+/*
+ Bacula(R) - The Network Backup Solution
+
+ Copyright (C) 2000-2015 Kern Sibbald
+ Copyright (C) 2000-2014 Free Software Foundation Europe e.V.
+
+ The original author of Bacula is Kern Sibbald, with contributions
+ from many others, a complete list can be found in the file AUTHORS.
+
+ You may use this file and others of this release according to the
+ license defined in the LICENSE file, which includes the Affero General
+ Public License, v3.0 ("AGPLv3") and some additional permissions and
+ terms pursuant to its AGPLv3 Section 7.
+
+ This notice must be preserved when any source code is
+ conveyed and/or propagated.
+
+ Bacula(R) is a registered trademark of Kern Sibbald.
+*/
/*
* Lexical scanner for Bacula configuration file
*
* Kern Sibbald, 2000
*
- * Version $Id$
- *
*/
-/*
- Bacula® - The Network Backup Solution
-
- Copyright (C) 2000-2006 Free Software Foundation Europe e.V.
-
- The main author of Bacula is Kern Sibbald, with contributions from
- many others, a complete list can be found in the file AUTHORS.
- This program is Free Software; you can redistribute it and/or
- modify it under the terms of version two of the GNU General Public
- License as published by the Free Software Foundation plus additions
- that are listed in the file LICENSE.
-
- This program is distributed in the hope that it will be useful, but
- WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- General Public License for more details.
-
- You should have received a copy of the GNU General Public License
- along with this program; if not, write to the Free Software
- Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
- 02110-1301, USA.
-
- Bacula® is a registered trademark of John Walker.
- The licensor of Bacula is the Free Software Foundation Europe
- (FSFE), Fiduciary Program, Sumatrastrasse 25, 8006 Zürich,
- Switzerland, email:ftf@fsfeurope.org.
-*/
#include "bacula.h"
#include "lex.h"
-extern int debug_level;
-
/* Debug level for this source file */
static const int dbglvl = 5000;
void scan_to_eol(LEX *lc)
{
int token;
- Dmsg0(dbglvl, "start scan to eof\n");
+ Dmsg0(dbglvl, "start scan to eol\n");
while ((token = lex_get_token(lc, T_ALL)) != T_EOL) {
if (token == T_EOB) {
lex_unget_char(lc);
return;
}
+ if (token == T_EOF) {
+ return;
+ }
}
}
_("Problem probably begins at line %d.\n"), lc->begin_line_no);
} else {
more[0] = 0;
- }
+ }
if (lc->line_no > 0) {
e_msg(file, line, lc->err_type, 0, _("Config error: %s\n"
" : line %d, col %d of file %s\n%s\n%s"),
Dmsg1(dbglvl, "Close lex file: %s\n", lf->fname);
of = lf->next;
- fclose(lf->fd);
+ if (lf->bpipe) {
+ close_bpipe(lf->bpipe);
+ lf->bpipe = NULL;
+ } else if (lf->fd) {
+ fclose(lf->fd);
+ }
Dmsg1(dbglvl, "Close cfg file %s\n", lf->fname);
- free(lf->fname);
+ if (lf->fname) {
+ free(lf->fname);
+ }
+ free_memory(lf->line);
+ lf->line = NULL;
+ free_memory(lf->str);
+ lf->str = NULL;
if (of) {
of->options = lf->options; /* preserve options */
memcpy(lf, of, sizeof(LEX));
of = lf;
lf = NULL;
}
- free(of);
+ if (of) {
+ free(of);
+ }
+ return lf;
+}
+
+/*
+ * Open a configuration in memory buffer. We push the
+ * state of the current file (lf) so that we
+ * can do includes. This is a bit of a hammer.
+ * Instead of passing back the pointer to the
+ * new packet, I simply replace the contents
+ * of the caller's packet with the new packet,
+ * and link the contents of the old packet into
+ * the next field.
