Attachment 627449 Details for Bug 227498 – Patch for wget-1.14 solving this issue

[patch] Patch for wget-1.14 solving this issue

0001-Fixed-error-when-downloading-SYMLINK-from-FTP.patch (text/plain), 13.89 KB, created by Tomáš Hozza on 2012-10-15 13:56:39 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Tomáš Hozza

Created: 2012-10-15 13:56:39 UTC

Size: 13.89 KB

patch

obsolete

>From 151a60419bea65d6a2d0b5022f04384ce81c63b9 Mon Sep 17 00:00:00 2001
>From: Tomas Hozza <thozza@redhat.com>
>Date: Mon, 15 Oct 2012 14:49:30 +0200
>Subject: [PATCH] Fixed error when downloading SYMLINK from FTP.
>
>Fixed problem with downloading single file from a FTP
>server, when timestamping (-N) is used and the target
>file is a SYMLINK. Now wget downloads the file to which
>the SYMLINK points and sets timestamp correctly (to timestamp
>of the target file to which the SYMLINK points).
>---
> src/ftp.c | 302 +++++++++++++++++++++++++++++++++++++++++++-------------------
> src/url.c |   4 +-
> src/url.h |   2 +
> 3 files changed, 211 insertions(+), 97 deletions(-)
>
>diff --git a/src/ftp.c b/src/ftp.c
>index 669e663..4ac8a69 100644
>--- a/src/ftp.c
>+++ b/src/ftp.c
>@@ -1725,6 +1725,113 @@ static uerr_t ftp_retrieve_dirs (struct url *, struct fileinfo *, ccon *);
> static uerr_t ftp_retrieve_glob (struct url *, ccon *, int);
> static struct fileinfo *delelement (struct fileinfo *, struct fileinfo **);
> static void freefileinfo (struct fileinfo *f);
>+static bool has_insecure_name_p (const char *);
>+
>+/* Do a safe check on given fileinfo list and remove entries that
>+   1. Do not conform with global rules given in opt.accepts or opt.rejects.
>+   2. Files with possible harmful names.
>+   3. Entries that do not match given globbing pattern if glob_action is
>+      GLOB_GLOBALL or GLOB_GETONE.
>+   
>+   If argument "file" is null, entries are not matched against any globbing
>+   pattern. */
>+static uerr_t
>+fileinfo_cleanup (struct fileinfo **start, const char *file, int glob_action)
>+{
>+  struct fileinfo *f;
>+
>+  /* First: weed out that do not conform the global rules given in
>+     opt.accepts and opt.rejects.  */
>+  if (opt.accepts || opt.rejects)
>+    {
>+      f = *start;
>+      while (f)
>+        {
>+          if (f->type != FT_DIRECTORY && !acceptable (f->name))
>+            {
>+              logprintf (LOG_VERBOSE, _("Rejecting %s.\n"),
>+                         quote (f->name));
>+              f = delelement (f, start);
>+            }
>+          else
>+            f = f->next;
>+        }
>+    }
>+  /* Remove all files with possible harmful names */
>+  f = *start;
>+  while (f)
>+    {
>+      if (has_insecure_name_p (f->name))
>+        {
>+          logprintf (LOG_VERBOSE, _("Rejecting %s.\n"),
>+                     quote (f->name));
>+          f = delelement (f, start);
>+        }
>+      else
>+        f = f->next;
>+    }
>+  /* Now weed out the files that do not match our globbing pattern.
>+     If we are dealing with a globbing pattern, that is.  */
>+  if (file)
>+    {
>+      if (glob_action == GLOB_GLOBALL)
>+        {
>+          int (*matcher) (const char *, const char *, int)
>+            = opt.ignore_case ? fnmatch_nocase : fnmatch;
>+          int matchres = 0;
>+
>+          f = *start;
>+          while (f)
>+            {
>+              matchres = matcher (file, f->name, 0);
>+              if (matchres == -1)
>+                {
>+                  logprintf (LOG_NOTQUIET, _("Error matching %s against %s: %s\n"),
>+                             file, quotearg_style (escape_quoting_style, f->name),
>+                             strerror (errno));
>+                  break;
>+                }
>+              if (matchres == FNM_NOMATCH)
