join: with --check-order print offending file name, line number and data

* src/join (g_names): New global (was main's "names").
(main): Update all uses of "names".
(line_no[2]): New globals.
(get_line): Increment after reading each line.
(check_order): Print the standard "file name:line_no: " prefix
as well as the offending line when reporting disorder.
Here is a sample old/new comparison:
  -join: file 1 is not in sorted order
  +join: in:4: is not sorted: contents-of-line-4
* tests/misc/join: Change the two affected tests to expect
the new diagnostic.
Add new tests for more coverage: mismatch in file 2,
two diagnostics, zero-length out-of-order line.
* NEWS (Improvements): Mention it.
Suggested by David Gast in http://debbugs.gnu.org/9236

diff --git a/NEWS b/NEWS
index 2e484979ea7939c0a4deb7aa8f0b0ddfed1a506a..6e24f5cb1a5462f7fde2e28d0b8691d4aa58c361 100644 (file)
--- a/NEWS
+++ b/NEWS
@@ -66,6 +66,9 @@ GNU coreutils NEWS                                    -*- outline -*-
   df now supports disk partitions larger than 4 TiB on MacOS X 10.5
   or newer and on AIX 5.2 or newer.
 
+  join --check-order now prints "join: FILE:LINE_NUMBER: bad_line" for an
+  unsorted input, rather than e.g., "join: file 1 is not in sorted order".
+
   shuf outputs small subsets of large permutations much more efficiently.
   For example `shuf -i1-$((2**32-1)) -n2` no longer exhausts memory.
 

diff --git a/THANKS.in b/THANKS.in
index 6353708f086524d0b411b7c0b8b5a2a2b7f4c7d3..23ac67920fe2aa11dcd352bb42c99538d5b4cbbd 100644 (file)
--- a/THANKS.in
+++ b/THANKS.in
@@ -139,6 +139,7 @@ David Alan Gilbert                  gilbertd@treblig.org
 David Dyck                          dcd@tc.fluke.COM
 David Eisner                        cradle@umd.edu
 David Flynn                         dav@chess.plus.com
+David Gast                          dgast@csulb.edu
 David Godfrey                       dave@delta.demon.co.uk
 David Luyer                         david_luyer@pacific.net.au
 David Madore                        david.madore@ens.fr

diff --git a/src/join.c b/src/join.c
index 99d918fd1a6a30a830e84454a0369e4f3b80f619..694fb55735991703881805079bbcec9046e85f43 100644 (file)
--- a/src/join.c
+++ b/src/join.c
@@ -86,9 +86,15 @@ struct seq
     struct line **lines;
   };
 
-/* The previous line read from each file. */
+/* The previous line read from each file.  */
 static struct line *prevline[2] = {NULL, NULL};
 
+/* The number of lines read from each file.  */
+static uintmax_t line_no[2] = {0, 0};
+
+/* The input file names.  */
+static char *g_names[2];
+
 /* This provides an extra line buffer for each file.  We need these if we
    try to read two consecutive lines into the same buffer, since we don't
    want to overwrite the previous buffer before we check order. */
@@ -384,12 +390,23 @@ check_order (const struct line *prev,
           size_t join_field = whatfile == 1 ? join_field_1 : join_field_2;
           if (keycmp (prev, current, join_field, join_field) > 0)
             {
+              /* Exclude any trailing newline. */
+              size_t len = current->buf.length;
+              if (0 < len && current->buf.buffer[len - 1] == '\n')
+                --len;
+
+              /* If the offending line is longer than INT_MAX, output
+                 only the first INT_MAX bytes in this diagnostic.  */
+              len = MIN (INT_MAX, len);
+
               error ((check_input_order == CHECK_ORDER_ENABLED
                       ? EXIT_FAILURE : 0),
-                     0, _("file %d is not in sorted order"), whatfile);
+                     0, _("%s:%ju: is not sorted: %.*s"),
+                     g_names[whatfile - 1], line_no[whatfile - 1],
+                     (int) len, current->buf.buffer);
 
-              /* If we get to here, the message was just a warning, but we
-                 want only to issue it once. */
+              /* If we get to here, the message was merely a warning.
+                 Arrange to issue it only once per file.  */
               issued_disorder_warning[whatfile-1] = true;
             }
         }
@@ -436,6 +453,7 @@ get_line (FILE *fp, struct line **linep, int which)
       freeline (line);
       return false;
     }
+  ++line_no[which - 1];
 
   xfields (line);
 
