code.delx.au - gnu-emacs-elpa/blob - packages/transcribe/transcribe.el

   1 ;;; transcribe.el --- Package for audio transcriptions
   2
   3 ;; Copyright 2014-2016  Free Software Foundation, Inc.
   4
   5 ;; Author: David Gonzalez Gandara <dggandara@member.fsf.org>
   6 ;; Version: 1.5.0
   7
   8 ;; This program is free software: you can redistribute it and/or modify
   9 ;; it under the terms of the GNU General Public License as published by
  10 ;; the Free Software Foundation, either version 3 of the License, or
  11 ;; (at your option) any later version.
  12 ;;
  13 ;; This program is distributed in the hope that it will be useful,
  14 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  15 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 ;; GNU General Public License for more details.
  17 ;;
  18 ;; You should have received a copy of the GNU General Public License
  19 ;; along with this program.  If not, see <http://www.gnu.org/licenses/>.
  20
  21 ;;; Commentary:
  22
  23 ;; REQUIRES:
  24 ;; -----------------------------
  25 ;; This module works without any requires, but in order to use the audio
  26 ;; functions, you need to install the Emacs package "emms", by Joe Drew,
  27 ;; and the external program "mpg321", by Jorgen Schafer and Ulrik Jensen,
  28 ;; both under GPL licenses.
  29 ;;
  30 ;; USAGE:
  31 ;; -------------------------
  32 ;; Transcribe is a tool to make audio transcriptions for discourse analysis
  33 ;; in the classroom.
  34 ;; It allows the transcriber to control the audio easily while typing, as well as
  35 ;; automate the insertion of xml tags, in case the transcription protocol
  36 ;; include them.
  37 ;; The analysis functions will search for a specific structure
  38 ;; of episodes that can be automatically added with the macro NewEpisode.
  39 ;; The function expects the speech acts to be transcribed inside a turn xml
  40 ;; tag with the identifier of the speaker with optional move attribute.
  41 ;; Each speech act is spected inside a <l1> or <l2> tag, depending
  42 ;; on the language used by the person. The attributes expected are the
  43 ;; number of clauses that form the utterance, the number of errors the
  44 ;; transcriber observes, and the function of the speech act. The parser will
  45 ;; work even if some attributes are missing.
  46 ;;
  47 ;;
  48 ;; AUDIO COMMANDS
  49 ;; ------------------------------
  50 ;;     C-x C-p ------> Play audio file. You will be prompted for the name
  51 ;;                     of the file. The recommended format is mp2.
  52 ;;     <f5> ---------> Pause or play audio.
  53 ;;     C-x <right> --> seek audio 10 seconds forward.
  54 ;;     C-x <left> --->seek audio 10 seconds backward.
  55 ;;     <f8> ---------> seek interactively: positive seconds go forward and
  56 ;;                       negative seconds go backward
  57 ;;
  58 ;; XML TAGGING COMMANDS
  59 ;; --------------------------------------------------
  60 ;;     C-x C-n ------> Create new episode structure. This is useful in case your
  61 ;;                 xml file structure requires it.
  62 ;;     <f2> ---------> Interactively insert a function attribute in a speech act
  63 ;;                 (l1 or l2) tag.
  64 ;;     <f3> ---------> Interactively insert a move attribute in a turn (person) tag
  65 ;;     <f4> ---------> Interactively insert an attribute (any kind)
  66 ;;     <f9> ---------> Insert turn (person) tag. Inserts a move attribute.
  67 ;;     <f10> --------> Insert a custom tag. Edit the function to adapt to your needs.
  68 ;;     <f11> --------> Insert speech act tag in L1, with clauses, errors and function
  69 ;;                     attributes.
  70 ;;     <f12> --------> Insert speech act tag in L2, with clauses, errors and function
  71 ;;                     attributes.
  72 ;;
  73 ;; AUTOMATIC PARSING
  74 ;; -----------------------------------------------------
  75 ;;     C-x C-a ------> Analyses the text for measurments of performance.
