thr3ads.net - Libguestfs - Re: [Libguestfs] [PATCH 3/3] lib: Add support for creating nodes (keys) and values with UTF-16LE-encoded names [Nov 2013]

If this information is useful, please help other people find it:
Share via:

Hilko Bengen

2013-Nov-24 22:25 UTC

[Libguestfs] [PATCH 1/3] lib: Further generalize iconv wrapper function.

---
 lib/hivex-internal.h |  8 +++++---
 lib/utf16.c          | 11 +++++++----
 2 files changed, 12 insertions(+), 7 deletions(-)

diff --git a/lib/hivex-internal.h b/lib/hivex-internal.h
index 4135f58..64fd49a 100644
--- a/lib/hivex-internal.h
+++ b/lib/hivex-internal.h
@@ -268,11 +268,13 @@ extern size_t * _hivex_return_offset_list (offset_list
*list);
 extern void _hivex_print_offset_list (offset_list *list, FILE *fp);
 
 /* utf16.c */
-extern char* _hivex_to_utf8 (/* const */ char *input, size_t len, char*
input_encoding);
+extern char* _hivex_recode (char *input_encoding,
+                            const char *input, size_t input_len,
+                            char *output_encoding, size_t *output_len);
 #define _hivex_windows_utf16_to_utf8(_input, _len) \
-  _hivex_to_utf8 (_input, _len, "UTF-16LE")
+  _hivex_recode ("UTF-16LE", _input, _len, "UTF-8", NULL)
 #define _hivex_windows_latin1_to_utf8(_input, _len) \
-  _hivex_to_utf8 (_input, _len, "LATIN1")
+  _hivex_recode ("LATIN1", _input, _len, "UTF-8", NULL)
 extern size_t _hivex_utf16_string_len_in_bytes_max (const char *str, size_t
len);
 
 /* util.c */
diff --git a/lib/utf16.c b/lib/utf16.c
index eca2343..6b8bf9a 100644
--- a/lib/utf16.c
+++ b/lib/utf16.c
@@ -29,18 +29,19 @@
 #include "hivex-internal.h"
 
 char *
-_hivex_to_utf8 (/* const */ char *input, size_t len, char* input_encoding)
+_hivex_recode (char *input_encoding, const char *input, size_t input_len,
+               char *output_encoding, size_t *output_len)
 {
-  iconv_t ic = iconv_open ("UTF-8", input_encoding);
+  iconv_t ic = iconv_open (output_encoding, input_encoding);
   if (ic == (iconv_t) -1)
     return NULL;
 
   /* iconv(3) has an insane interface ... */
 
-  size_t outalloc = len;
+  size_t outalloc = input_len;
 
  again:;
-  size_t inlen = len;
+  size_t inlen = input_len;
   size_t outlen = outalloc;
   char *out = malloc (outlen + 1);
   if (out == NULL) {
@@ -79,6 +80,8 @@ _hivex_to_utf8 (/* const */ char *input, size_t len, char*
input_encoding)
 
   *outp = '\0';
   iconv_close (ic);
+  if (output_len != NULL)
+    *output_len = outp - out;
 
   return out;
 }
-- 
1.8.4.4

Hilko Bengen

2013-Nov-24 22:25 UTC

head link

[Libguestfs] [PATCH 2/3] lib: Add function that encodes UTF-8 to "packed" strings or UTF-16LE as Windows does

---
 lib/hivex-internal.h |  1 +
 lib/utf16.c          | 19 +++++++++++++++++++
 2 files changed, 20 insertions(+)

diff --git a/lib/hivex-internal.h b/lib/hivex-internal.h
index 64fd49a..7a548c0 100644
--- a/lib/hivex-internal.h
+++ b/lib/hivex-internal.h
@@ -275,6 +275,7 @@ extern char* _hivex_recode (char *input_encoding,
   _hivex_recode ("UTF-16LE", _input, _len, "UTF-8", NULL)
 #define _hivex_windows_latin1_to_utf8(_input, _len) \
   _hivex_recode ("LATIN1", _input, _len, "UTF-8", NULL)
+extern char* _hivex_encode_string(const char *str, size_t *size, int *utf16);
 extern size_t _hivex_utf16_string_len_in_bytes_max (const char *str, size_t
len);
 
 /* util.c */
diff --git a/lib/utf16.c b/lib/utf16.c
index 6b8bf9a..437613b 100644
--- a/lib/utf16.c
+++ b/lib/utf16.c
@@ -24,6 +24,7 @@
 #include <stdlib.h>
 #include <stdint.h>
 #include <iconv.h>
+#include <string.h>
 
