From Newsgroup: comp.lang.scheme


i'd not use Gauche for this, but maybe someone can change my mind.


_______________________
From JoyceUlysses.txt -- words occurring exactly once


Given a text file of a novel (JoyceUlysses.txt) ...

could someone give me a pretty fast (and simple) program that'd give me
a list of all words occurring exactly once?

-- Also, a list of words occurring once, twice or 3 times



re: hyphenated words (you can treat it anyway you like)

ideally, i'd treat [editor-in-chief]
[go-ahead] [pen-knife]
[know-how] [far-fetched] ...
as one unit.
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: comp.lang.scheme

On 5/30/2024 2:09 PM, HenHanna wrote:

i'd not use Gauche for this, but maybe someone can change my mind.

_______________________
From JoyceUlysses.txt -- words occurring exactly once

Given a text file of a novel (JoyceUlysses.txt) ...

could someone give me a pretty fast (and simple) program that'd give me
a list of all words occurring exactly once?

-a-a-a-a-a-a-a-a-a-a-a-a-a -- Also, a list of words occurring once, twice or 3 times

re: hyphenated words-a-a-a-a-a-a-a (you can treat it anyway you like)

-a-a-a-a-a-a ideally, i'd treat-a [editor-in-chief]
-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a [go-ahead]-a [pen-knife]
-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a [know-how]-a [far-fetched] ...
-a-a-a-a-a-a as one unit.

Make a list (or array) of the individual words (as strings or symbols in
a special package) of the original document then sort the list using the Lisp-supplied sort function. You than write a loop using your favorite
tools and look for interior sequences of the required length. This gives
you a program that is asymptotically efficient as the theoretical
run-time will look something like (* c N (log N)), where N is the length
of the list produced by the first step and c is some constant.

Note, any solution resembling this one is not really what you want. For example it would think "Snark" and "Snarks" are different words. Some differences such as capitalization can be suppressed by choosing a sort predicate that is case insensitive. You can, of course, write your own
sort predicate. The thing to note is that the predicate (the <= operator
used by sort) will not access the words or maintain state between
invocations; otherwise, the complexity can become arbitrarily large.
--
Jeff Barnett

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: comp.lang.scheme

Given a text file of a novel (JoyceUlysses.txt) ...
could someone give me a pretty fast (and simple) program that'd give me
a list of all words occurring exactly once?

tr ' .;:,?!' '\n' | sort | uniq -u

?


- Stefan
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: comp.lang.scheme

On 2024-05-30, Stefan Monnier <monnier@iro.umontreal.ca> wrote:

Given a text file of a novel (JoyceUlysses.txt) ...
could someone give me a pretty fast (and simple) program that'd give me
a list of all words occurring exactly once?

tr ' .;:,?!' '\n' | sort | uniq -u

Yep, that's pretty much how Doug McIlroy famously shut down Knuth.
--
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal
Mastodon: @Kazinator@mstdn.ca
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: comp.lang.scheme

could someone give me a pretty fast (and simple) program that'd give
me a list of all words occurring exactly once?

To first approximation, this works for me (bash command):

tr -c "[a-zA-Z-]" "\n" < ulysses.txt |sort|uniq -c|sort -n
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: comp.lang.scheme

On 5/30/2024, HenHanna wrote:

i'd not use Gauche for this, but maybe someone can change my mind.

_______________________
From JoyceUlysses.txt -- words occurring exactly once

Given a text file of a novel (JoyceUlysses.txt) ...

could someone give me a pretty fast (and simple) program that'd give me a list of all words occurring exactly once?

-- Also, a list of words occurring once, twice or 3 times

re: hyphenated words (you can treat it anyway you like)

ideally, i'd treat [editor-in-chief]
[go-ahead] [pen-knife]
[know-how] [far-fetched] ...
as one unit.

Gauche Scheme

(use file.util) ;; file->string
(use srfi-13) ;; character sets
(use srfi-14) ;; string-tokenize

(define h (make-hash-table 'string=?))

(dolist
(s
(string-tokenize (file->string "Alice.txt")
(char-set-adjoin char-set:letter #\-)))
(hash-table-update! h
(regexp-replace* (string-upcase s) #/^-+/ "" #/-+$/ "")
(pa$ + 1) 0))

(filter (lambda(kv) (< (cdr kv) 3))
(hash-table->alist h))

===>

(("LASTED" . 2) ("WAY--NEVER" . 1) ("VISIT" . 1) ("CHANCED" . 1)
("WILDLY" . 2) ("BEHEAD" . 1) ("PROMISE" . 1) ("MEANWHILE" . 1)
("ENGAGED" . 1) ("KNIFE" . 2) ("ROARED" . 1) ("RETIRE" . 1)
("BLACKING" . 1) ("HATED" . 1) ("BRIGHT-EYED" . 1)
("SHEEP-BELLS" . 1) ("PROTECTION" . 1) ("CRIES" . 1) ("ADA" . 1)
("ENJOY" . 1) ("WRITHING" . 1) ("RAW" . 1) ("APPEALED" . 1)
("RELIEVED" . 1) ("CHILDHOOD" . 1) ("WEPT" . 1) ("RACE-COURSE" . 1)
("THEIRS" . 1) ("MAD--AT" . 1) ("SPOKEN" . 1) ("PENCILS" . 1)
("CLEAR" . 2) ("TREADING" . 2) ("RETURNED" . 2) ("CHERRY-TART" . 1)
("UNEASY" . 1) ("LOW-SPIRITED" . 1) ("BONE" . 1) ("PROMISED" . 1)
("HAPPENING" . 1) ("OYSTER" . 1) ("PATIENTLY" . 2) ("NEEDS" . 1)
("LESSON-BOOK" . 1) ("PITIED" . 1) ("UNCOMFORTABLY" . 1)
("ANTIPATHIES" . 1) ("PICTURED" . 1) ("DESPERATE" . 1)
("ENGRAVED" . 1)
...
)
--- Synchronet 3.22a-Linux NewsLink 1.2