guestfs_ls: Reimplement to avoid protocol limits.

3 files changed, 119 insertions, 64 deletions

diff --git a/daemon/ls.c b/daemon/ls.c
index 0f60a4ea..5adf5ef3 100644
--- a/daemon/ls.c
+++ b/daemon/ls.c
@@ -93,49 +93,6 @@ do_ls0 (const char *path)
   return 0;
 }
 
-char **
-do_ls (const char *path)
-{
-  DECLARE_STRINGSBUF (ret);
-  DIR *dir;
-  struct dirent *d;
-
-  CHROOT_IN;
-  dir = opendir (path);
-  CHROOT_OUT;
-
-  if (!dir) {
-    reply_with_perror ("opendir: %s", path);
-    return NULL;
-  }
-
-  while ((d = readdir (dir)) != NULL) {
-    if (STREQ (d->d_name, ".") || STREQ (d->d_name, ".."))
-      continue;
-
-    if (add_string (&ret, d->d_name) == -1) {
-      closedir (dir);
-      return NULL;
-    }
-  }
-
-  if (ret.size > 0)
-    sort_strings (ret.argv, ret.size);
-
-  if (end_stringsbuf (&ret) == -1) {
-    closedir (dir);
-    return NULL;
-  }
-
-  if (closedir (dir) == -1) {
-    reply_with_perror ("closedir: %s", path);
-    free_stringslen (ret.argv, ret.size);
-    return NULL;
-  }
-
-  return ret.argv;
-}
-
 /* Because we can't chroot and run the ls command (since 'ls' won't
  * necessarily exist in the chroot), this command can be used to escape
  * from the sysroot (eg. 'll /..').  This command is not meant for
diff --git a/generator/generator_actions.ml b/generator/generator_actions.ml
index be2db82a..66eb54f6 100644
--- a/generator/generator_actions.ml
+++ b/generator/generator_actions.ml
@@ -2219,6 +2219,23 @@ symbolic links already (albeit slightly less efficient).
 This call is intended for programs that want to efficiently
 list a directory contents without making many round-trips." };
 
+  { defaults with
+    name = "ls";
+    style = RStringList "listing", [Pathname "directory"], [];
+    tests = [
+      InitScratchFS, Always, TestOutputList (
+        [["mkdir"; "/ls"];
+         ["touch"; "/ls/new"];
+         ["touch"; "/ls/newer"];
+         ["touch"; "/ls/newest"];
+         ["ls"; "/ls"]], ["new"; "newer"; "newest"])
+    ];
+    shortdesc = "list the files in a directory";
+    longdesc = "\
+List the files in C<directory> (relative to the root directory,
+there is no cwd).  The '.' and '..' entries are not returned, but
+hidden files are shown." };
+
 ]
 
 (* daemon_functions are any functions which cause some action
@@ -2310,27 +2327,6 @@ This command is mostly useful for interactive sessions.  It
 is I<not> intended that you try to parse the output string." };
 
   { defaults with
-    name = "ls";
-    style = RStringList "listing", [Pathname "directory"], [];
-    proc_nr = Some 6;
-    tests = [
-      InitScratchFS, Always, TestOutputList (
-        [["mkdir"; "/ls"];
-         ["touch"; "/ls/new"];
-         ["touch"; "/ls/newer"];
-         ["touch"; "/ls/newest"];
-         ["ls"; "/ls"]], ["new"; "newer"; "newest"])
-    ];
-    shortdesc = "list the files in a directory";
-    longdesc = "\
-List the files in C<directory> (relative to the root directory,
-there is no cwd).  The '.' and '..' entries are not returned, but
-hidden files are shown.
-
-This command is mostly useful for interactive sessions.  Programs
-should probably use C<guestfs_readdir> instead." };
-
-  { defaults with
     name = "list_devices";
     style = RStringList "devices", [], [];
     proc_nr = Some 7;
diff --git a/src/file.c b/src/file.c
index 71b208d3..9257b857 100644
--- a/src/file.c
+++ b/src/file.c
@@ -535,3 +535,105 @@ guestfs__readlinklist (guestfs_h *g, const char *dir, char *const *names)
 
   return ret;
 }
+
+char **
+guestfs__ls (guestfs_h *g, const char *directory)
+{
+  int fd = -1;
+  struct stat statbuf;
+  char *tmpfile = NULL, *buf = NULL;
+  char **ret = NULL;
+  size_t i, count, size;
+
+  tmpfile = safe_asprintf (g, "%s/ls%d", g->tmpdir, ++g->unique);
+
+  if (guestfs_ls0 (g, directory, tmpfile) == -1)
+    goto err;
+
+  fd = open (tmpfile, O_RDONLY|O_CLOEXEC);
+  if (fd == -1) {
+    perrorf (g, "open: %s", tmpfile);
+    goto err;
+  }
+
+  unlink (tmpfile);
+  free (tmpfile);
+  tmpfile = NULL;
+
+  /* Read the whole file into memory. */
+  if (fstat (fd, &statbuf) == -1) {
+    perrorf (g, "stat: %s", tmpfile);
+    goto err;
+  }
+
+  /* Don't use safe_malloc, because we want to return an errno to the caller. */
+  size = statbuf.st_size;
+  buf = malloc (size);
+  if (!buf) {
+    perrorf (g, "malloc: %zu bytes", size);
+    goto err;
+  }
+
+  if (full_read (fd, buf, size) != size) {
+    perrorf (g, "full-read: %s: %zu bytes", tmpfile, size);
+    goto err;
+  }
+
+  if (close (fd) == -1) {
+    perrorf (g, "close: %s", tmpfile);
+    goto err;
+  }
+  fd = -1;
+
+  /* 'buf' contains the list of strings, separated (and terminated) by
+   * '\0' characters.  Convert this to a list of lines.  Note we
+   * handle the case where buf is completely empty (size == 0).
+   */
+  count = 0;
+  for (i = 0; i < size; ++i)
+    if (buf[i] == '\0')
+      count++;
+
+  ret = malloc ((count + 1) * sizeof (char *));
+  if (!ret) {
+    perrorf (g, "malloc");
+    goto err;
+  }
+
+  count = 0;
+  ret[count++] = buf;
+  for (i = 0; i < size; ++i) {
+    if (buf[i] == '\0')
+      ret[count++] = &buf[i+1];
+  }
+  ret[--count] = NULL;
+
+  /* Finally we have to duplicate and sort the strings, since that's
+   * what the caller is expecting.
+   */
+  for (i = 0; ret[i] != NULL; ++i) {
+    ret[i] = strdup (ret[i]);
+    if (ret[i] == NULL) {
+      perrorf (g, "strdup");
+      while (i > 0)
+        free (ret[--i]);
+      goto err;
+    }
+  }
+  free (buf);
+
+  sort_strings (ret, count);
+
+  return ret;                   /* caller frees */
+
+ err:
+  free (buf);
+  free (ret);
+  if (fd >= 0)
+    close (fd);
+  if (tmpfile) {
+    unlink (tmpfile);
+    free (tmpfile);
+  }
+  return NULL;
+}