./konko CONFIGURATION
./konko-server CONFIGURATION

Example:

./konko example/ice.json
./konko-server example/ice.json

The server is needed so that you can click hyperlinks in the Excel files. Press ctrl-c to stop the server.

Python and XlsxWriter.

If you have OS X and Homebrew, use the following commands to install everything:

brew install python
pip install xlsxwriter

You can also use PyPy for extra speed.

If you have OS X and Homebrew, use the following commands to install everything:

brew install pypy
pip_pypy install xlsxwriter

Then use the tool as follows:

pypy konko CONFIGURATION
pypy konko-server CONFIGURATION

See subdirectory example for some examples.

Which input files to read.

Where to store output: concordance tables, summary tables, log files, and HTML versions of the input files.

The character encoding of the input files.

How many characters of context to show in the concordance table.

(The tool picks words until it has at least this many characters of context, both before the match and after the match.)

What is considered a "word". A regular expession. Matching is case insensitive if word-ignore-case is true.

What is considered a "tag". A regular expession. Matching is case insensitive if tag-ignore-case is true.

Tags have a higher priority than words, so if something looks like both a tag and a word, it will be considered a tag.

Everything that is neither a word nor a tag is considered a separator (typically whitespace or punctuation).

What to search. Each search term is a regular expression. Matching is case insensitive if search-ignore-case is true.

The search terms are matched agains "words" only.

Text identifier. Regular expression, matched againts "tags" only. Matching is case insensitive if tag-ignore-case is true.

May contain parentheses that capture the text identifier. Otherwise the entire tag is considered the text identifier.

Multiple occurrences of the same text identifier are supported. By default there is no text separator, and each file is considered to be a text.

Sample identifier. Regular expression, matched againts "tags" only. Matching is case insensitive if tag-ignore-case is true.

May contain parentheses that capture the sample identifier. Otherwise the entire tag is considered the sample identifier.

Multiple occurrences of the same sample identifier are supported. By default there is no sample separator, and each text is considered to be a sample.

What to delete. A list of regular expressions and/or pairs of regular expressions. Each regular expression is matched againts "tags" only. Matching is case insensitive if tag-ignore-case is true.

A pair of regular expressions can be used to indicate a region that is to be deleted. For example, the following rules delete regions of text that are indicated with <O>...</O> or <X>...</X>

"tag": "<[^<>]+>",
"delete": [
    ["<O>", "</O>"],
    ["<X>", "</X>"],
]

Deletion is conservative if there are non-matching delimiters. In the following two examples, "a" is kept and "b" is deleted:

a <X> a <X> b </X> a

a <X> b </X> a </X> a

The delimiters do not need to be properly nested. In the following examples, "a" is kept and "b" is deleted:

a <X> b <O> b </X> b </O> a

Deletion takes place on the level of texts. Nothing is remembered between texts.

Deletion takes place after we have identified words, tags, texts, and samples.

Deletion takes place before searching.

Deleted words are not reported as matches, and they are not included in word counts. Deleted tags are not shown in the context. Deleted parts are shown with a gray colour in the HTML files.

TCP/IP port number for konko-server (default: 8000).