little-things/archivebot-jobs

#!/usr/bin/env python3
import argparse
import datetime
import json
import math
import re
import sys
import time
import urllib.request

# Column definitions
columns = {
	'jobid': (lambda job, pipelines: job["job_data"]["ident"], ()),
	'url': (lambda job, pipelines: job["job_data"]["url"], ()),
	'user': (lambda job, pipelines: job["job_data"]["started_by"], ()),
	'pipenick': (lambda job, pipelines: pipelines[job["job_data"]["pipeline_id"]] if job["job_data"]["pipeline_id"] in pipelines else "unknown", ()),
	'queued': (lambda job, pipelines: job["job_data"]["queued_at"], ('date',)),
	'started': (lambda job, pipelines: job["job_data"]["started_at"], ('date',)),
	'last active': (lambda job, pipelines: int(job["ts"]), ('date', 'coloured')),
	'dl urls': (lambda job, pipelines: job["job_data"]["items_downloaded"], ()),
	'dl size': (lambda job, pipelines: job["job_data"]["bytes_downloaded"], ('size',)),
	'queue': (lambda job, pipelines: job["job_data"]["items_queued"] - job["job_data"]["items_downloaded"], ()),
	'con': (lambda job, pipelines: job["job_data"]["concurrency"], ()),
	'delay': (lambda job, pipelines: str(int(job["job_data"]["delay_min"])) + '-' + str(int(job["job_data"]["delay_max"])) if job["job_data"]["delay_min"] != job["job_data"]["delay_max"] else str(int(job["job_data"]["delay_min"])), ()),
}
defaultSort = 'jobid'

# Parse arguments
class FilterAction(argparse.Action):
	def __call__(self, parser, namespace, values, optionString = None):
		global columns
		match = re.match(r"^(?P<column>[A-Za-z ]+)(?P<op>[=<>^*$~])(?P<value>.*)$", values[0])
		if not match:
			raise argparse.ArgumentError('Invalid filter')
		filterDict = match.groupdict()
		filterDict["column"] = filterDict["column"].lower()
		assert filterDict["column"] in columns
		transform = (lambda x: x.lower() if isinstance(x, str) else x) if optionString in ('--ifilter', '-i') else (lambda x: x)
		setattr(namespace, self.dest, (filterDict, transform))

def parse_sort(value):
	global columns
	sortDesc = value.startswith('-')
	if sortDesc:
		value = value[1:]
	value = value.lower()
	if value not in columns:
		raise argparse.ArgumentError('Invalid column name')
	return (value, sortDesc)

class SortAction(argparse.Action):
	def __call__(self, parser, namespace, values, optionString = None):
		result = parse_sort(values[0])
		if getattr(namespace, self.dest, None) is None:
			setattr(namespace, self.dest, [])
		getattr(namespace, self.dest).append(result)

parser = argparse.ArgumentParser(formatter_class = argparse.RawTextHelpFormatter)
parser.add_argument('--sort', '-s', nargs = 1, type = str, action = SortAction, help = "Sort the table by a COLUMN (descending if preceded by '-'). This can be used multiple times to refine the sorting.")
parser.add_argument('--filter', '-f', nargs = 1, type = str, action = FilterAction, help = '\n'.join([
	'Filter the table for rows where a COLUMN has a certain VALUE. If specified multiple times, only the last value is used.',
	'FILTER has the format COLUMN{=|<|>|^|*|$|~}VALUE',
	'  = means the value must be exactly as specified.',
	'  < and > mean it must be less/greater than the specified.',
	'  ^ and $ mean it must start/end with the specified.',
	'  * means it must contain the specified.',
	'  ~ means it must match the specified regex.',
  ]))
parser.add_argument('--ifilter', '-i', nargs = 1, type = str, action = FilterAction, dest = 'filter', help = 'Like --filter but case-insensitive')
parser.add_argument('--no-colours', '--no-colors', action = 'store_true', help = "Don't colourise the last activity column if it's been a while.")
parser.add_argument('--no-table', action = 'store_true', help = 'Raw output without feeding through column(1); columns are separated by tabs.')
parser.add_argument('--dates', action = 'store_true', help = 'Print dates instead of elapsed times for queued/started/last active columns.')
parser.add_argument('--dashboard-regex', action = 'store_true', help = 'Instead of the normal output, compose a regular expression that can be used on the dashboard to actively watch the jobs matched by the filter. (--sort, --no-colours, --no-table, and --dates have no effect when this option is used.)')
args = parser.parse_args()

if not args.sort:
	args.sort = [parse_sort(defaultSort)]