>+                f = delelement (f, start); /* delete the element from the list */
>+              else
>+                f = f->next;        /* leave the element in the list */
>+            }
>+          if (matchres == -1)
>+            {
>+              freefileinfo (*start);
>+              return RETRBADPATTERN;
>+            }
>+        }
>+      else if (glob_action == GLOB_GETONE)
>+        {
>+#ifdef __VMS
>+          /* 2009-09-09 SMS.
>+           * Odd-ball compiler ("HP C V7.3-009 on OpenVMS Alpha V7.3-2")
>+           * bug causes spurious %CC-E-BADCONDIT complaint with this
>+           * "?:" statement.  (Different linkage attributes for strcmp()
>+           * and strcasecmp().)  Converting to "if" changes the
>+           * complaint to %CC-W-PTRMISMATCH on "cmp = strcmp;".  Adding
>+           * the senseless type cast clears the complaint, and looks
>+           * harmless.
>+           */
>+          int (*cmp) (const char *, const char *)
>+            = opt.ignore_case ? strcasecmp : (int (*)())strcmp;
>+#else /* def __VMS */
>+          int (*cmp) (const char *, const char *)
>+            = opt.ignore_case ? strcasecmp : strcmp;
>+#endif /* def __VMS [else] */
>+          f = *start;
>+          while (f)
>+            {
>+              if (0 != cmp(file, f->name))
>+                f = delelement (f, start);
>+              else
>+                f = f->next;
>+            }
>+        }
>+    }
>+  return RETROK;
>+}
> 
> /* Retrieve a list of files given in struct fileinfo linked list.  If
>    a file is a symbolic link, do not retrieve it, but rather try to
>@@ -1734,7 +1841,7 @@ static void freefileinfo (struct fileinfo *f);
>    If opt.recursive is set, after all files have been retrieved,
>    ftp_retrieve_dirs will be called to retrieve the directories.  */
> static uerr_t
>-ftp_retrieve_list (struct url *u, struct fileinfo *f, ccon *con)
>+ftp_retrieve_list (struct url *u, struct fileinfo *f, ccon *con, int action)
> {
>   static int depth = 0;
>   uerr_t err;
>@@ -1841,11 +1948,102 @@ The sizes do not match (local %s) -- retrieving.\n\n"),
>       switch (f->type)
>         {
>         case FT_SYMLINK:
>+          /* In case we are downloading just one file using timestamping
>+             and this file is a SYMLINK, then we need to traverse this 
>+             SYMLINK to the destination file, get its fileinfo and match
>+             the timestamp with existing file and finaly download it if
>+             everything is OK! */ 
>+          if (action == GLOB_GETONE)
>+            {
>+              struct fileinfo *finfo_list;
>+              struct url *tmp_url;
>+              char *dir, *file, *tmp;
>+
>+              tmp_url = xnew0 (struct url);
>+
>+              split_path (f->linkto, &dir, &file);
>+              if (!strcmp (dir, ""))
>+                {
>+                  xfree (dir);
>+                  dir = xstrdup (u->dir);
>+                }
>+              tmp = concat_strings (u->host, ":", dir, "/", file, NULL);
>+              
>+              tmp_url->url = rewrite_shorthand_url (tmp);
>+              tmp_url->scheme = u->scheme;
>+              tmp_url->host = xstrdup (u->host);
>+              tmp_url->port = u->port;
>+              tmp_url->path = concat_strings (dir, "/", file, NULL);
>+              if (u->params)
>+                tmp_url->params = xstrdup (u->params);
>+              if (u->query)
>+                tmp_url->query = xstrdup (u->query);
>+              if (u->fragment)
>+                tmp_url->fragment = xstrdup (u->fragment);
>+              tmp_url->dir = dir;
>+              tmp_url->file = file;
>+              if (u->user)
>+                tmp_url->user = xstrdup (u->user);
>+              if (u->passwd)
>+                tmp_url->passwd = xstrdup (u->passwd);