@@ -980,7 +998,6 @@ main (int argc, char **argv)
   int prev_optc_status = MUST_BE_OPERAND;
   int operand_status[2];
   int joption_count[2] = { 0, 0 };
-  char *names[2];
   FILE *fp1, *fp2;
   int optc;
   int nfiles = 0;
@@ -1100,7 +1117,7 @@ main (int argc, char **argv)
           break;
 
         case 1:                /* Non-option argument.  */
-          add_file_name (optarg, names, operand_status, joption_count,
+          add_file_name (optarg, g_names, operand_status, joption_count,
                          &nfiles, &prev_optc_status, &optc_status);
           break;
 
@@ -1122,7 +1139,7 @@ main (int argc, char **argv)
   /* Process any operands after "--".  */
   prev_optc_status = MUST_BE_OPERAND;
   while (optind < argc)
-    add_file_name (argv[optind++], names, operand_status, joption_count,
+    add_file_name (argv[optind++], g_names, operand_status, joption_count,
                    &nfiles, &prev_optc_status, &optc_status);
 
   if (nfiles != 2)
@@ -1148,20 +1165,20 @@ main (int argc, char **argv)
   if (join_field_2 == SIZE_MAX)
     join_field_2 = 0;
 
-  fp1 = STREQ (names[0], "-") ? stdin : fopen (names[0], "r");
+  fp1 = STREQ (g_names[0], "-") ? stdin : fopen (g_names[0], "r");
   if (!fp1)
-    error (EXIT_FAILURE, errno, "%s", names[0]);
-  fp2 = STREQ (names[1], "-") ? stdin : fopen (names[1], "r");
+    error (EXIT_FAILURE, errno, "%s", g_names[0]);
+  fp2 = STREQ (g_names[1], "-") ? stdin : fopen (g_names[1], "r");
   if (!fp2)
-    error (EXIT_FAILURE, errno, "%s", names[1]);
+    error (EXIT_FAILURE, errno, "%s", g_names[1]);
   if (fp1 == fp2)
     error (EXIT_FAILURE, errno, _("both files cannot be standard input"));
   join (fp1, fp2);
 
   if (fclose (fp1) != 0)
-    error (EXIT_FAILURE, errno, "%s", names[0]);
+    error (EXIT_FAILURE, errno, "%s", g_names[0]);
   if (fclose (fp2) != 0)
-    error (EXIT_FAILURE, errno, "%s", names[1]);
+    error (EXIT_FAILURE, errno, "%s", g_names[1]);
 
   if (issued_disorder_warning[0] || issued_disorder_warning[1])
     exit (EXIT_FAILURE);

diff --git a/tests/misc/join b/tests/misc/join
index eae3f18a8eb8105a2f2d7c3965256c404680e23f..a892a10d24870ef63ea9ebb80bff87132b79b2c5 100755 (executable)
--- a/tests/misc/join
+++ b/tests/misc/join
@@ -196,7 +196,29 @@ my @tv = (
 # With check, both inputs out of order (in fact, in reverse order)
 ['chkodr-5', '--check-order',
  [" b 1\n a 2\n", " b Y\n a Z\n"], "", 1,
- "$prog: file 1 is not in sorted order\n"],
+ "$prog: chkodr-5.1:2: is not sorted:  a 2\n"],
+
+# Similar, but with only file 2 not sorted.
+['chkodr-5b', '--check-order',
+ [" a 2\n b 1\n", " b Y\n a Z\n"], "", 1,
+ "$prog: chkodr-5b.2:2: is not sorted:  a Z\n"],
+
+# Similar, but with the offending line having length 0 (excluding newline).
+['chkodr-5c', '--check-order',
+ [" a 2\n b 1\n", " b Y\n\n"], "", 1,
+ "$prog: chkodr-5c.2:2: is not sorted: \n"],
+
+# Similar, but elicit a warning for each input file (without --check-order).
+['chkodr-5d', '',
+ ["a\nx\n\n", "b\ny\n\n"], "", 1,
+ "$prog: chkodr-5d.1:3: is not sorted: \n" .
+ "$prog: chkodr-5d.2:3: is not sorted: \n"],
+
+# Similar, but make it so each offending line has no newline.
+['chkodr-5e', '',
+ ["a\nx\no", "b\ny\np"], "", 1,
+ "$prog: chkodr-5e.1:3: is not sorted: o\n" .
+ "$prog: chkodr-5e.2:3: is not sorted: p\n"],
 
 # Without order check, both inputs out of order and some lines
 # unpairable.  This is NOT supported by the GNU extension.  All that
@@ -229,7 +251,7 @@ my @tv = (
 # actual data out-of-order. This join should fail.
 ['header-3', '--header --check-order',
  ["ID Name\n2 B\n1 A\n", "ID Color\n2 blue\n"], "ID Name Color\n", 1,
- "$prog: file 1 is not in sorted order\n"],
+ "$prog: header-3.1:3: is not sorted: 1 A\n"],
 
 # '--header' with specific output format '-o'.
 # output header line should respect the requested format