# Retrieve
def fetch(url):
	req = urllib.request.Request(url)
	req.add_header('Accept', 'application/json')
	with urllib.request.urlopen(req) as f:
		if f.getcode() != 200:
			raise RuntimeError('Could not fetch job data')
		return json.load(f)

jobdata = fetch('http://dashboard.at.ninjawedding.org/logs/recent?count=1')
pipelinedata = fetch('http://dashboard.at.ninjawedding.org/pipelines')
currentTime = time.time()

# Process
pipelines = {p["id"]: p["nickname"] for p in pipelinedata["pipelines"]}

jobs = []
for job in jobdata:
	jobs.append({column: columnFunc(job, pipelines) for column, (columnFunc, _) in columns.items()})

if not jobs:
	# Nothing to do
	sys.exit(0)

# Filter
if args.filter:
	filterDict, transform = args.filter
	compFunc = {
		"=": lambda a, b: a == b,
		"<": lambda a, b: a < b,
		">": lambda a, b: a > b,
		"^": lambda a, b: a.startswith(b),
		"*": lambda a, b: b in a,
		"$": lambda a, b: a.endswith(b),
		"~": lambda a, b: re.search(b, a) is not None,
	}[filterDict["op"]]
	if isinstance(jobs[0][filterDict["column"]], (int, float)):
		filterDict["value"] = float(filterDict["value"])
	jobs = [job for job in jobs if compFunc(transform(job[filterDict["column"]]), transform(filterDict["value"]))]

	if not jobs:
		sys.exit(0)

if args.dashboard_regex:
	print('^(' + '|'.join(re.escape(job['url']) for job in jobs) + ')$')
	sys.exit(0)

# Sort
class reversor: # https://stackoverflow.com/a/56842689
	def __init__(self, obj):
		self.obj = obj

	def __eq__(self, other):
		return other.obj == self.obj

	def __lt__(self, other):
		return other.obj < self.obj

sortColumns = tuple((column, descending, columns[column]) for column, descending in args.sort)
if not args.dates:
	# Reverse sorting order for columns which have a date attribute since the column will have elapsed time
	sortColumns = tuple((column, not descending if 'date' in columnInfo[1] else descending, columnInfo) for column, descending, columnInfo in sortColumns)
jobs = sorted(jobs, key = lambda job: tuple(job[column] if not descending else reversor(job[column]) for column, descending, _ in sortColumns))

# Renderers
def render_date(ts, coloured = False):
	global args, currentTime
	diff = currentTime - ts
	colourStr = f"\x1b[{0 if diff < 6 * 3600 else 7};31m" if coloured and diff >= 300 else ""
	colourEndStr = "\x1b[0m" if colourStr else ""
	if args.dates:
		return (colourStr, datetime.datetime.fromtimestamp(ts).isoformat(sep = " "), colourEndStr)
	if diff <= 0:
		return "now"
	elif diff < 60:
		return "<1 min ago"
	elif diff < 86400:
		return (colourStr, (f"{diff // 3600:.0f}h " if diff >= 3600 else "") + f"{(diff % 3600) // 60:.0f}mn ago", colourEndStr)
	else:
		return (colourStr, f"{diff // 86400:.0f}d {(diff % 86400) // 3600:.0f}h ago", colourEndStr)

def render_size(size):
	units = ('B', 'KiB', 'MiB', 'GiB', 'TiB')
	unitIdx = min(int(math.log(size, 1024)), len(units) - 1) if size >= 1 else 0
	if unitIdx == 0:
		return f'{size} B' # No decimal places
	return f'{size / 1024 ** unitIdx:.1f} {units[unitIdx]}'

renderers = {}
for column, (_, columnAttr) in columns.items():
	if "date" in columnAttr:
		if "coloured" in columnAttr:
			renderers[column] = lambda x: render_date(x, coloured = not args.no_colours)
		else:
			renderers[column] = render_date
	elif "size" in columnAttr:
		renderers[column] = render_size
	elif isinstance(jobs[0][column], (int, float)):
		renderers[column] = str

# Print
output = []
output.append(tuple(column.upper() for column in columns))
for job in jobs:
	for column in renderers:
		job[column] = renderers[column](job[column])
	output.append(tuple(job[column] for column in columns))

if not args.no_table:
	widths = tuple(max(len(field) if isinstance(field, str) else len(field[1]) for field in column) for column in zip(*output))
	for row in output:
		print('  '.join((value.ljust(width) if isinstance(value, str) else ''.join((value[0], value[1], value[2], ' ' * (width - len(value[1]))))) for value, width in zip(row, widths)))
else:
	for row in output:
		print('\t'.join(field if isinstance(field, str) else ''.join(field) for field in row))