static int
parse_backtick(char **word, size_t * word_length, size_t * max_length,
const char *words, size_t * offset, int flags,
wordexp_t * pwordexp, const char *ifs,
const char *ifs_white)
{
/* We are poised just after "`" */
int error;
int squoting = 0;
size_t comm_length;
size_t comm_maxlen;
char *comm = w_newword(&comm_length, &comm_maxlen);
for (; words[*offset]; ++(*offset)) {
switch (words[*offset]) {
case '`':
/* Go -- give the script to the shell */
error = exec_comm(comm, word, word_length, max_length, flags,
pwordexp, ifs, ifs_white);
free(comm);
return error;
case '\\':
if (squoting) {
error = parse_qtd_backslash(&comm, &comm_length, &comm_maxlen,
words, offset);
if (error) {
free(comm);
return error;
}
break;
}
++(*offset);
error = parse_backslash(&comm, &comm_length, &comm_maxlen, words,
offset);
if (error) {
free(comm);
return error;
}
break;
case '\'':
squoting = 1 - squoting;
default:
comm = w_addchar(comm, &comm_length, &comm_maxlen,
words[*offset]);
if (comm == NULL)
return WRDE_NOSPACE;
}
}
/* Premature end */
free(comm);
return WRDE_SYNTAX;
}
static int
parse_glob(char **word, size_t * word_length, size_t * max_length,
const char *words, size_t * offset, int flags,
wordexp_t * pwordexp, const char *ifs, const char *ifs_white)
{
/* We are poised just after a '*', a '[' or a '?'. */
int error = WRDE_NOSPACE;
int quoted = 0; /* 1 if singly-quoted, 2 if doubly */
int i;
wordexp_t glob_list; /* List of words to glob */
glob_list.we_wordc = 0;
glob_list.we_wordv = NULL;
glob_list.we_offs = 0;
for (; words[*offset] != '\0'; ++*offset) {
if ((ifs && strchr(ifs, words[*offset])) ||
(!ifs && strchr(" \t\n", words[*offset])))
/* Reached IFS */
break;
/* Sort out quoting */
if (words[*offset] == '\'') {
if (quoted == 0) {
quoted = 1;
continue;
} else if (quoted == 1) {
quoted = 0;
continue;
}
} else if (words[*offset] == '"') {
if (quoted == 0) {
quoted = 2;
continue;
} else if (quoted == 2) {
quoted = 0;
continue;
}
}
/* Sort out other special characters */
if (quoted != 1 && words[*offset] == '$') {
error = parse_dollars(word, word_length, max_length, words,
offset, flags, &glob_list, ifs,
ifs_white, quoted == 2);
if (error)
goto tidy_up;
continue;
} else if (words[*offset] == '\\') {
if (quoted)
error = parse_qtd_backslash(word, word_length, max_length,
words, offset);
else
error = parse_backslash(word, word_length, max_length,
words, offset);
if (error)
goto tidy_up;
continue;
}
*word = w_addchar(*word, word_length, max_length, words[*offset]);
if (*word == NULL)
goto tidy_up;
}
/* Don't forget to re-parse the character we stopped at. */
--*offset;
/* Glob the words */
error = w_addword(&glob_list, *word);
*word = w_newword(word_length, max_length);
for (i = 0; error == 0 && i < glob_list.we_wordc; i++)
error = do_parse_glob(glob_list.we_wordv[i], word, word_length,
max_length, pwordexp, ifs /*, ifs_white*/);
/* Now tidy up */
tidy_up:
wordfree(&glob_list);
return error;
}
int wordexp(const char *words, wordexp_t * we, int flags)
{
size_t words_offset;
size_t word_length;
size_t max_length;
char *word = w_newword(&word_length, &max_length);
int error;
char *ifs;
char ifs_white[4];
wordexp_t old_word = *we;
if (flags & WRDE_REUSE) {
/* Minimal implementation of WRDE_REUSE for now */
wordfree(we);
old_word.we_wordv = NULL;
}
if ((flags & WRDE_APPEND) == 0) {
we->we_wordc = 0;
if (flags & WRDE_DOOFFS) {
we->we_wordv = calloc(1 + we->we_offs, sizeof(char *));
if (we->we_wordv == NULL) {
error = WRDE_NOSPACE;
goto do_error;
}
} else {
we->we_wordv = calloc(1, sizeof(char *));
if (we->we_wordv == NULL) {
error = WRDE_NOSPACE;
goto do_error;
}
we->we_offs = 0;
}
}
/* Find out what the field separators are.
