The code for blindliteratur.de, a dummy text generator in love with German literature classics.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
data
img
README.md
about.php
header.php
impressum.php
index.php
script.js
style.css

README.md

Blindliteratur

The code for blindliteratur.de, a dummy text generator in love with German literature classics.

Installation

Blindliteratur runs on PHP. Simply copy the files to a server. This repository does not come with any literature. You will have to supply your own, e.g. download your favourite books from one of various online ressources.

How to fill with literature

Blindliteratur feeds on books to be placed in the /data/content/ subfolder. First, create a data.txt file in this folder and fill it with pairs of unique author IDs and author names like this:

goethe: Johann Wolfgang v. Goethe
heine: Heinrich Heine
schiller: Friedrich Schiller

The chosen author IDs should correspond with further subfolders in that same directory, e.g. /data/content/goethe. Fill these folders with further text files containing all the content available to the generator:

Die Leiden des jungen Werther.txt
Faust.txt
Iphigenie auf Tauris.txt

Or alternatively, if you like to include chapter data, create subfolders – e.g. /data/content/goethe/Faust etc. - and put separate text files – Kapitel I.txt, Der Tragödie Erster Teil.txt and so on – there.

Some recommendations on text quality

Evidently, the longer your text files, the better. For best results, they should contain no less than 5.000 characters to make sure that the snippet choice is random enough. You may tweak variables $minsentence and $minwords in /data/fetch-data.php to adjust the minimum number of sentences or words, respectively, a snippet will have. Sanitize your texts (e.g. replace triple full stops with ellipses, trim line breaks and repeating spaces etc.) for better results. Due to their line rhythm, drama and poetry don't work as well as prose when its content is regrouped in paragraphs.

Known issues

  • Text that is too short may lead to infinite loops in getSnippet()
  • quotes in direct speech not recognized as part of their respective sentences
  • remains untested in Internet Explorer

Features that might or might not be included in a better, future version

  • an API for direct reference in external projects
  • an option to generate random HTML markup
  • more sophisticated sentence recognition
  • deep writing algorithm to mimic famous authors' styles