int
pdf_lex(fz_stream *f, pdf_lexbuf *buf)
{
while (1)
{
int c = fz_read_byte(f);
switch (c)
{
case EOF:
return PDF_TOK_EOF;
case IS_WHITE:
lex_white(f);
break;
case '%':
lex_comment(f);
break;
case '/':
lex_name(f, buf);
return PDF_TOK_NAME;
case '(':
return lex_string(f, buf);
case ')':
fz_warn(f->ctx, "lexical error (unexpected ')')");
continue;
case '<':
c = fz_read_byte(f);
if (c == '<')
{
return PDF_TOK_OPEN_DICT;
}
else
{
fz_unread_byte(f);
return lex_hex_string(f, buf);
}
case '>':
c = fz_read_byte(f);
if (c == '>')
{
return PDF_TOK_CLOSE_DICT;
}
fz_warn(f->ctx, "lexical error (unexpected '>')");
continue;
case '[':
return PDF_TOK_OPEN_ARRAY;
case ']':
return PDF_TOK_CLOSE_ARRAY;
case '{':
return PDF_TOK_OPEN_BRACE;
case '}':
return PDF_TOK_CLOSE_BRACE;
case IS_NUMBER:
return lex_number(f, buf, c);
default: /* isregular: !isdelim && !iswhite && c != EOF */
fz_unread_byte(f);
lex_name(f, buf);
return pdf_token_from_keyword(buf->scratch);
}
}
}
pdf_token
pdf_lex(fz_stream *f, pdf_lexbuf *buf)
{
while (1)
{
int c = fz_read_byte(f);
switch (c)
{
case EOF:
return PDF_TOK_EOF;
case IS_WHITE:
lex_white(f);
break;
case '%':
lex_comment(f);
break;
case '/':
lex_name(f, buf);
return PDF_TOK_NAME;
case '(':
return lex_string(f, buf);
case ')':
fz_warn(f->ctx, "lexical error (unexpected ')')");
continue;
case '<':
c = fz_read_byte(f);
if (c == '<')
{
return PDF_TOK_OPEN_DICT;
}
else
{
fz_unread_byte(f);
return lex_hex_string(f, buf);
}
case '>':
c = fz_read_byte(f);
if (c == '>')
{
return PDF_TOK_CLOSE_DICT;
}
fz_warn(f->ctx, "lexical error (unexpected '>')");
if (c == EOF)
{
return PDF_TOK_EOF;
}
fz_unread_byte(f);
continue;
case '[':
return PDF_TOK_OPEN_ARRAY;
case ']':
return PDF_TOK_CLOSE_ARRAY;
case '{':
return PDF_TOK_OPEN_BRACE;
case '}':
return PDF_TOK_CLOSE_BRACE;
case IS_NUMBER:
/* cf. https://code.google.com/p/sumatrapdf/issues/detail?id=2231 */
{
int tok = lex_number(f, buf, c);
while (1)
{
c = fz_peek_byte(f);
switch (c)
{
case IS_NUMBER:
fz_warn(f->ctx, "ignoring invalid character after number: '%c'", c);
fz_read_byte(f);
continue;
default:
return tok;
}
}
}
default: /* isregular: !isdelim && !iswhite && c != EOF */
fz_unread_byte(f);
lex_name(f, buf);
return pdf_token_from_keyword(buf->scratch);
}
}
}
/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
*
* LEXICAL SCANNER
*/
LOCAL te_dp_event
yylex(void)
{
lastToken = DP_EV_INVALID;
scanAgain:
/*
* Start the process of locating a token.
* We branch here after skipping over a comment
* or processing a directive (which may change our context).
*/
if (IS_WHITESPACE_CHAR(*pCurCtx->pzScan))
trim_whitespace();
switch (*pCurCtx->pzScan) {
case NUL:
/*
* IF we are not inside an include context,
* THEN go finish.
*/
if (pCurCtx->pCtx == NULL)
goto lex_done;
pop_context();
goto scanAgain;
case '#':
{
extern char * processDirective(char*);
char * pz = processDirective(pCurCtx->pzScan+1);
/*
* Ensure that the compiler doesn't try to save a copy of
* "pCurCtx" in a register. It must be reloaded from memory.
