Create .nfo-files recursively - find

Question

I have a collection of concert-media (audio and video). They're all organised using the same pattern:
./ARTISTNAME/[YYYY-MM-DD] VENUE, CITY
Basically I want to write a script which goes through the folders, looks for the [YYYY-MM-DD]-folders and get the information regarding artist (one folder level above), date, venue and location (using the name of the folder it just found) and writing the information into a .nfo-file which it saves into the found folder.
I already did quite a research on this topic, found a similar script but I am stuck because it searches files instead of folders:
#!/bin/bash
find . -path "*[????-??-??]*" | while read folder; do # the script is supposed to look for the concert-folders
-> band_name=`echo $file_name | sed 's/ *-.*$//'` # Get rid of song name
-> band_name=`echo $band_name | sed 's/^.*\///'` # Get rid of file path
-> song_name=`echo $file_name | sed 's/^.*- *//'` # Get rid of band name
-> song_name=`echo $song_name | sed 's/.avi//'` # Get rid of file extension
-> new_file_name=`echo $file_name | sed 's/.avi/.nfo/'` # Make new filename
-> echo "Making $new_file_name..."
echo -en "<musicvideo>\n<venue>$venue_name</venue>\n<city>$city_name</city><date>$date>\date>\n<artist>$band_name</artist>\n</musicvideo>\n" > "$new_file_name"
done
After changing the first part of the script (making it look for the folders with "[YYYY-MM-DD]") I understand that the second part of the script allocates the "tags" (such as artist, date, location, etc.). But I don't know how to make the script take the tags from folder names. Basically help is needed after the "->".
At the last part of the script it is supposed to write the collected information for this folder into a .nfo-file (e.g. FOLDERNAME.nfo).

Answer 1

Here is a simple Python example to get you started. If you want to port it to shell you can.
First, my setup:
test
test/BOB
test/BOB/[3011-01-01] Lollapalooza 213, Saturn Base 5
test/THE WHO
test/THE WHO/[1969-08-17] Woodstock, Woodstock
The code:
#!/usr/bin/env python
import os
import os.path
import re
import sys
def handle_concert(dirname, artist, date, venue, city):
"""Create a NFO file in the directory."""
# the {} are replaced by each argument in turn, like printf()
# Using a triple quote is a lot like a here document in shell, i.e.
# cat <EOF
# foo
# EOF
nfo_data = """<musicvideo>
<venue>{}</venue>
<city>{}</city>
<date>{}</date>
<artist>{}</artist>
</musicvideo>
""".format(venue, city, date, artist)
nfo_file = "[{}] {}, {}.nfo".format(date, venue, city)
# when the with statement is done the file is closed and fully
# written.
with open(os.path.join(dirname, nfo_file), "w") as fp:
fp.write(nfo_data)
# This is a regular expression which matches:
# */FOO/[YYYY-MM-DD] VENUE, CITY
# Where possible, white space is left intact
concert_re = re.compile(r'.*/(?P<artist>.+)/\[(?P<date>\d{4}-\d{2}-\d{2})\]\s+(?P<venue>.+),\s+(?P<city>.+)')
def handle_artist(dirname):
"""Found an ARTIST directory. Look for concerts.
If a subdirectory is found, see if it matches a concert.
When a concert is found, handle it.
"""
for i in os.listdir(dirname):
subdir = os.path.join(dirname, i)
m = concert_re.match(subdir)
if m:
print subdir # to watch the progress
handle_concert(subdir, m.group("artist"), m.group("date"),
m.group("venue"), m.group("city"))
def walk_collection(start_dir):
"""Examine contents of start_dir.
If a directory is found, assume it is an ARTIST.
"""
for i in os.listdir(start_dir):
# os.path.join ensures paths are right regardless of OS
dirname = os.path.join(start_dir, i)
if os.path.isdir(dirname):
print dirname # to watch the progress
handle_artist(dirname)
if __name__ == "__main__":
collection_dir = sys.argv[1] # equiv of $1
walk_collection(collection_dir)
How to run it:
$ nfo-creator.py /path/to/your/collection
The results:
<musicvideo>
<venue>Woodstock</venue>
<city>Woodstock</city>
<date>1969-08-17</date>
<artist>THE WHO</artist>
</musicvideo>
and
<musicvideo>
<venue>Lollapalooza 213</venue>
<city>Saturn Base 5</city>
<date>3011-01-01</date>
<artist>BOB</artist>
</musicvideo>
You can add more handle_ functions to handle differing formats. Just define a new regular expression and a handler for it. I kept this really simple for learning purposes with plenty of shell oriented comments.
This could totally be done in the shell. But writing this Python code was way easier and allows for more growth in the future.
Enjoy, and feel free to ask questions.