Looking for generic text transformation utility library

Looking for generic text transformation utility library

[Phil Eaton]

I can't seem to find a way to do simple programmatic transformations
without writing a good portion of code (that I imagine must have been
written before). I want to be able to make a text selection and run
`(map-selection #'prettify)` and have my `prettify` function called with
the selection and the result of `prettify` written back in place of the
selection.

Same thing goes for a hypothetical `map-buffer`, `map-line`, etc.

Is this a (builtin) thing? Or does every package re-implement this
functionality manually? Is there a library that provides these basic
transformation utilities?

Thank you!

-- 
Phil Eaton

Re: Looking for generic text transformation utility library

[Emanuel Berg]

Phil Eaton wrote:

> I can't seem to find a way to do simple
> programmatic transformations without writing
> a good portion of code

"generic text transformation utility", you mean
like the Unix tool sed(1), the "stream editor
for filtering and transforming text", or tr(1),
to "translate or delete characters"?

Can you provide me/us with an example of
a text, as well as the end result you want it
to be translated into?

-- 
underground experts united
http://user.it.uu.se/~embe8573

rx use case for ETL (was: Looking for generic text transformation utility library)

[Van L]

>> I can't seem to find a way to do simple
>> programmatic transformations without writing
>> a good portion of code
> 
> "generic text transformation utility", you mean
> like the Unix tool sed(1), the "stream editor
> for filtering and transforming text", or tr(1),
> to "translate or delete characters"?
> 
> Can you provide me/us with an example of
> a text, as well as the end result you want it
> to be translated into?

A killer-mode in Emacs for the 
Extract Transform Load (or ETL) industry would be 
for a way to use rx (not regexp) to step by step 
transform a source buffer to desired target form.

Step 1. begin with two buffers, the source and target
which are identical

Step 2. an rx machine assisted composition buffer does
what sed(1), tr(1), regexp do by showing in two mini buffers
candidate patterns for source to target change which are 
easy human readable and writable; imagine this kind of 
workload isn’t done more than four times each year so 
the readability has to be there after long stretches
and for first timers

Step 3. stop when there are no 
undesired outliers in the target form