X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?p=pspp;a=blobdiff_plain;f=src%2Flibpspp%2Fline-reader.c;fp=src%2Flibpspp%2Fline-reader.c;h=6f90b502cd856dd64db366ce1bb36561141648f0;hp=0000000000000000000000000000000000000000;hb=20af4ead4d4c440c5bc269274cd5a87fa9e7056d;hpb=bde3425960865b3740bca3e883864c7597dd59a3

diff --git a/src/libpspp/line-reader.c b/src/libpspp/line-reader.c
new file mode 100644
index 0000000000..6f90b502cd
--- /dev/null
+++ b/src/libpspp/line-reader.c
@@ -0,0 +1,350 @@
+/* PSPP - a program for statistical analysis.
+   Copyright (C) 2010, 2011 Free Software Foundation, Inc.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 3 of the License, 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 General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+#include "line-reader.h"
+
+#include <assert.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+
+#include "libpspp/assertion.h"
+#include "libpspp/encoding-guesser.h"
+#include "libpspp/i18n.h"
+#include "libpspp/str.h"
+
+#include "gl/minmax.h"
+#include "gl/xalloc.h"
+
+enum line_reader_state
+  {
+    S_UNIBYTE,                  /* Known stream encoding, 1-byte unit. */
+    S_MULTIBYTE,                /* Known stream encoding, multibyte unit. */
+    S_AUTO                      /* Encoding autodetection in progress. */
+  };
+
+struct line_reader
+  {
+    int fd;
+    enum line_reader_state state;
+    struct encoding_info encoding_info;
+
+    char *encoding;             /* Current encoding. */
+    char *auto_encoding;        /* In S_AUTO mode, user-specified encoding. */
+
+    char *buffer;
+    char *head;
+    size_t length;
+
+    int error;
+    bool eof;
+  };
+
+static ssize_t fill_buffer (struct line_reader *);
+
+/* Opens FILENAME, which is encoded in ENCODING, for reading line by line,
+   passing FLAGS to the open() function.  Returns a new line_reader if
+   successful, otherwise returns NULL and sets errno to an appropriate value.
+
+   The accepted forms for ENCODING are listed at the top of
+   encoding-guesser.h. */
+struct line_reader *
+line_reader_for_file (const char *encoding, const char *filename, int flags)
+{
+  struct line_reader *r;
+  int fd;
+
+  assert (!(flags & O_CREAT));
+
+  fd = open (filename, flags);
+  if (fd < 0)
+    return NULL;
+
+  r = line_reader_for_fd (encoding, fd);
+  if (r == NULL)
+    {
+      int save_errno = errno;
+      close (fd);
+      errno = save_errno;
+    }
+
+  return r;
+}
+
+/* Creates and returns a new line_reader that reads its input from FD.  Returns
+   a new line_reader if successful, otherwise returns NULL and sets errno to an
+   appropriate value.
+
+   The accepted forms for ENCODING are listed at the top of
+   encoding-guesser.h. */
+struct line_reader *
+line_reader_for_fd (const char *encoding, int fd)
+{
+  struct line_reader *r;
+
+  r = calloc (1, sizeof *r);
+  if (r == NULL)
+    return NULL;
+
+  r->fd = fd;
+  r->buffer = malloc (LINE_READER_BUFFER_SIZE);
+  if (r->buffer == NULL)
+    goto error;
+  r->head = r->buffer;
+  r->length = 0;
+
+  if (fill_buffer (r) < 0)
+    goto error;
+
+  r->encoding = xstrdup (encoding_guess_head_encoding (
+                           encoding, r->buffer, r->length));
+  if (!get_encoding_info (&r->encoding_info, r->encoding))
+    {
+      errno = EINVAL;
+      goto error;
+    }
+
+  if (encoding_guess_encoding_is_auto (encoding)
+      && !strcmp (r->encoding, "ASCII"))
+    {
+      r->state = S_AUTO;
+      r->auto_encoding = xstrdup (encoding);
+    }
+  else
+    r->state = r->encoding_info.unit == 1 ? S_UNIBYTE : S_MULTIBYTE;
+
+  return r;
+
+error:
+  line_reader_free (r);
+  return NULL;
+}
+
+/* Closes R and its underlying file descriptor and frees all associated
+   resources.  Returns the return value from close(). */
+int
+line_reader_close (struct line_reader *r)
+{
+  if (r != NULL)
+    {
+      int fd = r->fd;
+      line_reader_free (r);
+      return close (fd);
+    }
+  return 0;
+}
+
+/* Frees R and associated resources, but does not close the underlying file
+   descriptor.  (Thus, the client must close the file descriptor when it is no
+   longer needed.) */
+void
+line_reader_free (struct line_reader *r)
+{
+  if (r != NULL)
+    {
+      free (r->buffer);
+      free (r->encoding);
+      free (r->auto_encoding);
+      free (r);
+    }
+}
+
+static ssize_t
+fill_buffer (struct line_reader *r)
+{
+  ssize_t n;
+
+  /* Move any unused bytes to the beginning of the input buffer. */
+  if (r->length > 0 && r->buffer != r->head)
+    memmove (r->buffer, r->head, r->length);
+  r->head = r->buffer;
+
+  /* Read more input. */
+  do
+    {
+      n = read (r->fd, r->buffer + r->length,
+                LINE_READER_BUFFER_SIZE - r->length);
+    }
+  while (n < 0 && errno == EINTR);
+  if (n > 0)
+    r->length += n;
+  else if (n < 0)
+    r->error = errno;
+  else
+    r->eof = true;
+  return n;
+}
+
+static void
+output_bytes (struct line_reader *r, struct string *s, size_t n)
+{
+  ds_put_substring (s, ss_buffer (r->head, n));
+  r->head += n;
+  r->length -= n;
+}
+
+static void
+output_line (struct line_reader *r, struct string *s, size_t n)
+{
+  int unit = r->encoding_info.unit;
+
+  output_bytes (r, s, n);
+
+  r->head += unit;
+  r->length -= unit;
+
+  ds_chomp (s, ss_buffer (r->encoding_info.cr, unit));
+}
+
+/* Reads a line of text, but no more than MAX_LENGTH bytes, from R and appends
+   it to S, omitting the final new-line and the carriage return that
+   immediately precedes it, if one is present.  The line is left in its
+   original encoding.
