JustAnotherArchivist
/
little-things
2
0
Fork 0
Code Issues 0 Pull Requests 0 Releases 0 Wiki Activity
Browse Source

Refactor the Bash/Python abomination into a pure Python script so I get to keep my sanity while editing

master
JustAnotherArchivist 4 years ago
parent
cf879c86c9
commit
3f7d84ab12
1 changed files with 163 additions and 219 deletions
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. +163
    -219
      archivebot-jobs

+ 163
- 219
archivebot-jobs View File

@@ -1,227 +1,171 @@
#!/bin/bash
declare -a columns columndefs columnattributes
columns+=('jobid'); columndefs+=('job["job_data"]["ident"]'); columnattributes+=('');
columns+=('url'); columndefs+=('job["job_data"]["url"]'); columnattributes+=('');
columns+=('user'); columndefs+=('job["job_data"]["started_by"]'); columnattributes+=('');
columns+=('pipenick'); columndefs+=('pipelines[job["job_data"]["pipeline_id"]] if job["job_data"]["pipeline_id"] in pipelines else "unknown"'); columnattributes+=('');
columns+=('queued'); columndefs+=('job["job_data"]["queued_at"]'); columnattributes+=('date');
columns+=('started'); columndefs+=('job["job_data"]["started_at"]'); columnattributes+=('date');
columns+=('last active'); columndefs+=('int(job["ts"])'); columnattributes+=('date,coloured');
function valid_column {
local candidate="$1"
local column
for column in "${columns[@]}"
do
[[ "${candidate}" == "${column}" ]] && return 0
done
return 1
#!/usr/bin/env python3
import argparse
import datetime
import json
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')),
}

sortcolumns=()
filter=
filtercaseinsensitive=
nocolours=
notable=
dates= # Whether to use full dates for the time columns rather than expired time strings
while [[ $# -gt 0 ]]
do
if [[ "$1" == "--help" || "$1" == "-h" ]]
then
echo "Usage: archivebot-jobs [options]" >&2
echo "Prints a table of current AB jobs" >&2
echo "Options:" >&2
echo " --help, -h Show this message and exit." >&2
echo " --sort [-]COLUMN, -s Sort the table by a column (descending if preceded by '-'). This can be used multiple times to refine the sorting." >&2
echo " --filter FILTER, -f Filter the table for rows where a COLUMN has a certain VALUE. If specified multiple times, only the last value is used." >&2
echo " The FILTER has this format: COLUMN{=|<|>|^|*|$|~}VALUE" >&2
echo " = 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." >&2
echo " --ifilter FILTER, -i Like --filter, but case-insensitive" >&2
echo " --no-colours, --no-colors Don't colourise the last activity column if it's been a while." >&2
echo " --no-table Raw output without feeding through column(1); columns are separated by tabs." >&2
echo " --dates Print dates instead of elapsed times for queued/started/last active columns." >&2
echo "The COLUMNs are the names of each column, printed in capital letters in the first line of the output." >&2
exit 0
elif [[ "$1" == "--sort" || "$1" == "-s" ]]
then
sortcolumns+=("$2")
shift
elif [[ "$1" == "--filter" || "$1" == "-f" ]]
then
filter="$2"
filtercaseinsensitive=
shift
elif [[ "$1" == "--ifilter" || "$1" == "-i" ]]
then
filter="$2"
filtercaseinsensitive=1
shift
elif [[ "$1" == "--no-colours" || "$1" == "--no-colors" ]]
then
nocolours=1
elif [[ "$1" == "--no-table" ]]
then
notable=1
elif [[ "$1" == "--dates" ]]
then
dates=1
else
echo "Unknown option: $1" >&2
exit 1
fi
shift
done