  76
  77 ;;; Code:
  78
  79 (require 'xml)
  80
  81 ;; (if t (require 'emms-setup))
  82 ;; (require 'emms-player-mpd)
  83 ;; (setq emms-player-mpd-server-name "localhost")
  84 ;; (setq emms-player-mpd-server-port "6600")
  85
  86 (emms-standard)
  87 (emms-default-players)
  88 (if t (require 'emms-player-mpg321-remote))
  89 (defvar emms-player-list)
  90 (push 'emms-player-mpg321-remote emms-player-list)
  91
  92 (if t (require 'emms-mode-line))
  93 (emms-mode-line 1)
  94 (if t (require 'emms-playing-time))
  95 (emms-playing-time 1)
  96
  97 (defvar transcribe-function-list '("initiating" "responding" "control" "expresive" "interpersonal"))
  98 (defvar transcribe-move-list '("initiation" "response" "follow-up"))
  99 (defvar transcribe-attribute-list '("clauses" "errors" "function" "move"))
 100 ;; (append transcribe-attribute-list transcribe-function-list transcribe-move-list)
 101
 102 (defun transcribe-analyze-episode (episode person)
 103   "This calls the external python package analyze_episodes2.py. The new
 104    function transcribe-analyze implements its role now."
 105   (interactive "sepisode: \nsperson:")
 106   (shell-command (concat (expand-file-name  "analyze_episodes2.py")
 107                   " -e " episode " -p " person " -i " buffer-file-name )))
 108
 109 (defun transcribe-raw-to-buffer ()
 110   "EXPERIMENTAL - Convert the xml tagged transcription to raw transcription, with the names
 111    and the persons and the utterances only. The raw transcription will be send to buffer called
 112    'Raw Output'"
 113   (interactive)
 114   (let* ((xml (xml-parse-region (point-min) (point-max)))
 115     (results (car xml))
 116     (episodes (xml-get-children results 'episode)))
 117
 118     (dolist (episode episodes)
 119       (let* ((transcription (xml-get-children episode 'transcription)))
 120
 121         (dolist (turn transcription)
 122           (dolist (intervention (xml-node-children turn))
 123             (if (listp intervention)
 124               (progn
 125                 (with-current-buffer "Raw Output"
 126                   (insert (format "%s\t" (line-number-at-pos)))
 127                   (insert (format "%s:\t" (car intervention)))
 128                   (dolist (utterance (nthcdr 2 intervention))
 129                     (if (listp utterance)
 130                        (progn
 131                          (insert (format "%s "  (nth 2 utterance))))
 132
 133                          (insert (format "%s" utterance))))))
 134
 135                        (with-current-buffer "Raw Output"
 136                          (insert (format "%s" (line-number-at-pos)))
 137                          (insert (format "%s" intervention))))))))))
 138
 139 (defun transcribe-analyze (episodenumber personid)
 140   "Extract from a given episode and person the number of asunits per
 141    second produced, and the number of clauses per asunits, for L2 and L1.
 142    It writes two output files, one for L2 utterances and one for L1
 143    utterances, so that they can be used with external programs. Output will
 144    be inserted in 'Statistics Output' buffer"
 145   (interactive "sepisodenumber: \nspersonid:")
 146   (let* ((interventionsl2 '())
 147      (interventionsl1 '())
 148      (xml (xml-parse-region (point-min) (point-max)))
 149      (results (car xml))
 150      (episodes (xml-get-children results 'episode))
 151      (asunitsl2 0.0000)
 152      (asunitsl1 0.0000)
 153      ;; (shifts 0.0000);; TODO implement
 154      (initiating 0.0000);; TODO implement
 155      (responding 0.0000);; TODO implement
 156      (control 0.0000);; TODO implement
 157      (expressive 0.0000);; TODO implement
 158      (interpersonal 0.0000);; TODO implement
 159      (clausesl1 0.0000)
 160      ;; (errorsl1 0.0000);; TODO implement
 161      (clausesl2 0.0000)
 162      (errorsl2 0.0000)
 163      (duration nil)
 164      (role nil)
 165      (context nil)
 166      (demand nil)
 167      ;; (clausesmessage nil)
 168      (number nil))
 169
 170      (dolist (episode episodes)
 171        (let*((numbernode (xml-get-children episode 'number))
 172          (tasknode (xml-get-children episode 'task)))
 173
 174          (setq number (nth 2 (car numbernode)))
 175          (when (equal episodenumber number)
 176            (let* ((durationnode (xml-get-children episode 'duration))
 177              (transcription (xml-get-children episode 'transcription)))
 178
 179              (setq duration (nth 2 (car durationnode)))
 180
 181              (dolist (task tasknode)
 182               (let* ((rolenode (xml-get-children task 'role))
 183                 (contextnode (xml-get-children task 'context))
 184                 (demandnode (xml-get-children task 'demand)))
 185
 186                 (setq role (nth 2 (car rolenode)))
 187                 (setq context (nth 2 (car contextnode)))
 188                 (setq demand (nth 2 (car demandnode)))
 189                 ;; (with-current-buffer "Statistics Output"
