From 7ddeecf637983993eda02740660f2e389040571a Mon Sep 17 00:00:00 2001 From: Sacha Chua Date: Tue, 13 Dec 2022 01:35:21 -0500 Subject: more extraction --- emacsconf-extract.el | 202 +++++++++++++++++++++++++++++++++++++++++++++++---- emacsconf-subed.el | 2 +- emacsconf.el | 31 +++++--- 3 files changed, 211 insertions(+), 24 deletions(-) diff --git a/emacsconf-extract.el b/emacsconf-extract.el index 2943826..be28033 100644 --- a/emacsconf-extract.el +++ b/emacsconf-extract.el @@ -308,20 +308,194 @@ (defun emacsconf-extract-wget-bbb (o) (when (plist-get o :bbb-playback) - (let ((meeting-id (when (string-match "meetingId=\\(.+\\)" - (plist-get o :bbb-playback)) - (match-string 1 (plist-get o :bbb-playback))))) - (concat "mkdir " (plist-get o :slug) "\n" - "cd " (plist-get o :slug) "\n" - (mapconcat - (lambda (file) - (concat - "wget https://bbb.emacsverse.org/presentation/" - meeting-id "/" file "\n")) - '("video/webcams.webm" "metadata.xml" "deskshare/deskshare.webm" "panzooms.xml" "cursor.xml" "deskshare.xml" "captions.json" "presentation_text.json" "slides_new.xml") - "") - "cd ..\n" - )))) + (let ((meeting-id (when (string-match "meetingId=\\(.+\\)" + (plist-get o :bbb-playback)) + (match-string 1 (plist-get o :bbb-playback))))) + (concat "mkdir " (plist-get o :slug) "\n" + "cd " (plist-get o :slug) "\n" + (mapconcat + (lambda (file) + (concat + "wget https://bbb.emacsverse.org/presentation/" + meeting-id "/" file "\n")) + '("video/webcams.webm" "metadata.xml" "deskshare/deskshare.webm" "panzooms.xml" "cursor.xml" "deskshare.xml" "captions.json" "presentation_text.json" "slides_new.xml") + "") + "cd ..\n")))) +(defun emacsconf-extract-bbb-events-xml (o) + "Copy the events.xml from the raw BBB directory copied from bbb@bbb.emacsverse.org." + (if (plist-get o :bbb-playback) + (let ((meeting-id (when (string-match "meetingId=\\(.+\\)" + (plist-get o :bbb-playback)) + (match-string 1 (plist-get o :bbb-playback))))) + (format "scp ~/current/bbb-raw/%s/events.xml orga@media.emacsconf.org:~/backstage/%s--bbb-events.xml\n" + meeting-id + (plist-get o :video-slug))) + "")) + +(defun emacsconf-extract-bbb-voice-events (file) + "Return a list of voice events. +(:name participant :start-clock start-time :start-ms ... :stop-clock stop-time :stop-ms)." + (let ((dom (xml-parse-file file)) + start-recording + stop-recording + start-ms stop-ms + participants results) + (setq start-recording + (date-to-time + (dom-text + (dom-by-tag + (dom-elements dom 'eventname "StartRecordingEvent") + 'date)))) + (setq stop-recording + (date-to-time + (dom-text + (dom-by-tag + (dom-elements dom 'eventname "StopRecordingEvent") + 'date)))) + (setq start-ms (* 1000 (time-to-seconds start-recording)) + stop-ms (* 1000 (time-to-seconds stop-recording))) + ;; get the participant names and put them in an alist + (setq participants + (mapcar (lambda (o) (list + (dom-text (dom-by-tag o 'userId)) + :name (dom-text (dom-by-tag o 'name)))) + (seq-filter + (lambda (node) (string= (dom-attr node 'eventname) + "ParticipantJoinEvent")) + (dom-by-tag dom 'event)))) + ;; get the voice events + (mapc (lambda (o) + (let ((participant (assoc-default + (dom-text (dom-by-tag o 'participant)) + participants)) + (time (date-to-time (dom-text (dom-by-tag o 'date)))) + o-start o-stop) + (if (string= (dom-text (dom-by-tag o 'talking)) + "true") + ;; start talking + (plist-put participant + ;; although maybe timestampUTC will be useful somehow + :start time) + ;; clamp it to start-recording and stop-recording + (when (and (time-less-p (plist-get participant :start) + stop-recording) + (time-less-p start-recording time)) + (setq o-start + (- (max (* 1000 (time-to-seconds (plist-get participant :start))) + start-ms) + start-ms) + o-stop + (- (min (* 1000 (time-to-seconds time)) + stop-ms) + start-ms)) + (setq results + (cons (list + :name + (plist-get participant :name) + :start-ms + o-start + :stop-ms + o-stop + :start-clock + (plist-get participant :start) + :stop-clock + time + :duration-ms + (- o-stop o-start)) + results)))))) + (seq-filter + (lambda (node) (string= (dom-attr