# Validate sortcolumns and filter
if [[ "${filter}" ]]
then
if [[ "${filter}" == *$'\n'* ]]
then
echo "Invalid filter: newlines not allowed" >&2
exit 1
fi
if [[ ! ( "${filter}" == *'='* || "${filter}" == *'<'* || "${filter}" == *'>'* || "${filter}" == *'^'* || "${filter}" == *'*'* || "${filter}" == *'$'* || "${filter}" == *'~'* ) ]]
then
echo "Invalid filter: ${filter}" >&2
exit 1
fi
column="${filter%%[=<>^*$~]*}"
if ! valid_column "${column,,}"
then
echo "Invalid filter column: ${column}" >&2
exit 1
fi
fi
if [[ ${#sortcolumns[@]} -gt 0 ]]
then
for column in "${sortcolumns[@]}"
do
column="${column#-}"
if ! valid_column "${column,,}"
then
echo "Invalid sort column: ${column}" >&2
exit 1
fi
done
else
# Default sort order
sortcolumns+=("jobid")
fi

if [[ "${notable}" ]]
then
column=("cat")
else
column=("column" "-t" $'-s\t')
fi

jobdata="$(curl -s -H "Accept: application/json" "http://dashboard.at.ninjawedding.org/logs/recent?count=1" 2>/dev/null)"
pipelinedata="$(curl -s -H "Accept: application/json" "http://dashboard.at.ninjawedding.org/pipelines" 2>/dev/null)"

if [[ -z "${jobdata}" || -z "${pipelinedata}" ]]
then
echo "Error retrieving job or pipeline data" >&2
exit 1
fi

{ echo "${jobdata}"; echo "${pipelinedata}"; echo "${filter}"; } | python3 -c \
'
if True: # For sensible indentation
import datetime
import json
import sys
import time

currentTime = time.time()
def render_date(ts, coloured = False):
global 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 "'${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

jobdata = json.loads(sys.stdin.readline())
pipelinedata = json.loads(sys.stdin.readline())
filter = sys.stdin.readline().strip()

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

jobs = []
for job in jobdata:
jobs.append({'"$(for i in ${!columns[@]}; do echo '
"'"${columns[$i]}"'": '"${columndefs[$i]}"','; done)"'
})

columns = ('"$(for column in "${columns[@]}"; do echo '"'"${column}"'", '; done)"')
columnAttributes = {'"$(for i in ${!columns[@]}; do echo -n '"'"${columns[$i]}"'": "'"${columnattributes[$i]}"'".split(","), '; done)"'}

# Filter
if filter:
import re
match = re.match(r"^(?P<column>[A-Za-z ]+)(?P<op>[=<>^*$~])(?P<value>.*)$", filter)
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
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,
"$": lambda a, b: a.endswith(b),
"~": lambda a, b: re.search(b, a) is not None,
}[filterDict["op"]]
if isinstance(jobs[0][filterDict["column"]], (int, float)):
filterDict["value"] = float(filterDict["value"])
transform = lambda x: x.lower() if "'${filtercaseinsensitive}'" and isinstance(x, str) else x
jobs = [job for job in jobs if compFunc(transform(job[filterDict["column"]]), transform(filterDict["value"]))]
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.')
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)

# 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

sortColumnsRaw = ('"$(printf "'%s', " "${sortcolumns[@]}")"')
sortColumns = tuple((column[1:] if column.startswith("-") else column).lower() for column in sortColumnsRaw)
sortColumnAsc = tuple(not column.startswith("-") for column in sortColumnsRaw)
assert all(column in columns for column in sortColumns)
if not "'${dates}'":
# Reverse sorting order for columns which have a date attribute since the column will have elapsed time
sortColumnAttrs = tuple(columnAttr for column, columnAttr in columnAttributes.items() if column in sortColumns)
sortColumnAsc = tuple(not columnAsc if "date" in columnAttr else columnAsc for columnAsc, column, columnAttr in zip(sortColumnAsc, sortColumns, sortColumnAttrs))
jobs = sorted(jobs, key = lambda job: tuple(job[column] if columnAsc else reversor(job[column]) for column, columnAsc in zip(sortColumns, sortColumnAsc)))

# Renderers
renderers = {}
for column, columnAttr in columnAttributes.items():
if "date" in columnAttr:
if "coloured" in columnAttr:
renderers[column] = lambda x: render_date(x, coloured = not "'${nocolours}'")
else:
renderers[column] = render_date

# Print
print("\t".join(column.upper() for column in columns))
for job in jobs:
for column in renderers:
job[column] = renderers[column](job[column])
print("\t".join(job[column] for column in columns))
' | "${column[@]}"
# 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)

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

# 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))

Write Preview
Loading…
Cancel
Save