I guess the above is false, so I'm trying to find a counter example. The only thing I found is the whole space should not be compact. If not, the statement is true directly from the Tietze extension.
Any help will be appreciated.
I guess the above is false, so I'm trying to find a counter example. The only thing I found is the whole space should not be compact. If not, the statement is true directly from the Tietze extension.
Any help will be appreciated.
On
No, this does not hold, as $\mathbb{R}$ in the lower limit topology $\mathcal{T}_l$, generated by all sets of the form $(a,b]$ (or homeomorphically by sets $[a,b)$ if one prefers) (a. k.a. the Sorgenfrey line) is zero-dimensional and thus totally disconnected. This implies that any continuous function $f : X \to \mathbb{R}$ is constant whenever $X$ is connected. So for connected normal spaces we don't have Urysohn functions to separate the closed sets.
I guess you mean the topology generated by the half-open intervals $[a,b)$. It's called the lower limit topology: https://en.wikipedia.org/wiki/Lower_limit_topology
You are asking if the Tietze extension theorem still holds on a space $X$ (I assume normal) if the topology on $R$ is not the standard topology (as it is in the usual Tietze extension theorem) but is instead the half-open interval topology.
I think the semi-open Tietze extension property will imply a semi-open Urysohn lemma property in the usual way: http://at.yorku.ca/cgi-bin/bbqa?forum=ask_a_topologist_2001&task=show_msg&msg=0263.0001
But R with the lower limit topology is completely disconnected. So any continuous function from a connected space $X$ to $R$ will be constant. So, if $X$ is connected, a semi-open Urysohn lemma cannot hold, hence a semi-open Tietze theorem cannot hold.