Revision: 787
http://www.exim.org/viewvc/pcre2?view=rev&revision=787
Author: zherczeg
Date: 2017-05-18 07:46:22 +0100 (Thu, 18 May 2017)
Log Message:
-----------
Support character classes in glob conversion.
Modified Paths:
--------------
code/trunk/src/pcre2_convert.c
code/trunk/testdata/testinput24
code/trunk/testdata/testoutput24
Modified: code/trunk/src/pcre2_convert.c
===================================================================
--- code/trunk/src/pcre2_convert.c 2017-05-17 17:55:11 UTC (rev 786)
+++ code/trunk/src/pcre2_convert.c 2017-05-18 06:46:22 UTC (rev 787)
@@ -58,7 +58,7 @@
#define ERROR_END_BACKSLASH 101
#define ERROR_MISSING_SQUARE_BRACKET 106
#define ERROR_MISSING_CLOSING_PARENTHESIS 114
-#define ERROR_TOO_DEEP_NESTING 119
+#define ERROR_UNKNOWN_POSIX_CLASS 130
#define ERROR_NO_UNICODE 132
/* Generated pattern fragments */
@@ -651,17 +651,17 @@
/* Write a character into the output.
Arguments:
- context the bash glob context
+ out output context
chr the next character
*/
static void
-convert_glob_bash_write(pcre2_output_context *context, PCRE2_UCHAR chr)
+convert_glob_bash_write(pcre2_output_context *out, PCRE2_UCHAR chr)
{
-context->output_size++;
+out->output_size++;
-if (context->output < context->output_end)
- *context->output++ = chr;
+if (out->output < out->output_end)
+ *out->output++ = chr;
}
@@ -668,17 +668,17 @@
/* Write a string into the output.
Arguments:
- context the bash glob context
- length length of context->out_str
+ out output context
+ length length of out->out_str
*/
static void
-convert_glob_bash_write_str(pcre2_output_context *context, PCRE2_SIZE length)
+convert_glob_bash_write_str(pcre2_output_context *out, PCRE2_SIZE length)
{
-uint8_t *out_str = context->out_str;
-PCRE2_UCHAR *output = context->output;
-PCRE2_SPTR output_end = context->output_end;
-PCRE2_SIZE output_size = context->output_size;
+uint8_t *out_str = out->out_str;
+PCRE2_UCHAR *output = out->output;
+PCRE2_SPTR output_end = out->output_end;
+PCRE2_SIZE output_size = out->output_size;
do
{
@@ -689,17 +689,219 @@
}
while (--length != 0);
-context->output = output;
-context->output_size = output_size;
+out->output = output;
+out->output_size = output_size;
}
+/* Parse a posix class.
+
+Arguments:
+ from starting point of scanning the range
+ pattern_end end of pattern
+ out output context
+
+Returns: TRUE => success
+ FALSE => malformed class
+*/
+
+static int
+convert_glob_bash_parse_class(PCRE2_SPTR *from, PCRE2_SPTR pattern_end,
+ pcre2_output_context *out)
+{
+static const char *posix_classes = "alnum:alpha:ascii:blank:cntrl:digit:"
+ "graph:lower:print:punct:space:upper:word:xdigit:";
+PCRE2_SPTR pattern = *from;
+PCRE2_SPTR start;
+const char *class_ptr;
+PCRE2_UCHAR c;
+
+out->out_str[0] = CHAR_LEFT_SQUARE_BRACKET;
+out->out_str[1] = CHAR_COLON;
+convert_glob_bash_write_str(out, 2);
+
+while (TRUE)
+ {
+ if (pattern >= pattern_end)
+ {
+ *from = pattern;
+ return ERROR_MISSING_SQUARE_BRACKET;
+ }
+
+ c = *pattern++;
+
+ if (c == CHAR_COLON && pattern < pattern_end &&
+ *pattern == CHAR_RIGHT_SQUARE_BRACKET)
+ {
+ break;
+ }
+
+ if (c < CHAR_a || c > CHAR_z)
+ {
+ /* All POSIX class is composed of lowercase characters */
+ *from = pattern;
+ return ERROR_MISSING_SQUARE_BRACKET;
+ }
+
+ convert_glob_bash_write(out, c);
+ }
+
+start = *from;
+*from = pattern + 1;
+class_ptr = posix_classes;
+
+while (TRUE)
+ {
+ if (*class_ptr == CHAR_NULL) return ERROR_UNKNOWN_POSIX_CLASS;
+
+ pattern = start;
+
+ while (*pattern == (PCRE2_UCHAR) *class_ptr)
+ {
+ if (*pattern == CHAR_COLON)
+ {
+ out->out_str[0] = CHAR_COLON;
+ out->out_str[1] = CHAR_RIGHT_SQUARE_BRACKET;
+ convert_glob_bash_write_str(out, 2);
+ return 0;
+ }
+ pattern++;
+ class_ptr++;
+ }
+
+ while (*class_ptr != CHAR_COLON) class_ptr++;
+ class_ptr++;
+ }
+}
+
+
+/* Parse a range of characters.
