User:Eleanorg/2.1/Prototypes/transcription: Difference between revisions

Revision as of 10:41, 26 September 2012

Asking people to transcribe text and submit it to a central document. Will they transcribe verbatim?

Feedback

From one of the guinea-pigs:

So you've asked me about my copy of your synthesis of piet zwart media design first year.

I did copy it, in the photocopying machine. I placed in the copying A3 glass surface as much material as if it would fit that space and made A3 copies of it.
As the 1st generation of copied pages had some white spaces, I choose to take away by riping little bits of the paper (to much bright surfaces).
I then placed the pages with holes on top of each-other, so that the ones on top could show behind the ones at the bottom.
As I am a bit restless and nobody was around I decided to move the pages while being scanned.
The 2nd generation appeared with a a motion effect and the same material often reoccurred.

Did I need to do all this? No. I could have just made a verbatim copy of what you gave, or (re)digitized the material and keep in some hard-drive. Maybe becasue you told me that some people decided to make verbatim copies I thought: "that's too easy".  Question is if what I did is any different that a verbatim copy only with less resolution, less linearity, harder reading, or does it manages to say something more about those works present in your pages?

Code

make db of imgs waiting to be transcribed

#!/usr/bin/python
#-*- coding:utf-8 -*-

import pymongo
from pymongo import Connection


#======== create db with img files

connection = Connection()
myDB = connection['consentTexts1']
collection = myDB.collection  

for x in range(0,11):
	title = "text" + str(x)
	fileName = "img" + str(x) + ".jpg"
	
	
	sentence = {'title': title, 'file': fileName, 'status': "waiting", 'text':" " }
	collection.insert(sentence)

input form

Show an img waiting to be transcribed, with input form for transcription.

#!/usr/bin/python
#-*- coding:utf-8 -*-

 
import cgi
import cgitb; cgitb.enable()
import pymongo
from pymongo import Connection
import random

#========= get untranscribed img file

connection = Connection()
myDB = connection['consentTexts1']
collection = myDB.collection 

waiting = []
for entry in myDB.collection.find({"status": "waiting"}):
    waiting.append(entry)			# add this hash to 'waiting' list
#print waiting
 
howMany = len(waiting)				# find out how many items in 'waiting' list
#print howMany
random = random.randint(0,howMany)		# pick a random number in this range
#print random
sentence = waiting[random]			# ...and choose the sentence at this index



#========== show input form

print "Content-Type: text/html"
print 
print """
<!DOCTYPE html>
<html>
  <head>
    <title></title>
    <style type="text/css">
    	.textImg { height: 300px; float: left;}
    	.textInput { height: 100px; width: 500px; margin: 0px 0px 30px 0px; }
    </style>

  </head>
  
<body>

<form action="updatedb.cgi" name="inputForm">	
	<img src=" """ + sentence['file'] + """ "  class="textImg"/> 
	<br />
	Transcription: <br />
	<textarea name="text" class="textInput" rows="3"></textarea><br />
	Your name (optional):<br />
	<input type="text" name="author" value="">
	<input style="" name="title" value=" """ + sentence['title'] + """ ">
	<input type="submit" value="Submit" />
</form>

</body>

</html>"""

proces input form

Change status of that sentence in db to 'finished', & add the transcribed text & author's name.

#!/usr/bin/python
#-*- coding:utf-8 -*-

import cgi
import cgitb; cgitb.enable()
import pymongo
from pymongo import Connection

#======== get text from input form

form = cgi.FieldStorage()			# Grabs whatever input comes from form
text = form.getvalue("text", "(blank transcription)")		
title = form.getvalue("title")
name = form.getvalue("author", "anon")


#========= put it in the db

connection = Connection()
myDB = connection['consentTexts1']
collection = myDB.collection  

collection.update( {'title': title}, {"$set":{'status': 'done', 'text': text, 'author': name}} )


#========= print thank you
print "Content-Type: text/html"
print 
print """
<!DOCTYPE html>
<html>
  <head>
    <title></title>
  </head>
  
<body>thanks, you added this transcription:<br/> <br />
<em>""" + text + """</em><br /><br />
for """ + title + """
View the other <a href="showTexts.cgi">transcribed text so far</a>.


</body>

</html>
""""

display transcribed texts

CGI displays an html doc of all db entries. Using a loop to stop entries printing in the strange order they're stored in db, starting with number 8 rather than 0 - ?

#!/usr/bin/python
#-*- coding:utf-8 -*-

import cgi
import cgitb; cgitb.enable()
import pymongo
from pymongo import Connection



connection = Connection()
myDB = connection['consentTexts1']
collection = myDB.collection 


#========= show all db entries
# tr appears blank for those not yet transcribed.

print "Content-Type: text/html"
print 
print """
<!DOCTYPE html>
<html>
  <head>
    <title></title>
    <style type="text/css">
    	table {margin: auto;}
    	tr {height:40px;}
    	.text {font-size: 30px; width: 450px; text-align:right;}
    	.name {font-size:9px; min-width: 200px; padding: 10px;}
    </style>

  </head>
  
<body>
<table>"""

# loop to make sure db entries are printed in sensible order, 0-10
for x in range(0,11):
	entry = myDB.collection.find_one({'title': 'text' + str(x)})
	print """<tr>
			<td class="text">""" + entry['text'] + """</td>
			<td class="name">(""" + entry['author'] + """)</td>
	      </tr>"""
	
	
print """</table>
</body>

</html>"""

@@ Line 1: / Line 1: @@
 Asking people to transcribe text and submit it to a central document. Will they transcribe verbatim?
+==Feedback==
+From one of the guinea-pigs:
+<source>
+So you've asked me about my copy of your synthesis of piet zwart media design first year.
+I did copy it, in the photocopying machine. I placed in the copying A3 glass surface as much material as if it would fit that space and made A3 copies of it.
+As the 1st generation of copied pages had some white spaces, I choose to take away by riping little bits of the paper (to much bright surfaces).
+I then placed the pages with holes on top of each-other, so that the ones on top could show behind the ones at the bottom.
+As I am a bit restless and nobody was around I decided to move the pages while being scanned.
+The 2nd generation appeared with a a motion effect and the same material often reoccurred.
+Did I need to do all this? No. I could have just made a verbatim copy of what you gave, or (re)digitized the material and keep in some hard-drive. Maybe becasue you told me that some people decided to make verbatim copies I thought: "that's too easy".  Question is if what I did is any different that a verbatim copy only with less resolution, less linearity, harder reading, or does it manages to say something more about those works present in your pages?
+</source>
+==Code==
 ===make db of imgs waiting to be transcribed===