 190                    ;; (insert (format "role: %s; context: %s; demand: %s\n" role context demand)))
 191                 ))
 192
 193              (dolist (turn transcription)
 194                (let* ((interventionnode (xml-get-children turn
 195                  (intern personid))))
 196
 197                  (dolist (intervention interventionnode)
 198                    (let* ((l2node (xml-get-children intervention 'l2))
 199                      (l1node (xml-get-children intervention 'l1)))
 200
 201                      (dolist (l2turn l2node)
 202                        (let* ((l2 (nth 2 l2turn))
 203                           (attrs (nth 1 l2turn))
 204                           (clausesl2nodeinc (cdr (assq 'clauses attrs)))
 205                           (errorsl2inc (cdr (assq 'errors attrs)))
 206                           (function (cdr (assq 'function attrs))))
 207
 208                           (when (string-equal function "initiating")
 209                             (setq initiating (+ initiating 1)))
 210                           (when (string-equal function "responding")
 211                             (setq responding (+ responding 1)))
 212                           (when (string-equal function "control")
 213                             (setq control (+ control 1)))
 214                           (when (string-equal function "expressive")
 215                             (setq expressive (+ expressive 1)))
 216                           (when (string-equal function "interpersonal")
 217                             (setq interpersonal (+ interpersonal 1)))
 218                           (when attrs
 219                             (setq clausesl2 (+ clausesl2 (string-to-number
 220                              clausesl2nodeinc)))
 221                             (setq errorsl2 (+ errorsl2 (string-to-number
 222                              errorsl2inc))))
 223                           (when l2
 224                             ;; (add-to-list 'interventionsl2 l2)
 225                             (cl-pushnew l2 interventionsl2 :test #'equal)
 226                             (setq asunitsl2 (1+ asunitsl2)))))
 227                      (dolist (l1turn l1node)
 228                        (let*((l1 (nth 2 l1turn))
 229                          (clausesl1node (nth 1 l1turn))
 230                          (clausesl1nodeinc (cdr (car clausesl1node))))
 231
 232                          (when (not (equal clausesl1node nil))
 233                            (setq clausesl1 (+ clausesl1 (string-to-number
 234                               clausesl1nodeinc))))
 235                          (when l1
 236                            ;; (add-to-list 'interventionsl1 l1)
 237                            (cl-pushnew l1 interventionsl1 :test #'equal)
 238                            (setq asunitsl1 (1+ asunitsl1)))))))))))))
 239   (reverse interventionsl2)
 240   ;; (write-region (format "%s" interventionsl2) nil (format "transcribe-output-%s-%s-l2.txt" episodenumber personid))
 241   ;; Write raw interventions to file will be supported by a different function
 242   (reverse interventionsl1)
 243   ;; (write-region (format "%s" interventionsl1) nil (format "transcribe-output-%s-%s-l1.txt" episodenumber personid))
 244   ;; (print interventionsl2) ;uncomment to display all the interventions on screen
 245   (let((asunitspersecondl2 (/ asunitsl2 (string-to-number duration)))
 246     (clausesperasunitl2 (/ clausesl2 asunitsl2))
 247     (errorsperasunitl2 (/ errorsl2 asunitsl2))
 248     (asunitspersecondl1 (/ asunitsl1 (string-to-number duration)))
 249     ;; (clausesperasunitl1 (/ clausesl1 asunitsl1))
 250     (initiatingperasunitl2 (/ initiating asunitsl2))
 251     (respondingperasunitl2 (/ responding asunitsl2))
 252     (controlperasunitl2 (/ control asunitsl2))
 253     (expressiveperasunitl2 (/ expressive asunitsl2))
 254     (interpersonalperasunitl2 (/ interpersonal asunitsl2)))
 255
 256     ;; (princ clausesmessage)
 257     (princ (format "episode: %s, duration: %s, person: %s\n" episodenumber duration personid))
 258     (with-current-buffer "Statistics Output"
 259       (insert (format "%s,%s,%s,0,0,%s,%s,%s,%s,%s,QUAN-L2,segmented,aux,level,subject,yearofclil,month\n" personid episodenumber duration role context demand asunitspersecondl2 asunitspersecondl1)))
 260     (princ (format "L2(Asunits/second): %s, L2(clauses/Asunit): %s, L2(errors/Asunit):%s, L1(Asunits/second): %s\n"
 261           asunitspersecondl2 clausesperasunitl2 errorsperasunitl2 asunitspersecondl1))
 262     (princ (format "Functions/unit: Initiating: %s, Responding: %s, Control: %s, Expressive: %s, Interpersonal: %s" initiatingperasunitl2 respondingperasunitl2 controlperasunitl2 expressiveperasunitl2 interpersonalperasunitl2)))))
 263
 264 (defun transcribe-analyze-all ()
 265   "Analyze all file and output to 'Statistics Output' buffer. The buffer will
 266    lost all previous data. The data in the buffer can be saved to a file and be
 267    passed to 'R' for statistical analysis."