+ *
+ */
+LEX *lex_open_buf(LEX *lf, const char *buffer, LEX_ERROR_HANDLER *scan_error)
+
+{
+ LEX *nf;
+
+ Dmsg0(400, "Open config buffer\n");
+ nf = (LEX *)malloc(sizeof(LEX));
+ if (lf) {
+ memcpy(nf, lf, sizeof(LEX));
+ memset(lf, 0, sizeof(LEX));
+ lf->next = nf; /* if have lf, push it behind new one */
+ lf->options = nf->options; /* preserve user options */
+ /*
+ * preserve err_type to prevent bacula exiting on 'reload'
+ * if config is invalid. Fixes bug #877
+ */
+ lf->err_type = nf->err_type;
+ } else {
+ lf = nf; /* start new packet */
+ memset(lf, 0, sizeof(LEX));
+ lex_set_error_handler_error_type(lf, M_ERROR_TERM);
+ }
+ if (scan_error) {
+ lf->scan_error = scan_error;
+ } else {
+ lex_set_default_error_handler(lf);
+ }
+ lf->fd = NULL;
+ lf->bpipe = NULL;
+ lf->fname = NULL;
+ lf->line = get_memory(5000);
+ pm_strcpy(lf->line, buffer);
+ pm_strcat(lf->line, "");
+ lf->state = lex_none;
+ lf->ch = 0;
+ lf->str = get_memory(5000);
return lf;
}
{
LEX *nf;
FILE *fd;
+ BPIPE *bpipe = NULL;
char *fname = bstrdup(filename);
-
- if ((fd = fopen(fname, "rb")) == NULL) {
+ if (fname[0] == '|') {
+ if ((bpipe = open_bpipe(fname+1, 0, "rb")) == NULL) {
+ free(fname);
+ return NULL;
+ }
+ fd = bpipe->rfd;
+ } else if ((fd = fopen(fname, "rb")) == NULL) {
+ free(fname);
return NULL;
}
Dmsg1(400, "Open config file: %s\n", fname);
memset(lf, 0, sizeof(LEX));
lf->next = nf; /* if have lf, push it behind new one */
lf->options = nf->options; /* preserve user options */
+ /*
+ * preserve err_type to prevent bacula exiting on 'reload'
+ * if config is invalid. Fixes bug #877
+ */
+ lf->err_type = nf->err_type;
} else {
lf = nf; /* start new packet */
memset(lf, 0, sizeof(LEX));
lex_set_default_error_handler(lf);
}
lf->fd = fd;
+ lf->bpipe = bpipe;
lf->fname = fname;
+ lf->line = get_memory(5000);
lf->state = lex_none;
lf->ch = L_EOL;
+ lf->str = get_memory(5000);
Dmsg1(dbglvl, "Return lex=%x\n", lf);
return lf;
}
int lex_get_char(LEX *lf)
{
if (lf->ch == L_EOF) {
- Emsg0(M_ABORT, 0, _("get_char: called after EOF\n"));
+ Emsg0(M_ABORT, 0, _("get_char: called after EOF."
+ " You may have a open double quote without the closing double quote.\n"));
}
- if (lf->ch == L_EOL) {
- if (bfgets(lf->line, MAXSTRING, lf->fd) == NULL) {
+ if (lf->fd && lf->ch == L_EOL) {
+ if (bfgets(lf->line, lf->fd) == NULL) {
lf->ch = L_EOF;
if (lf->next) {
lex_close_file(lf);
lf->line_no++;
lf->col_no = 0;
Dmsg2(1000, "fget line=%d %s", lf->line_no, lf->line);
+ } else if (lf->ch == L_EOL) {
+ lf->line_no++;
+ lf->col_no++;
}
lf->ch = (uint8_t)lf->line[lf->col_no];
- if (lf->ch == 0) {
- lf->ch = L_EOL;
+ if (lf->fd) {
+ if (lf->ch == 0) {
+ lf->ch = L_EOL; /* reached end of line, force bfgets */
+ } else {
+ lf->col_no++;
+ }
} else {
- lf->col_no++;
+ if (lf->ch == 0) { /* End of buffer, stop scan */
+ lf->ch = L_EOF;
+ if (lf->next) {
+ lex_close_file(lf);
+ }
+ return lf->ch;
+ } else if (lf->ch == '\n') { /* End of line */
+ Dmsg0(dbglvl, "Found newline return L_EOL\n");
+ lf->ch = L_EOL;
+ } else {
+ lf->col_no++;
+ }
}
- Dmsg2(dbglvl, "lex_get_char: %c %d\n", lf->ch, lf->ch);
+ Dmsg3(dbglvl, "lex_get_char: %c %d col=%d\n", lf->ch, lf->ch, lf->col_no);
return lf->ch;
}
} else {
lf->col_no--; /* Backup to re-read char */
}
-
}
*/
static void add_str(LEX *lf, int ch)
