+ switch (ch)
+ {
+ case 'b':
+ return '\b';
+ case 'f':
+ return '\f';
+ case 'n':
+ return '\n';
+ case 'r':
+ return '\r';
+ case 't':
+ return '\t';
+ case '\'':
+ return '\'';
+ case '"':
+ return '\"';
+ default:
+ return ch;
+ }
+}
+
+/* This function is called to process input characters. The GET
+ parameter is used to retrieve more input characters. GET should
+ set its parameter to point to a buffer, and return the length of
+ the buffer; it should return 0 at end of file. The scrubbed output
+ characters are put into the buffer starting at TOSTART; the TOSTART
+ buffer is TOLEN bytes in length. The function returns the number
+ of scrubbed characters put into TOSTART. This will be TOLEN unless
+ end of file was seen. This function is arranged as a state
+ machine, and saves its state so that it may return at any point.
+ This is the way the old code used to work. */
+
+size_t
+do_scrub_chars (size_t (*get) (char *, size_t), char *tostart, size_t tolen)
+{
+ char *to = tostart;
+ char *toend = tostart + tolen;
+ char *from;
+ char *fromend;
+ size_t fromlen;
+ register int ch, ch2 = 0;
+ /* Character that started the string we're working on. */
+ static char quotechar;
+
+ /*State 0: beginning of normal line
+ 1: After first whitespace on line (flush more white)
+ 2: After first non-white (opcode) on line (keep 1white)
+ 3: after second white on line (into operands) (flush white)
+ 4: after putting out a .linefile, put out digits
+ 5: parsing a string, then go to old-state
+ 6: putting out \ escape in a "d string.
+ 7: no longer used
+ 8: no longer used
+ 9: After seeing symbol char in state 3 (keep 1white after symchar)
+ 10: After seeing whitespace in state 9 (keep white before symchar)
+ 11: After seeing a symbol character in state 0 (eg a label definition)
+ -1: output string in out_string and go to the state in old_state
+ -2: flush text until a '*' '/' is seen, then go to state old_state
+#ifdef TC_V850
+ 12: After seeing a dash, looking for a second dash as a start
+ of comment.
+#endif
+#ifdef DOUBLEBAR_PARALLEL
+ 13: After seeing a vertical bar, looking for a second
+ vertical bar as a parallel expression separator.
+#endif
+#ifdef TC_PREDICATE_START_CHAR
+ 14: After seeing a predicate start character at state 0, looking
+ for a predicate end character as predicate.
+ 15: After seeing a predicate start character at state 1, looking
+ for a predicate end character as predicate.
+#endif
+#ifdef TC_Z80
+ 16: After seeing an 'a' or an 'A' at the start of a symbol
+ 17: After seeing an 'f' or an 'F' in state 16
+#endif
+ */
+
+ /* I added states 9 and 10 because the MIPS ECOFF assembler uses
+ constructs like ``.loc 1 20''. This was turning into ``.loc
+ 120''. States 9 and 10 ensure that a space is never dropped in
+ between characters which could appear in an identifier. Ian
+ Taylor, ian@cygnus.com.
+
+ I added state 11 so that something like "Lfoo add %r25,%r26,%r27" works
+ correctly on the PA (and any other target where colons are optional).
+ Jeff Law, law@cs.utah.edu.
+
+ I added state 13 so that something like "cmp r1, r2 || trap #1" does not
+ get squashed into "cmp r1,r2||trap#1", with the all important space
+ between the 'trap' and the '#1' being eliminated. nickc@cygnus.com */
+
+ /* This macro gets the next input character. */
+
+#define GET() \
+ (from < fromend \
+ ? * (unsigned char *) (from++) \
+ : (saved_input = NULL, \
+ fromlen = (*get) (input_buffer, sizeof input_buffer), \
+ from = input_buffer, \
+ fromend = from + fromlen, \
+ (fromlen == 0 \
+ ? EOF \
+ : * (unsigned char *) (from++))))
+
+ /* This macro pushes a character back on the input stream. */
+
+#define UNGET(uch) (*--from = (uch))
+
+ /* This macro puts a character into the output buffer. If this
+ character fills the output buffer, this macro jumps to the label
+ TOFULL. We use this rather ugly approach because we need to
+ handle two different termination conditions: EOF on the input
+ stream, and a full output buffer. It would be simpler if we
+ always read in the entire input stream before processing it, but
+ I don't want to make such a significant change to the assembler's
+ memory usage. */
+
+#define PUT(pch) \
+ do \
+ { \
+ *to++ = (pch); \
+ if (to >= toend) \
+ goto tofull; \
+ } \
+ while (0)
+
+ if (saved_input != NULL)
+ {
+ from = saved_input;
+ fromend = from + saved_input_len;
+ }
+ else
+ {
+ fromlen = (*get) (input_buffer, sizeof input_buffer);
+ if (fromlen == 0)
+ return 0;
+ from = input_buffer;
+ fromend = from + fromlen;
+ }
+
+ while (1)
+ {
+ /* The cases in this switch end with continue, in order to
+ branch back to the top of this while loop and generate the
+ next output character in the appropriate state. */
+ switch (state)
+ {
+ case -1:
+ ch = *out_string++;
+ if (*out_string == '\0')
+ {
+ state = old_state;
+ old_state = 3;
+ }
+ PUT (ch);
+ continue;
+