125 lines
3.0 KiB
Python
Executable File
125 lines
3.0 KiB
Python
Executable File
#! /usr/bin/env python
|
|
|
|
"""Consolidate a bunch of CVS or RCS logs read from stdin.
|
|
|
|
Input should be the output of a CVS or RCS logging command, e.g.
|
|
|
|
cvs log -rrelease14
|
|
|
|
which dumps all log messages from release1.4 upwards (assuming that
|
|
release 1.4 was tagged with tag 'release14').
|
|
|
|
This collects all the revision records and outputs them sorted by date
|
|
rather than by file, collapsing duplicate revision record, i.e.,
|
|
records with the same message for different files.
|
|
|
|
The -t option causes it to truncate (discard) the last revision log
|
|
entry; this is useful when using something like the above cvs log
|
|
command, which shows the revisions including the given tag, while you
|
|
probably want everything *since* that tag.
|
|
|
|
XXX This code was created by reverse engineering CVS 1.9 and RCS 5.7.
|
|
|
|
"""
|
|
|
|
import os, sys, getopt, string, re
|
|
|
|
sep1 = '='*77 + '\n' # file separator
|
|
sep2 = '-'*28 + '\n' # revision separator
|
|
|
|
def main():
|
|
"""Main program"""
|
|
truncate_last = 0
|
|
opts, args = getopt.getopt(sys.argv[1:], "-t")
|
|
for o, a in opts:
|
|
if o == '-t':
|
|
truncate_last = 1
|
|
database = []
|
|
while 1:
|
|
chunk = read_chunk(sys.stdin)
|
|
if not chunk:
|
|
break
|
|
records = digest_chunk(chunk)
|
|
if truncate_last:
|
|
del records[-1]
|
|
database[len(database):] = records
|
|
database.sort()
|
|
database.reverse()
|
|
format_output(database)
|
|
|
|
def read_chunk(fp):
|
|
"""Read a chunk -- data for one file, ending with sep1.
|
|
|
|
Split the chunk in parts separated by sep2.
|
|
|
|
"""
|
|
chunk = []
|
|
lines = []
|
|
while 1:
|
|
line = fp.readline()
|
|
if not line:
|
|
break
|
|
if line == sep1:
|
|
if lines:
|
|
chunk.append(lines)
|
|
break
|
|
if line == sep2:
|
|
if lines:
|
|
chunk.append(lines)
|
|
lines = []
|
|
else:
|
|
lines.append(line)
|
|
return chunk
|
|
|
|
def digest_chunk(chunk):
|
|
"""Digest a chunk -- extrach working file name and revisions"""
|
|
lines = chunk[0]
|
|
key = 'Working file:'
|
|
keylen = len(key)
|
|
for line in lines:
|
|
if line[:keylen] == key:
|
|
working_file = string.strip(line[keylen:])
|
|
break
|
|
else:
|
|
working_file = None
|
|
records = []
|
|
for lines in chunk[1:]:
|
|
revline = lines[0]
|
|
dateline = lines[1]
|
|
text = lines[2:]
|
|
words = string.split(dateline)
|
|
if len(words) >= 3 and words[0] == 'date:':
|
|
dateword = words[1]
|
|
timeword = words[2]
|
|
if timeword[-1:] == ';':
|
|
timeword = timeword[:-1]
|
|
date = dateword + ' ' + timeword
|
|
else:
|
|
date = None
|
|
text.insert(0, revline)
|
|
words = string.split(revline)
|
|
if len(words) >= 2 and words[0] == 'revision':
|
|
rev = words[1]
|
|
else:
|
|
rev = None
|
|
text.insert(0, revline)
|
|
records.append((date, working_file, rev, text))
|
|
return records
|
|
|
|
def format_output(database):
|
|
prevtext = None
|
|
prev = []
|
|
database.append((None, None, None, None)) # Sentinel
|
|
for (date, working_file, rev, text) in database:
|
|
if text != prevtext:
|
|
if prev:
|
|
print sep2,
|
|
for (date, working_file, rev) in prev:
|
|
print date, working_file
|
|
sys.stdout.writelines(prevtext)
|
|
prev = []
|
|
prev.append((date, working_file, rev))
|
|
prevtext = text
|
|
|
|
main()
|