node 'eventname) + "ParticipantTalkingEvent")) + (dom-by-tag dom 'event))) + (nreverse results))) +;; (emacsconf-extract-bbb-voice-events "~/proj/emacsconf/cache/emacsconf-2022-sqlite--using-sqlite-as-a-data-source-a-framework-and-an-example--andrew-hyatt--bbb-events.xml") +;; Okay, now that we have voice events, what can we do with them? +;; We can insert notes into the VTT for now to try to guess the speaker, when the speaker changes +;; The audio is not split up by speaker, so the transcript is also not very split up +;; Some speech-to-text systems can do speaker diarization, which also tries to identify speakers +;; huh, is the StartRecordingEvent timestamp reliable? Am I misreading it? + +(defvar emacsconf-extract-irc-speaker-nick nil) +(defun emacsconf-extract-irc-copy-line-to-other-window-as-list-item () + (interactive) + (goto-char (line-beginning-position)) + (when (looking-at "\\[[0-9:]+\\] <\\(.*?\\)> \\([^ ]+?:\\)?\\(.+\\)$") + (let ((line (string-trim (match-string 3)))) + (setq line + (if (string= (or emacsconf-extract-irc-speaker-nick "") + (match-string 1)) + (concat " - A: " line "\n") + (concat "- " line "\n"))) + (other-window 1) + (insert line) + (other-window 1) + (forward-line 1)))) + +(defvar emacsconf-extract-irc-map (make-sparse-keymap)) +(defalias 'emacsconf-extract-irc-other-window #'other-window) +(defalias 'emacsconf-extract-irc-next-line #'next-line) +(defalias 'emacsconf-extract-irc-previous-line #'previous-line) +(defun emacsconf-extract-irc-open-talk-in-other-window (talk) + (interactive (list (emacsconf-complete-talk-info))) + (other-window 1) + (emacsconf-edit-wiki-page talk)) + +(define-key emacsconf-extract-irc-map "c" #'emacsconf-extract-irc-copy-line-to-other-window-as-list-item) +(define-key emacsconf-extract-irc-map "o" #'emacsconf-extract-irc-other-window) +(define-key emacsconf-extract-irc-map "t" #'emacsconf-extract-irc-open-talk-in-other-window) +(define-key emacsconf-extract-irc-map "n" #'emacsconf-extract-irc-next-line) +(define-key emacsconf-extract-irc-map "p" #'emacsconf-extract-irc-previous-line) +(mapc (lambda (sym) + (put sym 'repeat-map 'emacsconf-extract-irc-map)) + '(emacsconf-extract-irc-copy-line-to-other-window-as-list-item + emacsconf-extract-irc-other-window + emacsconf-extract-irc-next-line + emacsconf-extract-irc-previous-line)) + +;; (local-set-key (kbd "C-c C-c") emacsconf-extract-irc-map) + +(defun emacsconf-extract-publish-qa (talk &optional note) + (interactive (list (emacsconf-complete-talk-info))) + (setq talk (emacsconf-resolve-talk talk)) + (let ((large-file-warning-threshold nil)) + ;; Copy the files + (unless (file-exists-p (expand-file-name (concat (plist-get talk :video-slug) "--answers.webm" emacsconf-cache-dir))) + (if (file-exists-p (expand-file-name (concat (plist-get talk :video-slug) "--bbb-deskshare.webm") emacsconf-cache-dir)) + ;; use the screenshare if available + (call-process "ffmpeg" nil (get-buffer-create "*ffmpeg*") t + "-y" + "-i" (expand-file-name (concat (plist-get talk :video-slug) "--bbb-deskshare.webm") emacsconf-cache-dir) + "-i" (expand-file-name (concat (plist-get talk :video-slug) "--bbb-webcams.opus") emacsconf-cache-dir) + "-c" "copy" + (expand-file-name (concat (plist-get talk :video-slug) "--answers.webm") emacsconf-cache-dir)) + (copy-file + (expand-file-name (concat (plist-get talk :video-slug) "--bbb-webcams.webm") emacsconf-cache-dir) + (expand-file-name (concat (plist-get talk :video-slug) "--answers.webm") emacsconf-cache-dir) + t))) + (unless (file-exists-p (expand-file-name (concat (plist-get talk :video-slug) "--answers.opus") emacsconf-cache-dir)) + (copy-file + (expand-file-name (concat (plist-get talk :video-slug) "--bbb-webcams.opus") emacsconf-cache-dir) + (expand-file-name (concat (plist-get talk :video-slug) "--answers.opus") emacsconf-cache-dir) + t)) + (dolist (suffix '("--answers.webm" "--answers.opus")) + (unless (file-exists-p (expand-file-name (concat (plist-get talk :video-slug) suffix) emacsconf-backstage-dir)) + (copy-file + (expand-file-name (concat (plist-get talk :video-slug) suffix) emacsconf-cache-dir) + (expand-file-name (concat (plist-get talk :video-slug) suffix) emacsconf-backstage-dir) + t)) + (unless (file-exists-p (expand-file-name (concat (plist-get talk :video-slug) suffix) emacsconf-public-media-directory)) + (copy-file + (expand-file-name (concat (plist-get talk :video-slug) suffix) emacsconf-backstage-dir) + (expand-file-name (concat (plist-get talk :video-slug) suffix) emacsconf-public-media-directory) + t))) + ;; Update the org entry + (save-window-excursion + (emacsconf-go-to-talk talk) + (org-entry-put (point) "QA_PUBLIC" "t") + (org-entry-put (point) "QA_NOTE" (or note (concat "Q&A posted publicly." (emacsconf-surround " " (plist-get talk :qa-note) "" ""))))))) +;; (kill-new (mapconcat #'emacsconf-extract-bbb-events-xml (emacsconf-get-talk-info) "")) +;; (dolist (slug '("haskell" "hyperorg" "health" "jupyter" "workflows" "wayland" "mail" "meetups" "orgsuperlinks" "rde" "science")) +;; (emacsconf-extract-publish-qa slug)) (provide 'emacsconf-extract) ;;; emacsconf-extract.el ends here diff --git a/emacsconf-subed.el b/emacsconf-subed.el index ec02a9d..250f707 100644 --- a/emacsconf-subed.el +++ b/emacsconf-subed.el @@ -68,7 +68,7 @@ TYPE can be 'end if you want the match end instead of the beginning." (let (pos (found (string-match regexp string after)) result) (while (and found (< found before)) (setq result (if (eq type 'end) (match-end 0) (match-beginning 0))) - (if (string-match regexp string (1+ found)) + (if (and (< found (length string)) (string-match regexp string (1+ found))) (setq found (match-end 0)) (setq found nil))) result)) diff --git a/emacsconf.el b/emacsconf.el index d09167b..897bc4d 100644 --- a/emacsconf.el +++ b/emacsconf.el @@ -241,9 +241,11 @@ (defun emacsconf-edit-wiki-page (search) (interactive (list (emacsconf-complete-talk))) - (setq search (emacsconf-get-slug-from-string search)) - (find-file (expand-file-name (concat search ".md") - (expand-file-name "talks" (expand-file-name emacsconf-year emacsconf-directory))))) + (setq search (if (stringp search) (emacsconf-get-slug-from-string search) + (plist-get search :slug))) + (find-file (expand-file-name + (concat search ".md") + (expand-file-name "talks" (expand-file-name emacsconf-year emacsconf-directory))))) (defun emacsconf-find-caption-directives-from-slug (search) (interactive (list (emacsconf-complete-talk))) @@ -321,7 +323,7 @@ (defun emacsconf-go-to-talk (search) (interactive (list (emacsconf-complete-talk))) - (pop-to-buffer (find-file-noselect emacsconf-org-file)) + (find-file emacsconf-org-file) (widen) (cond ((plist-get search :slug) @@ -382,7 +384,6 @@ (let ((heading (org-heading-components)) (field-props '( ;; Initial creation - (:title "ITEM") (:track "TRACK") (:slug "SLUG") (:speakers "NAME") @@ -444,6 +445,7 @@ o (list :point (point) + :title (org-entry-get (point) "ITEM") :year emacsconf-year :conf-year emacsconf-year :type (if (org-entry-get (point) "SLUG") 'talk 'headline) @@ -475,10 +477,13 @@ (org-entry-get (point) "SCHEDULED")) t))) emacsconf-timezone-offset)))) - (let* ((entry-props (org-entry-properties))) - (mapcar - (lambda (o) (list (car o) (assoc-default (cadr o) entry-props))) - field-props))))) + (mapcar + (lambda (prop) + (list + (or (car (rassoc (list (car prop)) field-props)) + (intern (concat ":" (replace-regexp-in-string "_" "-" (downcase (car prop)))))) + (cdr prop))) + (org-entry-properties))))) (defvar emacsconf-abstract-heading-regexp "abstract" "Regexp matching heading for talk abstract.") @@ -1454,5 +1459,13 @@ tracks with the ID in the cdr of that list." (format-time-string "%-I:%M %P %Z" (plist-get talk :start-time) (plist-get talk :timezone)) " in " (plist-get talk :timezone))))) + +(defun emacsconf-collect-prop (prop list) + (mapcar (lambda (o) (plist-get o prop)) list)) + +(defun emacsconf-talk-file (talk suffix) + (and (file-exists-p (expand-file-name (concat (plist-get o :video-slug) suffix) emacsconf-cache-dir)) + (expand-file-name (concat (plist-get o :video-slug) suffix) emacsconf-cache-dir))) + (provide 'emacsconf) ;;; emacsconf.el ends here -- cgit v1.2.3