// Counts the number of words (separted by separator charactors: newlines, tabs,
// return, space). See r_util.h for more details of the isseparator macro.
R_API int r_str_word_count(const char *string) {
const char *text, *tmp;
int word;
for (text = tmp = string; *text && isseparator (*text); text++);
for (word = 0; *text; word++) {
for (;*text && !isseparator (*text); text++);
for (tmp = text; *text && isseparator (*text); text++);
}
return word;
}
/*
* Sets the given char in the string to an '\0' char to end the previous string
* in the argument list. Then adds the next non-empty string to the argument
* list.
*/
inline int
add_arg_to_list (char *cmd, int i, char **args, int *count)
{
cmd[i] = '\0';
if (!isseparator(cmd[i+1])) { // the next string isn't empty
args[++(*count)] = cmd + i + 1;
}
}
/* Return whole word at specified position in string */
char *get_word_at(const char *str, int pos, char **startpos)
{
const char *start, *end;
g_return_val_if_fail(str != NULL, NULL);
g_return_val_if_fail(pos >= 0, NULL);
/* get previous word if char at `pos' is space */
start = str+pos;
while (start > str && isseparator(start[-1])) start--;
end = start;
while (start > str && !isseparator(start[-1])) start--;
while (*end != '\0' && !isseparator(*end)) end++;
while (*end != '\0' && isseparator_notspace(*end)) end++;
*startpos = (char *) start;
return g_strndup(start, (int) (end-start));
}
R_API char *r_str_word_get_first(const char *text) {
char *ret;
int len = 0;
for (;*text && isseparator (*text); text++);
/* strdup */
len = strlen (text);
ret = (char *)malloc (len + 1);
if (!ret) {
eprintf ("Cannot allocate %d bytes.\n", len+1);
exit (1);
}
strncpy (ret, text, len);
ret[len] = '\0';
return ret;
}
/* manual word completion - called when TAB is pressed */
char *word_complete(WINDOW_REC *window, const char *line, int *pos)
{
static int startpos = 0, wordlen = 0;
GString *result;
char *word, *wordstart, *linestart, *ret;
int want_space;
g_return_val_if_fail(line != NULL, NULL);
g_return_val_if_fail(pos != NULL, NULL);
if (complist != NULL && *pos == last_line_pos &&
strcmp(line, last_line) == 0) {
/* complete from old list */
complist = complist->next != NULL ? complist->next :
g_list_first(complist);
want_space = last_want_space;
} else {
/* get new completion list */
free_completions();
/* get the word we want to complete */
word = get_word_at(line, *pos, &wordstart);
startpos = (int) (wordstart-line);
wordlen = strlen(word);
/* get the start of line until the word we're completing */
if (isseparator(*line)) {
/* empty space at the start of line */
if (wordstart == line)
wordstart += strlen(wordstart);
} else {
while (wordstart > line && isseparator(wordstart[-1]))
wordstart--;
}
linestart = g_strndup(line, (int) (wordstart-line));
/* completions usually add space after the word, that makes
things a bit harder. When continuing a completion
"/msg nick1 "<tab> we have to cycle to nick2, etc.
BUT if we start completion with "/msg "<tab>, we don't
want to complete the /msg word, but instead complete empty
word with /msg being in linestart. */
if (*pos > 0 && line[*pos-1] == ' ') {
char *old;
old = linestart;
linestart = *linestart == '\0' ?
