struct token *lexer_read_literal_in_expandable_until(struct lexer_book *bk, char end_marker)
{
char end_markers[7] = { end_marker, '$', '\\', '"', '\'', '#', CHAR_EOF };
int count = 0;
do {
count += lexer_read_until(bk, end_markers);
if(bk->eof)
break;
char c = lexer_next_peek(bk);
if(c == '\\') {
lexer_next_char(bk); /* Jump the slash */
char n = lexer_next_char(bk);
count += 2;
if(lexer_special_escape(n))
lexer_add_to_lexeme(bk, lexer_special_to_code(n));
else
lexer_add_to_lexeme(bk, n);
} else if(c == '#') {
if(end_marker == '\n') {
lexer_discard_comments(bk);
break;
}
} else
break;
} while(!bk->eof);
if(bk->eof && strchr(")\"'", end_marker))
lexer_report_error(bk, "Missing closing %c.\n", end_marker);
return lexer_pack_token(bk, LITERAL);
}
/* As lexer_read_until, but elements of char_set preceded by \ are
ignored as stops, with \n replaced with spaces. */
int lexer_read_escaped_until(struct lexer *lx, char *char_set)
{
char *char_set_slash = string_format("\\%s", char_set);
int count = 0;
do {
count += lexer_read_until(lx, char_set_slash);
if(!lx->eof && lexer_next_peek(lx) == '\\') {
lexer_next_char(lx); /* Jump the slash */
char c = lexer_next_char(lx);
count += 2;
if(lexer_next_peek(lx) != CHAR_EOF) {
if(c == '\n') {
lexer_add_to_lexeme(lx, ' ');
} else {
lexer_add_to_lexeme(lx, c);
}
}
} else
break;
} while(!lx->eof);
free(char_set_slash);
if(lx->eof && !strchr(char_set, CHAR_EOF))
lexer_report_error(lx, "Missing %s\n", char_set);
return count;
}
accept_t lexer_read_variable(struct lexer_book * bk, struct token * name)
{
lexer_discard_white_space(bk);
if(lexer_next_peek(bk) == '=') {
lexer_next_char(bk);
lexer_add_to_lexeme(bk, '=');
} else {
int c = lexer_next_char(bk);
if(lexer_next_peek(bk) != '=')
return NO;
lexer_add_to_lexeme(bk, c);
lexer_next_char(bk); /* Jump = */
}
lexer_push_token(bk, lexer_pack_token(bk, VARIABLE));
lexer_push_token(bk, name);
lexer_discard_white_space(bk);
lexer_read_expandable(bk, '\n');
lexer_roll_back(bk, 1); //Recover '\n'
lexer_discard_white_space(bk);
if(lexer_next_char(bk) != '\n')
return NO;
return YES;
}
int lexer_read_variable(struct lexer *lx, struct token *name)
{
lexer_discard_white_space(lx);
if(lexer_next_peek(lx) == '=') {
lexer_next_char(lx);
lexer_add_to_lexeme(lx, '=');
} else {
int c = lexer_next_char(lx);
if(lexer_next_peek(lx) != '=')
lexer_report_error(lx, "Missing = in variable definition.");
lexer_add_to_lexeme(lx, c);
lexer_next_char(lx); /* Jump = */
}
lexer_push_token(lx, lexer_pack_token(lx, TOKEN_VARIABLE));
lexer_push_token(lx, name);
lexer_discard_white_space(lx);
//Read variable value
lexer_push_token(lx, lexer_read_expandable(lx, '\n'));
lexer_roll_back(lx, 1); //Recover '\n'
lexer_discard_white_space(lx);
if(lexer_next_char(lx) != '\n')
lexer_report_error(lx, "Missing newline at end of variable definition.");
return 1;
}
struct token *lexer_read_command_argument(struct lexer_book *bk)
{
int c = lexer_next_peek(bk);
switch(c)
{
case CHAR_EOF:
