cut: ensure responsive input processing

* gl/lib/mbbuf.h (fill_buf): Switch from fread() to read()
as the former retries read() internally to fill the buffer.
* src/cut.c: Adjust accordingly, and avoid getc() interface entirely.
* bootstrap.h: Depend explicitly on fseterr.  This is already depended
on transitively, so should not introduce new build portability issues.

diff --git a/bootstrap.conf b/bootstrap.conf
index d0e3b8842bedf9aea312da6d25256651555b2224..bc31a58623ad80538d01f479a5666128bd18a264 100644 (file)
--- a/bootstrap.conf
+++ b/bootstrap.conf
@@ -119,6 +119,7 @@ gnulib_modules="
   freopen
   freopen-safer
   fseeko
+  fseterr
   fstatat
   fsusage
   fsync

diff --git a/gl/lib/mbbuf.h b/gl/lib/mbbuf.h
index 84e1fdecf58dbc7c4cf5ef7a291d3be2f9974b84..ed02a83d54e361a1929da0785f144bf8d6a02abe 100644 (file)
--- a/gl/lib/mbbuf.h
+++ b/gl/lib/mbbuf.h
@@ -25,7 +25,9 @@
 
 #include <stdio.h>
 #include <stddef.h>
+#include <unistd.h>
 
+#include "fseterr.h"
 #include "mcel.h"
 #include "idx.h"
 
@@ -47,6 +49,7 @@ typedef struct
   idx_t size;      /* Number of bytes allocated for BUFFER.  */
   idx_t length;    /* Number of bytes with data in BUFFER.  */
   idx_t offset;    /* Current position in BUFFER.  */
+  bool eof;        /* Whether at End Of File.  */
 } mbbuf_t;
 
 MBBUF_INLINE idx_t
@@ -68,16 +71,19 @@ mbbuf_init (mbbuf_t *mbbuf, char *buffer, idx_t size, FILE *fp)
   mbbuf->size = size;
   mbbuf->length = 0;
   mbbuf->offset = 0;
+  mbbuf->eof = false;
 }
 
 /* Fill the input buffer with at least MCEL_LEN_MAX bytes if possible.
-   Return the number of bytes available from the current offset.  */
+   Return the number of bytes available from the current offset.
+   At end of file, MBBUF.EOF is set, and zero will eventually be returned.
+   Note feof() will _NOT_ be set on the MBBUF.FP.  */
 MBBUF_INLINE idx_t
 mbbuf_fill (mbbuf_t *mbbuf)
 {
   idx_t available = mbbuf_avail (mbbuf);
 
-  if (available < MCEL_LEN_MAX && ! feof (mbbuf->fp))
+  if (available < MCEL_LEN_MAX && ! mbbuf->eof)
     {
       idx_t start;
       if (!(0 < available))
@@ -87,8 +93,20 @@ mbbuf_fill (mbbuf_t *mbbuf)
           memmove (mbbuf->buffer, mbbuf->buffer + mbbuf->offset, available);
           start = available;
         }
-      mbbuf->length = fread (mbbuf->buffer + start, 1, mbbuf->size - start,
-                             mbbuf->fp) + start;
+      ssize_t  read_ret = read (fileno (mbbuf->fp), mbbuf->buffer + start,
+                                mbbuf->size - start);
+      if (read_ret < 0)
+        {
+          fseterr (mbbuf->fp);
+          mbbuf->eof = true;  /* Avoid any more reads().  */
+          mbbuf->length = start;
+        }
+      else
+        {
+          mbbuf->eof = read_ret == 0;
+          mbbuf->length = read_ret + start;
+        }
+
       mbbuf->offset = 0;
       available = mbbuf_avail (mbbuf);
     }

diff --git a/src/cut.c b/src/cut.c
index 3a941dadd77c9eb7afa155be6711ce2f5b92539b..3ea85c4661a657dbb37df4dd00553f41dc2b4036 100644 (file)
--- a/src/cut.c
+++ b/src/cut.c
@@ -727,17 +727,22 @@ reset_field_line (uintmax_t *field_idx, bool *found_any_selected_field,
    callers to hand off mid-line.  */
 