g_strdup(word) :
g_strconcat(linestart, " ", word, NULL);
g_free(old);
g_free(word);
word = g_strdup("");
startpos = strlen(linestart)+1;
wordlen = 0;
}
want_space = TRUE;
signal_emit("complete word", 5, &complist, window, word, linestart, &want_space);
last_want_space = want_space;
g_free(linestart);
g_free(word);
}
if (complist == NULL)
return NULL;
/* word completed */
*pos = startpos+strlen(complist->data);
/* replace the word in line - we need to return
a full new line */
result = g_string_new(line);
g_string_erase(result, startpos, wordlen);
g_string_insert(result, startpos, complist->data);
if (want_space) {
if (!isseparator(result->str[*pos]))
g_string_insert_c(result, *pos, ' ');
(*pos)++;
}
wordlen = strlen(complist->data);
last_line_pos = *pos;
g_free_not_null(last_line);
last_line = g_strdup(result->str);
ret = result->str;
g_string_free(result, FALSE);
return ret;
}
token_type get_token(token *token)
{
token_type current_token_t;
lexer_status status = INITIAL_STATUS;
char cc;
char only_one_char_sign = TRUE;
char skip = FALSE;
unsigned char token_pos = 0;
while (status != INFINISH_STATUS) {
skip = FALSE;
cc = get_next_char();
switch (status) {
case INITIAL_STATUS:
if ('"' == cc) {
status = INSTRING_STATUS;
} else if (isdigit((int)cc)) {
status = INNUM_INT_STATUS;
} else if (isalpha((int)cc) || ('_' == cc)) {
status = INIDENTIFIER_STATUS;
} else if ('`' == cc) {
status = INREGEXP_STATUS;
} else if ('#' == cc) {
skip = TRUE;
status = INCOMMENT_STATUS;
} else if (isseparator(cc)) {
skip = TRUE;
} else {
status = INFINISH_STATUS;
switch (cc) {
case '+':
current_token_t = PLUS;
break;
case '-': current_token_t = MINUS;
break;
case '*':
current_token_t = TIMES;
break;
case '/':
current_token_t = OVER;
break;
case '=':
current_token_t = ASSIGN;
status = INSIGN_STATUS;
break;
case '<':
current_token_t = LT;
status = INSIGN_STATUS;
break;
case '>':
current_token_t = GT;
status = INSIGN_STATUS;
break;
case '(':
current_token_t = LPAREN;
break;
case ')':
current_token_t = RPAREN;
break;
case ';':
current_token_t = SEMI;
break;
case EOF:
current_token_t = END_OF_FILE_TOKEN;
break;
default:
current_token_t = ERROR;
break;
}
}
break;
case INSTRING_STATUS:
if ('"' == cc) {
status = INFINISH_STATUS;
}
break;
case INSIGN_STATUS:
if (current_token_t == LT) {
if ('<' == cc) {
current_token_t = LSHIFT;
} else if ('=' == cc) {
current_token_t = LTOREQ;
} else {
skip = TRUE;
unget_next_char();
}
} else if (current_token_t == GT) {
if ('>' == cc) {
current_token_t = RSHIFT;
} else if ('=' == cc) {
current_token_t = GTOREQ;
} else {
skip = TRUE;
unget_next_char();
}
} else if (current_token_t == ASSIGN) {
if ('=' == cc) {
current_token_t = EQUAL;
} else {
skip = TRUE;
unget_next_char();
}
} else {
/* no way to get here */
}
status = INFINISH_STATUS;
break;
case INCOMMENT_STATUS:
skip = TRUE;
if ('\n' == cc) {
status = INITIAL_STATUS;
} else if (EOF == cc) {
status == INFINISH_STATUS;
current_token_t = END_OF_FILE_TOKEN;
} else {
status = INCOMMENT_STATUS;
}
break;
case INNUM_INT_STATUS:
if ('.' == cc) {
status = INNUM_DOT_STATUS;
} else if (!isdigit(cc)) {
status = INFINISH_STATUS;
skip = TRUE;
unget_next_char();
}
break;
case INNUM_DOT_STATUS:
if (isdigit(cc)) {
} else {
}
break;
case INNUM_DEC_STATUS:
if (!isdigit(cc)) {
status = INFINISH_STATUS;
}
break;
case INIDENTIFIER_STATUS:
if (isdigit((int)cc) || isalpha((int)cc) || ('_' == cc)) {
current_token_t = IDENTIFIER_TOKEN;
} else {
skip = TRUE;
status = INFINISH_STATUS;
unget_next_char();
}
break;
case INREGEXP_STATUS:
if ('`' == cc) {
status = INFINISH_STATUS;
}
break;
case INFINISH_STATUS:
break;
case INERROR_STATUS:
/* point to where is wrong */
exit(-1);
default:
break;
}
if (!skip && (token_pos <= MAX_TOKEN_SIZE)) {
token_buf[token_pos++] = cc;
} else {
/* skip or token length is too long*/
}
if (INFINISH_STATUS == status) {
/* Do some saving token work */
token_buf[token_pos] = '\0';
if (IDENTIFIER_TOKEN == current_token_t) {
if (NONE != is_key_word(token_buf, token_pos)) {
current_token_t = KEYWORD_TOKEN;
}
}
printf("token = %d = %s\n", token_pos, token_buf);
}
}
return current_token_t;
}
/* manual word completion - called when TAB is pressed */
char *word_complete(WINDOW_REC *window, const char *line, int *pos, int erase, int backward)
{
static int startpos = 0, wordlen = 0;
int old_startpos, old_wordlen;
GString *result;
const char *cmdchars;
char *word, *wordstart, *linestart, *ret, *data;
int continue_complete, want_space, expand_escapes;
g_return_val_if_fail(line != NULL, NULL);
g_return_val_if_fail(pos != NULL, NULL);
continue_complete = complist != NULL && *pos == last_line_pos &&
g_strcmp0(line, last_line) == 0;
if (erase && !continue_complete)
return NULL;
old_startpos = startpos;
old_wordlen = wordlen;
if (!erase && continue_complete) {
word = NULL;
linestart = NULL;
} else {
char* old_wordstart;
/* get the word we want to complete */
word = get_word_at(line, *pos, &wordstart);
old_wordstart = wordstart;
startpos = (int) (wordstart-line);
wordlen = strlen(word);
/* remove trailing spaces from linestart */
while (wordstart > line && isseparator_space(wordstart[-1]))
wordstart--;
/* unless everything was spaces */
if (old_wordstart > line && wordstart == line)
wordstart = old_wordstart - 1;
linestart = g_strndup(line, (int) (wordstart-line));
/* completions usually add space after the word, that makes
things a bit harder. When continuing a completion
"/msg nick1 "<tab> we have to cycle to nick2, etc.
