hyung-hwan/qse
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

enhanced sed escape handling

Browse Source 
fixed an issue of adding an extra newline after matching $ in do_subst() of sed.
changed tre slightly for QSE's own needs
This commit is contained in:
hyung-hwan
2011-09-05 10:21:54 +00:00
parent 3fe32033cb
commit 2f15ca2335
8 changed files with 599 additions and 561 deletions
Download Patch File Download Diff File Expand all files Collapse all files

225
qse/lib/sed/sed.c
View File

@ -1,5 +1,5 @@
/*
* $Id: sed.c 558 2011-09-02 15:27:44Z hyunghwan.chung $
* $Id: sed.c 559 2011-09-04 16:21:54Z hyunghwan.chung $
*
Copyright 2006-2011 Chung, Hyung-Hwan.
This file is part of QSE.
@ -177,7 +177,11 @@ static qse_tre_t* maketre (
if (qse_tre_compx (tre, str->ptr, str->len, QSE_NULL,
((sed->option & QSE_SED_EXTENDEDREX)? QSE_TRE_EXTENDED: 0)) <= -1)
{
SETERR1 (sed, QSE_SED_EREXBL, str->ptr, str->len, loc);
qse_sed_errnum_t errnum;
errnum = (QSE_TRE_ERRNUM(tre) == QSE_TRE_ENOMEM)?
QSE_TRE_ENOMEM: QSE_SED_EREXBL;
SETERR1 (sed, errnum, str->ptr, str->len, loc);
qse_tre_close (tre);
return QSE_NULL;
}
@ -201,8 +205,13 @@ static int matchtre (
n = qse_tre_execx (tre, str->ptr, str->len, match, 10, opt);
if (n <= -1)
{
qse_sed_errnum_t errnum;
if (QSE_TRE_ERRNUM(tre) == QSE_TRE_ENOMATCH) return 0;
SETERR0 (sed, QSE_SED_EREXMA, loc);
errnum = (QSE_TRE_ERRNUM(tre) == QSE_TRE_ENOMEM)?
QSE_TRE_ENOMEM: QSE_SED_EREXMA;
SETERR0 (sed, errnum, loc);
return -1;
}
@ -374,6 +383,39 @@ static void free_command (qse_sed_t* sed, qse_sed_cmd_t* cmd)
}
}
static qse_cint_t trans_escaped (qse_cint_t c)
{
switch (c)
{
case QSE_T('a'):
c = QSE_T('\a');
break;
/*
Omitted for clash with regular expression \b.
case QSE_T('b'):
c = QSE_T('\b');
break;
*/
case QSE_T('f'):
c = QSE_T('\f');
case QSE_T('n'):
c = QSE_T('\n');
break;
case QSE_T('r'):
c = QSE_T('\r');
break;
case QSE_T('t'):
c = QSE_T('\t');
break;
case QSE_T('v'):
c = QSE_T('\v');
break;
}
return c;
}
static void* compile_rex (qse_sed_t* sed, qse_char_t rxend)
{
#ifdef USE_REX
@ -401,8 +443,10 @@ static void* compile_rex (qse_sed_t* sed, qse_char_t rxend)
if (c == QSE_T('\\'))
{
c = NXTSC (sed);
if (c == QSE_CHAR_EOF || c == QSE_T('\n'))
qse_cint_t nc;
nc = NXTSC (sed);
if (nc == QSE_CHAR_EOF /*|| nc == QSE_T('\n')*/)
{
SETERR1 (
sed, QSE_SED_EREXIC,
@ -413,8 +457,25 @@ static void* compile_rex (qse_sed_t* sed, qse_char_t rxend)
return QSE_NULL;
}
if (c == QSE_T('n')) c = QSE_T('\n');
/* TODO: support more escaped characters?? */
if (nc == QSE_T('\n')) c = nc;
else
{
qse_cint_t ec;
ec = trans_escaped (nc);
if (ec == nc)
{
/* if the character after a backslash is not special at the
* this layer, add the backslash into the regular expression
* buffer as it is. */
if (qse_str_ccat (&sed->tmp.rex, QSE_T('\\')) == (qse_size_t)-1)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
return QSE_NULL;
}
}
c = ec;
}
}
if (qse_str_ccat (&sed->tmp.rex, c) == (qse_size_t)-1)
@ -824,6 +885,16 @@ do { \
} \
} while (0)
#define CHECK_CMDIC_ESCAPED(sed,cmd,c,action) \
do { \
if (c == QSE_CHAR_EOF) \
{ \
SETERR1 (sed, QSE_SED_ECMDIC, \
&cmd->type, 1, &sed->src.loc); \
action; \
} \
} while (0)
static int get_subst (qse_sed_t* sed, qse_sed_cmd_t* cmd)
{
qse_cint_t c, delim;
@ -859,14 +930,39 @@ static int get_subst (qse_sed_t* sed, qse_sed_cmd_t* cmd)
{
CHECK_CMDIC (sed, cmd, c, goto oops);
#if 0
if (c == QSE_T('\\'))
{
c = NXTSC (sed);
CHECK_CMDIC (sed, cmd, c, goto oops);
if (c == QSE_T('n')) c = QSE_T('\n');
qse_cint_t nc;
nc = NXTSC (sed);
CHECK_CMDIC_ESCAPED (sed, cmd, nc, goto oops);
if (nc == QSE_T('\n')) c = nc;
else
{
qse_cint_t ec;
/* Escaping a known speical character for the regular expression
* part is done here. However, Escaping a special character for
* the replacement part is done in do_subst() except '\n' because
* it has more special characters like '&'. */
ec = trans_escaped (nc);
if (ec == nc)
{
/* if the character after a backslash is not special at the