/* Found end of file while completing command */
bk->lexeme_end++;
bk->eof = 1;
if( bk->stream )
return lexer_pack_token(bk, NEWLINE);
else
return NULL;
break;
case '\n' :
lexer_next_char(bk); /* Jump \n */
lexer_add_to_lexeme(bk, c);
return lexer_pack_token(bk, NEWLINE);
break;
case '#' :
lexer_discard_comments(bk);
lexer_add_to_lexeme(bk, '\n');
return lexer_pack_token(bk, NEWLINE);
case ' ' :
case '\t':
return lexer_read_white_space(bk);
break;
case '$' :
return lexer_read_substitution(bk);
break;
case '"' :
return lexer_read_expandable(bk, '"');
break;
case '<' :
case '>' :
lexer_next_char(bk); /* Jump <, > */
lexer_add_to_lexeme(bk, c);
return lexer_pack_token(bk, IO_REDIRECT);
break;
case '\'':
lexer_add_to_lexeme(bk, '\'');
lexer_read_literal(bk);
lexer_add_to_lexeme(bk, '\'');
return lexer_pack_token(bk, LITERAL);
break;
default:
lexer_read_literal(bk);
return lexer_pack_token(bk, LITERAL);
break;
}
}
struct token *lexer_read_file(struct lexer *lx)
{
int c = lexer_next_peek(lx);
switch (c) {
case CHAR_EOF:
lx->lexeme_end++;
lx->eof = 1;
if(lx->depth == 0)
lexer_report_error(lx, "Found end of file while completing file list.\n");
return NULL;
break;
case '\n':
lexer_next_char(lx); /* Jump \n */
lexer_add_to_lexeme(lx, c);
return lexer_pack_token(lx, TOKEN_NEWLINE);
break;
case '#':
lexer_discard_comments(lx);
lexer_add_to_lexeme(lx, '\n');
return lexer_pack_token(lx, TOKEN_NEWLINE);
case ':':
lexer_next_char(lx); /* Jump : */
return lexer_pack_token(lx, TOKEN_COLON);
break;
case ' ':
case '\t':
/* Discard white-space and add space token. */
lexer_discard_white_space(lx);
return lexer_pack_token(lx, TOKEN_SPACE);
break;
case '$':
return lexer_read_substitution(lx);
break;
case '\'':
lexer_add_to_lexeme(lx, '\'');
lexer_read_literal_quoted(lx);
lexer_add_to_lexeme(lx, '\'');
return lexer_pack_token(lx, TOKEN_LITERAL);
break;
case '-':
if(lexer_peek_remote_rename_syntax(lx)) {
lexer_next_char(lx); /* Jump -> */
lexer_next_char(lx);
return lexer_pack_token(lx, TOKEN_REMOTE_RENAME);
}
/* Else fall through */
default:
return lexer_read_filename(lx);
break;
}
}
struct token *lexer_read_command_argument(struct lexer *lx)
{
int c = lexer_next_peek(lx);
switch (c) {
case CHAR_EOF:
/* Found end of file while completing command */
lx->lexeme_end++;
lx->eof = 1;
if(lx->depth == 0)
lexer_report_error(lx, "Found end of file while completing command.\n");
return NULL;
break;
case '\n':
lexer_next_char(lx); /* Jump \n */
lexer_add_to_lexeme(lx, c);
return lexer_pack_token(lx, TOKEN_NEWLINE);
break;
case '#':
lexer_discard_comments(lx);
lexer_add_to_lexeme(lx, '\n');
return lexer_pack_token(lx, TOKEN_NEWLINE);
case ' ':
case '\t':
return lexer_read_white_space(lx);
break;
case '$':
return lexer_read_substitution(lx);
break;
case '"':
return lexer_read_expandable(lx, '"');
break;
case '<':
case '>':
lexer_next_char(lx); /* Jump <, > */
lexer_add_to_lexeme(lx, c);
return lexer_pack_token(lx, TOKEN_IO_REDIRECT);
break;
case '\'':
lexer_add_to_lexeme(lx, '\'');
