Optimize `peek-char'.

This makes `peek-char' 40x faster on a port whose encoding is
faster on a UTF-8 port containing multi-byte codepoints.

The `xml->sxml' procedure is 4x faster on a 2.7 MiB XML file.

* libguile/ports.c (get_codepoint): New procedure, moved here from
  `scm_getc', with the additional BUF and LEN parameters.
  (scm_getc): Use it.
  (scm_peek_char): Use it instead of the `scm_getc'/`scm_ungetc'
  sequence.

* test-suite/tests/ports.test ("string ports")["peek-char [latin-1]",
  "peek-char [utf-8]"]: New tests.

* benchmark-suite/Makefile.am (SCM_BENCHMARKS): Add
  `benchmarks/ports.bm'.

* benchmark-suite/benchmarks/ports.bm: New file.

benchmark-suite/Makefile.am

@@ -4,6 +4,7 @@ SCM_BENCHMARKS = benchmarks/0-reference.bm		\
 		 benchmarks/continuations.bm		\
                  benchmarks/if.bm			\
                  benchmarks/logand.bm			\
+		 benchmarks/ports.bm			\
 		 benchmarks/read.bm			\
 		 benchmarks/srfi-1.bm			\
 		 benchmarks/srfi-13.bm			\

benchmark-suite/benchmarks/ports.bm

@@ -0,0 +1,67 @@
+;;; ports.bm --- Port I/O.         -*- mode: scheme; coding: utf-8; -*-
+;;;
+;;; Copyright (C) 2010 Free Software Foundation, Inc.
+;;;
+;;; This program is free software; you can redistribute it and/or
+;;; modify it under the terms of the GNU Lesser General Public License
+;;; as published by the Free Software Foundation; either version 3, or
+;;; (at your option) any later version.
+;;;
+;;; This program is distributed in the hope that it will be useful,
+;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
+;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+;;; GNU Lesser General Public License for more details.
+;;;
+;;; You should have received a copy of the GNU Lesser General Public
+;;; License along with this software; see the file COPYING.LESSER.  If
+;;; not, write to the Free Software Foundation, Inc., 51 Franklin
+;;; Street, Fifth Floor, Boston, MA 02110-1301 USA
+
+(define-module (benchmarks ports)
+  #:use-module (benchmark-suite lib))
+
+(define %latin1-port
+  (with-fluids ((%default-port-encoding #f))
+    (open-input-string "hello, world")))
+
+(define %utf8/ascii-port
+  (with-fluids ((%default-port-encoding "UTF-8"))
+    (open-input-string "hello, world")))
+
+(define %utf8/wide-port
+  (with-fluids ((%default-port-encoding "UTF-8"))
+    (open-input-string "안녕하세요")))
+
+
+(with-benchmark-prefix "peek-char"
+
+  (benchmark "latin-1 port" 700000
+    (peek-char %latin1-port))
+
+  (benchmark "utf-8 port, ascii character" 700000
+    (peek-char %utf8/ascii-port))
+
+  (benchmark "utf-8 port, Korean character" 700000
+    (peek-char %utf8/wide-port)))
+
+(with-benchmark-prefix "read-char"
+
+  (benchmark "latin-1 port" 10000000
+    (read-char %latin1-port))
+
+  (benchmark "utf-8 port, ascii character" 10000000
+    (read-char %utf8/ascii-port))
+
+  (benchmark "utf-8 port, Korean character" 10000000
+    (read-char %utf8/wide-port)))
+
+(with-benchmark-prefix "char-ready?"
+
+  (benchmark "latin-1 port" 10000000
+    (char-ready? %latin1-port))
+
+  (benchmark "utf-8 port, ascii character" 10000000
+    (char-ready? %utf8/ascii-port))
+
+  (benchmark "utf-8 port, Korean character" 10000000
+    (char-ready? %utf8/wide-port)))

libguile/ports.c

@@ -1023,13 +1023,15 @@ SCM_DEFINE (scm_read_char, "read-char", 0, 1, 0,
 
 #define SCM_MBCHAR_BUF_SIZE (4)
 