* this layer, add the backslash into the regular expression
* buffer as it is. */
if (qse_str_ccat (t[i], QSE_T('\\')) == (qse_size_t)-1)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
goto oops;
}
}
c = ec;
}
}
#endif
if (qse_str_ccat (t[i], c) == (qse_size_t)-1)
{
@ -1017,8 +1113,8 @@ static int get_transet (qse_sed_t* sed, qse_sed_cmd_t* cmd)
if (c == QSE_T('\\'))
{
c = NXTSC (sed);
CHECK_CMDIC (sed, cmd, c, goto oops);
if (c == QSE_T('n')) c = QSE_T('\n');
CHECK_CMDIC_ESCAPED (sed, cmd, c, goto oops);
c = trans_escaped (c);
}
b[0] = c;
@ -1039,8 +1135,8 @@ static int get_transet (qse_sed_t* sed, qse_sed_cmd_t* cmd)
if (c == QSE_T('\\'))
{
c = NXTSC (sed);
CHECK_CMDIC (sed, cmd, c, goto oops);
if (c == QSE_T('n')) c = QSE_T('\n');
CHECK_CMDIC_ESCAPED (sed, cmd, c, goto oops);
c = trans_escaped (c);
}
if (pos >= QSE_STR_LEN(t))
@ -1659,10 +1755,21 @@ static int write_char (qse_sed_t* sed, qse_char_t c)
static int write_str (qse_sed_t* sed, const qse_char_t* str, qse_size_t len)
{
qse_size_t i;
int flush_needed = 0;
for (i = 0; i < len; i++)
{
if (write_char (sed, str[i]) <= -1) return -1;
/*if (write_char (sed, str[i]) <= -1) return -1;*/
sed->e.out.buf[sed->e.out.len++] = str[i];
if (sed->e.out.len >= QSE_COUNTOF(sed->e.out.buf))
{
if (flush (sed) <= -1) return -1;
flush_needed = 0;
}
else if (str[i] == QSE_T('\n')) flush_needed = 1;
}
if (flush_needed && flush(sed) <= -1) return -1;
return 0;
}
@ -1913,7 +2020,7 @@ static int do_subst (qse_sed_t* sed, qse_sed_cmd_t* cmd)
/* TODO: support different line end convension */
if (str.len > 0 && str.ptr[str.len-1] == QSE_T('\n')) str.len--;
str_end = str.ptr + str.len;
cur = str;
@ -1984,27 +2091,33 @@ static int do_subst (qse_sed_t* sed, qse_sed_cmd_t* cmd)
if (max_count > 0 && sub_count + 1 != max_count)
{
m = qse_str_ncat (
&sed->e.txt.subst,
cur.ptr, mat.ptr-cur.ptr+mat.len
);
if (m == (qse_size_t)-1)
if (cur.ptr < str_end)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
return -1;
m = qse_str_ncat (
&sed->e.txt.subst,
cur.ptr, mat.ptr-cur.ptr+mat.len
);
if (m == (qse_size_t)-1)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
return -1;
}
}
}
else
{
repl = 1;
m = qse_str_ncat (
&sed->e.txt.subst, cur.ptr, mat.ptr-cur.ptr);
if (m == (qse_size_t)-1)
if (cur.ptr < str_end)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
return -1;
m = qse_str_ncat (
&sed->e.txt.subst, cur.ptr, mat.ptr-cur.ptr
);
if (m == (qse_size_t)-1)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
return -1;
}
}
for (i = 0; i < cmd->u.subst.rpl.len; i++)
@ -2018,40 +2131,20 @@ static int do_subst (qse_sed_t* sed, qse_sed_cmd_t* cmd)
if (nc >= QSE_T('1') && nc <= QSE_T('9'))
{
int smi = nc - QSE_T('1');
m = qse_str_ncat (&sed->e.txt.subst, submat[smi].ptr, submat[smi].len);
m = qse_str_ncat (
&sed->e.txt.subst,
submat[smi].ptr, submat[smi].len
);
}
else
{
#endif
switch (nc)
{
case QSE_T('n'):
nc = QSE_T('\n');
break;
case QSE_T('r'):
nc = QSE_T('\r');
break;
case QSE_T('t'):
nc = QSE_T('\t');
break;
case QSE_T('f'):
nc = QSE_T('\f');
break;
case QSE_T('b'):
nc = QSE_T('\b');
break;
case QSE_T('v'):
nc = QSE_T('\v');
break;
case QSE_T('a'):
nc = QSE_T('\a');
break;
#ifndef USE_REX
}
#endif
/* the know speical characters have been escaped
* in get_subst(). so i don't call trans_escaped() here */
m = qse_str_ccat (&sed->e.txt.subst, nc);
#ifndef USE_REX
}
#endif
i++;
}
@ -2085,13 +2178,15 @@ static int do_subst (qse_sed_t* sed, qse_sed_cmd_t* cmd)
if (mat.len == 0)
{
skip_one_char:
/* special treament is need if the match length is 0 */
m = qse_str_ncat (&sed->e.txt.subst, cur.ptr, 1);
if (m == (qse_size_t)-1)
if (cur.ptr < str_end)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
return -1;
/* special treament is needed if the match length is 0 */
m = qse_str_ncat (&sed->e.txt.subst, cur.ptr, 1);
if (m == (qse_size_t)-1)
{
SETERR0 (sed, QSE_SED_ENOMEM, QSE_NULL);
return -1;
}
}
cur.ptr++; cur.len--;