lexer_read_literal(lx);
lexer_add_to_lexeme(lx, '\'');
return lexer_pack_token(lx, TOKEN_LITERAL);
break;
default:
lexer_read_literal(lx);
return lexer_pack_token(lx, TOKEN_LITERAL);
break;
}
}
struct token *lexer_read_file(struct lexer_book *bk)
{
int c = lexer_next_peek(bk);
switch(c)
{
case CHAR_EOF:
lexer_report_error(bk, "Found end of file while completing file list.\n");
return NULL;
break;
case '\n' :
lexer_next_char(bk); /* Jump \n */
lexer_add_to_lexeme(bk, c);
return lexer_pack_token(bk, NEWLINE);
break;
case '#' :
lexer_discard_comments(bk);
lexer_add_to_lexeme(bk, '\n');
return lexer_pack_token(bk, NEWLINE);
case ':' :
lexer_next_char(bk); /* Jump : */
return lexer_pack_token(bk, COLON);
break;
case ' ' :
case '\t':
/* Discard white-space and try again */
lexer_discard_white_space(bk);
return lexer_read_file(bk);
break;
case '$' :
return lexer_read_substitution(bk);
break;
case '\'':
lexer_add_to_lexeme(bk, '\'');
lexer_read_literal_quoted(bk);
lexer_add_to_lexeme(bk, '\'');
return lexer_pack_token(bk, LITERAL);
break;
case '-' :
if( lexer_peek_remote_rename_syntax(bk) )
{
lexer_next_char(bk); /* Jump -> */
lexer_next_char(bk);
return lexer_pack_token(bk, REMOTE_RENAME);
}
/* Else fall through */
default:
return lexer_read_filename(bk);
break;
}
}
/* Read everything between single quotes */
int lexer_read_literal_quoted(struct lexer * lx)
{
int c = lexer_next_peek(lx);
if(c != '\'')
lexer_report_error(lx, "Missing opening quote.\n");
lexer_add_to_lexeme(lx, lexer_next_char(lx)); /* Add first ' */
int count = lexer_read_escaped_until(lx, "'");
lexer_add_to_lexeme(lx, lexer_next_char(lx)); /* Add second ' */
return count;
}
/* Returns the count of characters that we would have to roll-back to
undo the read. */
int lexer_read_until(struct lexer_book *bk, char *char_set)
{
int count = 0;
char c;
do
{
c = lexer_next_peek(bk);
if( strchr(char_set, c) )
{
return count;
}
if( c != CHAR_EOF )
lexer_add_to_lexeme(bk, c);
lexer_next_char(bk);
count++;
} while( c != CHAR_EOF );
bk->eof = 1;
return count;
}
struct token *lexer_read_literal_in_expandable_until(struct lexer *lx, char end_marker)
{
const char end_markers[8] = { end_marker, '$', '\\', '"', '\'', '#', CHAR_EOF ,0};
int count = 0;
do {
count += lexer_read_until(lx, end_markers);
if(lx->eof)
break;
char c = lexer_next_peek(lx);
if(c == '\\') {
lexer_next_char(lx); /* Jump the slash */
char n = lexer_next_char(lx);
count += 2;
if(lexer_special_escape(n)) {
lexer_add_to_lexeme(lx, lexer_special_to_code(n));
} else if(n == '\n') {
lexer_add_to_lexeme(lx, ' ');
} else {
lexer_add_to_lexeme(lx, n);
}
} else if(c == '#') {
if(end_marker == '\n') {
lexer_discard_comments(lx);
break;
}
} else
break;
} while(!lx->eof);
if(lx->eof && strchr(")\"'", end_marker))
lexer_report_error(lx, "Missing closing %c.\n", end_marker);
return lexer_pack_token(lx, TOKEN_LITERAL);
}
//opened tracks whether it is the opening (opened = 0) or closing (opened = 1) double quote we encounter.
