Blender truncates a long multibyte character object's name to an invalid utf-8 string #50305

New Issue

meeya · 2016-12-25T00:41:43+01:00

meeya commented

2016-12-25 00:41:43 +01:00

System Information
Linux x64 (Ubuntu 16.04)
Windows 7 x86

Blender Version
blender-2.78-b47c912-linux-glibc219-x86_64 (Python 3.5.2)
blender 2.78a linux x64 official
blender 2.78a Windows x86 official

Short description of error
error: in the python console, creating an object with a long multibyte character name
Blender truncates a long name to 63 bytes
But does not handle properly utf-8 character boundary

Exact steps for others to reproduce the error

open python console

63 byte string (No Error)

>>> mesh = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€')
>>> mesh.name

'€€€€€€€€€€€€€€€€€€€€€'

>>> mesh

bpy.data.meshes['€€€€€€€€€€€€€€€€€€€€€']

64 byte string

>>> mesh = bpy.data.meshes.new('a€€€€€€€€€€€€€€€€€€€€€')
>>> mesh.name

UnicodeDecodeError: 'utf-8' codec can't decode byte 0x82 in position 61: invalid start byte

>>> mesh

# crash

63 byte string with an suffix ".001"

>>> mesh = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€')
>>> mesh.name

'€€€€€€€€€€€€€€€€€€€€€'

>>> mesh = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€') # Blender appends an suffix
>>> mesh.name

UnicodeDecodeError: 'utf-8' codec can't decode bytes in position 57-58: invalid continuation byte

>>> mesh

# crash

**System Information** Linux x64 (Ubuntu 16.04) Windows 7 x86 **Blender Version** blender-2.78-b47c912-linux-glibc219-x86_64 (Python 3.5.2) blender 2.78a linux x64 official blender 2.78a Windows x86 official **Short description of error** error: in the python console, creating an object with a long multibyte character name Blender truncates a long name to 63 bytes But does not handle properly utf-8 character boundary **Exact steps for others to reproduce the error** open python console 63 byte string (No Error) ``` >>> mesh = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€') >>> mesh.name '€€€€€€€€€€€€€€€€€€€€€' >>> mesh bpy.data.meshes['€€€€€€€€€€€€€€€€€€€€€'] ``` 64 byte string ``` >>> mesh = bpy.data.meshes.new('a€€€€€€€€€€€€€€€€€€€€€') >>> mesh.name UnicodeDecodeError: 'utf-8' codec can't decode byte 0x82 in position 61: invalid start byte >>> mesh # crash ``` 63 byte string with an suffix ".001" ``` >>> mesh = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€') >>> mesh.name '€€€€€€€€€€€€€€€€€€€€€' >>> mesh = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€') # Blender appends an suffix >>> mesh.name UnicodeDecodeError: 'utf-8' codec can't decode bytes in position 57-58: invalid continuation byte >>> mesh # crash ```

meeya commented

2016-12-25 00:41:43 +01:00

Changed status to: 'Open'

meeya commented

2016-12-25 00:41:43 +01:00

Added subscriber: @meeya

blender-admin commented

2016-12-27 16:39:19 +01:00

This issue was referenced by 752a783fa4

This issue was referenced by 752a783fa431565d7d964871aceb90bfaa7c9234

Bastien Montagne self-assigned this 2016-12-27 16:39:48 +01:00

Bastien Montagne commented

2016-12-27 16:41:29 +01:00

Changed status from 'Open' to: 'Resolved'

Bastien Montagne closed this issue

2016-12-27 16:41:29 +01:00

meeya commented

2016-12-30 17:36:41 +01:00

Changed status from 'Resolved' to: 'Open'

meeya reopened this issue

2016-12-30 17:36:41 +01:00

meeya commented

2016-12-30 17:36:41 +01:00

Thank you for the quick response.

I still got the error in the case of the same ID.

Blender Version : blender-2.78-bf7d7bc-linux-glibc219-x86_64

>>> mesh1 = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€')
>>> mesh2 = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€')
>>> mesh2.name

UnicodeDecodeError: 'utf-8' codec can't decode bytes in position 57-58: invalid continuation byte

>>> mesh2

# crash

I'm not sure, but this bug might be caused by check_for_dupid function.
After BLI_utf8_invalid_strip in new_id function, it seems that check_for_dupid function truncates an ID ignoring utf-8 characters.

new_id (source/blender/blenkernel/intern/library.c)
check_for_dupid (source/blender/blenkernel/intern/library.c)

Thank you for the quick response. I still got the error in the case of the same ID. **Blender Version** : blender-2.78-bf7d7bc-linux-glibc219-x86_64 ``` >>> mesh1 = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€') >>> mesh2 = bpy.data.meshes.new('€€€€€€€€€€€€€€€€€€€€€€') >>> mesh2.name UnicodeDecodeError: 'utf-8' codec can't decode bytes in position 57-58: invalid continuation byte >>> mesh2 # crash ``` I'm not sure, but this bug might be caused by `check_for_dupid` function. After `BLI_utf8_invalid_strip` in `new_id` function, it seems that `check_for_dupid` function truncates an ID ignoring utf-8 characters. * new_id (source/blender/blenkernel/intern/library.c) * check_for_dupid (source/blender/blenkernel/intern/library.c)

Bastien Montagne commented

2016-12-30 18:20:07 +01:00

Eeek! Good catch (though this is actually another bug…).

blender-admin commented

2017-01-01 02:27:03 +01:00

This issue was referenced by 7924c84c44

This issue was referenced by 7924c84c44c70fefa64a872d6d767913fba636d2

Bastien Montagne commented

2017-01-01 02:30:00 +01:00

Changed status from 'Open' to: 'Resolved'

Bastien Montagne closed this issue

2017-01-01 02:30:00 +01:00

Sign in to join this conversation.

No Label

Download

What's New

Blender Studio

Manual

Developers Blog

Documentation

Benchmark

Blender Conference

Development Fund

One-time Donations

Blender truncates a long multibyte character object's name to an invalid utf-8 string #50305