BUT if we start completion with "/msg "<tab>, we don't
want to complete the /msg word, but instead complete empty
word with /msg being in linestart. */
if (!erase && *pos > 0 && isseparator_space(line[*pos-1]) &&
(*linestart == '\0' || !isseparator_space(wordstart[-1]))) {
char *old;
old = linestart;
/* we want to move word into linestart */
if (*linestart == '\0') {
linestart = g_strdup(word);
} else {
GString *str = g_string_new(linestart);
if (old_wordstart[-1] != str->str[str->len - 1]) {
/* do not accidentally duplicate the word separator */
g_string_append_c(str, old_wordstart[-1]);
}
g_string_append(str, word);
linestart = g_string_free(str, FALSE);
}
g_free(old);
g_free(word);
word = g_strdup("");
startpos = *linestart == '\0' ? 0 :
strlen(linestart)+1;
wordlen = 0;
}
}
if (erase) {
signal_emit("complete erase", 3, window, word, linestart);
/* jump to next completion */
startpos = old_startpos;
wordlen = old_wordlen;
}
if (continue_complete) {
/* complete from old list */
if (backward)
complist = complist->prev != NULL ? complist->prev :
g_list_last(complist);
else
complist = complist->next != NULL ? complist->next :
g_list_first(complist);
want_space = last_want_space;
} else {
int keep_word = settings_get_bool("completion_keep_word");
/* get new completion list */
free_completions();
want_space = TRUE;
signal_emit("complete word", 5, &complist, window, word, linestart, &want_space);
last_want_space = want_space;
if (complist != NULL) {
/* Remove all nulls (from the signal) before doing further processing */
complist = g_list_remove_all(g_list_first(complist), NULL);
if (keep_word) {
complist = g_list_append(complist, g_strdup(word));
}
if (backward) {
complist = g_list_last(complist);
if (keep_word) {
complist = complist->prev;
}
}
}
}
g_free(linestart);
g_free(word);
if (complist == NULL)
return NULL;
/* get the cmd char */
cmdchars = settings_get_str("cmdchars");
/* get the expand_escapes setting */
expand_escapes = settings_get_bool("expand_escapes");
/* escape if the word doesn't begin with '/' and expand_escapes are turned on */
data = strchr(cmdchars, *line) == NULL && expand_escapes ?