 #include "hivex.h"
 #include "hivex-internal.h"
@@ -86,6 +87,24 @@ _hivex_recode (char *input_encoding, const char *input,
size_t input_len,
   return out;
 }
 
+/* Encode a given UTF-8 string to Latin1 (preferred) or UTF-16 for
+ * storing in the hive file, as needed.
+ */
+char*
+_hivex_encode_string(const char *str, size_t *size, int *utf16)
+{
+  char* outstr;
+  *utf16 = 0;
+  outstr = _hivex_recode ("UTF-8", str, strlen(str),
+                          "LATIN1", size);
+  if (outstr != NULL)
+    return outstr;
+  *utf16 = 1;
+  outstr = _hivex_recode ("UTF-8", str, strlen(str),
+                          "UTF-16LE", size);
+  return outstr;
+}
+
 /* Get the length of a UTF-16 format string.  Handle the string as
  * pairs of bytes, looking for the first \0\0 pair.  Only read up to
  * 'len' maximum bytes.
-- 
1.8.4.4

Hilko Bengen

2013-Nov-24 22:25 UTC

head link

[Libguestfs] [PATCH 3/3] lib: Add support for creating nodes (keys) and values with UTF-16LE-encoded names

---
 lib/write.c | 49 ++++++++++++++++++++++++++++++++++---------------
 1 file changed, 34 insertions(+), 15 deletions(-)

diff --git a/lib/write.c b/lib/write.c
index dbb8292..72b1f8a 100644
--- a/lib/write.c
+++ b/lib/write.c
@@ -608,9 +608,17 @@ hivex_node_add_child (hive_h *h, hive_node_h parent, const
char *name)
     return 0;
   }
 
+  size_t recoded_name_len;
+  int use_utf16 = 0;
+  char* recoded_name = _hivex_encode_string (name, &recoded_name_len,
&use_utf16);
+  if (recoded_name == NULL) {
+    SET_ERRNO (EINVAL, "malformed name");
+    return 0;
+  }
+
   /* Create the new nk-record. */
   static const char nk_id[2] = { 'n', 'k' };
-  size_t seg_len = sizeof (struct ntreg_nk_record) + strlen (name);
+  size_t seg_len = sizeof (struct ntreg_nk_record) + recoded_name_len;
   hive_node_h nkoffset = allocate_block (h, seg_len, nk_id);
   if (nkoffset == 0)
     return 0;
@@ -619,14 +627,18 @@ hivex_node_add_child (hive_h *h, hive_node_h parent, const
char *name)
 
   struct ntreg_nk_record *nk      (struct ntreg_nk_record *) ((char *)
h->addr + nkoffset);
-  nk->flags = htole16 (0x0020); /* key is ASCII. */
+  if (use_utf16)
+    nk->flags = htole16 (0x0000);
+  else
+    nk->flags = htole16 (0x0020);
   nk->parent = htole32 (parent - 0x1000);
   nk->subkey_lf = htole32 (0xffffffff);
   nk->subkey_lf_volatile = htole32 (0xffffffff);
   nk->vallist = htole32 (0xffffffff);
   nk->classname = htole32 (0xffffffff);
-  nk->name_len = htole16 (strlen (name));
-  strcpy (nk->name, name);
+  nk->name_len = htole16 (recoded_name_len);
+  memcpy (nk->name, recoded_name, recoded_name_len);
+  free(recoded_name);
 
   /* Inherit parent sk. */
   struct ntreg_nk_record *parent_nk @@ -719,9 +731,9 @@ hivex_node_add_child
(hive_h *h, hive_node_h parent, const char *name)
   parent_nk->nr_subkeys = htole32 (nr_subkeys_in_parent_nk);
 
   /* Update max_subkey_name_len in parent nk. */
-  uint16_t max = le16toh (parent_nk->max_subkey_name_len);
-  if (max < strlen (name) * 2)  /* *2 because "recoded" in
UTF16-LE. */
-    parent_nk->max_subkey_name_len = htole16 (strlen (name) * 2);
+  size_t utf16_len = use_utf16 ? recoded_name_len : recoded_name_len * 2;
+  if (le16toh (parent_nk->max_subkey_name_len) < utf16_len)
+    parent_nk->max_subkey_name_len = htole16 (utf16_len);
 