>+
>+              dir = NULL;
>+              file = NULL;
>+              xfree(tmp);
>+
>+              err = ftp_get_listing (tmp_url, con, &finfo_list);
>+              if (err != RETROK)
>+                {
>+                  url_free (tmp_url);
>+                  return err;
>+                }
>+
>+              err = fileinfo_cleanup (&finfo_list, tmp_url->file, action);
>+              if (err != RETROK)
>+                {
>+                  url_free (tmp_url);
>+                  return err;
>+                }
>+
>+              url_free (tmp_url);
>+
>+              /* Copy the real file fileinfo */
>+              f->type = finfo_list->type;
>+              f->size = finfo_list->size;
>+              f->tstamp = finfo_list->tstamp;
>+              f->ptype = finfo_list->ptype;
>+              f->perms = finfo_list->perms;
>+              xfree (f->linkto);
>+              if (finfo_list->linkto)
>+                f->linkto = xstrdup (finfo_list->linkto);
>+              else
>+                f->linkto = NULL;
>+              
>+              freefileinfo (finfo_list);
>+              
>+              /* Reset FTP connection status and command
>+                 to the state like it was in the beginning */
>+              con->st &= ~ON_YOUR_OWN;
>+              if (!(con->st & DONE_CWD))
>+                con->cmd |= DO_CWD;
>+              else
>+                con->cmd &= ~DO_CWD;
>+              con->cmd |= (DO_RETR | LEAVE_PENDING);
>+
>+              if (con->csock < 0)
>+                con->cmd |= DO_LOGIN;
>+              else
>+                con->cmd &= ~DO_LOGIN;
>+
>+              /* Now we have correct fileinfo and can check timestamp
>+                 of the real file. */
>+              continue;
>+            }
>           /* If opt.retr_symlinks is defined, we treat symlinks as
>              if they were normal files.  There is currently no way
>              to distinguish whether they might be directories, and
>              follow them.  */
>-          if (!opt.retr_symlinks)
>+          else if (!opt.retr_symlinks)
>             {
> #ifdef HAVE_SYMLINK
>               if (!f->linkto)
>@@ -2060,8 +2258,9 @@ has_insecure_name_p (const char *s)
> 
> /* A near-top-level function to retrieve the files in a directory.
>    The function calls ftp_get_listing, to get a linked list of files.
>-   Then it weeds out the file names that do not match the pattern.
>-   ftp_retrieve_list is called with this updated list as an argument.
>+   Then it weeds out the file names that do not match the pattern
>+   using fileinfo_cleanup function. ftp_retrieve_list is called with
>+   this updated list as an argument.
> 
>    If the argument ACTION is GLOB_GETONE, just download the file (but
>    first get the listing, so that the time-stamp is heeded); if it's
>@@ -2078,100 +2277,15 @@ ftp_retrieve_glob (struct url *u, ccon *con, int action)
>   res = ftp_get_listing (u, con, &start);
>   if (res != RETROK)
>     return res;
>-  /* First: weed out that do not conform the global rules given in
>-     opt.accepts and opt.rejects.  */
>-  if (opt.accepts || opt.rejects)
>-    {
>-      f = start;
>-      while (f)
>-        {
>-          if (f->type != FT_DIRECTORY && !acceptable (f->name))
>-            {
>-              logprintf (LOG_VERBOSE, _("Rejecting %s.\n"),
>-                         quote (f->name));
>-              f = delelement (f, &start);
>-            }
>-          else
>-            f = f->next;
>-        }
>-    }
>-  /* Remove all files with possible harmful names */
>-  f = start;
>-  while (f)
>-    {
>-      if (has_insecure_name_p (f->name))
>-        {
>-          logprintf (LOG_VERBOSE, _("Rejecting %s.\n"),
>-                     quote (f->name));
