Thanks touliloup, it extract the English part in friendly_name as entity_id indeed. If there is no English part just like all Chinese characters, it shows Invalid entity id.
So I wonder if it will be fix in the future version or is there some alternative way to use Chinese as friendly_name.
There were some scenarios that friendly name were not set by the configuration.yaml , but by the ‘homeassistant.components’ during the hass initialization.
e.g. homeassistant.components.tado will do so ,it fetches the home name from tado.com (maker’s site)
It is UGLY to extend this INTERNAL obligation of ‘NO chinese /asia name in configuration’ to innocent third party.
Same problem here with 0.60.
I’m newbie to python but for unicode system, there should not be CJK issues any more.
As to the current implementation, how about using python-pinyin to extract pinyin chars from Chinese string as entity id ?
Personally thinking, of course it is a bug. Some platform offers the possibility to setup friendly_name inside component’s column, others are not. It all depends on component’s creators. Like broadlink, it actually open this possibility. If it doesn’t or decide to remove this approach, they should correct relevant docs to avoid other language speaker misunderstanding about this issue and let users only go to customize part.
For now, try to use customize for Chinese friendly_name and input English name in name part of component.
Furthermore, actually it raises a issue that if we can figure out a way to transform none-English name into and entity_ID.
Chinese translation:
我认为这是一个 bug,理由是博联组件本身的描述不清。正常情况下昵称应该在 customize 模块设置,但是博联的 Doc 又给出了组件内设置的路径,导致大家误解。应该是组件本身没有对字符编码解析导致的,错误将昵称作为名称(只能英文)。目前建议还是在 customize 里设中文昵称,在博联组件内的设置使用英文。