   return nkoffset;
 }
@@ -942,7 +954,12 @@ hivex_node_set_values (hive_h *h, hive_node_h node,
   for (i = 0; i < nr_values; ++i) {
     /* Allocate vk record to store this (key, value) pair. */
     static const char vk_id[2] = { 'v', 'k' };
-    seg_len = sizeof (struct ntreg_vk_record) + strlen (values[i].key);
+    size_t name_len = strlen (values[i].key);
+    size_t recoded_name_len;
+    int use_utf16;
+    char* recoded_name = _hivex_encode_string (values[i].key,
&recoded_name_len,
+                                               &use_utf16);
+    seg_len = sizeof (struct ntreg_vk_record) + recoded_name_len;
     size_t vk_offs = allocate_block (h, seg_len, vk_id);
     if (vk_offs == 0)
       return -1;
@@ -957,15 +974,17 @@ hivex_node_set_values (hive_h *h, hive_node_h node,
 
     struct ntreg_vk_record *vk        (struct ntreg_vk_record *) ((char *)
h->addr + vk_offs);
-    size_t name_len = strlen (values[i].key);
-    vk->name_len = htole16 (name_len);
-    strcpy (vk->name, values[i].key);
+    vk->name_len = htole16 (recoded_name_len);
+    memcpy (vk->name, recoded_name, recoded_name_len);
     vk->data_type = htole32 (values[i].t);
     uint32_t len = values[i].len;
     if (len <= 4)               /* store it inline => set MSB flag */
       len |= 0x80000000;
     vk->data_len = htole32 (len);
-    vk->flags = name_len == 0 ? 0 : 1;
+    if (recoded_name_len == 0)
+      vk->flags = 0;
+    else
+      vk->flags = htole16 (!use_utf16);
 
     if (values[i].len <= 4)     /* store it inline */
       memcpy (&vk->data_offset, values[i].value, values[i].len);
@@ -985,9 +1004,9 @@ hivex_node_set_values (hive_h *h, hive_node_h node,
       vk->data_offset = htole32 (offs - 0x1000);
     }
 
-    if (name_len * 2 > le32toh (nk->max_vk_name_len))
-      /* * 2 for UTF16-LE "reencoding" */
-      nk->max_vk_name_len = htole32 (name_len * 2);
+    size_t utf16_len = use_utf16 ? recoded_name_len : recoded_name_len * 2;
+    if (utf16_len > le32toh (nk->max_vk_name_len))
+      nk->max_vk_name_len = htole32 (utf16_len);
     if (values[i].len > le32toh (nk->max_vk_data_len))
       nk->max_vk_data_len = htole32 (values[i].len);
   }
-- 
1.8.4.4

Richard W.M. Jones

2013-Nov-25 08:44 UTC

head link

Re: [Libguestfs] [PATCH 1/3] lib: Further generalize iconv wrapper function.

On Sun, Nov 24, 2013 at 11:25:51PM +0100, Hilko Bengen
wrote:> ---
>  lib/hivex-internal.h |  8 +++++---
>  lib/utf16.c          | 11 +++++++----
>  2 files changed, 12 insertions(+), 7 deletions(-)
> 
> diff --git a/lib/hivex-internal.h b/lib/hivex-internal.h
> index 4135f58..64fd49a 100644
> --- a/lib/hivex-internal.h
> +++ b/lib/hivex-internal.h
> @@ -268,11 +268,13 @@ extern size_t * _hivex_return_offset_list
(offset_list *list);
>  extern void _hivex_print_offset_list (offset_list *list, FILE *fp);
>  
>  /* utf16.c */
> -extern char* _hivex_to_utf8 (/* const */ char *input, size_t len, char*
input_encoding);
> +extern char* _hivex_recode (char *input_encoding,
> +                            const char *input, size_t input_len,
> +                            char *output_encoding, size_t *output_len);
>  #define _hivex_windows_utf16_to_utf8(_input, _len) \
> -  _hivex_to_utf8 (_input, _len, "UTF-16LE")
> +  _hivex_recode ("UTF-16LE", _input, _len, "UTF-8",
NULL)
>  #define _hivex_windows_latin1_to_utf8(_input, _len) \
> -  _hivex_to_utf8 (_input, _len, "LATIN1")
> +  _hivex_recode ("LATIN1", _input, _len, "UTF-8",
NULL)
>  extern size_t _hivex_utf16_string_len_in_bytes_max (const char *str,
size_t len);
>  
>  /* util.c */
> diff --git a/lib/utf16.c b/lib/utf16.c
> index eca2343..6b8bf9a 100644
> --- a/lib/utf16.c
> +++ b/lib/utf16.c
> @@ -29,18 +29,19 @@
>  #include "hivex-internal.h"
>  
>  char *
> -_hivex_to_utf8 (/* const */ char *input, size_t len, char* input_encoding)
> +_hivex_recode (char *input_encoding, const char *input, size_t input_len,
> +               char *output_encoding, size_t *output_len)
>  {
> -  iconv_t ic = iconv_open ("UTF-8", input_encoding);
> +  iconv_t ic = iconv_open (output_encoding, input_encoding);
>    if (ic == (iconv_t) -1)
>      return NULL;
>  
>    /* iconv(3) has an insane interface ... */
>  
> -  size_t outalloc = len;
> +  size_t outalloc = input_len;
>  
>   again:;
> -  size_t inlen = len;
> +  size_t inlen = input_len;
>    size_t outlen = outalloc;
>    char *out = malloc (outlen + 1);
>    if (out == NULL) {
> @@ -79,6 +80,8 @@ _hivex_to_utf8 (/* const */ char *input, size_t len,
char* input_encoding)
>  
>    *outp = '\0';
>    iconv_close (ic);
> +  if (output_len != NULL)
> +    *output_len = outp - out;
>  
>    return out;
>  }
> -- 
> 1.8.4.4
ACK.

