It's a python script that takes a PDF (usually an academic article) and creates a plain text document (specifically, markdown) with sections for each page of the PDF. Each section has a hi-res image of that page, the extracted text from that page (to facilitate searching) and space for entering notes about that page. The idea is to make it easy to take "in-situ" plain text notes interwoven with the content of an article. This makes it easy to share notes with co-authors and keep everything version-controlled.
The script also creates a folder with all the plaintext snippets from the passed pdf (if they can be
extracted). If each PDF is your pdf library gets this treatment, it
would easy to grep over the plain text of the doc and extract what you need.
And because each text file is named after the page it came from, finding the exact location in the text is straightforward.
This is a Python program for creating the directory structure and some of the files needed for a new project.
Here is an example of the PDF created by create_project.py.
- Creates a high-res image from every page in the passed pdf and crops it to remove whitespace
- Extracts the text from that page
- Creates a markdown document with the same name as your passed pdf,
but the markdown extension,
.mdwith the images, text and space for notes interwoven in the document.
You need to have the following command line tools accessible:
imagemagickimagemagickpdftotexticonv
To set it up, download the package
git clone git@github.com:johnjosephhorton/litfisk.git
cd litfisk
sudo python setup.py install
Once you save your changes, you can simply run:
litfisk test.pdf
The script will create a new directory test with the subdirectories
images and texts.
See the licence file.
grep 'cost' *_ascii.txt
file_15_ascii.txt:with little effort required on our part. The cost was also far less than that of standard
file_15_ascii.txt:lab experiments, at an average cost of less than $1 per subject. However, even this low
file_15_ascii.txt:per-subject cost vastly understates the comparative efficiency of online experiments.
file_15_ascii.txt:We entirely avoided both the costs associated with hiring full-time assistants and the
file_15_ascii.txt:costs of maintaining a laboratory. We also avoided the high initial costs of setting up
file_15_ascii.txt:Of course, low costs would be irrelevant if the results were not informative. And
file_16_ascii.txt:or flat fees for fund transfers, both of which raise the costs of keeping multiple accounts.
file_17_ascii.txt:cost of a more elaborate experimental design.
file_18_ascii.txt:13 Physical laboratory experiments essentially create the same pattern of costs, implying incentives not to
file_20_ascii.txt:are needed, which gives the online laboratory an advantage due to its low costs and
file_24_ascii.txt:norms could raise the cost of bad behavior, with the effect of both fostering honesty
file_24_ascii.txt:is that it would reduce costly duplication of programming effort and design. As a
file_24_ascii.txt:set.20 The Internet makes such sharing a low-cost chore, since the data are invariably
file_25_ascii.txt:as valid as other kinds of experiments, while greatly reducing cost, time and inconvenience. Our replications of well-known experiments relied on MTurk, as MTurk
file_6_ascii.txt:scammers cannot amortize script-writing costs over a larger volume of work). With
file_9_ascii.txt:kinds of experiments that would be very difficult and costly to conduct in offline
The names of the files correspond to the page number, which makes finding the exact usage easy.
- Nowhere but my own personal linux 12.00 LTS machine.