{
- if (lf->str_len >= MAXSTRING-3) {
+ if (lf->str_len >= sizeof_pool_memory(lf->str)) {
Emsg3(M_ERROR_TERM, 0, _(
_("Config token too long, file: %s, line %d, begins at line %d\n")),
lf->fname, lf->line_no, lf->begin_line_no);
case lex_identifier: return _("identifier");
case lex_string: return _("string");
case lex_quoted_string: return _("quoted_string");
+ case lex_include: return _("include");
+ case lex_include_quoted_string: return _("include_quoted_string");
+ case lex_utf8_bom: return _("UTF-8 Byte Order Mark");
+ case lex_utf16_le_bom: return _("UTF-16le Byte Order Mark");
default: return "??????";
}
}
case T_EOF: return "T_EOF";
case T_COMMA: return "T_COMMA";
case T_EOL: return "T_EOL";
+ case T_UTF8_BOM: return "T_UTF8_BOM";
+ case T_UTF16_BOM: return "T_UTF16_BOM";
default: return "??????";
}
}
return (uint32_t)val;
}
+static uint64_t scan_pint64(LEX *lf, char *str)
+{
+ uint64_t val = 0;
+ if (!is_a_number(str)) {
+ scan_err1(lf, _("expected a positive integer number, got: %s"), str);
+ /* NOT REACHED */
+ } else {
+ errno = 0;
+ val = str_to_uint64(str);
+ if (errno != 0) {
+ scan_err1(lf, _("expected a positive integer number, got: %s"), str);
+ /* NOT REACHED */
+ }
+ }
+ return val;
+}
+
/*
*
* Get the next token from the input
int ch;
int token = T_NONE;
bool esc_next = false;
- int unicode_count = 0;
-
- Dmsg0(dbglvl, "enter lex_get_token\n");
+ /* Unicode files, especially on Win32, may begin with a "Byte Order Mark"
+ to indicate which transmission format the file is in. The codepoint for
+ this mark is U+FEFF and is represented as the octets EF-BB-BF in UTF-8
+ and as FF-FE in UTF-16le(little endian) and FE-FF in UTF-16(big endian).
+ We use a distinct state for UTF-8 and UTF-16le, and use bom_bytes_seen
+ to tell which byte we are expecting. */
+ int bom_bytes_seen = 0;
+
+ Dmsg1(dbglvl, "enter lex_get_token state=%s\n", lex_state_to_str(lf->state));
while (token == T_NONE) {
ch = lex_get_char(lf);
switch (lf->state) {
case lex_none:
- Dmsg2(dbglvl, "Lex state lex_none ch=%d,%x\n", ch, ch);
+ Dmsg2(dbglvl, "Lex state lex_none ch=%c,%d\n", ch, ch);
if (B_ISSPACE(ch))
break;
if (B_ISALPHA(ch)) {
}
break;
case '@':
- lf->state = lex_include;
- begin_str(lf, 0);
+ /* In NO_EXTERN mode, @ is part of a string */
+ if (lf->options & LOPT_NO_EXTERN) {
+ lf->state = lex_string;
+ begin_str(lf, ch);
+ } else {
+ lf->state = lex_include;
+ begin_str(lf, 0);
+ }
break;
- case 0xEF:
+ case 0xEF: /* probably a UTF-8 BOM */
+ case 0xFF: /* probably a UTF-16le BOM */
+ case 0xFE: /* probably a UTF-16be BOM (error)*/
if (lf->line_no != 1 || lf->col_no != 1)
{
lf->state = lex_string;
begin_str(lf, ch);
- break;
+ } else {
+ bom_bytes_seen = 1;
+ if (ch == 0xEF) {
+ lf->state = lex_utf8_bom;
+ } else if (ch == 0xFF) {
+ lf->state = lex_utf16_le_bom;
+ } else {
+ scan_err0(lf, _("This config file appears to be in an "
+ "unsupported Unicode format (UTF-16be). Please resave as UTF-8\n"));
+ return T_ERROR;
+ }
}
- lf->state = lex_unicode_mark;
- unicode_count = 1;
break;
default:
lf->state = lex_string;
}
if (ch == '"') {
token = T_QUOTED_STRING;
+ /*
+ * Since we may be scanning a quoted list of names,
+ * we get the next character (a comma indicates another
+ * one), then we put it back for rescanning.