* There are two types: whitespace and non-whitespace.
*/
ifs = getenv("IFS");
if (!ifs)
/* IFS unset - use <space><tab><newline>. */
ifs = strcpy(ifs_white, " \t\n");
else {
char *ifsch = ifs;
char *whch = ifs_white;
/* Start off with no whitespace IFS characters */
ifs_white[0] = '\0';
while (*ifsch != '\0') {
if ((*ifsch == ' ') || (*ifsch == '\t') || (*ifsch == '\n')) {
/* Whitespace IFS. See first whether it is already in our
collection. */
char *runp = ifs_white;
while (runp < whch && *runp != '\0' && *runp != *ifsch)
++runp;
if (runp == whch)
*whch++ = *ifsch;
}
++ifsch;
}
*whch = '\0';
}
for (words_offset = 0; words[words_offset]; ++words_offset)
switch (words[words_offset]) {
case '\\':
error = parse_backslash(&word, &word_length, &max_length, words,
&words_offset);
if (error)
goto do_error;
break;
case '$':
error = parse_dollars(&word, &word_length, &max_length, words,
&words_offset, flags, we, ifs, ifs_white,
0);
if (error)
goto do_error;
break;
case '`':
if (flags & WRDE_NOCMD) {
error = WRDE_CMDSUB;
goto do_error;
}
++words_offset;
error = parse_backtick(&word, &word_length, &max_length, words,
&words_offset, flags, we, ifs,
ifs_white);
if (error)
goto do_error;
break;
case '"':
++words_offset;
error = parse_dquote(&word, &word_length, &max_length, words,
&words_offset, flags, we, ifs, ifs_white);
if (error)
goto do_error;
if (!word_length) {
error = w_addword(we, NULL);
if (error)
return error;
}
break;
case '\'':
++words_offset;
error = parse_squote(&word, &word_length, &max_length, words,
&words_offset);
if (error)
goto do_error;
if (!word_length) {
error = w_addword(we, NULL);
if (error)
return error;
}
break;
case '~':
error = parse_tilde(&word, &word_length, &max_length, words,
&words_offset, we->we_wordc);
if (error)
goto do_error;
break;
case '*':
case '[':
case '?':
error = parse_glob(&word, &word_length, &max_length, words,
&words_offset, flags, we, ifs, ifs_white);
if (error)
goto do_error;
break;
default:
/* Is it a word separator? */
if (strchr(" \t", words[words_offset]) == NULL) {
char ch = words[words_offset];
/* Not a word separator -- but is it a valid word char? */
if (strchr("\n|&;<>(){}", ch)) {
/* Fail */
error = WRDE_BADCHAR;
goto do_error;
}
/* "Ordinary" character -- add it to word */
word = w_addchar(word, &word_length, &max_length, ch);
if (word == NULL) {
error = WRDE_NOSPACE;
goto do_error;
}
break;
}
/* If a word has been delimited, add it to the list. */
if (word != NULL) {
error = w_addword(we, word);
if (error)
goto do_error;
}
word = w_newword(&word_length, &max_length);
}
/* End of string */
/* There was a word separator at the end */
if (word == NULL) /* i.e. w_newword */
return 0;
/* There was no field separator at the end */
return w_addword(we, word);
do_error:
/* Error:
* free memory used (unless error is WRDE_NOSPACE), and
* set we members back to what they were.