struct list *lexer_read_expandable_recursive(struct lexer *lx, char end_marker, int opened)
{
lexer_discard_white_space(lx);
struct list *tokens = list_create();
while(!lx->eof) {
int c = lexer_next_peek(lx);
if(c == '$') {
list_push_tail(tokens, lexer_read_substitution(lx));
}
if(c == '\'') {
lexer_read_literal(lx);
list_push_tail(tokens, lexer_pack_token(lx, TOKEN_LITERAL));
} else if(c == '"' && opened == 0) {
lexer_add_to_lexeme(lx, lexer_next_char(lx));
list_push_tail(tokens, lexer_pack_token(lx, TOKEN_LITERAL)); // Add first "
tokens = list_splice(tokens, lexer_read_expandable_recursive(lx, '"', 1));
lexer_add_to_lexeme(lx, '"');
list_push_tail(tokens, lexer_pack_token(lx, TOKEN_LITERAL)); // Add closing "
if(end_marker == '"')
return tokens;
} else if(c == '#' && end_marker != '"') {
lexer_discard_comments(lx);
} else if(c == end_marker) {
lexer_next_char(lx); /* Jump end_marker */
return tokens;
} else {
list_push_tail(tokens, lexer_read_literal_in_expandable_until(lx, end_marker));
}
}
lexer_report_error(lx, "Found EOF before end marker: %c.\n", end_marker);
return NULL;
}
/* Read a filename, adding '-' to names when - is not followed by
>. The 'recursive' comes because the function calls itself when
completing a name when it added a -. */
int lexer_read_filename_recursive(struct lexer_book *bk)
{
int count = lexer_read_escaped_until(bk, FILENAME_LIMITS);
if(count < 1)
return count;
if(lexer_next_peek(bk) == '-' && !lexer_peek_remote_rename_syntax(bk)) {
lexer_add_to_lexeme(bk, '-');
count++;
count += lexer_read_filename_recursive(bk);
}
return count;
}
/* Read a filename, adding '-' to names when - is not followed by
>. The 'recursive' comes because the function calls itself when
completing a name when it added a -. */
int lexer_read_filename_recursive(struct lexer *lx)
{
int count = lexer_read_escaped_until(lx, FILENAME_LIMITS);
if(count < 1)
return count;
if(lexer_next_peek(lx) == '-' && !lexer_peek_remote_rename_syntax(lx)) {
lexer_add_to_lexeme(lx, '-');
lexer_next_char(lx);
count++;
count += lexer_read_filename_recursive(lx);
}
return count;
}
/* Consolidates a sequence of white space into a single SPACE token */
struct token *lexer_read_white_space(struct lexer_book *bk)
{
int count = 0;
while(strchr(WHITE_SPACE, lexer_next_peek(bk))) {
count++;
lexer_next_char(bk);
}
if(count > 0) {
lexer_add_to_lexeme(bk, ' ');
return lexer_pack_token(bk, SPACE);
} else
lexer_report_error(bk, "Expecting white space.");
return NULL;
}
/* Consolidates a sequence of white space into a single SPACE token */
struct token *lexer_read_white_space(struct lexer *lx)
{
int count = lexer_discard_white_space(lx);
while(strchr(WHITE_SPACE, lexer_next_peek(lx))) {
count++;
lexer_next_char(lx);
}
if(count > 0) {
lexer_add_to_lexeme(lx, ' ');
return lexer_pack_token(lx, TOKEN_SPACE);
} else
lexer_report_error(lx, "Expecting white space.");
return NULL;
}
int lexer_read_variable_list(struct lexer * lx)
{
int c;
while((c = lexer_next_peek(lx)) != '\n') {
lexer_discard_white_space(lx);
if(c == '#') {
lexer_discard_comments(lx);
lexer_roll_back(lx, 1); //Recover the newline
break;
}
lexer_push_token(lx, lexer_read_syntax_name(lx));
}
lexer_add_to_lexeme(lx, lexer_next_char(lx)); //Drop the newline
lexer_push_token(lx, lexer_pack_token(lx, TOKEN_NEWLINE));
return 1;
}
accept_t lexer_read_variable_list(struct lexer_book * bk)
{
int c;
while((c = lexer_next_peek(bk)) != '\n') {
lexer_discard_white_space(bk);
if(c == '#') {
lexer_discard_comments(bk);
lexer_roll_back(bk, 1); //Recover the newline
break;
}
lexer_push_token(bk, lexer_read_syntax_name(bk));
}
lexer_add_to_lexeme(bk, lexer_next_char(bk)); //Drop the newline
lexer_push_token(bk, lexer_pack_token(bk, NEWLINE));
return YES;
}
/* Returns the count of characters that we would have to roll-back to
undo the read. */
int lexer_read_until(struct lexer *lx, const char *char_set )
{
int count = 0;
char c;
do {
c = lexer_next_peek(lx);
if(strchr(char_set, c)) {
return count;
}
if(c != CHAR_EOF)
lexer_add_to_lexeme(lx, c);
lexer_next_char(lx);
count++;
} while(c != CHAR_EOF);
lx->eof = 1;
return count;
}