 static void
-cut_bytes_buffered (FILE *stream, uintmax_t *byte_idx, bool *print_delimiter)
+cut_bytes (FILE *stream)
 {
   static char bytes_in[IO_BUFSIZE];
+  uintmax_t byte_idx = 0;
+  bool print_delimiter = false;
+
+  current_rp = frp;
 
   while (true)
     {
-      idx_t available = fread (bytes_in, sizeof *bytes_in, sizeof bytes_in,
-                               stream);
-      if (available == 0)
+      idx_t available = read (fileno (stream), bytes_in, sizeof bytes_in);
+      if (available <= 0)
         {
-          write_pending_line_delim (*byte_idx);
+          if (available < 0)
+            fseterr (stream);
+          write_pending_line_delim (byte_idx);
           break;
         }
 
@@ -754,22 +759,22 @@ cut_bytes_buffered (FILE *stream, uintmax_t *byte_idx, bool *print_delimiter)
 
           while (p < end)
             {
-              sync_byte_selection (*byte_idx);
+              sync_byte_selection (byte_idx);
 
-              if (*byte_idx + 1 < current_rp->lo)
+              if (byte_idx + 1 < current_rp->lo)
                 {
-                  idx_t skip = MIN (end - p, current_rp->lo - (*byte_idx + 1));
+                  idx_t skip = MIN (end - p, current_rp->lo - (byte_idx + 1));
                   p += skip;
-                  *byte_idx += skip;
+                  byte_idx += skip;
                 }
               else
                 {
-                  idx_t n = MIN (end - p, current_rp->hi - *byte_idx);
-                  write_selected_item (print_delimiter,
-                                       is_range_start_index (*byte_idx + 1),
+                  idx_t n = MIN (end - p, current_rp->hi - byte_idx);
+                  write_selected_item (&print_delimiter,
+                                       is_range_start_index (byte_idx + 1),
                                        p, n);
                   p += n;
-                  *byte_idx += n;
+                  byte_idx += n;
                 }
             }
 
@@ -777,49 +782,7 @@ cut_bytes_buffered (FILE *stream, uintmax_t *byte_idx, bool *print_delimiter)
           if (line_end)
             {
               processed++;
-              reset_item_line (byte_idx, print_delimiter);
-            }
-        }
-    }
-}
-
-/* Read from stream STREAM, printing to standard output any selected bytes.
-   This avoids data copies and function calls for short lines,
-   and will defer to cut_bytes_buffered() once a longer line is encountered.  */
-
-static void
-cut_bytes (FILE *stream)
-{
-  uintmax_t byte_idx = 0;
-  bool print_delimiter = false;
-
-  current_rp = frp;
-
-  while (true)
-    {
-      int c = getc (stream);
-
-      if (c == line_delim)
-        reset_item_line (&byte_idx, &print_delimiter);
-      else if (c == EOF)
-        {
-          write_pending_line_delim (byte_idx);
-          break;
-        }
-      else
-        {
-          next_item (&byte_idx);
-          if (print_kth (byte_idx))
-            {
-              char ch = c;
-              write_selected_item (&print_delimiter,
-                                   is_range_start_index (byte_idx), &ch, 1);
-            }
-
-          if (SMALL_BYTE_THRESHOLD < byte_idx)
-            {
-              cut_bytes_buffered (stream, &byte_idx, &print_delimiter);
-              break;
+              reset_item_line (&byte_idx, &print_delimiter);
             }
         }
     }
@@ -1023,7 +986,7 @@ cut_fields_bytesearch (FILE *stream)
       idx_t safe = mbbuf_fill (&mbbuf);
       if (safe == 0)
         break;
-      search.at_eof = feof (mbbuf.fp);
+      search.at_eof = mbbuf.eof;
       search.line_end_known = false;
 
       char *chunk = mbbuf.buffer + mbbuf.offset;