Best way to find/replace Unicode characters

I want to replace a number of Unicode characters (about 250) with different ones.

For example, these
ἐἑἒἓἔἕὲέἢἣἤἥἦ

with these
εεεεεεεεηηηηη

What is the best way to do it with a macro? I know this way, but it seems too repetitive...

Please Login or Register  to view this content.

Hi,

Just create two arrays for each set of characters and then create a for next loop and use the loop counter as the index to the two arrays,

You could use a macro like:

Please Login or Register  to view this content.

So far, the code will replace all accented lower-case α, ε & η characters with their unaccented versions. I'll leave you to add the rest and/or exclude any you don't want converted.

Each new line beginning with StrFR has two parts - the wildcard Find string and a Replacement character. There is one new line beginning with StrFR per replacement character. Each Find string is bounded by the '[' and ']' characters, whilst the '-' characters denote a contiguous character range. The '|' characters differentiate the Find string and a Replacement characters and each part of the StrFR 'array'. The '&H????' strings are the Unicode values of the characters, which you can find via the Insert|Symbol dialogue.

Although it looks complicated, the code as posted does all of the α, ε & η replacements with just three Find/Replace executions - one for each letter.

Paul, wicked

Just trying to understand how it works.

I see in the line 3 StrFR = , then in lines 5 and 6 StrFR = StrFR . Why is that?

Then I see at the beginning of a line

"[" & ChrW(&H3AC) & ChrW(&H1F00)

But further down the next character being separated via & "-" & ChrW(&H1F07)

You said these characters denote a contiguous character range, but how do I figure out which of the characters I want to replace are contiguous or not.

By the way, I use this to find the Unicode values of the characters (UTF-16 code units):
http://rishida.net/tools/conversion/

The first 'StrFR = ' line starts building the StrFR string. The next two, which use 'StrFR = StrFR &' add to that string. As I've coded it, each StrFR caters for one replacement letter; it doesn't have to be that way, but I thought it would make the code a tad easier to follow.

If we look at the second StrFR line:
StrFR = StrFR & "[" & ChrW(&H3AD) & ChrW(&H1F10) & "-" & ChrW(&H1F15) & ChrW(&H1F72) & ChrW(&H1F73) & "]|" & ChrW(&H3B5) & "|"
the first thing you'll notice is that it's adding to the exsiting string.
From the '[' to the ']' is all one wildcard Find expression. These square brackets tell Word to find any instance of whatever's between them. Usually, you might see something like '[A-Z0-9]', which would tell Word to find any capital letter or number. In this case, because we're looking for Unicode characters that you can't just type into the VBE, we have to tell Word what character values to look for. The 'ChrW(&H3AD) & ChrW(&H1F10)' is just two characters that are not part of a range of character values, just as 'Z0' in '[A-Z0-9]' are two unrelated characters. Where 'ChrW(&H1F10) & "-" & ChrW(&H1F15)', though, that's telling Word to look for anything in that range of character values, just as 'A-Z' is a range of character values.

As for finding whether a series of characters occupies a contiguous value range, that's easily seen in the Insert|Symbol dialogue. If you see a bunch of accented α characters there, one after the other, it's a contiguous range. The dialogue also shows what values you need to use for them.