 268   (interactive)
 269   (let* ((xml (xml-parse-region (point-min) (point-max)))
 270      (results (car xml))
 271      (episodes (xml-get-children results 'episode)))
 272
 273      (with-current-buffer "Statistics Output"
 274        (erase-buffer)
 275        (insert "person,episode,duration,C-UNITS(L2),C-UNITS(L1),role,context,demand,QUAN-L2,QUAN-L1,QUAL-L2,segmented,aux,level,subjects,yearofCLIL,month\n"))
 276      (dolist (episode episodes)
 277        (let* ((numbernode (xml-get-children episode 'number))
 278          (participantsnode (xml-get-children episode 'participants))
 279          ;; (transcription (xml-get-children episode 'transcription))
 280          (number (nth 2 (car numbernode)))
 281          (participantsstring (nth 2 (car participantsnode)))
 282          (participants (split-string participantsstring)))
 283
 284          (dolist (participant participants)
 285            (transcribe-analyze number participant))))))
 286
 287
 288 (defun transcribe-xml-tag-person (xmltag)
 289   "This function allows the automatic insetion of a speaker xml tag and places the cursor."
 290   (interactive "stag:")
 291   (insert (format "<%s move=\"\"></%s>" xmltag xmltag))
 292   (backward-char 3)
 293   (backward-char (string-width xmltag)))
 294
 295 (defun transcribe-xml-tag (xmltag)
 296   "This function allows the automatic insetion of a custom xml tag and places the cursor."
 297   (interactive "stag:")
 298   (insert (format "<%s></%s>" xmltag xmltag))
 299   (backward-char 3)
 300   (backward-char (string-width xmltag)))
 301
 302 (defun transcribe-region-xml-tag (xmltag)
 303   "This function encapsulates the marked region in the given tag."
 304   (interactive "stag:")
 305   (let ((beginning (region-beginning))
 306        (end (region-end)))
 307   (goto-char beginning)
 308   (insert (format "<%s>" xmltag))
 309   (goto-char end)
 310   (insert (format "</%s>" xmltag))))
 311
 312 (defun transcribe-add-attribute (att val)
 313   "Adds a xml attribute at cursor with the name and value specified (autocompletion possible)"
 314   (interactive (list(completing-read "attibute name:" transcribe-attribute-list)(read-string "value:")))
 315   (insert (format "%s=\"%s\"" att val)))
 316
 317 (defun transcribe-add-attribute-function (val)
 318   "Adds the xml attribute 'function' at cursor with the name specified (autocompletion possible)"
 319   (interactive (list(completing-read "function name:" transcribe-function-list)))
 320   (insert (format "function=\"%s\"" val)))
 321
 322 (defun transcribe-add-attribute-move (val)
 323   "Adds the xml attribute 'move' at cursor with the name specified (autocompletion possible"
 324   (interactive (list(completing-read "move name:" transcribe-move-list)))
 325   (insert (format "move=\"%s\"" val)))
 326
 327 (defun transcribe-xml-tag-l1 ()
 328   "Inserts a l1 tag and places the cursor"
 329   (interactive)
 330   (insert "<l1 clauses=\"1\" errors=\"0\" function=\"\"></l1>")
 331   (backward-char 5))
 332
 333 (defun transcribe-xml-tag-l2 ()
 334   "Inserts a l2 tag and places the cursor"
 335   (interactive)
 336   (insert "<l2 clauses=\"1\" errors=\"0\" function=\"\"></l2>")
 337   (backward-char 5))
 338
 339 (defun transcribe-xml-tag-break (xmltag)
