X-Git-Url: https://git.xandkar.net/?a=blobdiff_plain;f=tt.rkt;h=7f663c2174e6b7b5675fd8eec1a8fa4c95241d24;hb=96ba4c69b4ed97b2999129380e8a4a2f51d69745;hp=b691fd0669da6ed0504868939722cea1eb88e064;hpb=dad4504d0e1c23bcd1de18eaa5d6615a228f9658;p=tt.git diff --git a/tt.rkt b/tt.rkt index b691fd0..7f663c2 100644 --- a/tt.rkt +++ b/tt.rkt @@ -2,13 +2,12 @@ ; TODO write ; TODO caching (use cache by default, unless explicitly asked for update) ; - [x] value --> cache -; - [ ] value <-- cache +; - [x] value <-- cache ; requires: commands ; TODO timeline limits ; TODO feed set operations (perhaps better done externally?) ; TODO timeline as a result of a query (feed set op + filter expressions) ; TODO named timelines -; TODO CLI params ; TODO config files ; TODO parse "following" from feed ; - following = @@ -34,6 +33,9 @@ ; - d | download ; - u | upload ; - calls user-configured command to upload user's own feed file to their server +; +; Looks like a better CLI parser than "racket/cmdline": +; https://docs.racket-lang.org/natural-cli/ #lang racket @@ -44,7 +46,7 @@ (require rfc3339-old) (module+ test - (require rackunit)) + (require rackunit)) (struct msg (ts_epoch ts_rfc3339 nick uri text)) (struct feed (nick uri)) @@ -59,25 +61,25 @@ (define (work) (thread-send parent (cons 'next self)) (match (thread-receive) - ['done (thread-send parent (cons 'exit id))] - [(cons 'unit x) (begin - (define y (f x)) - (when y (thread-send parent (cons 'result y))) - (work))])) + ['done (thread-send parent (cons 'exit id))] + [(cons 'unit x) (begin + (define y (f x)) + (when y (thread-send parent (cons 'result y))) + (work))])) (work))) (define (dispatch ws xs ys) (if (empty? ws) - ys - (match (thread-receive) - [(cons 'exit w) (dispatch (remove w ws =) xs ys)] - [(cons 'result y) (dispatch ws xs (cons y ys))] - [(cons 'next thd) (match xs - ['() (begin - (thread-send thd 'done) - (dispatch ws xs ys))] - [(cons x xs) (begin - (thread-send thd (cons 'unit x)) - (dispatch ws xs ys))])]))) + ys + (match (thread-receive) + [(cons 'exit w) (dispatch (remove w ws =) xs ys)] + [(cons 'result y) (dispatch ws xs (cons y ys))] + [(cons 'next thd) (match xs + ['() (begin + (thread-send thd 'done) + (dispatch ws xs ys))] + [(cons x xs) (begin + (thread-send thd (cons 'unit x)) + (dispatch ws xs ys))])]))) (define workers (range num_workers)) (define threads (map (λ (id) (thread (make-worker id f))) workers)) (define results (dispatch workers xs '())) @@ -85,20 +87,18 @@ results) (module+ test - (define n-workers 10) - (define given (list - (λ (x) (if (even? x) x #f)) - (range 11))) - (check-equal? - (sort (apply concurrent-filter-map (cons n-workers given)) <) - (sort (apply filter-map given ) <))) + (let* ([f (λ (x) (if (even? x) x #f))] + [xs (range 11)] + [actual (sort (concurrent-filter-map 10 f xs) <)] + [expected (sort ( filter-map f xs) <)]) + (check-equal? actual expected "concurrent-filter-map"))) (define (msg-print out-format odd msg) (printf (match out-format - ['single-line "~a \033[1;37m<~a ~a>\033[0m \033[0;~am~a\033[0m~n"] - ['multi-line "~a~n\033[1;37m<~a ~a>\033[0m~n\033[0;~am~a\033[0m~n~n"] - [_ (raise (format "Invalid output format: ~a" out-format))]) + ['single-line "~a \033[1;37m<~a ~a>\033[0m \033[0;~am~a\033[0m~n"] + ['multi-line "~a~n\033[1;37m<~a ~a>\033[0m~n\033[0;~am~a\033[0m~n~n"] + [_ (raise (format "Invalid output format: ~a" out-format))]) (date->string (seconds->date [msg-ts_epoch msg]) #t) (msg-nick msg) (msg-uri msg) @@ -111,30 +111,63 @@ (define (str->msg nick uri str) (if (not (regexp-match? re-msg-begin str)) - (begin - (log-debug "Non-msg