
+ lstrip_next = true;
+ state = TEMPLATE_BLOCK_END;
+ break;
+ }
+ state = TEMPLATE_BLOCK_END;
+
+ case TEMPLATE_BLOCK_END:
if (c == '%') {
state = TEMPLATE_CLOSE_BRACKET;
break;
}
+ if (c == '-') {
+ *err = blogc_error_parser(BLOGC_ERROR_TEMPLATE_PARSER, src,
+ src_len, current,
+ "Invalid statement syntax. Duplicated whitespace "
+ "cleaner after statement.");
+ break;
+ }
*err = blogc_error_parser(BLOGC_ERROR_TEMPLATE_PARSER, src,
src_len, current,
"Invalid statement syntax. Must end with '%%}'.");
@@ -502,6 +572,7 @@ blogc_template_parse(const char *src, size_t src_len, blogc_error_t **err)
end2 = 0;
}
stmts = b_slist_append(stmts, stmt);
+ previous = stmt;
stmt = NULL;
state = TEMPLATE_START;
type = BLOGC_TEMPLATE_CONTENT_STMT;
diff --git a/src/template-parser.h b/src/template-parser.h
index 5add574..6cd2c80 100644
--- a/src/template-parser.h
+++ b/src/template-parser.h
@@ -12,6 +12,11 @@
#include "utils/utils.h"
#include "error.h"
+/*
+ * note: whitespace cleaners are NOT added to ast. we fix strings right during
+ * template parsing. renderer does not need to care about it, for the sake of
+ * simplicity.
+ */
typedef enum {
BLOGC_TEMPLATE_IFDEF_STMT = 1,
BLOGC_TEMPLATE_IFNDEF_STMT,
diff --git a/src/utils/strings.c b/src/utils/strings.c
index 6f10d56..846ae95 100644
--- a/src/utils/strings.c
+++ b/src/utils/strings.c
@@ -103,22 +103,44 @@ b_str_ends_with(const char *str, const char *suffix)
char*
-b_str_strip(char *str)
+b_str_lstrip(char *str)
{
if (str == NULL)
- return str;
+ return NULL;
+ int i;
+ size_t str_len = strlen(str);
+ for (i = 0; i < str_len; i++) {
+ if ((str[i] != ' ') && (str[i] != '\t') && (str[i] != '\n') &&
+ (str[i] != '\r') && (str[i] != '\t'))
+ {
+ str += i;
+ break;
+ }
+ if (i == str_len - 1) {
+ str += str_len;
+ break;
+ }
+ }
+ return str;
+}
+
+
+char*
+b_str_rstrip(char *str)
+{
+ if (str == NULL)
+ return NULL;
int i;
size_t str_len = strlen(str);
for (i = str_len - 1; i >= 0; i--) {
- if (!isspace(str[i])) {
+ if ((str[i] != ' ') && (str[i] != '\t') && (str[i] != '\n') &&
+ (str[i] != '\r') && (str[i] != '\t'))
+ {
str[i + 1] = '\0';
break;
}
- }
- str_len = strlen(str);
- for (i = 0; i < str_len; i++) {
- if (!isspace(str[i])) {
- str = str + i;
+ if (i == 0) {
+ str[0] = '\0';
break;
}
}
@@ -126,6 +148,13 @@ b_str_strip(char *str)
}
+char*
+b_str_strip(char *str)
+{
+ return b_str_lstrip(b_str_rstrip(str));
+}
+
+
char**
b_str_split(const char *str, char c, unsigned int max_pieces)
{
diff --git a/src/utils/utils.h b/src/utils/utils.h
index 49a7735..dc67497 100644
--- a/src/utils/utils.h
+++ b/src/utils/utils.h
@@ -50,6 +50,8 @@ char* b_strdup_vprintf(const char *format, va_list ap);
char* b_strdup_printf(const char *format, ...);
bool b_str_starts_with(const char *str, const char *prefix);
bool b_str_ends_with(const char *str, const char *suffix);
+char* b_str_lstrip(char *str);
+char* b_str_rstrip(char *str);
char* b_str_strip(char *str);
char** b_str_split(const char *str, char c, unsigned int max_pieces);
char* b_str_replace(const char *str, const char search, const char *replace);
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2024-12-31 20:24:16 +0000