 340   "This function breaks an unit into two. That is, insert a closing and an opening equal tags"
 341   (interactive "stag:")
 342   (insert (format "</%s><%s>" xmltag xmltag)))
 343
 344 (defun transcribe-display-audio-info ()
 345   (interactive)
 346   (emms-player-mpg321-remote-proc)
 347   (shell-command "/usr/bin/mpg321 -R - &"))
 348
 349
 350 (fset 'NewEpisode
 351       "<episode>\n<number>DATE-NUMBER</number>\n<duration></duration>\n<comment></comment>\n<subject>Subject (level)</subject>\n<participants></participants>\n<task>\n\t<role>low or high</role>\n<context>low or high</context>\n<demand>low or high</demand>\r</task>\n<auxiliar>Yes/no</auxiliar>\n<transcription>\n</transcription>\n</episode>");Inserts a new episode structure
 352
 353
 354 (defvar transcribe-mode-map
 355    (let ((map (make-sparse-keymap)))
 356     (define-key map (kbd "C-x C-p") 'emms-play-file)
 357     (define-key map (kbd "C-x C-a") 'transcribe-analyze)
 358     (define-key map (kbd "C-x C-n") 'NewEpisode)
 359     (define-key map (kbd "C-x <down>") 'emms-stop)
 360     (define-key map (kbd "C-x <right>") 'emms-seek-forward)
 361     (define-key map (kbd "C-x <left>") 'emms-seek-backward)
 362     (define-key map (kbd "<f2>") 'transcribe-add-attribute-move)
 363     (define-key map (kbd "<f3>") 'transcribe-add-attribute-function)
 364     (define-key map (kbd "<f4>") 'transcribe-add-attribute)
 365     (define-key map (kbd "<f5>") 'emms-pause)
 366     (define-key map (kbd "<f8>") 'emms-seek)
 367     (define-key map (kbd "<f9>") 'transcribe-xml-tag)
 368     (define-key map (kbd "<f10>") 'transcribe-xml-tag-person)
 369     (define-key map (kbd "<f11>") 'transcribe-xml-tag-l1)
 370     (define-key map (kbd "<f12>") 'transcribe-xml-tag-l2)
 371     map)
 372   "Keymap for Transcribe minor mode.")
 373
 374
 375 (easy-menu-define transcribe-mode-menu transcribe-mode-map
 376   "Menu for Transcribe mode"
 377   '("Transcribe"
 378     ["Raw Output" transcribe-raw-to-buffer]
 379     "---"
 380     ["Analyze" transcribe-analyze]
 381     ["Analyze all" arbitools-analyze-all]
 382     "---"
 383     ["Add transcription header" NewEpisode]
 384     ["Add move attribute" transcribe-add-attribute-move]
 385     ["Add function attribute" transcribe-add-attribute-function]
 386     ["Add L1 intervention" transcribe-xml-tag-l1]
 387     ["Add L2 intervention" transcribe-xml-tag-l2]
 388     ["Add move" transcribe-xml-tag-person]
 389     "---"
 390     ["Play audio file" emms-play-file]
 391     ))
 392
 393
 394 ;;;###autoload
 395 (define-minor-mode transcribe-mode
 396  "Toggle transcribe-mode"
 397   nil
 398   " Trans"
 399   transcribe-mode-map
 400   (generate-new-buffer "Statistics Output")
 401   (generate-new-buffer "Raw Output")
 402 ;;  (with-current-buffer "Raw Output"
 403 ;;    (linum-mode t)
 404 ;;    (setq linum-format "%d "))
 405   (with-current-buffer "Statistics Output"
 406     ;; (insert "person,episode,duration,C-UNITS(L2),C-UNITS(L1),role,context,demand,QUAN-L2,QUAN-L1,QUAL-L2,segmented,aux,level,subjects,yearofCLIL,month\n")
 407   )
 408   ;; TODO: save the students present in transcription in list so that we can use that list for transcribe-analyze-all
 409 )
 410
 411 (provide 'transcribe)
 412
 413 ;;; transcribe.el ends here