*/
free(word);
if (error == WRDE_NOSPACE)
return WRDE_NOSPACE;
if ((flags & WRDE_APPEND) == 0)
wordfree(we);
*we = old_word;
return error;
}
bool assembly_lexer(std::string const & file_name, unsigned int & line, std::string::const_iterator & begin, std::string::const_iterator const & end, lexeme & output)
{
std::string input;
for(; begin != end; ++begin)
{
char current_char = *begin;
char type = type_lookup_table[current_char];
switch(type)
{
case char_type_illegal:
{
lexer_exception(file_name, line, "Illegal character");
}
case char_type_name:
{
std::string::const_iterator name_begin = begin;
for(++begin;
(begin != end)
&&
(is_name_char(*begin) == true);
++begin);
output = lexeme(lexeme_name, std::string(name_begin, begin), line);
return true;
}
case char_type_digit:
{
std::string::const_iterator number_begin = begin;
for(++begin;
(begin != end)
&&
(nil::string::is_digit(*begin) == true)
; ++begin);
output = lexeme(lexeme_number, std::string(number_begin, begin), line);
return true;
}
case char_type_zero:
{
std::string::const_iterator number_begin = begin;
++begin;
if(begin != end)
{
char second_character = *begin;
if(
(second_character == 'x')
||
(second_character == 'X')
)
{
for(++begin;
(begin != end)
&&
(nil::string::is_digit(*begin) == true)
; ++begin);
}
else if(is_binary_digit(second_character) == true)
{
for(++begin;
(begin != end)
&&
(is_binary_digit(*begin) == true)
; ++begin);
}
}
output = lexeme(lexeme_number, std::string(number_begin, begin), line);
return true;
}
case char_type_string:
{
std::string string;
for(++begin; begin != end;)
{
char current_char = *begin;
switch(current_char)
{
case '"':
{
++begin;
output = lexeme(lexeme_string, string, line);
return true;
}
case '\\':
{
try
{
parse_backslash(begin, end, string);
}
catch(std::exception & exception)
{
lexer_exception(file_name, line, exception.what());
}
break;
}
case '\n':
{
lexer_exception(file_name, line, "Newline in string");
}
default:
{
string += current_char;
++begin;
break;
}
}
}
lexer_exception(file_name, line, "Incomplete string at the end of file");
}
case char_type_operator:
{
++begin;
output = lexeme(lexeme_operator, std::string(1, current_char), line);
return true;
}
case char_type_operator_extended:
{
++begin;
if(begin == end)
{
output = lexeme(lexeme_operator, std::string(1, current_char), line);
return false;
}
else
{
char second_char = *begin;
bool is_extended = false;
switch(current_char)
{
case '&':
{
if(second_char == '&')
{
is_extended = true;
}
break;
}
case '|':
{
if(second_char == '|')
{
is_extended = true;
}
break;
}
case '=':
{
if(second_char == '=')
{
is_extended = true;
}
break;
}
case '!':
{
if(second_char == '=')
{
is_extended = true;
}
break;
}
case '<':
{
if(second_char == '=')
{
is_extended = true;
}
break;
}
case '>':
{
if(second_char == '=')
{
is_extended = true;
}
break;
}
}
if(is_extended == true)
{
++begin;
output = lexeme(lexeme_operator, std::string(1, current_char) + second_char, line);
}
else
{
output = lexeme(lexeme_operator, std::string(1, current_char), line);
}
return true;
}
}
case char_type_newline:
{
++begin;
++line;
output = lexeme(lexeme_newline, line);
return true;
}
case char_type_comment:
{
for(++begin; begin != end; ++begin)
{
if(*begin == '\n')
{
++begin;
++line;
output = lexeme(lexeme_newline, line);
return true;
}
}
return false;
}
}
}
return false;
}