>-          f = delelement (f, &start);
>-        }
>-      else
>-        f = f->next;
>-    }
>-  /* Now weed out the files that do not match our globbing pattern.
>-     If we are dealing with a globbing pattern, that is.  */
>-  if (*u->file)
>-    {
>-      if (action == GLOB_GLOBALL)
>-        {
>-          int (*matcher) (const char *, const char *, int)
>-            = opt.ignore_case ? fnmatch_nocase : fnmatch;
>-          int matchres = 0;
>+  
>+  res = fileinfo_cleanup (&start, u->file, action);
>+  if (res != RETROK)
>+    return res;
> 
>-          f = start;
>-          while (f)
>-            {
>-              matchres = matcher (u->file, f->name, 0);
>-              if (matchres == -1)
>-                {
>-                  logprintf (LOG_NOTQUIET, _("Error matching %s against %s: %s\n"),
>-                             u->file, quotearg_style (escape_quoting_style, f->name),
>-                             strerror (errno));
>-                  break;
>-                }
>-              if (matchres == FNM_NOMATCH)
>-                f = delelement (f, &start); /* delete the element from the list */
>-              else
>-                f = f->next;        /* leave the element in the list */
>-            }
>-          if (matchres == -1)
>-            {
>-              freefileinfo (start);
>-              return RETRBADPATTERN;
>-            }
>-        }
>-      else if (action == GLOB_GETONE)
>-        {
>-#ifdef __VMS
>-          /* 2009-09-09 SMS.
>-           * Odd-ball compiler ("HP C V7.3-009 on OpenVMS Alpha V7.3-2")
>-           * bug causes spurious %CC-E-BADCONDIT complaint with this
>-           * "?:" statement.  (Different linkage attributes for strcmp()
>-           * and strcasecmp().)  Converting to "if" changes the
>-           * complaint to %CC-W-PTRMISMATCH on "cmp = strcmp;".  Adding
>-           * the senseless type cast clears the complaint, and looks
>-           * harmless.
>-           */
>-          int (*cmp) (const char *, const char *)
>-            = opt.ignore_case ? strcasecmp : (int (*)())strcmp;
>-#else /* def __VMS */
>-          int (*cmp) (const char *, const char *)
>-            = opt.ignore_case ? strcasecmp : strcmp;
>-#endif /* def __VMS [else] */
>-          f = start;
>-          while (f)
>-            {
>-              if (0 != cmp(u->file, f->name))
>-                f = delelement (f, &start);
>-              else
>-                f = f->next;
>-            }
>-        }
>-    }
>   if (start)
>     {
>       /* Just get everything.  */
>-      res = ftp_retrieve_list (u, start, con);
>+      res = ftp_retrieve_list (u, start, con, action);
>     }
>   else
>     {
>diff --git a/src/url.c b/src/url.c
>index e44dfcd..4ae7d33 100644
>--- a/src/url.c
>+++ b/src/url.c
>@@ -585,8 +585,6 @@ rewrite_shorthand_url (const char *url)
>     }
>   return ret;
> }
>-
>-static void split_path (const char *, char **, char **);
> 
> /* Like strpbrk, with the exception that it returns the pointer to the
>    terminating zero (end-of-string aka "eos") if no matching character
>@@ -980,7 +978,7 @@ url_error (const char *url, int error_code)
> 
>    DIR and FILE are freshly allocated.  */
> 
>-static void
>+void
> split_path (const char *path, char **dir, char **file)
> {
>   char *last_slash = strrchr (path, '/');
>diff --git a/src/url.h b/src/url.h
>index edb6b06..da13e2c 100644
>--- a/src/url.h
>+++ b/src/url.h
>@@ -111,4 +111,6 @@ bool schemes_are_similar_p (enum url_scheme a, enum url_scheme b);
> 
> bool are_urls_equal (const char *u1, const char *u2);
> 
>+void split_path (const char *path, char **dir, char **file);
>+
> #endif /* URL_H */
>-- 
>1.7.11.7
>

Flags:

thozza: review?

Actions: View | Diff

Attachments on bug 227498: 627449