+
+Arguments:
+ from starting point of scanning the range
+ pattern_end end of pattern
+ out output context
+ separator glob separator
+
+Returns: 0 => success
+ !0 => error code
+*/
+
+static int
+convert_glob_bash_parse_range(PCRE2_SPTR *from, PCRE2_SPTR pattern_end,
+ pcre2_output_context *out, PCRE2_UCHAR separator)
+{
+PCRE2_SPTR pattern = *from;
+PCRE2_UCHAR c;
+int result, len;
+
+if (pattern >= pattern_end)
+ {
+ *from = pattern;
+ return ERROR_MISSING_SQUARE_BRACKET;
+ }
+
+c = *pattern;
+
+if (c == CHAR_EXCLAMATION_MARK
+ || c == CHAR_CIRCUMFLEX_ACCENT)
+ {
+ out->out_str[0] = CHAR_LEFT_SQUARE_BRACKET;
+ out->out_str[1] = CHAR_CIRCUMFLEX_ACCENT;
+ len = 2;
+ }
+else
+ {
+ out->out_str[0] = CHAR_LEFT_PARENTHESIS;
+ out->out_str[1] = CHAR_QUESTION_MARK;
+ out->out_str[2] = CHAR_EXCLAMATION_MARK;
+ len = 3;
+ }
+
+if (separator < 128 && strchr(pcre2_escaped_literals, separator) != NULL)
+ {
+ out->out_str[len] = CHAR_BACKSLASH;
+ len++;
+ }
+
+convert_glob_bash_write_str(out, len);
+convert_glob_bash_write(out, separator);
+
+if (c == CHAR_EXCLAMATION_MARK
+ || c == CHAR_CIRCUMFLEX_ACCENT)
+ {
+ pattern++;
+ if (pattern >= pattern_end)
+ {
+ *from = pattern;
+ return ERROR_MISSING_SQUARE_BRACKET;
+ }
+ c = *pattern;
+ }
+else
+ {
+ out->out_str[0] = CHAR_RIGHT_PARENTHESIS;
+ out->out_str[1] = CHAR_LEFT_SQUARE_BRACKET;
+ convert_glob_bash_write_str(out, 2);
+ }
+
+if (c == CHAR_MINUS || c == CHAR_RIGHT_SQUARE_BRACKET)
+ {
+ convert_glob_bash_write(out, CHAR_BACKSLASH);
+ convert_glob_bash_write(out, c);
+ pattern++;
+ }
+
+while (pattern < pattern_end)
+ {
+ c = *pattern++;
+
+ if (c == CHAR_RIGHT_SQUARE_BRACKET)
+ {
+ convert_glob_bash_write(out, c);
+ *from = pattern;
+ return 0;
+ }
+
+ if (c == CHAR_LEFT_SQUARE_BRACKET && pattern < pattern_end &&
+ *pattern == CHAR_COLON)
+ {
+ *from = pattern + 1;
+
+ result = convert_glob_bash_parse_class(from, pattern_end, out);
+ if (result != 0) return result;
+
+ pattern = *from;
+ continue;
+ }
+
+ if (c == CHAR_BACKSLASH)
+ {
+ if (pattern >= pattern_end)
+ {
+ *from = pattern;
+ return ERROR_END_BACKSLASH;
+ }
+ c = *pattern++;
+ }
+
+ if (c == CHAR_LEFT_SQUARE_BRACKET || c == CHAR_RIGHT_SQUARE_BRACKET ||
+ c == CHAR_BACKSLASH || c == CHAR_MINUS)
+ convert_glob_bash_write(out, CHAR_BACKSLASH);
+
+ convert_glob_bash_write(out, c);
+ }
+
+*from = pattern;
+return ERROR_MISSING_SQUARE_BRACKET;
+}
+
+
/* Prints a wildcard into the output.