+ */
+ lex_get_char(lf);
+ lex_unget_char(lf);
lf->state = lex_none;
break;
}
add_str(lf, ch);
break;
- case lex_include: /* scanning a filename */
+ case lex_include_quoted_string:
if (ch == L_EOF) {
token = T_ERROR;
break;
}
- if (B_ISSPACE(ch) || ch == '\n' || ch == L_EOL || ch == '}' || ch == '{' ||
- ch == ';' || ch == ',' || ch == '"' || ch == '#') {
+ if (esc_next) {
+ add_str(lf, ch);
+ esc_next = false;
+ break;
+ }
+ if (ch == '\\') {
+ esc_next = true;
+ break;
+ }
+ if (ch == '"') {
/* Keep the original LEX so we can print an error if the included file can't be opened. */
LEX* lfori = lf;
-
+ /* Skip the double quote when restarting parsing */
+ lex_get_char(lf);
+
lf->state = lex_none;
lf = lex_open_file(lf, lf->str, lf->scan_error);
if (lf == NULL) {
berrno be;
scan_err2(lfori, _("Cannot open included config file %s: %s\n"),
- lfori->str, be.strerror());
+ lfori->str, be.bstrerror());
return T_ERROR;
}
break;
}
add_str(lf, ch);
break;
- case lex_unicode_mark:
+ case lex_include: /* scanning a filename */
if (ch == L_EOF) {
token = T_ERROR;
break;
}
- unicode_count++;
- if (unicode_count == 2) {
- if (ch != 0xBB) {
- token = T_ERROR;
- break;
- }
- } else if (unicode_count == 3) {
- if (ch != 0xBF) {
- token = T_ERROR;
- break;
- }
- token = T_UNICODE_MARK;
+ if (ch == '"') {
+ lf->state = lex_include_quoted_string;
+ break;
+ }
+
+
+ if (B_ISSPACE(ch) || ch == '\n' || ch == L_EOL || ch == '}' || ch == '{' ||
+ ch == ';' || ch == ',' || ch == '"' || ch == '#') {
+ /* Keep the original LEX so we can print an error if the included file can't be opened. */
+ LEX* lfori = lf;
+
lf->state = lex_none;
+ lf = lex_open_file(lf, lf->str, lf->scan_error);
+ if (lf == NULL) {
+ berrno be;
+ scan_err2(lfori, _("Cannot open included config file %s: %s\n"),
+ lfori->str, be.bstrerror());
+ return T_ERROR;
+ }
break;
}
+ add_str(lf, ch);
+ break;
+ case lex_utf8_bom:
+ /* we only end up in this state if we have read an 0xEF
+ as the first byte of the file, indicating we are probably
+ reading a UTF-8 file */
+ if (ch == 0xBB && bom_bytes_seen == 1) {
+ bom_bytes_seen++;
+ } else if (ch == 0xBF && bom_bytes_seen == 2) {
+ token = T_UTF8_BOM;
+ lf->state = lex_none;
+ } else {
+ token = T_ERROR;
+ }
+ break;
+ case lex_utf16_le_bom:
+ /* we only end up in this state if we have read an 0xFF
+ as the first byte of the file -- indicating that we are
+ probably dealing with an Intel based (little endian) UTF-16 file*/
+ if (ch == 0xFE) {
+ token = T_UTF16_BOM;
+ lf->state = lex_none;
+ } else {
+ token = T_ERROR;
+ }
break;
}
Dmsg4(dbglvl, "ch=%d state=%s token=%s %c\n", ch, lex_state_to_str(lf->state),
}
break;
+ case T_PINT64_RANGE:
+ if (token == T_NUMBER) {
+ lf->pint64_val = scan_pint64(lf, lf->str);
+ lf->pint64_val2 = lf->pint64_val;
+ token = T_PINT64;
+ } else {
+ char *p = strchr(lf->str, '-');
+ if (!p) {
+ scan_err2(lf, _("expected an integer or a range, got %s: %s"),
+ lex_tok_to_str(token), lf->str);
+ token = T_ERROR;
+ break;
+ }
+ *p++ = 0; /* terminate first half of range */
+ lf->pint64_val = scan_pint64(lf, lf->str);
+ lf->pint64_val2 = scan_pint64(lf, p);
+ token = T_PINT64_RANGE;
+ }
+ break;
+
case T_NAME:
if (token != T_IDENTIFIER && token != T_UNQUOTED_STRING && token != T_QUOTED_STRING) {
scan_err2(lf, _("expected a name, got %s: %s"),