line from nick:~a, line:~a" nick str) - #f) - (let ([toks (string-split str (regexp "\t+"))]) - (if (not (= 2 (length toks))) - (begin - (log-warning "Invalid msg line from nick:~a, msg:~a" nick str) - #f) - (let* - ([ts_rfc3339 (first toks)] - [text (second toks)] - [t (string->rfc3339-record ts_rfc3339)] - ; TODO handle tz offset - [ts_epoch (find-seconds [rfc3339-record:second t] - [rfc3339-record:minute t] - [rfc3339-record:hour t] - [rfc3339-record:mday t] - [rfc3339-record:month t] - [rfc3339-record:year t])]) - (msg ts_epoch ts_rfc3339 nick uri text)))))) + (begin + (log-debug "Non-msg line from nick:~a, line:~a" nick str) + #f) + (let ([toks (string-split str (regexp "\t+"))]) + (if (not (= 2 (length toks))) + (begin + (log-warning "Invalid msg line from nick:~a, msg:~a" nick str) + #f) + (let* + ([ts_rfc3339 (first toks)] + [text (second toks)] + [t (string->rfc3339-record ts_rfc3339)] + ; TODO handle tz offset + [ts_epoch (find-seconds [rfc3339-record:second t] + [rfc3339-record:minute t] + [rfc3339-record:hour t] + [rfc3339-record:mday t] + [rfc3339-record:month t] + [rfc3339-record:year t])]) + (msg ts_epoch ts_rfc3339 nick uri text)))))) + +(module+ test + (let* ([ts "2020-11-18T22:22:09-0500"] + [tab " "] + [text "Lorem ipsum"] + [nick "foo"] + [uri "bar"] + [actual (str->msg nick uri (string-append ts tab text))] + [expected (msg 1605756129 ts nick uri text)]) + ; FIXME re-enable after handling tz offset + ;(check-equal? + ; (msg-ts_epoch actual) + ; (msg-ts_epoch expected) + ; "str->msg ts_epoch") + (check-equal? + (msg-ts_rfc3339 actual) + (msg-ts_rfc3339 expected) + "str->msg ts_rfc3339") + (check-equal? + (msg-nick actual) + (msg-nick expected) + "str->msg nick") + (check-equal? + (msg-uri actual) + (msg-uri expected) + "str->msg uri") + (check-equal? + (msg-text actual) + (msg-text expected) + "str->msg text"))) (define (str->lines str) (string-split str (regexp "[\r\n]+"))) +(module+ test + (check-equal? (str->lines "abc\ndef\n\nghi") '("abc" "def" "ghi"))) + (define (str->msgs nick uri str) (filter-map (λ (line) (str->msg nick uri line)) (str->lines str))) @@ -144,33 +177,34 @@ (close-input-port in) digest) -(define (uri-fetch uri) - (log-info "GET ~a" uri) - (define resp (http-get uri)) - (define status (http-response-code resp)) - (define body (http-response-body resp)) - (log-debug "finished GET ~a status:~a body length:~a" - uri status (string-length body)) - ; TODO Handle redirects - (if (= status 200) - (let* - ([url-digest - (hash-sha1 uri)] - [cache-file-path - (expand-user-path (string-append "~/.tt/cache/" url-digest))]) - (display-to-file - body cache-file-path - #:exists 'replace) - body) - ; TODO A more-informative exception - (raise status))) +(define (uri-fetch use-cache uri) + (define cache-file-path + (expand-user-path (string-append "~/.tt/cache/" (hash-sha1 uri)))) + (if (and use-cache (file-exists? cache-file-path)) + (begin + (log-info "uri-fetch cached ~a" uri) + (file->string cache-file-path)) + (begin + (log-info "uri-fetch new ~a" uri) + (let* ([resp (http-get uri)] + [status (http-response-code resp)] + [body (http-response-body resp)]) + (log-debug "finished GET ~a status:~a body length:~a" + uri status (string-length body)) + ; TODO Handle redirects + (if (= status 200) + (begin + (display-to-file body cache-file-path #:exists 'replace) + body) + ; TODO A more-informative exception + (raise status)))))) (define (timeline-print out-format timeline) (for ([msg timeline] [i (in-naturals)]) - (msg-print out-format (odd? i) msg))) + (msg-print out-format (odd? i) msg))) -(define (feed->msgs