CONTRIBUTING.md

Contributing

Code Style

odsearcher has a strict code style. It is based on PEP 8, but with a few extra rules:

• Indentation is always tabs. No spaces. No exceptions. Ever.
• There should be a blank line between every function, and two blank lines between every class.
• There should be a blank line between every import (not __import__) and the code that follows it.
• Use. Markdownlint. If your markdown isn't compliant with the rules (except no-duplicate-header), I will not accept your PR until you fix it. (Quite frankly, it's not that hard to follow the rules.)

Pull Requests

When making a pull request, please make sure to follow these guidelines:

• Make sure your code follows the code style guidelines (above).
• Make sure your changes are documented in the pull request description.
• Make sure your pull request is up to date with the latest version of the master branch.
• When dealing with merge conflicts, make sure to resolve them in favor of the master branch, unless they are critical to your changes.

Issues

If you find a bug or a missing key feature, please report it on the issues page on GitHub. Please make sure to follow these guidelines:

• Make sure your issue is not a duplicate of another issue.
• Ensure that your issue is related to odsearcher, and not the indexer.
• Always include a command or example index that reproduces the issue, if applicable.

Adding Support for a Directory

In the /utils folder, you'll find a list of supporting modules/scripts for the script as a whole. In order to add your own support for a directory that uses a different format for indexes (and is too large to index otherwise), as was the case with the-eye.ru, create a new file there and implement the following:

• register_args(parser) - Takes the argument parser from the main script and adds its own command-line arguments to it. If this isn't required, you can skip it.
• do_search(target_string, args) - Takes the target search string and the passed command-line arguments and searches the directory for the file. You can use any form of fuzzy search here (for example, if the site already provides an integrated search feature), but fuzzysearch is perferred. This code should do the trick for checking a given record, record, and the target string, target_string:

if fuzzysearch.find_near_matches(target_string, record.lower(), max_l_dist=1) or fuzzysearch.find_near_matches(target_string.replace(" ", "."), record.lower(), max_l_dist=1) or fuzzysearch.find_near_matches(target_string.replace(" ", "_"), record.lower(), max_l_dist=1):
   # matched, do something

After finding all records on the directory, do_search should return an tuple of a list of strings and an int: the matches, and the number of matches found.

After creating your module, you'll need to add it to the main searcher.py file. Just look for a line towards the top of the file defining a list called utils. Just add the name of your python script to the end of that list, and it should be automatically imported. Now, look in the main search() function. You'll need to add your module's search to the chain of search. As of commit f1fc3fc, that chain looks like this:

    total = 0
    matches = []
    if not args.eye_only:
        for file in os.listdir():
            if file.endswith(".txt"):
                print("\nSearching in " + file + " ...")
                with open(file, "r", errors='replace') as f:
                    text = f.read()
                    i = 0
                    for line in text.splitlines():
                        if fuzzysearch.find_near_matches(name, line.lower(), max_l_dist=1) or fuzzysearch.find_near_matches(name.replace(" ", "."), line.lower(), max_l_dist=1) or fuzzysearch.find_near_matches(name.replace(" ", "_"), line.lower(), max_l_dist=1):
                            if check_filetype(line):
                                print(line)
                                total += 1
                                matches.append(line)
                        i += 1
                print("Done searching in " + file + ".\n")
                
    # extra utils here
    _matches, _total = the_eye.do_search(name, args)
    matches += _matches
    total += _total

At the very bottom, you'll want to duplicate the call to the_eye.do_search() and replace the_eye with the name of your directory. Hopefully, everything will work first-try, and you'll have your new directory up and running in no time!