How degrotesque works

degrotesque is quite a simple tool.

degrotesque workflow

For each found file, degrotesque performs the following steps:

  • load the file;
  • compute a mask of what to process in dependence to the file type;
  • apply the chosen actions to the unmasked part of the file contents;
  • write the file, optionally building a backup of the file, first.

Loading

The file is plainly loaded. Nothing special, despite the fact that you may define the file's encoding.

Masking

When processing plain text files, degrotesque is applied to the whole content. This cannot be done when processing XML, HTML, or Markdown files as some parts - elements, attributes in XML/HTML-files or code parts in Markdown files - must be omitted.

As such, after loading a file, degrotesque computes a mask, first. The mask shows which parts of the file contents shall be processed and which ones shall be omitted.

The mask is a string of the same length as the file contents. Parts that shall be processed are represented by a '0' ("zero" character), parts to omit by a '1' ("one" character). degrotesque will simply skip the parts marked with a '1' when applying the actions.

Applying actions ("degrotesquing")

degrotesque builds a copy of the chosen actions, first.

When processing the document, the next occurences of each thing to change is determined for each action, first. The nearest of those occurences is taken. If an action consists of two parts (e.g. " should occure twice), the second part is searched. If an action has a second part, but it cannot be found in the document, the action is skipped. Actions for which no occurence has been found are removed from the current list of action. All found occurences are sorted by their position within the document. The first one is then applied by replacing the initial characters by their prettier representation.

Then, the search for the action to process is repeated, starting after the performed replacement (the first part of it).

Saving the result

Nothing special about this as well. Just backup the original file, save the results using the defined encoding.

No, there are no checks whether the encoding understands the changed contents.