*/
pCurCtx->pzScan = pz;
goto scanAgain;
}
case '{': SET_LIT_TKN(O_BRACE); break;
case '=': SET_LIT_TKN(EQ); break;
case '}': SET_LIT_TKN(C_BRACE); break;
case '[': SET_LIT_TKN(OPEN_BKT); break;
case ']': SET_LIT_TKN(CLOSE_BKT); break;
case ';': SET_LIT_TKN(SEMI); break;
case ',': SET_LIT_TKN(COMMA); break;
case '\'':
case '"':
{
char* pz = ao_string_cook(pCurCtx->pzScan, &(pCurCtx->lineNo));
if (pz == NULL)
goto NUL_error;
pz_token = pCurCtx->pzScan;
lastToken = DP_EV_STRING;
pCurCtx->pzScan = pz;
break;
}
case '<':
switch (lex_here_string()) {
case SUCCESS: break;
case FAILURE: goto BrokenToken;
case PROBLEM: return DP_EV_INVALID;
}
break;
case '(':
loadScheme();
break;
case '\\':
if (strncmp(pCurCtx->pzScan+1, "'(", (size_t)2) == 0) {
alist_to_autogen_def();
goto scanAgain;
}
lex_escaped_char();
break;
case '`':
switch (lex_backquote()) {
case FAILURE: goto NUL_error;
case PROBLEM: goto scanAgain;
case SUCCESS: break;
}
break;
case '/':
switch (lex_comment()) {
case SUCCESS: goto scanAgain;
default: break;
}
/* FALLTHROUGH */ /* to Invalid input char */
default:
BrokenToken:
pCurCtx->pzScan = assembleName(pCurCtx->pzScan, &lastToken);
break;
} /* switch (*pCurCtx->pzScan) */
return lastToken;
NUL_error:
AG_ABEND(aprf(zErrMsg, pzProg, "unterminated quote in definition",
pCurCtx->pzCtxFname, pCurCtx->lineNo));
return DP_EV_INVALID;
lex_done:
/*
* First time through, return the DP_EV_END token.
* Second time through, we really finish.
*/
if (pCurCtx->pzScan == zNil) {
pCurCtx->pCtx = pDoneCtx;
pDoneCtx = pCurCtx;
return DP_EV_INVALID;
}
pCurCtx->pzScan = (char*)zNil;
return DP_EV_END;
}
fz_error
pdf_lex(int *tok, fz_stream *f, char *buf, int n, int *sl)
{
while (1)
{
int c = fz_read_byte(f);
switch (c)
{
case EOF:
*tok = PDF_TOK_EOF;
return fz_okay;
case IS_WHITE:
lex_white(f);
break;
case '%':
lex_comment(f);
break;
case '/':
lex_name(f, buf, n);
*sl = strlen(buf);
*tok = PDF_TOK_NAME;
return fz_okay;
case '(':
*sl = lex_string(f, buf, n);
*tok = PDF_TOK_STRING;
return fz_okay;
case ')':
*tok = PDF_TOK_ERROR;
goto cleanuperror;
case '<':
c = fz_read_byte(f);
if (c == '<')
{
*tok = PDF_TOK_OPEN_DICT;
}
else
{
fz_unread_byte(f);
*sl = lex_hex_string(f, buf, n);
*tok = PDF_TOK_STRING;
}
return fz_okay;
case '>':
c = fz_read_byte(f);
if (c == '>')
{
*tok = PDF_TOK_CLOSE_DICT;
return fz_okay;
}
*tok = PDF_TOK_ERROR;
goto cleanuperror;
case '[':
*tok = PDF_TOK_OPEN_ARRAY;
return fz_okay;
case ']':
*tok = PDF_TOK_CLOSE_ARRAY;
return fz_okay;
case '{':
*tok = PDF_TOK_OPEN_BRACE;
return fz_okay;
case '}':
*tok = PDF_TOK_CLOSE_BRACE;
return fz_okay;
case IS_NUMBER:
fz_unread_byte(f);
*sl = lex_number(f, buf, n, tok);
return fz_okay;
default: /* isregular: !isdelim && !iswhite && c != EOF */
fz_unread_byte(f);
lex_name(f, buf, n);
*sl = strlen(buf);
*tok = pdf_token_from_keyword(buf);
return fz_okay;
}
}
cleanuperror:
*tok = PDF_TOK_ERROR;
return fz_throw("lexical error");
}