+
+   Returns true if anything was successfully read from the file.  (If an empty
+   line was read, then nothing is appended to S.)  Returns false if end of file
+   was reached or a read error occurred before any text could be read. */
+bool
+line_reader_read (struct line_reader *r, struct string *s, size_t max_length)
+{
+  size_t original_length = ds_length (s);
+  int unit = r->encoding_info.unit;
+
+  do
+    {
+      size_t max_out = max_length - (ds_length (s) - original_length);
+      size_t max_in = r->length;
+      size_t max = MIN (max_in, max_out);
+      size_t n;
+      char *p;
+
+      if (max_out < unit)
+        break;
+
+      switch (r->state)
+        {
+        case S_UNIBYTE:
+          p = memchr (r->head, r->encoding_info.lf[0], max);
+          if (p != NULL)
+            {
+              output_line (r, s, p - r->head);
+              return true;
+            }
+          n = max;
+          break;
+
+        case S_MULTIBYTE:
+          for (n = 0; n + unit <= max; n += unit)
+            if (!memcmp (r->head + n, r->encoding_info.lf, unit))
+              {
+                output_line (r, s, n);
+                return true;
+              }
+          break;
+
+        case S_AUTO:
+          for (n = 0; n < max; n++)
+            if (!encoding_guess_is_ascii_text (r->head[n]))
+              {
+                char *encoding;
+
+                output_bytes (r, s, n);
+                fill_buffer (r);
+                r->state = S_UNIBYTE;
+
+                encoding = xstrdup (encoding_guess_tail_encoding (
+                                      r->auto_encoding, r->head, r->length));
+                free (r->encoding);
+                r->encoding = encoding;
+
+                free (r->auto_encoding);
+                r->auto_encoding = NULL;
+
+                n = 0;
+                break;
+              }
+            else if (r->head[n] == '\n')
+              {
+                output_line (r, s, n);
+                return true;
+              }
+          break;
+
+        default:
+          NOT_REACHED ();
+        }
+
+      output_bytes (r, s, n);
+    }
+  while (r->length >= unit || fill_buffer (r) > 0);
+
+  return ds_length (s) > original_length;
+}
+
+/* Returns the file descriptor underlying R. */
+int
+line_reader_fileno (const struct line_reader *r)
+{
+  return r->fd;
+}
+
+/* Returns the offset in the file of the next byte to be read from R, or -1 on
+   error (e.g. if the file is not seekable). */
+off_t
+line_reader_tell (const struct line_reader *r)
+{
+  off_t pos = lseek (r->fd, 0, SEEK_CUR);
+  if (pos >= 0)
+    pos = MAX (0, pos - r->length);
+  return pos;
+}
+
+/* Returns true if end of file has been encountered reading R. */
+bool
+line_reader_eof (const struct line_reader *r)
+{
+  return r->eof && !r->length;
+}
+
+/* Returns an nonzero errno value if an error has been encountered reading
+   R, zero otherwise. */
+int
+line_reader_error (const struct line_reader *r)
+{
+  return !r->length ? r->error : 0;
+}
+
+/* Returns the encoding of R.  If line_reader_is_auto(R) returns true, the
+   encoding might change as more lines are read. */
+const char *
+line_reader_get_encoding (const struct line_reader *r)
+{
+  return r->encoding;
+}
+
+/* Returns true if the encoding of the file being read by R is not yet
+   completely known.  If this function returns true, then the encoding returned
+   by line_reader_get_encoding() might change as more lines are read (and after
+   the change, this function will return false). */
+bool
+line_reader_is_auto (const struct line_reader *r)
+{
+  return r->state == S_AUTO;
+}