escape_string_backslashes(complist->data) : g_strdup(complist->data);
/* word completed */
*pos = startpos + strlen(data);
/* replace the word in line - we need to return
a full new line */
result = g_string_new(line);
g_string_erase(result, startpos, wordlen);
g_string_insert(result, startpos, data);
if (want_space) {
if (!isseparator(result->str[*pos]))
g_string_insert_c(result, *pos, ' ');
(*pos)++;
}
wordlen = strlen(data);
last_line_pos = *pos;
g_free_not_null(last_line);
last_line = g_strdup(result->str);
ret = result->str;
g_string_free(result, FALSE);
/* free the data */
g_free(data);
return ret;
}
int MyRequest::Lexer(char *buf, int length) {
int consumed = 0;
if(length == 0) {
valid = methodValid && urlValid && httpVersionValid;
if(data.size() == contentLength)
done = lexerState != LEXER_init;
return 0;
}
if(parserState == PARSER_body) {
consumed = AppendData(buf, length);
return consumed;
}
for(int i = 0; i < length; i++, consumed++) {
if(parserState == PARSER_body) {
consumed += AppendData(buf + i, length - i);
break;
}
char ch = buf[i];
rawHeader.push_back(ch);
again:
if(parserState == PARSER_body) {
consumed += AppendData(buf + i, length - i);
break;
}
if(lexerState > LEXER_error) {
aborted = true;
break;
}
switch(lexerState) {
case LEXER_init:
if((isctl(ch) || isseparator(ch) || ch == ' ' || ch == '\t') && !lexerToken.empty()) {
ProcessTokenIfAvailable();
goto again;
}
if(ch == '\r') lexerState = LEXER_cr_read;
else if(ch == '\n') ProcessLF();
else if(opaqueEnabled) {
lexerOpaqueString.clear();
lexerState = LEXER_opaque;
goto again;
}
else if(parseURLEnabled && !isspace(ch)) {
lexerUrl.clear();
lexerState = LEXER_url;
goto again;
}
else if(ch == '"' && quotedStringsEnabled) {
lexerQuotedString.clear();
lexerState = LEXER_quoted_string;
}
else if(ch == '(' && commentsEnabled) {
lexerComment.clear();
lexerComment.push_back(ch);
lexerState = LEXER_comment;
lexerCommentNesting = 1;
}
else if(ch == ' ' || ch == '\t') lexerState = LEXER_LWS_read;
else if(isseparator(ch)) ProcessSeparator(ch);
else if(!isctl(ch)) lexerToken.push_back(ch);
else lexerState = LEXER_token_text_error;
break;
case LEXER_opaque:
ProcessTokenIfAvailable();
if(ch == '\r') {
if(foldingEnabled) lexerState = LEXER_opaque_cr_read;
else {
ProcessOpaque(lexerOpaqueString);
lexerState = LEXER_init;
goto again;
}
}
else lexerOpaqueString.push_back(ch);
break;
case LEXER_opaque_cr_read:
ProcessTokenIfAvailable();
if(ch == '\n') lexerState = LEXER_opaque_newline_read;
else {
ProcessOpaque(lexerOpaqueString);
lexerState = LEXER_init;
goto again;
}
break;
case LEXER_opaque_newline_read:
if(ch == ' ' || ch == '\t') {
lexerOpaqueString.push_back(' ');
lexerState = LEXER_opaque;
}
else {
ProcessOpaque(lexerOpaqueString);
ProcessNewline();
lexerState = LEXER_init;
goto again;
}
break;
case LEXER_url:
ProcessTokenIfAvailable();
if(isspace(ch)) {
ProcessURL(lexerUrl);
lexerState = LEXER_init;
goto again;
}
else {
if(ch == '%') lexerState = LEXER_url_encoded_1;
else if(ch == '+') lexerUrl.push_back(' ');
else lexerUrl.push_back(ch);
}
break;
case LEXER_url_encoded_1:
if(ch == '%') {
lexerUrl.push_back('%');
lexerState = LEXER_url;
}
else if(isxdigit(ch)) {
lexerTmpHex[0] = ch;
lexerTmpHex[2] = '\0';
lexerState = LEXER_url_encoded_2;
}
else lexerState = LEXER_url_error;
break;
case LEXER_url_encoded_2:
if(isxdigit(ch)) {
lexerTmpHex[1] = ch;
char newch = strtol(lexerTmpHex, NULL, 16);
lexerUrl.push_back(newch);
lexerState = LEXER_url;
}
else lexerState = LEXER_url_error;
break;
case LEXER_comment:
ProcessTokenIfAvailable();
if(ch == '\\') lexerState = LEXER_quoted_pair_comment;
else if(ch == '(') {
lexerCommentNesting++;
lexerComment.push_back(ch);
}
else if(ch == ')') {
lexerCommentNesting--;
lexerComment.push_back(ch);
if(lexerCommentNesting == 0) {
ProcessComment(lexerComment);
lexerState = LEXER_init;
}
}
