C&C 108 fcm-database.py

#!/usr/bin/env python
import os, json, datetime, re, csv
# only necessary if you are using the drive CLI tool.
from contextlib import contextmanager
from subprocess import call
# only necessary if you are using the drive CLI tool.

# general settings
txtFile = os.path.join(os.path.expanduser("~"), "article-list.txt") # Target file name list.
databaseFile = "fcm-database.json"

# article-list creation
searchPaths = ['/home/lswest/Documents/GDrive/FCM', '/home/lswest/Dropbox/Articles/Full Circle Magazine/C&C'] # Must be list!

# for database creation
dateStart = {100: datetime.date(2015, 8, 28)}
entryTemplate = {-1: {}} # Template for the entry format - used in the update_database function.

if os.stat(databaseFile).st_size == 0:
    db = {} # create a blank database if the json file is empty.
else:
    with open(databaseFile, 'r') as jsonFile:
        db = json.load(jsonFile) # load the json file into a database, if not empty.
    print("Database Loaded")

@contextmanager
def cd(newdir):
    """ This function is used to call a python equivalent of cd - just required for update_drive """
    prevdir = os.getcwd()
    os.chdir(os.path.expanduser(newdir))
    try:
        yield
    finally:
        os.chdir(prevdir)

def update_drive(drvPath):
    """ This function is used to call the drive pull command - to update any changes to the folder.  If you don't use drive, just comment out the call to this function in main() """
    with cd(drvPath):
        call(["drive", "pull"])
    return True

def dateFind(issue, startPoint):
    """This function calculates the date based off the number issues to/from 100.  I.e. subtracts 2 months if considering issue 98."""
    dateDiff = 100 - int(issue.replace("FCM", ""))
    return (startPoint[100] - datetime.timedelta((dateDiff * 365) / 12)).isoformat()

def createArticleList(target, paths):
    """This generates the text file containing all file names that should be included in the database.  Also does some basic cleanup (deletes .odt and changes Command & Conquer into C&C)"""
    regexp = re.compile("FCM[0-9]+")
    output = []
    for path in paths:
        files = [f for f in os.listdir(path) if os.path.isfile(os.path.join(path, f)) and regexp.search(f) is not None and ".desktop" not in f]
        output.append(files)
    output = [item for sublist in output for item in sublist]
    output = [w.replace(".odt", "") for w in output]
    output = [w.replace("Command & Conquer", "C&C") for w in output]
    output = set(output)
    with open(target, 'w') as f:
        for line in output:
            f.write(line+"\n")
    f.close()
    return True


def update_database(database, txtPath, entry):
    """This function uses the text file created in createArticleList, and splits the file names into a dict object, which is added to the JSON database."""
    with open(txtPath, 'r') as f:
        for line in f.readlines():
            if line != "\n":
                items = line.split(" - ")
                number = items[0].replace("FCM", "")
                entry[number] = {}
                entry[number]["issue"] = number
                if len(items) > 2:
                    entry[number]["article"] = items[2].replace("\n", "")
                else:
                    entry[number]["article"] = "Unknown"
                entry[number]["column"] = items[1]
                entry[number]["date"] = dateFind(items[0], dateStart)
                database.update(entry)
        f.close()
        del database[-1]
        return database

def write_database(database, dbPath):
    """Writes the database to the json file."""
    with open(dbPath, 'w') as f:
        f.write(json.dumps(database, indent=4))
    return True

def write_csv_database(database, dbPath):
    """For easy import into some spreadsheets programs, you can save the database as a CSV file as well."""
    f = csv.writer(open(dbPath, "w"))

    allKeys = list(database["100"].keys()) # used to ensure values and keys match up.
    f.writerow(allKeys)

    for key in database:
        f.writerow([database[key][allKeys[0]], database[key][allKeys[1]], database[key][allKeys[2]], database[key][allKeys[3]]])

def main():
    #update_drive(searchPaths[0])
    createArticleList(txtFile, searchPaths)
    dbN = update_database(db, txtFile, entryTemplate)
    write_database(dbN, databaseFile)
    #write_csv_database(db, "fcm-database.csv")

if __name__ == "__main__":
    main() # Function is called, instead of inserting methods directly into this section.  Useful for eventual imports of the script.