-/* Get one codepoint from a file, using the port's encoding.  */
+/* Read a codepoint from PORT and return it.  Fill BUF with the byte
-scm_t_wchar
+   representation of the codepoint in PORT's encoding, and set *LEN to
-scm_getc (SCM port)
+   the length in bytes of that representation.  Raise an error on
+   failure.  */
+static scm_t_wchar
+get_codepoint (SCM port, char buf[SCM_MBCHAR_BUF_SIZE], size_t *len)
 {
   int c;
-  unsigned int bufcount = 0;
+  size_t bufcount = 0;
-  char buf[SCM_MBCHAR_BUF_SIZE];
   scm_t_uint32 result_buf;
   scm_t_wchar codepoint = 0;
   scm_t_uint32 *u32;
@@ -1133,6 +1135,8 @@ scm_getc (SCM port)
       break;
     }
 
+  *len = bufcount;
+
   return codepoint;
 
  failure:
@@ -1155,6 +1159,15 @@ scm_getc (SCM port)
   return 0;
 }
 
+/* Read a codepoint from PORT and return it.  */
+scm_t_wchar
+scm_getc (SCM port)
+{
+  size_t len;
+  char buf[SCM_MBCHAR_BUF_SIZE];
+
+  return get_codepoint (port, buf, &len);
+}
 
 /* this should only be called when the read buffer is empty.  it
    tries to refill the read buffer.  it returns the first char from
@@ -1635,18 +1648,37 @@ SCM_DEFINE (scm_peek_char, "peek-char", 0, 1, 0,
 	    "to @code{read-char} would have hung.")
 #define FUNC_NAME s_scm_peek_char
 {
-  scm_t_wchar c, column;
+  SCM result;
+  scm_t_wchar c;
+  char bytes[SCM_MBCHAR_BUF_SIZE];
+  long column, line;
+  size_t len;
+
   if (SCM_UNBNDP (port))
     port = scm_current_input_port ();
   else
     SCM_VALIDATE_OPINPORT (1, port);
-  column = SCM_COL(port);
+
-  c = scm_getc (port);
+  column = SCM_COL (port);
-  if (EOF == c)
+  line = SCM_LINUM (port);
-    return SCM_EOF_VAL;
+
-  scm_ungetc (c, port);
+  c = get_codepoint (port, bytes, &len);
-  SCM_COL(port) = column;
+  if (c == EOF)
-  return SCM_MAKE_CHAR (c);
+    result = SCM_EOF_VAL;
+  else
+    {
+      long i;
+
+      result = SCM_MAKE_CHAR (c);
+
+      for (i = len - 1; i >= 0; i--)
+	scm_unget_byte (bytes[i], port);
+
+      SCM_COL (port) = column;
+      SCM_LINUM (port) = line;
+    }
+
+  return result;
 }
 #undef FUNC_NAME
 

test-suite/tests/ports.test

@@ -422,7 +422,25 @@
           (and (eq? faulty-str str)
                (string=? from "UTF-32")
                (string=? to "ISO-8859-1")
-               (string? (strerror errno))))))))
+               (string? (strerror errno)))))))
+
+  (pass-if "peek-char [latin-1]"
+    (let ((p (with-fluids ((%default-port-encoding #f))
+               (open-input-string "hello, world"))))
+      (and (char=? (peek-char p) #\h)
+           (char=? (peek-char p) #\h)
+           (char=? (peek-char p) #\h)
+           (= (port-line p) 0)
+           (= (port-column p) 0))))
+
+  (pass-if "peek-char [utf-8]"
+    (let ((p (with-fluids ((%default-port-encoding "UTF-8"))
+               (open-input-string "안녕하세요"))))
+      (and (char=? (peek-char p) #\안)
+           (char=? (peek-char p) #\안)
+           (char=? (peek-char p) #\안)
+           (= (port-line p) 0)
+           (= (port-column p) 0)))))
 
 (with-test-prefix "call-with-output-string"