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Fedora Windows cross-compiler. Compile Windows programs, test, and
build Windows installers. Over 100 libraries supported.
http://fedoraproject.org/wiki/MinGW

Richard W.M. Jones

2013-Nov-25 09:06 UTC

head link

Re: [Libguestfs] [PATCH 3/3] lib: Add support for creating nodes (keys) and values with UTF-16LE-encoded names

On Sun, Nov 24, 2013 at 11:25:53PM +0100, Hilko Bengen
wrote:> ---
>  lib/write.c | 49 ++++++++++++++++++++++++++++++++++---------------
>  1 file changed, 34 insertions(+), 15 deletions(-)
> 
> diff --git a/lib/write.c b/lib/write.c
> index dbb8292..72b1f8a 100644
> --- a/lib/write.c
> +++ b/lib/write.c
> @@ -608,9 +608,17 @@ hivex_node_add_child (hive_h *h, hive_node_h parent,
const char *name)
>      return 0;
>    }
>  
> +  size_t recoded_name_len;
> +  int use_utf16 = 0;
> +  char* recoded_name = _hivex_encode_string (name, &recoded_name_len,
&use_utf16);
> +  if (recoded_name == NULL) {
> +    SET_ERRNO (EINVAL, "malformed name");
> +    return 0;
> +  }
> +
>    /* Create the new nk-record. */
>    static const char nk_id[2] = { 'n', 'k' };
> -  size_t seg_len = sizeof (struct ntreg_nk_record) + strlen (name);
> +  size_t seg_len = sizeof (struct ntreg_nk_record) + recoded_name_len;
>    hive_node_h nkoffset = allocate_block (h, seg_len, nk_id);
>    if (nkoffset == 0)
>      return 0;
> @@ -619,14 +627,18 @@ hivex_node_add_child (hive_h *h, hive_node_h parent,
const char *name)
>  
>    struct ntreg_nk_record *nk >      (struct ntreg_nk_record *) ((char
*) h->addr + nkoffset);
> -  nk->flags = htole16 (0x0020); /* key is ASCII. */
> +  if (use_utf16)
> +    nk->flags = htole16 (0x0000);
> +  else
> +    nk->flags = htole16 (0x0020);
>    nk->parent = htole32 (parent - 0x1000);
>    nk->subkey_lf = htole32 (0xffffffff);
>    nk->subkey_lf_volatile = htole32 (0xffffffff);
>    nk->vallist = htole32 (0xffffffff);
>    nk->classname = htole32 (0xffffffff);
> -  nk->name_len = htole16 (strlen (name));
> -  strcpy (nk->name, name);
> +  nk->name_len = htole16 (recoded_name_len);
> +  memcpy (nk->name, recoded_name, recoded_name_len);
> +  free(recoded_name);
Please put spaces after function names!  It improves readability:

http://www1.psych.purdue.edu/~zpizlo/rsteinma/Bob-FOR%20CV/Epelboim%20et%20al%201997%20Fillers%20in%20Reading.pdf
>    /* Inherit parent sk. */
>    struct ntreg_nk_record *parent_nk > @@ -719,9 +731,9 @@
hivex_node_add_child (hive_h *h, hive_node_h parent, const char *name)
>    parent_nk->nr_subkeys = htole32 (nr_subkeys_in_parent_nk);
>  
>    /* Update max_subkey_name_len in parent nk. */
> -  uint16_t max = le16toh (parent_nk->max_subkey_name_len);
> -  if (max < strlen (name) * 2)  /* *2 because "recoded" in
UTF16-LE. */
> -    parent_nk->max_subkey_name_len = htole16 (strlen (name) * 2);
> +  size_t utf16_len = use_utf16 ? recoded_name_len : recoded_name_len * 2;
* 2 is probably wrong here for non-BMP characters, but the original
code makes the same mistake ...  Could we get the true length from the
hivex_encode_string function?
> +  if (le16toh (parent_nk->max_subkey_name_len) < utf16_len)
> +    parent_nk->max_subkey_name_len = htole16 (utf16_len);
>    return nkoffset;
>  }
> @@ -942,7 +954,12 @@ hivex_node_set_values (hive_h *h, hive_node_h node,
>    for (i = 0; i < nr_values; ++i) {
>      /* Allocate vk record to store this (key, value) pair. */
>      static const char vk_id[2] = { 'v', 'k' };
> -    seg_len = sizeof (struct ntreg_vk_record) + strlen (values[i].key);
> +    size_t name_len = strlen (values[i].key);
> +    size_t recoded_name_len;
> +    int use_utf16;
> +    char* recoded_name = _hivex_encode_string (values[i].key,
&recoded_name_len,
> +                                               &use_utf16);
> +    seg_len = sizeof (struct ntreg_vk_record) + recoded_name_len;
>      size_t vk_offs = allocate_block (h, seg_len, vk_id);
>      if (vk_offs == 0)
>        return -1;
> @@ -957,15 +974,17 @@ hivex_node_set_values (hive_h *h, hive_node_h node,
>  
>      struct ntreg_vk_record *vk >        (struct ntreg_vk_record *)
((char *) h->addr + vk_offs);
> -    size_t name_len = strlen (values[i].key);
> -    vk->name_len = htole16 (name_len);
> -    strcpy (vk->name, values[i].key);
> +    vk->name_len = htole16 (recoded_name_len);
> +    memcpy (vk->name, recoded_name, recoded_name_len);
>      vk->data_type = htole32 (values[i].t);
>      uint32_t len = values[i].len;
>      if (len <= 4)               /* store it inline => set MSB flag
*/
>        len |= 0x80000000;
>      vk->data_len = htole32 (len);
> -    vk->flags = name_len == 0 ? 0 : 1;
> +    if (recoded_name_len == 0)
> +      vk->flags = 0;
> +    else
> +      vk->flags = htole16 (!use_utf16);
>      if (values[i].len <= 4)     /* store it inline */
>        memcpy (&vk->data_offset, values[i].value, values[i].len);
> @@ -985,9 +1004,9 @@ hivex_node_set_values (hive_h *h, hive_node_h node,
>        vk->data_offset = htole32 (offs - 0x1000);
>      }
>  
> -    if (name_len * 2 > le32toh (nk->max_vk_name_len))
> -      /* * 2 for UTF16-LE "reencoding" */
> -      nk->max_vk_name_len = htole32 (name_len * 2);
> +    size_t utf16_len = use_utf16 ? recoded_name_len : recoded_name_len *
2;
* 2 - see above.
> +    if (utf16_len > le32toh (nk->max_vk_name_len))
> +      nk->max_vk_name_len = htole32 (utf16_len);
>      if (values[i].len > le32toh (nk->max_vk_data_len))
>        nk->max_vk_data_len = htole32 (values[i].len);
>    }
> -- 
> 1.8.4.4
2/3 & 3/3 are generally good, so ACK.  Any comment on *2 issue above?

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming blog: http://rwmj.wordpress.com
Fedora now supports 80 OCaml packages (the OPEN alternative to F#)

Apparently Analagous Threads

Search for more reasonably related threads

Libguestfs - Nov 2013 - Re: [PATCH 3/3] lib: Add support for creating nodes (keys) and values with UTF-16LE-encoded names

[Libguestfs] [PATCH 1/3] lib: Further generalize iconv wrapper function.

[Libguestfs] [PATCH 2/3] lib: Add function that encodes UTF-8 to "packed" strings or UTF-16LE as Windows does

[Libguestfs] [PATCH 3/3] lib: Add support for creating nodes (keys) and values with UTF-16LE-encoded names

Re: [Libguestfs] [PATCH 1/3] lib: Further generalize iconv wrapper function.

Re: [Libguestfs] [PATCH 3/3] lib: Add support for creating nodes (keys) and values with UTF-16LE-encoded names

Apparently Analagous Threads