else if(istext(ch)) {
lexerComment.push_back(ch);
}
else lexerState = LEXER_comment_text_error;
break;
case LEXER_quoted_pair_comment:
ProcessTokenIfAvailable();
lexerComment.push_back(ch);
lexerState = LEXER_comment;
break;
case LEXER_quoted_string:
ProcessTokenIfAvailable();
if(ch == '\\') lexerState = LEXER_quoted_pair;
else if(ch == '"') {
ProcessQuotedString(lexerQuotedString);
lexerState = LEXER_init;
}
else if(istext(ch)) {
lexerQuotedString.push_back(ch);
}
else lexerState = LEXER_quoted_string_text_error;
break;
case LEXER_quoted_pair:
ProcessTokenIfAvailable();
lexerQuotedString.push_back(ch);
lexerState = LEXER_quoted_string;
break;
case LEXER_cr_read:
ProcessTokenIfAvailable();
if(ch == '\n') {
if(foldingEnabled) lexerState = LEXER_newline_read;
else {
ProcessNewline();
lexerState = LEXER_init;
}
}
else {
ProcessCR();
lexerState = LEXER_init;
goto again;
}
break;
case LEXER_newline_read: // LWS folding state
ProcessTokenIfAvailable();
if(ch == ' ' || ch == '\t') lexerState = LEXER_LWS_read;
else {
ProcessNewline();
lexerState = LEXER_init;
goto again;
}
break;
case LEXER_LWS_read: // partial LWS
ProcessTokenIfAvailable();
if(ch == '\r') lexerState = LEXER_LWS_cr_read;
else if(ch != ' ' && ch != '\t') {
lexerState = LEXER_init;
ProcessLWS();
goto again;
}
break;
case LEXER_LWS_cr_read:
ProcessTokenIfAvailable();
if(ch == '\n') {
if(foldingEnabled) lexerState = LEXER_LWS_newline_read;
else {
ProcessLWS();
ProcessNewline();
lexerState = LEXER_init;
}
}
else {
ProcessCR();
lexerState = LEXER_init;
goto again;
}
break;
case LEXER_LWS_newline_read:
ProcessTokenIfAvailable();
if(ch == ' ' || ch == '\t') lexerState = LEXER_LWS_read;
else {
ProcessLWS();
ProcessNewline();
lexerState = LEXER_init;
goto again;
}
break;
}
}
if(parserState == PARSER_body) {
valid = methodValid && urlValid && httpVersionValid;
if(data.size() == contentLength)
done = true;
}
return consumed;
}
/* Reads next symbol */
static
void InSymbol(LPIT8 it8)
{
register char *idptr;
register int k;
SYMBOL key;
int sng;
do {
while (isseparator(it8->ch))
NextCh(it8);
if (isfirstidchar(it8->ch)) { /* Identifier */
k = 0;
idptr = it8->id;
do {
if (++k < MAXID) *idptr++ = (char) it8->ch;
NextCh(it8);
} while (isidchar(it8->ch));
*idptr = '\0';
xstrupr(it8->id);
key = BinSrchKey(it8->id);
if (key == SNONE) it8->sy = SIDENT;
else it8->sy = key;
}
else /* Is a number? */
if (isdigit(it8->ch) || it8->ch == '.' || it8->ch == '-' || it8->ch == '+')
{
int sign = 1;
if (it8->ch == '-') {
sign = -1;
NextCh(it8);
}
it8->inum = 0;
it8->sy = SINUM;
while (isdigit(it8->ch))
{
if ((long) it8->inum * 10L > (long) INT_MAX)
{
ReadReal(it8, it8->inum);
it8->sy = SDNUM;
it8->dnum *= sign;
return;
}
it8->inum = it8->inum * 10 + (it8->ch - '0');
NextCh(it8);
}
if (it8->ch == '.') {
ReadReal(it8, it8->inum);
it8->sy = SDNUM;
it8->dnum *= sign;
return;
}
it8 -> inum *= sign;
return;
}
else
switch ((int) it8->ch) {
case '\0':
case '\x1a':
it8->sy = SEOF;
break;
case '\n':
NextCh(it8);
it8->sy = SEOLN;
it8->lineno++;
break;
/* Comment */
case '#':
NextCh(it8);
while (it8->ch && it8->ch != '\n')
NextCh(it8);
it8->sy = SCOMMENT;
break;
/* String. I will support \", \n, \t and \\. */
/* But otherwise I hardly doubt these will be used ... */
case '\'':
case '\"':
idptr = it8->str;
sng = it8->ch;
k = 0;
NextCh(it8);
while (k < MAXSTR && it8->ch != sng) {
if (it8->ch == '\n'|| it8->ch == '\r') k = MAXSTR+1;
else {
if (it8->ch == '\\')
{
NextCh(it8);
switch (it8->ch) {
case 'n': *idptr++ = '\n'; break;
case 'r': *idptr++ = '\r'; break;
case 't': *idptr++ = '\t'; break;
case '\\': *idptr++ = '\\'; break;
default:
*idptr++ = (char) it8->ch;
}
NextCh(it8);
}
else
{
*idptr++ = (char) it8->ch;
NextCh(it8);
}
k++;
}
}
it8->sy = SSTRING;
*idptr = '\0';
NextCh(it8);
break;
default:
it8->sy = SSYNERROR;
NextCh(it8);
}
} while (it8->sy == SCOMMENT);
}
