[Bug-wget] bug #28541: Faulty time information in FTP directory listing

[John Trengrove]

This is a patch to change the behaviour for FTP directory listing.
Currently the hours are printed only if the hour is non-zero and does
not account for if we received the hours & minutes or not. A
simplistic patch would be to just create an else if statement in
ftp_index to check if the minutes are non-zero also. This fails though
for the case 00:00. I am uncertain how much this matters.

I created a more complicated patch (below) that alters the struct
fileinfo in ftp.h to hold whether hours:minutes were stored or not. It
is assumes only UNIX FTP servers fail to provide the time for old
entries.

This would be my first contribution to wget. Criticism/feedback encouraged.

ChangeLog

2010-07-25  John Trengrove  <address@hidden>

  * ftp.h: Modified struct to hold parsetype.
  Added enum for parsetype.
  * ftp-ls.c:
  (ftp_parse_unix_ls): Default to TT_DAY. Change to TT_HOUR_MIN if
hours/minutes parsed.
  (ftp_parse_winnt_ls): Default to TT_HOUR_MIN.
  (ftp_parse_vms_ls): Default to TT_HOUR_MIN.
  (ftp_index): Print only if fileinfo struct value ttype set to TT_HOUR_MIN.

Patch

=== modified file 'src/ftp-ls.c'
--- src/ftp-ls.c 2010-05-08 19:56:15 +0000
+++ src/ftp-ls.c 2010-07-25 06:00:04 +0000
@@ -100,7 +100,7 @@
  };
  int next, len, i, error, ignore;
  int year, month, day;         /* for time analysis */
-  int hour, min, sec;
+  int hour, min, sec, ptype;
  struct tm timestruct, *tnow;
  time_t timenow;

@@ -183,6 +183,7 @@
                                   treated equally for now.  */
      year = hour = min = sec = 0; /* Silence the compiler.  */
      month = day = 0;
+      ptype = TT_DAY;
      next = -1;
      /* While there are tokens on the line, parse them.  Next is the
         number of tokens left until the filename.
@@ -262,6 +263,7 @@
                      /* This means these were hours!  */
                      hour = year;
                      year = 0;
+                      ptype = TT_HOUR_MIN;
                      ++tok;
                      /* Get the minutes...  */
                      for (; c_isdigit (*tok); tok++)
@@ -414,6 +416,7 @@
      timestruct.tm_yday  = 0;
      timestruct.tm_isdst = -1;
      l->tstamp = mktime (&timestruct); /* store the time-stamp */
+      l->ptype = ptype;

      xfree (line);
    }
@@ -501,6 +504,7 @@
      timestruct.tm_yday  = 0;
      timestruct.tm_isdst = -1;
      cur.tstamp = mktime (&timestruct); /* store the time-stamp */
+      cur.ptype = TT_HOUR_MIN;

      DEBUGP(("Timestamp: %ld\n", cur.tstamp));

@@ -987,6 +991,7 @@
        }
      cur.tstamp = timenow; /* Store the time-stamp. */
      DEBUGP(("Timestamp: %ld\n", cur.tstamp));
+      cur.ptype = TT_HOUR_MIN;

      /* Add the data for this item to the linked list, */
      if (!dir)
@@ -1134,7 +1139,7 @@

          fprintf (fp, "%d %s %02d ", ptm->tm_year + 1900, months[ptm->tm_mon],
                  ptm->tm_mday);
-          if (ptm->tm_hour)
+          if (f->ptype == TT_HOUR_MIN)
            fprintf (fp, "%02d:%02d  ", ptm->tm_hour, ptm->tm_min);
          else
            fprintf (fp, "       ");

=== modified file 'src/ftp.h'
--- src/ftp.h 2010-05-08 19:56:15 +0000
+++ src/ftp.h 2010-07-25 05:58:22 +0000
@@ -87,6 +87,12 @@
  GLOB_GLOBALL, GLOB_GETALL, GLOB_GETONE
};

+/* Used by to test if time parsed includes hours and minutes. */
+enum parsetype
+{
+  TT_HOUR_MIN, TT_DAY
+};
+
/* Information about one filename in a linked list.  */
struct fileinfo
{
@@ -94,6 +100,7 @@
  char *name; /* file name */
  wgint size; /* file size */
  long tstamp; /* time-stamp */
+  enum parsetype ptype; /* time parsing */
  int perms; /* file permissions */
  char *linkto; /* link to which file points */
  struct fileinfo *prev; /* previous... */