Arguments:
- context the bash glob context
- separator glob separator
- after_sep whether the wildcard is right after a separator
+ out output context
+ separator glob separator
*/
static void
@@ -711,7 +913,7 @@
out->out_str[0] = CHAR_LEFT_SQUARE_BRACKET;
out->out_str[1] = CHAR_CIRCUMFLEX_ACCENT;
-if (separator == CHAR_BACKSLASH)
+if (separator < 128 && strchr(pcre2_escaped_literals, separator) != NULL)
{
out->out_str[2] = CHAR_BACKSLASH;
len = 3;
@@ -749,8 +951,8 @@
pcre2_output_context out;
PCRE2_SPTR pattern_start = pattern;
PCRE2_SPTR pattern_end = pattern + plength;
+PCRE2_UCHAR c;
int result;
-PCRE2_UCHAR c;
/* Initialize default for error offset as end of input. */
out.output = use_buffer;
@@ -800,6 +1002,14 @@
continue;
}
+ if (c == CHAR_LEFT_SQUARE_BRACKET)
+ {
+ result = convert_glob_bash_parse_range(&pattern, pattern_end,
+ &out, ccontext->glob_separator);
+ if (result != 0) break;
+ continue;
+ }
+
if (c == CHAR_BACKSLASH)
{
if (pattern >= pattern_end)
Modified: code/trunk/testdata/testinput24
===================================================================
--- code/trunk/testdata/testinput24 2017-05-17 17:55:11 UTC (rev 786)
+++ code/trunk/testdata/testinput24 2017-05-18 06:46:22 UTC (rev 787)
@@ -228,7 +228,23 @@
/?a?\/?b?/
+/[a\\b\c][]][-][\]\-]/
+/[^a\\b\c][!]][!-][^\]\-]/
+
+/[[:alpha:][:xdigit:][:word:]]/
+
+/[[:alpha:/
+
+/[[:alpha:]/
+
+/[[:alphaa:]]/
+
+/[[:xdigi:]]/
+
+/[[:xdigit::]]/
+
+
#pattern convert=unset
#pattern convert=posix_extended
Modified: code/trunk/testdata/testoutput24
===================================================================
--- code/trunk/testdata/testoutput24 2017-05-17 17:55:11 UTC (rev 786)
+++ code/trunk/testdata/testoutput24 2017-05-18 06:46:22 UTC (rev 787)
@@ -359,7 +359,31 @@
/?a?\/?b?/
(?s)\A[^/]a[^/]/[^/]b[^/]\z
+/[a\\b\c][]][-][\]\-]/
+(?s)\A(?!/)[a\\bc](?!/)[\]](?!/)[\-](?!/)[\]\-]\z
+/[^a\\b\c][!]][!-][^\]\-]/
+(?s)\A[^/a\\bc][^/\]][^/\-][^/\]\-]\z
+
+/[[:alpha:][:xdigit:][:word:]]/
+(?s)\A(?!/)[[:alpha:][:xdigit:][:word:]]\z
+
+/[[:alpha:/
+** Pattern conversion error at offset 9: missing terminating ] for character class
+
+/[[:alpha:]/
+** Pattern conversion error at offset 10: missing terminating ] for character class
+
+/[[:alphaa:]]/
+** Pattern conversion error at offset 11: unknown POSIX class name
+
+/[[:xdigi:]]/
+** Pattern conversion error at offset 10: unknown POSIX class name
+
+/[[:xdigit::]]/
+** Pattern conversion error at offset 10: missing terminating ] for character class
+
+
#pattern convert=unset
#pattern convert=posix_extended
