cut: optimize when no delimiter in input

This is about 20x faster.
Note we only do the delimiter search once per chunk,
and it's usually quick as delimiters wouldn't be too far
into the a chunk if present, so we don't bother
to cache the found delimiter.

diff --git a/src/cut.c b/src/cut.c
index 15f2c0fa9ded71e114d374cb67cbfb71b66cd556..208fa6e4a62901acf89e62527deb08107596a06e 100644 (file)
--- a/src/cut.c
+++ b/src/cut.c
@@ -936,6 +936,31 @@ cut_fields_bytesearch (FILE *stream)
           .blank_delimited = whitespace_delimited
         };
 
+      /* Shortcut the case were there is no delimiter in input,
+         as directly outputting without parsing is 20x faster.  */
+      if (field_idx == 1
+          && !suppress_non_delimited && !whitespace_delimited
+          && !field_delim_is_line_delim ()
+          && !have_pending_line
+          && field_1_n_bytes == 0
+          && !skip_line_remainder
+          && !find_bytesearch_field_delim (chunk, safe))
+        {
+          char *last_line_delim = feof (mbbuf.fp) ? chunk + safe - 1
+                                  : memrchr ((void *) chunk, line_delim, safe);
+          if (last_line_delim)
+            {
+              idx_t n = last_line_delim - chunk + 1;
+              write_bytes (chunk, n);
+              if (feof (mbbuf.fp) && chunk[n - 1] != line_delim)
+                write_line_delim ();
+              mbbuf_advance (&mbbuf, n);
+              if (feof (mbbuf.fp))
+                return;
+              continue;
+            }
+        }
+
       while (processed < safe)
         {
           char *terminator = NULL;

diff --git a/tests/cut/cut.pl b/tests/cut/cut.pl
index 8c3c06653ef4754840df8ed68994a3d0310627c2..bfccdbe85f05e52c7dcc7f07ccd8051ed6c31f5a 100755 (executable)
--- a/tests/cut/cut.pl
+++ b/tests/cut/cut.pl
@@ -170,6 +170,7 @@ my @Tests =
   ['newline-4', '-d:', '-f1', {IN=>"a:1\nb:2"}, {OUT=>"a\nb\n"}],
   ['newline-5', '-d:', '-f2', {IN=>"a:1\nb:2\n"}, {OUT=>"1\n2\n"}],
   ['newline-6', '-d:', '-f2', {IN=>"a:1\nb:2"}, {OUT=>"1\n2\n"}],
+  ['newline-6a', '-d:', '-f2', {IN=>"a\nb"}, {OUT=>"a\nb\n"}],
   ['newline-7', '-s', '-d:', '-f1', {IN=>"a:1\nb:2"}, {OUT=>"a\nb\n"}],
   ['newline-8', '-s', '-d:', '-f1', {IN=>"a:1\nb:2\n"}, {OUT=>"a\nb\n"}],
   ['newline-9', '-s', '-d:', '-f1', {IN=>"a1\nb2"}, {OUT=>""}],