feed) +(define (feed->msgs use-cache feed) (log-info "downloading feed nick:~a uri:~a" (feed-nick feed) (feed-uri feed)) @@ -190,11 +224,11 @@ status) #f)]) (define uri (feed-uri feed)) - (str->msgs [feed-nick feed] uri [uri-fetch uri]))) + (str->msgs [feed-nick feed] uri [uri-fetch use-cache uri]))) ; TODO timeline contract : time-sorted list of messages -(define (timeline num_workers feeds) - (sort (append* (concurrent-filter-map num_workers feed->msgs feeds)) +(define (timeline use-cache num_workers feeds) + (sort (append* (concurrent-filter-map num_workers (curry feed->msgs use-cache) feeds)) (λ (a b) [< (msg-ts_epoch a) (msg-ts_epoch b)]))) (define (str->feed str) @@ -208,50 +242,59 @@ (define (file->feeds filename) (str->feeds (file->string filename))) -(define (we-are-twtxt) - (define uri - "https://raw.githubusercontent.com/mdom/we-are-twtxt/master/we-are-twtxt.txt") - (str->feeds (uri-fetch uri))) - -(define user-agent +(define (user-agent prog-name prog-version) (let* - ([prog-name "tt"] - [prog-version "0.3.4"] - [prog-uri "https://github.com/xandkar/tt"] + ([prog-uri "https://github.com/xandkar/tt"] [user-feed-file (expand-user-path "~/twtxt-me.txt")] [user (if (file-exists? user-feed-file) - (let ([user (first (file->feeds user-feed-file))]) - (format "+~a; @~a" (feed-uri user) (feed-nick user))) - (format "+~a" prog-uri))] + (let ([user (first (file->feeds user-feed-file))]) + (format "+~a; @~a" (feed-uri user) (feed-nick user))) + (format "+~a" prog-uri))] ) (format "~a/~a (~a)" prog-name prog-version user))) (module+ main - (define (setup-logging) - (define logger (make-logger #f #f 'debug #f)) - (define log-chan (make-log-receiver logger 'debug)) - (void (thread (λ () - [date-display-format 'iso-8601] - [let loop () - (define data (sync log-chan)) - (define level (vector-ref data 0)) - (define msg (vector-ref data 1)) - (define ts (date->string (current-date) #t)) - (eprintf "~a [~a] ~a~n" ts level msg) - (loop)]))) - (current-logger logger)) + (require setup/getinfo) + + (let* ([level 'info] + [logger (make-logger #f #f level #f)] + [log-receiver (make-log-receiver logger level)]) + (void (thread (λ () + [date-display-format 'iso-8601] + [let loop () + (define data (sync log-receiver)) + (define level (vector-ref data 0)) + (define msg (vector-ref data 1)) + (define ts (date->string (current-date) #t)) + (eprintf "~a [~a] ~a~n" ts level msg) + (loop)]))) + (current-logger logger)) + (current-http-response-auto #f) + (let* ([prog-name "tt"] + [prog-version ((get-info (list prog-name)) 'version)] + [user-agent (user-agent prog-name prog-version)]) + (current-http-user-agent user-agent)) + (date-display-format 'rfc2822) + (let* ([use-cache + #f] + [out-format + 'multi-line] + [num_workers + 15]) ; 15 was fastest out of the tried 1, 5, 10, 15 and 20. + (command-line + #:once-each + [("-c" "--cached") + "Read cached data instead of downloading." + (set! use-cache #t)] + + [("-j" "--jobs") + njobs "Number of concurrent jobs." + (set! num_workers (string->number njobs))] - (setup-logging) - (current-http-response-auto #f) - (current-http-user-agent user-agent) - (date-display-format 'rfc2822) + #:args (filename) - (define args (current-command-line-arguments)) - (define feeds - (if (vector-empty? args) - (we-are-twtxt) - (file->feeds (vector-ref args 0)))) - (define out-format 'multi-line) - (define num_workers 15) ; 15 was fastest out of the tried 1, 5, 10, 15 and 20. - (timeline-print out-format (timeline num_workers feeds))) + (timeline-print out-format + (timeline use-cache + num_workers + (file->feeds filename))))))