Simplify Zero and One Words construction and Address compression

[ChihChengLiang]

Description

While reviewing #1699, I noticed some room to improve the ergonomics of the words.

Issue Link

Type of change

New feature (non-breaking change which adds functionality)

Contents

• Replaced the address_word_to_expr function with the compress method so that we can compress a word directly without finding that function to use.
• introduced zero_f and one_f to create zero and one for Word<F>.

Rationale

I noticed that it is hard to create a one() and zero() method for both Word<Expression<F>> and Word<F>. The compiler complains about implementing duplicated methods. Since we already use one() and zero() in many places for Word<Expression<F>>, I named zero_f and one_f for the methods for Word<F>.

[KimiWu123]

LGTM! The code looks nicer and cleaner now!

However, one_f and zero_f are not straightforward to me and it's easy to get confused. Give couples of hours to see if I can find a better solution. If no, then I'll approve it. 🙏

[KimiWu123]

LGTM! The code looks nicer and cleaner now!

However, one_f and zero_f are not straightforward to me and it's easy to get confused. Give couples of hours to see if I can find a better solution. If no, then I'll approve it. 🙏

@ChihChengLiang, I didn't find any good solutions. But I was thinking to use zero_expr() for Word<Expression<F>> and zero() for Word<F>. The methond name is more aligned with the generic type (_expr maps <Expression<F>>).

[hero78119]

LGTM!

[ChihChengLiang]

@KimiWu123 I think you're reasoning makes sense to me. I'll merge this then submit a new one for the remaining.
Thank you @KimiWu123 and @hero78119 for the review.